Emily Peterson, PharmD, BCACP

Migraine is a headache disorder that often causes unilateral pain, photophobia, phonophobia, nausea, and vomiting. More than 70% of office visits for migraine are made to primary care physicians.¹ Recent data suggest migraine may be caused primarily by neuronal dysfunction and only secondarily by vasodilation.² Although there are numerous classes of drugs used for migraine prevention and treatment, their success has been limited by inadequate efficacy, tolerability, and patient adherence.³ The discovery of pro-inflammatory markers such as calcitonin gene-related peptide (CGRP) has led to the development of new medications to prevent and treat migraine.⁴

JFP07209292_f.jpg

Pathophysiology, Dx and triggers, indications for pharmacotherapy

Pathophysiology. A migraine is thought to be caused by cortical spreading depression (CSD), a depolarization of glial and neuronal cell membranes.⁵This results in increased cortical excitability, central trigeminal-thalamic sensitization, and defective descending pain modulatory activity.⁶The activation of the trigeminal sensory pathways, primarily the ophthalmic branch, sends nociceptive signals to second-order neurons mediated by the release of neurotransmitters, such as CGRPs.⁵ This activation explains in part the primary location for a migraine, which is around the eye and the neighboring cranial regions. The pain perceived by the patient is caused by these second-order neurons.

It has been theorized that gepants bind to calcitonin gene-related peptide receptors, resulting in decreased blood flow to the brain, inhibition of neurogenic inflammation, and reduced pain signaling.

Dx and triggers. In 2018, the International Headache Society revised its guidelines for the diagnosis of migraine.⁷ According to the 3rd edition of The International Classification of Headache Disorders (ICHD-3), the diagnosis of migraine is made when a patient has at least 5 headache attacks that last 4 to 72 hours and have at least 2 of the following characteristics: (1) unilateral location, (2) pulsating quality, (3) moderate-to-severe pain intensity, and (4) aggravated by or causing avoidance of routine physical activity.⁷ The headache attacks also should have (1) associated nausea or vomiting or (2) photophobia and phonophobia.⁷ The presence of atypical signs or symptoms as indicated by the SNNOOP10 mnemonic raises concerns for secondary headaches and the need for further investigation into the cause of the headache (TABLE 1).⁸ It is not possible to detect every secondary headache with standard neuroimaging, but the SNNOOP10 red flags can help determine when imaging may be indicated.⁸ Potential triggers for migraine can be found in TABLE 2.⁹

JFP07209292_t1.jpg

Indications for pharmacotherapy. All patients receiving a diagnosis of migraine should be offered acute pharmacologic treatment. Consider preventive therapy anytime there are ≥ 4 headache days per month, debilitating attacks despite acute therapy, overuse of acute medication (> 2 d/wk), difficulty tolerating acute medication, patient preference, or presence of certain migraine subtypes.^7,10

JFP07209292_t2.jpg

Acute treatments

Abortive therapies for migraine include analgesics such as nonsteroidal anti-­inflammatory drugs (NSAIDs) and acetaminophen, and ergot alkaloids, triptans, or small-molecule CGRP receptor antagonists (gepants). Prompt administration increases the chance of success with acute therapy. Medications with the highest levels of efficacy based on the 2015 guidelines from the American Headache Society (AHS) are given in TABLE 3.¹¹ Lasmiditan (Reyvow) is not included in the 2015 guidelines, as it was approved after publication of the guidelines.

JFP07209292_t3a.jpg

JFP07209292_t3b.jpg

Non-CGRP first-line therapies

NSAIDs and acetaminophen. NSAIDs such as aspirin, diclofenac, ibuprofen, and naproxen have a high level of evidence to support their use as first-line treatments for mild-to-moderate migraine attacks. Trials consistently demonstrate their superiority to placebo in headache relief and complete pain relief at 2 hours. There is no recommendation for selecting one NSAID over another; however, consider their frequency of dosing and ­adverse effect profiles. The number needed to treat for complete pain relief at 2 hours ranges from 7 to 10 for most NSAIDs.^11,12 In some placebo-controlled studies, acetaminophen was less effective than NSAIDs, but was safer because it did not cause gastric irritation or antiplatelet effects.¹²

Triptans inhibit 5-HT_1B/1D receptors. Consider formulation, route of administration, cost, and pharmacokinetics when selecting a triptan. Patients who do not respond well to one triptan may respond favorably to another. A meta-analysis of the effectiveness of the 7 available agents found that triptans at standard doses provided pain relief within 2 hours in 42% to 76% of patients, and sustained freedom from pain for 2 hours in 18% to 50% of patients.¹³ Lasmiditan is a selective serotonin receptor (5-HT_1F) agonist that lacks vasoconstrictor activity. This is an option for patients with relative contraindications to triptans due to cardiovascular risk factors.¹⁰

Continue to: Second-line therapies

Intranasal dihydroergotamine has a favorable adverse event profile and greater evidence for efficacy compared with ergotamine. Compared with triptans, intranasal dihydroergotamine has a high level of efficacy but causes more adverse effects.¹⁴ Severe nausea is common, and dihydroergotamine often is used in combination with an antiemetic drug. Dihydroergotamine should not be used within 24 hours of taking a triptan, and it is contraindicated for patients who have hypertension or ischemic heart disease or who are pregnant or breastfeeding. There is also the potential for adverse drug interactions.¹⁵

A meta-analysis found that triptans at standard doses provided pain relief within 2 hours in 42% to 76% of patients, and sustained freedom from pain for 2 hours in 18% to 50% of patients.

Antiemetics may be helpful for migraine associated with severe nausea or vomiting. The dopamine antagonists metoclopramide, prochlorperazine, and chlorpromazine have demonstrated benefit in randomized ­placebo-controlled trials.¹¹ Ondansetron has not been studied extensively, but sometimes is used in clinical practice. Nonoral routes of administration may be useful in patients having trouble swallowing medications or in those experiencing significant nausea or vomiting early during migraine attacks.

Due to the high potential for abuse, opioids should not be used routinely for the treatment of migraine.¹² There is no high-quality evidence supporting the efficacy of barbiturates (ie, butalbital-containing compounds) for acute migraine treatment.¹¹ Moreover, use of these agents may increase the likelihood of progression from episodic to chronic migraine.¹⁶

Gepants for acute migraine treatment

Neuropeptide CGRP is released from trigeminal nerves and is a potent dilator of cerebral and dural vessels, playing a key role in regulating blood flow to the brain. Other roles of CGRP include the release of inflammatory agents from mast cells and the transmission of painful stimuli from intracranial vessels.¹⁷ The CGRP receptor or ligand can be targeted by small-molecule receptor antagonists for acute and preventive migraine treatment (and by monoclonal antibodies solely for prevention, discussed later). It has been theorized that gepants bind to CGRP receptors, resulting in decreased blood flow to the brain, inhibition of neurogenic inflammation, and reduced pain signaling.¹⁷ Unlike triptans and ergotamine derivatives, these novel treatments do not constrict blood vessels and may have a unique role in patients with contraindications to triptans.

The 3 gepants approved for acute treatment—ubrogepant (Ubrelvy),¹⁸ rimegepant (Nurtec),¹⁹ and zavegepant (Zavzpret)²⁰—were compared with placebo in clinical trials and were shown to increase the number of patients who were completely pain free at 2 hours, were free of the most bothersome associated symptom (photophobia, phonophobia, or nausea) at 2 hours, and remained pain free at 24 hours (TABLE 4^18-24).

JFP07209292_t4.jpg

Continue to: Ubrogrepant

Ubrogepant, in 2 Phase 3 trials (ACHIEVE I and ACHIEVE II) demonstrated effectiveness compared with placebo.^21,22 The most common adverse effects reported were nausea and somnolence at very low rates. Pain-relief rates at 2 hours post dose (> 60% of participants) were higher than pain-free rates, and a significantly higher percentage (> 40%) of ubrogepant-treated participants reported ability to function normally on the Functional Disability Scale.²⁵

Rimegepant was also superior to placebo (59% vs 43%) in pain relief at 2 hours post dose and other secondary endpoints.²³ Rimegepant also has potential drug interactions and dose adjustments (TABLE 4^18-24).

Zavegepant, approved in March 2023, is administered once daily as a 10-mg nasal spray. In its Phase 3 trial, zavegepant was significantly superior to placebo at 2 hours post dose in freedom from pain (24% v 15%), and in freedom from the most bothersome symptom (40% v 31%).²⁴ Dosage modifications are not needed with mild-to-moderate renal or hepatic disease.²⁰

Worth noting. The safety of using ubrogepant to treat more than 8 migraine episodes in a 30-day period has not been established. The safety of using more than 18 doses of zavegepant in a 30-day period also has not been established. With ubrogepant and rimegepant, there are dosing modifications for concomitant use with specific drugs (CYP3A4 inhibitors and inducers) due to potential interactions and in patients with hepatic or renal impairment.^18,19

There are no trials comparing efficacy of CGRP antagonists to triptans. Recognizing that these newer medications would be costly, the AHS position statement released in 2019 recommends that gepants be considered for those with contraindications to triptans or for whom at least 2 oral triptans have failed (as determined by a validated patient outcome questionnaire).¹⁰ Step therapy with documentation of previous trials and therapy failures is often required by insurance companies prior to gepant coverage.

Continue to: Preventive therapies

Preventive therapies

Preventive migraine therapies are used to reduce duration, frequency, and severity of attacks, the need for acute treatment, and overall headache disability.²⁶ Medications typically are chosen based on efficacy, adverse effect profile, and patient comorbidities. Barriers to successful use include poor patient adherence and tolerability, the need for slow dose titration, and long-term use (minimum of 2 months) at maximum tolerated or minimum effective doses. Medications with established efficacy (Level A^a) based on the 2012 guidelines from the American Academy of Neurology (AAN) and the AHS are given in TABLE 5.^27-29

JFP07209292_t5.jpg

Drugs having received the strongest level of evidence for migraine prevention are metoprolol, propranolol, timolol, topiramate, valproate sodium, divalproex sodium, and onabotulinumtoxinA (Botox), and frovatriptan for menstrual migraine prevention. Because these guidelines were last updated in 2012, they did not cover gepants (which will be discussed shortly). The AHS released a position statement in 2019 supporting the use of anti-CGRP monoclonal antibodies (mAbs) in those who cannot tolerate or have had an inadequate response to a 6-week trial of at least 2 AAN/AHA Level A or B^b treatments.¹⁰ No head-to-head trials exist between non-CGRP preventive therapies and the CGRP antagonists.

CGRP-targeted prevention

Four anti-CGRP mAbs and 2 gepants have been approved for migraine prevention in the United States. Differences between products include targets (ligand vs receptor), antibody IgG subtype, bioavailability, route of administration, and frequency of administration.²⁸ As noted in the Phase 3 studies (TABLE 6^19,30-47), these therapies are highly efficacious, safe, and tolerable.

Gepants. Rimegepant, discussed earlier for migraine treatment, is one of the CGRP receptor antagonists approved for prevention. The other is atogepant (Qulipta), approved only for prevention. Ubrogepant is not approved for prevention.

Anti-CGRP mAb is the only medication class specifically created for migraine prevention.^10,26 As already noted, several efficacious non-CGRP treatment options are available for migraine prevention. However, higher doses of those agents, if needed, can lead to intolerable adverse effects for some patients, thereby limiting overall efficacy. ­Anti-CGRP mAbs, a targeted, highly efficacious treatment option, offer efficacy comparable to non-CGRP agents with a more favorable adverse effect profile for those who cannot tolerate or achieve only minimal efficacy with traditional preventive therapies.¹⁰

Continue to: The targeted anti-CGRP approach...

The targeted anti-CGRP approach, which can be used by patients with liver or kidney disease, results in decreased toxicity and minimal drug interactions. Long half-lives allow for monthly or quarterly injections, possibly resulting in increased compliance.²⁸ Dose titration is not needed, allowing for more rapid symptom management. The large molecular size of a mAb limits its transfer across the blood-brain barrier, making central nervous system adverse effects unlikely.²⁸ Despite the compelling mAb pharmacologic properties, their use may be limited by a lack of long-term safety data and the need for parenteral administration. Although ­immunogenicity—the development of neutralizing antibodies—can limit long-term tolerability or efficacy of mAbs generally,^26,28 anti-CGRP mAbs were engineered to minimally activate the immune system and have not been associated with immune suppression, opportunistic infections, malignancies, or decreased efficacy.²⁸

Unlike triptans and ergotamine derivatives, gepants do not constrict blood vessels and may have a unique role in patients with contraindications to triptans.

A pooled meta-analysis including 4 trials (3166 patients) found that CGRP mAbs compared with placebo significantly improved patient response rates, defined as at least a 50% and 75% reduction in monthly ­headache/migraine days from baseline to Weeks 9 to 12.⁴⁸ Another meta-analysis including 8 trials (2292 patients) found a significant reduction from baseline in monthly migraine days and monthly acute migraine medication consumption among patients taking CGRP mAbs compared with those taking placebo.⁴⁹ Open-label extension studies have shown progressive and cumulative benefits in individuals who respond to anti-CGRP mAbs. Therefore, several treatment cycles may be necessary to determine overall efficacy of therapy.^10,28

Cost initially can be a barrier. Insurance companies often require step therapy before agreeing to cover mAb therapy, which aligns with the 2019 AHS position statement.¹⁰Due to differences in insurance coverage, out-of-pocket expenses can vary greatly. However, options are available through online manufacturer assistance to reduce cost, making it comparable to other migraine treatments. Safety and efficacy studies of anti-CGRP mAbs use in pregnant individuals are limited. At this time, they should not be prescribed for those who are pregnant, planning to become pregnant, or breastfeeding. Counsel nonpregnant patients on appropriate contraception while using a mAb due to possible teratogenicity and negative pregnancy outcomes.^28,50

When combination treatment may be appropriate

Monotherapy is the usual approach to preventing migraine due to advantages of efficacy, simplified regimens, lower cost, and reduced adverse effects.⁵¹ However, if a patient does not benefit from monotherapy even after trying dose titrations as tolerated or switching therapies, trying complementary combination therapy is appropriate. Despite a shortage of clinical trials supporting the use of 2 or more preventive medications with different mechanisms of action, this strategy is used clinically.¹⁰ Consider combination therapy in those with refractory disease, partial responses, or intolerance to recommended doses.⁵² Articles reporting on case study reviews have rationalized the combined use of onabotulinumtoxinA and anti-CGRP mAbs, noting better migraine control.^51,53 The 2019 AHS position statement recommends adding a mAb to an existing preventive treatment regimen with no other changes until mAb effectiveness is determined, as the risk for drug interactions on dual therapy is low.¹⁰ Safety and efficacy also have been demonstrated with the combination of preventive anti-CGRP mAbs and acute treatment with gepants as needed.⁵⁴

Monoclonal antibodies are highly efficacious options for those who cannot tolerate or achieve only minimal efficacy with traditional preventive therapies.

Overall, gepants and mAbs are as effective as traditional acute and preventive treatments for migraine, and they cause fewer adverse effects and often allow a more simplified regimen. Gepants and mAbs are viable options in the primary care setting. Due to limited long-term data and high cost, however, they routinely are used for refractory migraine rather than as first-line agents. These therapies are especially favorable options for patients when traditional migraine therapies yield inadequate efficacy, cause intolerable adverse effects, are contraindicated, or introduce the risk for medication interactions.

CORRESPONDENCE
Emily Peterson, PharmD, BCACP, 3640 Middlebury Road, Iowa City, IA 52242; Emily-a-peterson@uiowa.edu

Migraine is a headache disorder that often causes unilateral pain, photophobia, phonophobia, nausea, and vomiting. More than 70% of office visits for migraine are made to primary care physicians.¹ Recent data suggest migraine may be caused primarily by neuronal dysfunction and only secondarily by vasodilation.² Although there are numerous classes of drugs used for migraine prevention and treatment, their success has been limited by inadequate efficacy, tolerability, and patient adherence.³ The discovery of pro-inflammatory markers such as calcitonin gene-related peptide (CGRP) has led to the development of new medications to prevent and treat migraine.⁴

JFP07209292_f.jpg

Pathophysiology, Dx and triggers, indications for pharmacotherapy

Pathophysiology. A migraine is thought to be caused by cortical spreading depression (CSD), a depolarization of glial and neuronal cell membranes.⁵This results in increased cortical excitability, central trigeminal-thalamic sensitization, and defective descending pain modulatory activity.⁶The activation of the trigeminal sensory pathways, primarily the ophthalmic branch, sends nociceptive signals to second-order neurons mediated by the release of neurotransmitters, such as CGRPs.⁵ This activation explains in part the primary location for a migraine, which is around the eye and the neighboring cranial regions. The pain perceived by the patient is caused by these second-order neurons.

It has been theorized that gepants bind to calcitonin gene-related peptide receptors, resulting in decreased blood flow to the brain, inhibition of neurogenic inflammation, and reduced pain signaling.

Dx and triggers. In 2018, the International Headache Society revised its guidelines for the diagnosis of migraine.⁷ According to the 3rd edition of The International Classification of Headache Disorders (ICHD-3), the diagnosis of migraine is made when a patient has at least 5 headache attacks that last 4 to 72 hours and have at least 2 of the following characteristics: (1) unilateral location, (2) pulsating quality, (3) moderate-to-severe pain intensity, and (4) aggravated by or causing avoidance of routine physical activity.⁷ The headache attacks also should have (1) associated nausea or vomiting or (2) photophobia and phonophobia.⁷ The presence of atypical signs or symptoms as indicated by the SNNOOP10 mnemonic raises concerns for secondary headaches and the need for further investigation into the cause of the headache (TABLE 1).⁸ It is not possible to detect every secondary headache with standard neuroimaging, but the SNNOOP10 red flags can help determine when imaging may be indicated.⁸ Potential triggers for migraine can be found in TABLE 2.⁹

JFP07209292_t1.jpg

Indications for pharmacotherapy. All patients receiving a diagnosis of migraine should be offered acute pharmacologic treatment. Consider preventive therapy anytime there are ≥ 4 headache days per month, debilitating attacks despite acute therapy, overuse of acute medication (> 2 d/wk), difficulty tolerating acute medication, patient preference, or presence of certain migraine subtypes.^7,10

JFP07209292_t2.jpg

Acute treatments

Abortive therapies for migraine include analgesics such as nonsteroidal anti-­inflammatory drugs (NSAIDs) and acetaminophen, and ergot alkaloids, triptans, or small-molecule CGRP receptor antagonists (gepants). Prompt administration increases the chance of success with acute therapy. Medications with the highest levels of efficacy based on the 2015 guidelines from the American Headache Society (AHS) are given in TABLE 3.¹¹ Lasmiditan (Reyvow) is not included in the 2015 guidelines, as it was approved after publication of the guidelines.

JFP07209292_t3a.jpg

JFP07209292_t3b.jpg

Non-CGRP first-line therapies

NSAIDs and acetaminophen. NSAIDs such as aspirin, diclofenac, ibuprofen, and naproxen have a high level of evidence to support their use as first-line treatments for mild-to-moderate migraine attacks. Trials consistently demonstrate their superiority to placebo in headache relief and complete pain relief at 2 hours. There is no recommendation for selecting one NSAID over another; however, consider their frequency of dosing and ­adverse effect profiles. The number needed to treat for complete pain relief at 2 hours ranges from 7 to 10 for most NSAIDs.^11,12 In some placebo-controlled studies, acetaminophen was less effective than NSAIDs, but was safer because it did not cause gastric irritation or antiplatelet effects.¹²

Triptans inhibit 5-HT_1B/1D receptors. Consider formulation, route of administration, cost, and pharmacokinetics when selecting a triptan. Patients who do not respond well to one triptan may respond favorably to another. A meta-analysis of the effectiveness of the 7 available agents found that triptans at standard doses provided pain relief within 2 hours in 42% to 76% of patients, and sustained freedom from pain for 2 hours in 18% to 50% of patients.¹³ Lasmiditan is a selective serotonin receptor (5-HT_1F) agonist that lacks vasoconstrictor activity. This is an option for patients with relative contraindications to triptans due to cardiovascular risk factors.¹⁰

Continue to: Second-line therapies

Intranasal dihydroergotamine has a favorable adverse event profile and greater evidence for efficacy compared with ergotamine. Compared with triptans, intranasal dihydroergotamine has a high level of efficacy but causes more adverse effects.¹⁴ Severe nausea is common, and dihydroergotamine often is used in combination with an antiemetic drug. Dihydroergotamine should not be used within 24 hours of taking a triptan, and it is contraindicated for patients who have hypertension or ischemic heart disease or who are pregnant or breastfeeding. There is also the potential for adverse drug interactions.¹⁵

A meta-analysis found that triptans at standard doses provided pain relief within 2 hours in 42% to 76% of patients, and sustained freedom from pain for 2 hours in 18% to 50% of patients.

Antiemetics may be helpful for migraine associated with severe nausea or vomiting. The dopamine antagonists metoclopramide, prochlorperazine, and chlorpromazine have demonstrated benefit in randomized ­placebo-controlled trials.¹¹ Ondansetron has not been studied extensively, but sometimes is used in clinical practice. Nonoral routes of administration may be useful in patients having trouble swallowing medications or in those experiencing significant nausea or vomiting early during migraine attacks.

Due to the high potential for abuse, opioids should not be used routinely for the treatment of migraine.¹² There is no high-quality evidence supporting the efficacy of barbiturates (ie, butalbital-containing compounds) for acute migraine treatment.¹¹ Moreover, use of these agents may increase the likelihood of progression from episodic to chronic migraine.¹⁶

Gepants for acute migraine treatment

Neuropeptide CGRP is released from trigeminal nerves and is a potent dilator of cerebral and dural vessels, playing a key role in regulating blood flow to the brain. Other roles of CGRP include the release of inflammatory agents from mast cells and the transmission of painful stimuli from intracranial vessels.¹⁷ The CGRP receptor or ligand can be targeted by small-molecule receptor antagonists for acute and preventive migraine treatment (and by monoclonal antibodies solely for prevention, discussed later). It has been theorized that gepants bind to CGRP receptors, resulting in decreased blood flow to the brain, inhibition of neurogenic inflammation, and reduced pain signaling.¹⁷ Unlike triptans and ergotamine derivatives, these novel treatments do not constrict blood vessels and may have a unique role in patients with contraindications to triptans.

The 3 gepants approved for acute treatment—ubrogepant (Ubrelvy),¹⁸ rimegepant (Nurtec),¹⁹ and zavegepant (Zavzpret)²⁰—were compared with placebo in clinical trials and were shown to increase the number of patients who were completely pain free at 2 hours, were free of the most bothersome associated symptom (photophobia, phonophobia, or nausea) at 2 hours, and remained pain free at 24 hours (TABLE 4^18-24).

JFP07209292_t4.jpg

Continue to: Ubrogrepant

Ubrogepant, in 2 Phase 3 trials (ACHIEVE I and ACHIEVE II) demonstrated effectiveness compared with placebo.^21,22 The most common adverse effects reported were nausea and somnolence at very low rates. Pain-relief rates at 2 hours post dose (> 60% of participants) were higher than pain-free rates, and a significantly higher percentage (> 40%) of ubrogepant-treated participants reported ability to function normally on the Functional Disability Scale.²⁵

Rimegepant was also superior to placebo (59% vs 43%) in pain relief at 2 hours post dose and other secondary endpoints.²³ Rimegepant also has potential drug interactions and dose adjustments (TABLE 4^18-24).

Zavegepant, approved in March 2023, is administered once daily as a 10-mg nasal spray. In its Phase 3 trial, zavegepant was significantly superior to placebo at 2 hours post dose in freedom from pain (24% v 15%), and in freedom from the most bothersome symptom (40% v 31%).²⁴ Dosage modifications are not needed with mild-to-moderate renal or hepatic disease.²⁰

Worth noting. The safety of using ubrogepant to treat more than 8 migraine episodes in a 30-day period has not been established. The safety of using more than 18 doses of zavegepant in a 30-day period also has not been established. With ubrogepant and rimegepant, there are dosing modifications for concomitant use with specific drugs (CYP3A4 inhibitors and inducers) due to potential interactions and in patients with hepatic or renal impairment.^18,19

There are no trials comparing efficacy of CGRP antagonists to triptans. Recognizing that these newer medications would be costly, the AHS position statement released in 2019 recommends that gepants be considered for those with contraindications to triptans or for whom at least 2 oral triptans have failed (as determined by a validated patient outcome questionnaire).¹⁰ Step therapy with documentation of previous trials and therapy failures is often required by insurance companies prior to gepant coverage.

Continue to: Preventive therapies

Preventive therapies

Preventive migraine therapies are used to reduce duration, frequency, and severity of attacks, the need for acute treatment, and overall headache disability.²⁶ Medications typically are chosen based on efficacy, adverse effect profile, and patient comorbidities. Barriers to successful use include poor patient adherence and tolerability, the need for slow dose titration, and long-term use (minimum of 2 months) at maximum tolerated or minimum effective doses. Medications with established efficacy (Level A^a) based on the 2012 guidelines from the American Academy of Neurology (AAN) and the AHS are given in TABLE 5.^27-29

JFP07209292_t5.jpg

Drugs having received the strongest level of evidence for migraine prevention are metoprolol, propranolol, timolol, topiramate, valproate sodium, divalproex sodium, and onabotulinumtoxinA (Botox), and frovatriptan for menstrual migraine prevention. Because these guidelines were last updated in 2012, they did not cover gepants (which will be discussed shortly). The AHS released a position statement in 2019 supporting the use of anti-CGRP monoclonal antibodies (mAbs) in those who cannot tolerate or have had an inadequate response to a 6-week trial of at least 2 AAN/AHA Level A or B^b treatments.¹⁰ No head-to-head trials exist between non-CGRP preventive therapies and the CGRP antagonists.

CGRP-targeted prevention

Four anti-CGRP mAbs and 2 gepants have been approved for migraine prevention in the United States. Differences between products include targets (ligand vs receptor), antibody IgG subtype, bioavailability, route of administration, and frequency of administration.²⁸ As noted in the Phase 3 studies (TABLE 6^19,30-47), these therapies are highly efficacious, safe, and tolerable.

Gepants. Rimegepant, discussed earlier for migraine treatment, is one of the CGRP receptor antagonists approved for prevention. The other is atogepant (Qulipta), approved only for prevention. Ubrogepant is not approved for prevention.

Anti-CGRP mAb is the only medication class specifically created for migraine prevention.^10,26 As already noted, several efficacious non-CGRP treatment options are available for migraine prevention. However, higher doses of those agents, if needed, can lead to intolerable adverse effects for some patients, thereby limiting overall efficacy. ­Anti-CGRP mAbs, a targeted, highly efficacious treatment option, offer efficacy comparable to non-CGRP agents with a more favorable adverse effect profile for those who cannot tolerate or achieve only minimal efficacy with traditional preventive therapies.¹⁰

Continue to: The targeted anti-CGRP approach...

The targeted anti-CGRP approach, which can be used by patients with liver or kidney disease, results in decreased toxicity and minimal drug interactions. Long half-lives allow for monthly or quarterly injections, possibly resulting in increased compliance.²⁸ Dose titration is not needed, allowing for more rapid symptom management. The large molecular size of a mAb limits its transfer across the blood-brain barrier, making central nervous system adverse effects unlikely.²⁸ Despite the compelling mAb pharmacologic properties, their use may be limited by a lack of long-term safety data and the need for parenteral administration. Although ­immunogenicity—the development of neutralizing antibodies—can limit long-term tolerability or efficacy of mAbs generally,^26,28 anti-CGRP mAbs were engineered to minimally activate the immune system and have not been associated with immune suppression, opportunistic infections, malignancies, or decreased efficacy.²⁸

Unlike triptans and ergotamine derivatives, gepants do not constrict blood vessels and may have a unique role in patients with contraindications to triptans.

A pooled meta-analysis including 4 trials (3166 patients) found that CGRP mAbs compared with placebo significantly improved patient response rates, defined as at least a 50% and 75% reduction in monthly ­headache/migraine days from baseline to Weeks 9 to 12.⁴⁸ Another meta-analysis including 8 trials (2292 patients) found a significant reduction from baseline in monthly migraine days and monthly acute migraine medication consumption among patients taking CGRP mAbs compared with those taking placebo.⁴⁹ Open-label extension studies have shown progressive and cumulative benefits in individuals who respond to anti-CGRP mAbs. Therefore, several treatment cycles may be necessary to determine overall efficacy of therapy.^10,28

Cost initially can be a barrier. Insurance companies often require step therapy before agreeing to cover mAb therapy, which aligns with the 2019 AHS position statement.¹⁰Due to differences in insurance coverage, out-of-pocket expenses can vary greatly. However, options are available through online manufacturer assistance to reduce cost, making it comparable to other migraine treatments. Safety and efficacy studies of anti-CGRP mAbs use in pregnant individuals are limited. At this time, they should not be prescribed for those who are pregnant, planning to become pregnant, or breastfeeding. Counsel nonpregnant patients on appropriate contraception while using a mAb due to possible teratogenicity and negative pregnancy outcomes.^28,50

When combination treatment may be appropriate

Monotherapy is the usual approach to preventing migraine due to advantages of efficacy, simplified regimens, lower cost, and reduced adverse effects.⁵¹ However, if a patient does not benefit from monotherapy even after trying dose titrations as tolerated or switching therapies, trying complementary combination therapy is appropriate. Despite a shortage of clinical trials supporting the use of 2 or more preventive medications with different mechanisms of action, this strategy is used clinically.¹⁰ Consider combination therapy in those with refractory disease, partial responses, or intolerance to recommended doses.⁵² Articles reporting on case study reviews have rationalized the combined use of onabotulinumtoxinA and anti-CGRP mAbs, noting better migraine control.^51,53 The 2019 AHS position statement recommends adding a mAb to an existing preventive treatment regimen with no other changes until mAb effectiveness is determined, as the risk for drug interactions on dual therapy is low.¹⁰ Safety and efficacy also have been demonstrated with the combination of preventive anti-CGRP mAbs and acute treatment with gepants as needed.⁵⁴

Monoclonal antibodies are highly efficacious options for those who cannot tolerate or achieve only minimal efficacy with traditional preventive therapies.

Overall, gepants and mAbs are as effective as traditional acute and preventive treatments for migraine, and they cause fewer adverse effects and often allow a more simplified regimen. Gepants and mAbs are viable options in the primary care setting. Due to limited long-term data and high cost, however, they routinely are used for refractory migraine rather than as first-line agents. These therapies are especially favorable options for patients when traditional migraine therapies yield inadequate efficacy, cause intolerable adverse effects, are contraindicated, or introduce the risk for medication interactions.

CORRESPONDENCE
Emily Peterson, PharmD, BCACP, 3640 Middlebury Road, Iowa City, IA 52242; Emily-a-peterson@uiowa.edu

Migraine is a headache disorder that often causes unilateral pain, photophobia, phonophobia, nausea, and vomiting. More than 70% of office visits for migraine are made to primary care physicians.¹ Recent data suggest migraine may be caused primarily by neuronal dysfunction and only secondarily by vasodilation.² Although there are numerous classes of drugs used for migraine prevention and treatment, their success has been limited by inadequate efficacy, tolerability, and patient adherence.³ The discovery of pro-inflammatory markers such as calcitonin gene-related peptide (CGRP) has led to the development of new medications to prevent and treat migraine.⁴

JFP07209292_f.jpg

Pathophysiology, Dx and triggers, indications for pharmacotherapy

Pathophysiology. A migraine is thought to be caused by cortical spreading depression (CSD), a depolarization of glial and neuronal cell membranes.⁵This results in increased cortical excitability, central trigeminal-thalamic sensitization, and defective descending pain modulatory activity.⁶The activation of the trigeminal sensory pathways, primarily the ophthalmic branch, sends nociceptive signals to second-order neurons mediated by the release of neurotransmitters, such as CGRPs.⁵ This activation explains in part the primary location for a migraine, which is around the eye and the neighboring cranial regions. The pain perceived by the patient is caused by these second-order neurons.

It has been theorized that gepants bind to calcitonin gene-related peptide receptors, resulting in decreased blood flow to the brain, inhibition of neurogenic inflammation, and reduced pain signaling.

Dx and triggers. In 2018, the International Headache Society revised its guidelines for the diagnosis of migraine.⁷ According to the 3rd edition of The International Classification of Headache Disorders (ICHD-3), the diagnosis of migraine is made when a patient has at least 5 headache attacks that last 4 to 72 hours and have at least 2 of the following characteristics: (1) unilateral location, (2) pulsating quality, (3) moderate-to-severe pain intensity, and (4) aggravated by or causing avoidance of routine physical activity.⁷ The headache attacks also should have (1) associated nausea or vomiting or (2) photophobia and phonophobia.⁷ The presence of atypical signs or symptoms as indicated by the SNNOOP10 mnemonic raises concerns for secondary headaches and the need for further investigation into the cause of the headache (TABLE 1).⁸ It is not possible to detect every secondary headache with standard neuroimaging, but the SNNOOP10 red flags can help determine when imaging may be indicated.⁸ Potential triggers for migraine can be found in TABLE 2.⁹

JFP07209292_t1.jpg

Indications for pharmacotherapy. All patients receiving a diagnosis of migraine should be offered acute pharmacologic treatment. Consider preventive therapy anytime there are ≥ 4 headache days per month, debilitating attacks despite acute therapy, overuse of acute medication (> 2 d/wk), difficulty tolerating acute medication, patient preference, or presence of certain migraine subtypes.^7,10

JFP07209292_t2.jpg

Acute treatments

Abortive therapies for migraine include analgesics such as nonsteroidal anti-­inflammatory drugs (NSAIDs) and acetaminophen, and ergot alkaloids, triptans, or small-molecule CGRP receptor antagonists (gepants). Prompt administration increases the chance of success with acute therapy. Medications with the highest levels of efficacy based on the 2015 guidelines from the American Headache Society (AHS) are given in TABLE 3.¹¹ Lasmiditan (Reyvow) is not included in the 2015 guidelines, as it was approved after publication of the guidelines.

JFP07209292_t3a.jpg

JFP07209292_t3b.jpg

Non-CGRP first-line therapies

NSAIDs and acetaminophen. NSAIDs such as aspirin, diclofenac, ibuprofen, and naproxen have a high level of evidence to support their use as first-line treatments for mild-to-moderate migraine attacks. Trials consistently demonstrate their superiority to placebo in headache relief and complete pain relief at 2 hours. There is no recommendation for selecting one NSAID over another; however, consider their frequency of dosing and ­adverse effect profiles. The number needed to treat for complete pain relief at 2 hours ranges from 7 to 10 for most NSAIDs.^11,12 In some placebo-controlled studies, acetaminophen was less effective than NSAIDs, but was safer because it did not cause gastric irritation or antiplatelet effects.¹²

Triptans inhibit 5-HT_1B/1D receptors. Consider formulation, route of administration, cost, and pharmacokinetics when selecting a triptan. Patients who do not respond well to one triptan may respond favorably to another. A meta-analysis of the effectiveness of the 7 available agents found that triptans at standard doses provided pain relief within 2 hours in 42% to 76% of patients, and sustained freedom from pain for 2 hours in 18% to 50% of patients.¹³ Lasmiditan is a selective serotonin receptor (5-HT_1F) agonist that lacks vasoconstrictor activity. This is an option for patients with relative contraindications to triptans due to cardiovascular risk factors.¹⁰

Continue to: Second-line therapies

Intranasal dihydroergotamine has a favorable adverse event profile and greater evidence for efficacy compared with ergotamine. Compared with triptans, intranasal dihydroergotamine has a high level of efficacy but causes more adverse effects.¹⁴ Severe nausea is common, and dihydroergotamine often is used in combination with an antiemetic drug. Dihydroergotamine should not be used within 24 hours of taking a triptan, and it is contraindicated for patients who have hypertension or ischemic heart disease or who are pregnant or breastfeeding. There is also the potential for adverse drug interactions.¹⁵

A meta-analysis found that triptans at standard doses provided pain relief within 2 hours in 42% to 76% of patients, and sustained freedom from pain for 2 hours in 18% to 50% of patients.

Antiemetics may be helpful for migraine associated with severe nausea or vomiting. The dopamine antagonists metoclopramide, prochlorperazine, and chlorpromazine have demonstrated benefit in randomized ­placebo-controlled trials.¹¹ Ondansetron has not been studied extensively, but sometimes is used in clinical practice. Nonoral routes of administration may be useful in patients having trouble swallowing medications or in those experiencing significant nausea or vomiting early during migraine attacks.

Due to the high potential for abuse, opioids should not be used routinely for the treatment of migraine.¹² There is no high-quality evidence supporting the efficacy of barbiturates (ie, butalbital-containing compounds) for acute migraine treatment.¹¹ Moreover, use of these agents may increase the likelihood of progression from episodic to chronic migraine.¹⁶

Gepants for acute migraine treatment

Neuropeptide CGRP is released from trigeminal nerves and is a potent dilator of cerebral and dural vessels, playing a key role in regulating blood flow to the brain. Other roles of CGRP include the release of inflammatory agents from mast cells and the transmission of painful stimuli from intracranial vessels.¹⁷ The CGRP receptor or ligand can be targeted by small-molecule receptor antagonists for acute and preventive migraine treatment (and by monoclonal antibodies solely for prevention, discussed later). It has been theorized that gepants bind to CGRP receptors, resulting in decreased blood flow to the brain, inhibition of neurogenic inflammation, and reduced pain signaling.¹⁷ Unlike triptans and ergotamine derivatives, these novel treatments do not constrict blood vessels and may have a unique role in patients with contraindications to triptans.

The 3 gepants approved for acute treatment—ubrogepant (Ubrelvy),¹⁸ rimegepant (Nurtec),¹⁹ and zavegepant (Zavzpret)²⁰—were compared with placebo in clinical trials and were shown to increase the number of patients who were completely pain free at 2 hours, were free of the most bothersome associated symptom (photophobia, phonophobia, or nausea) at 2 hours, and remained pain free at 24 hours (TABLE 4^18-24).

JFP07209292_t4.jpg

Continue to: Ubrogrepant

Ubrogepant, in 2 Phase 3 trials (ACHIEVE I and ACHIEVE II) demonstrated effectiveness compared with placebo.^21,22 The most common adverse effects reported were nausea and somnolence at very low rates. Pain-relief rates at 2 hours post dose (> 60% of participants) were higher than pain-free rates, and a significantly higher percentage (> 40%) of ubrogepant-treated participants reported ability to function normally on the Functional Disability Scale.²⁵

Rimegepant was also superior to placebo (59% vs 43%) in pain relief at 2 hours post dose and other secondary endpoints.²³ Rimegepant also has potential drug interactions and dose adjustments (TABLE 4^18-24).

Zavegepant, approved in March 2023, is administered once daily as a 10-mg nasal spray. In its Phase 3 trial, zavegepant was significantly superior to placebo at 2 hours post dose in freedom from pain (24% v 15%), and in freedom from the most bothersome symptom (40% v 31%).²⁴ Dosage modifications are not needed with mild-to-moderate renal or hepatic disease.²⁰

Worth noting. The safety of using ubrogepant to treat more than 8 migraine episodes in a 30-day period has not been established. The safety of using more than 18 doses of zavegepant in a 30-day period also has not been established. With ubrogepant and rimegepant, there are dosing modifications for concomitant use with specific drugs (CYP3A4 inhibitors and inducers) due to potential interactions and in patients with hepatic or renal impairment.^18,19

There are no trials comparing efficacy of CGRP antagonists to triptans. Recognizing that these newer medications would be costly, the AHS position statement released in 2019 recommends that gepants be considered for those with contraindications to triptans or for whom at least 2 oral triptans have failed (as determined by a validated patient outcome questionnaire).¹⁰ Step therapy with documentation of previous trials and therapy failures is often required by insurance companies prior to gepant coverage.

Continue to: Preventive therapies

Preventive therapies

Preventive migraine therapies are used to reduce duration, frequency, and severity of attacks, the need for acute treatment, and overall headache disability.²⁶ Medications typically are chosen based on efficacy, adverse effect profile, and patient comorbidities. Barriers to successful use include poor patient adherence and tolerability, the need for slow dose titration, and long-term use (minimum of 2 months) at maximum tolerated or minimum effective doses. Medications with established efficacy (Level A^a) based on the 2012 guidelines from the American Academy of Neurology (AAN) and the AHS are given in TABLE 5.^27-29

JFP07209292_t5.jpg

Drugs having received the strongest level of evidence for migraine prevention are metoprolol, propranolol, timolol, topiramate, valproate sodium, divalproex sodium, and onabotulinumtoxinA (Botox), and frovatriptan for menstrual migraine prevention. Because these guidelines were last updated in 2012, they did not cover gepants (which will be discussed shortly). The AHS released a position statement in 2019 supporting the use of anti-CGRP monoclonal antibodies (mAbs) in those who cannot tolerate or have had an inadequate response to a 6-week trial of at least 2 AAN/AHA Level A or B^b treatments.¹⁰ No head-to-head trials exist between non-CGRP preventive therapies and the CGRP antagonists.

CGRP-targeted prevention

Four anti-CGRP mAbs and 2 gepants have been approved for migraine prevention in the United States. Differences between products include targets (ligand vs receptor), antibody IgG subtype, bioavailability, route of administration, and frequency of administration.²⁸ As noted in the Phase 3 studies (TABLE 6^19,30-47), these therapies are highly efficacious, safe, and tolerable.

Gepants. Rimegepant, discussed earlier for migraine treatment, is one of the CGRP receptor antagonists approved for prevention. The other is atogepant (Qulipta), approved only for prevention. Ubrogepant is not approved for prevention.

Anti-CGRP mAb is the only medication class specifically created for migraine prevention.^10,26 As already noted, several efficacious non-CGRP treatment options are available for migraine prevention. However, higher doses of those agents, if needed, can lead to intolerable adverse effects for some patients, thereby limiting overall efficacy. ­Anti-CGRP mAbs, a targeted, highly efficacious treatment option, offer efficacy comparable to non-CGRP agents with a more favorable adverse effect profile for those who cannot tolerate or achieve only minimal efficacy with traditional preventive therapies.¹⁰

Continue to: The targeted anti-CGRP approach...

The targeted anti-CGRP approach, which can be used by patients with liver or kidney disease, results in decreased toxicity and minimal drug interactions. Long half-lives allow for monthly or quarterly injections, possibly resulting in increased compliance.²⁸ Dose titration is not needed, allowing for more rapid symptom management. The large molecular size of a mAb limits its transfer across the blood-brain barrier, making central nervous system adverse effects unlikely.²⁸ Despite the compelling mAb pharmacologic properties, their use may be limited by a lack of long-term safety data and the need for parenteral administration. Although ­immunogenicity—the development of neutralizing antibodies—can limit long-term tolerability or efficacy of mAbs generally,^26,28 anti-CGRP mAbs were engineered to minimally activate the immune system and have not been associated with immune suppression, opportunistic infections, malignancies, or decreased efficacy.²⁸

Unlike triptans and ergotamine derivatives, gepants do not constrict blood vessels and may have a unique role in patients with contraindications to triptans.

A pooled meta-analysis including 4 trials (3166 patients) found that CGRP mAbs compared with placebo significantly improved patient response rates, defined as at least a 50% and 75% reduction in monthly ­headache/migraine days from baseline to Weeks 9 to 12.⁴⁸ Another meta-analysis including 8 trials (2292 patients) found a significant reduction from baseline in monthly migraine days and monthly acute migraine medication consumption among patients taking CGRP mAbs compared with those taking placebo.⁴⁹ Open-label extension studies have shown progressive and cumulative benefits in individuals who respond to anti-CGRP mAbs. Therefore, several treatment cycles may be necessary to determine overall efficacy of therapy.^10,28

Cost initially can be a barrier. Insurance companies often require step therapy before agreeing to cover mAb therapy, which aligns with the 2019 AHS position statement.¹⁰Due to differences in insurance coverage, out-of-pocket expenses can vary greatly. However, options are available through online manufacturer assistance to reduce cost, making it comparable to other migraine treatments. Safety and efficacy studies of anti-CGRP mAbs use in pregnant individuals are limited. At this time, they should not be prescribed for those who are pregnant, planning to become pregnant, or breastfeeding. Counsel nonpregnant patients on appropriate contraception while using a mAb due to possible teratogenicity and negative pregnancy outcomes.^28,50

When combination treatment may be appropriate

Monotherapy is the usual approach to preventing migraine due to advantages of efficacy, simplified regimens, lower cost, and reduced adverse effects.⁵¹ However, if a patient does not benefit from monotherapy even after trying dose titrations as tolerated or switching therapies, trying complementary combination therapy is appropriate. Despite a shortage of clinical trials supporting the use of 2 or more preventive medications with different mechanisms of action, this strategy is used clinically.¹⁰ Consider combination therapy in those with refractory disease, partial responses, or intolerance to recommended doses.⁵² Articles reporting on case study reviews have rationalized the combined use of onabotulinumtoxinA and anti-CGRP mAbs, noting better migraine control.^51,53 The 2019 AHS position statement recommends adding a mAb to an existing preventive treatment regimen with no other changes until mAb effectiveness is determined, as the risk for drug interactions on dual therapy is low.¹⁰ Safety and efficacy also have been demonstrated with the combination of preventive anti-CGRP mAbs and acute treatment with gepants as needed.⁵⁴

Monoclonal antibodies are highly efficacious options for those who cannot tolerate or achieve only minimal efficacy with traditional preventive therapies.

Overall, gepants and mAbs are as effective as traditional acute and preventive treatments for migraine, and they cause fewer adverse effects and often allow a more simplified regimen. Gepants and mAbs are viable options in the primary care setting. Due to limited long-term data and high cost, however, they routinely are used for refractory migraine rather than as first-line agents. These therapies are especially favorable options for patients when traditional migraine therapies yield inadequate efficacy, cause intolerable adverse effects, are contraindicated, or introduce the risk for medication interactions.

CORRESPONDENCE
Emily Peterson, PharmD, BCACP, 3640 Middlebury Road, Iowa City, IA 52242; Emily-a-peterson@uiowa.edu

From the University of Iowa Hospitals and Clinics, Iowa City, IA.

Abstract

• Objective: To discuss the variability in response to tumor necrosis factor inhibitors (TNFis) observed in patients with rheumatoid arthritis (RA) and discuss therapeutic options for patients who do not respond to initial TNFi therapy.
• Methods: Review of the literature.
• Results: Optimal treatment of RA aims at achieving and then maintaining remission or low disease activity. In a patient with an inadequate response to initial biologic therapy, several therapeutic options exist. Current evidence supports TNFi dose escalation for only infliximab; optimization of concurrent conventional synthetic disease-modifying antirheumatic drug (csDMARD) or switching to a different csDMARD are other options. Cycling (switching to an alternative TNFi) and swapping (switching to a therapy with a different mode of action) strategies are other alternate approaches supported by many observational studies. While no head-to-head trials exist directly comparing the 2 strategies, data suggest superiority of the swapping strategy over the cycling approach. Also, several studies have shown that switching to a drug with a different mechanism of action is associated with higher treatment persistence and lower health care costs than TNFi cycling.
• Conclusion: Physicians have a growing list of treatment options to help their patients with RA achieve disease remission. The choice of best treatment for a given patient needs to be individualized, keeping in mind other factors, including comorbidities.

Keywords: biologics; rheumatoid arthritis; swapping strategy; cycling strategy; TNF inhibitors.

Following the discovery of tumor necrosis factor (TNF) as a proinflammatory cytokine 30 years ago, the use of TNF antagonists has revolutionized the treatment of rheumatoid arthritis (RA). Although TNF inhibitors (TNFIs) are frequently used as a first-line biologic disease-modifying antirheumatic drug (bDMARD), they are not uniformly efficacious in achieving remission in all patients with RA. This article highlights the reasons for such variability in observed response and discusses therapeutic options for patients who do not respond to TNFi therapy.

Case Presentation

A 60-year-old woman is evaluated in the clinic for complaints of pain in her hands, morning stiffness lasting 2 hours, and swelling in her wrists, all of which have been ongoing for 3 months. Physical exam reveals evidence of active inflammation, with synovitis in her second, third, and fourth metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints bilaterally, swelling over both wrists, and a weak grip. Inflammatory markers are elevated, and rheumatoid factor and anti-cyclic citrullinated peptide (anti-CCP) are both positive at high titer. Radiographs reveal evidence of small erosions at the third and fourth MCPs and PIPs bilaterally and periarticular osteopenia. The patient is diagnosed with seropositive, erosive RA based on history, physical exam, laboratory studies, and imaging. She is started on 20 mg of prednisone for acute treatment of her symptoms along with methotrexate, and, initially, her symptoms are well controlled. A few months after starting treatment, she develops voluminous diarrhea that necessitates cessation of methotrexate. Leflunomide also causes similar symptoms. The combination of sulfasalazine and hydroxychloroquine does not adequately control her symptoms, and ongoing use of low-dose glucocorticoids is required to improve functionality in all joints. Using the treat-to-target (T2T) strategy, adalimumab is initiated. However, she continues to report persistent swelling and pain and still requests oral glucocorticoids to help decrease inflammation. The 28-joint Disease Activity Score (DAS28) is 4.8, suggestive of moderate disease activity.

Why are TNFi agents sometimes ineffective?

The introduction of monoclonal antibodies and fusion proteins to block TNF and other cytokines was a remarkable development in the treatment of RA that revolutionized patient care. Despite the efficacy of TNFis, clinical response to these agents is not universal and only some patients achieve complete remission. In targeting the eventual goal of remission or low disease activity in patients with RA, the concept of “TNF failure” becomes extremely relevant. These inadequate responses to anti-TNF therapy may be due to primary failures, or complete lack of clinical response after initiation of the bDMARD, and secondary failures, or the loss of initially achieved clinical response to therapy. Other reasons for discontinuation of a given TNFi include partial disease control and intolerance to the medication (possible injection-site or infusion reactions). Keystone and Kavanaugh¹ divided causes of failure of TNF agents into 2 broad categories: perceptual (related to natural variations in disease course like hormonal variation and physical and emotional stress) and pathophysiological failures (genetic variations, high body mass index, concomitant cigarette use).

Another important consideration in patients treated with a TNFi is the consequent formation of anti-drug antibodies (ADAs). TNFi agents are immunogenic and normally elicit an immune response. The appearance of such ADAs may reduce the bioavailability of free drug, resulting in a decreased clinical response,² or may lead to serious adverse effects.

Several studies have reported that the prevalence of primary failure, secondary failure, and intolerance to TNFis ranges from 30% to 40%.^3-6 Female sex,⁷ concurrent prednisone use,⁸ high disease activity scores,^6,8,9 and the absence of treatment with low-dose methotrexate^7,8 have all been shown to be negative predictors of bDMARD retention and response.¹⁰

Are there any factors that predict TNFi failure?

There are no specific parameters to accurately predict responses to TNFI therapy.¹¹ Several clinical and molecular biomarkers in synovium (initial TNF levels, macrophages, T cells)¹² and peripheral blood (serum myeloid-related protein 8 and 14 complex levels,¹³ prealbumin, platelet factor 4, and S100A12)¹⁴ have been described as predictors of clinical response to TNFis, but their utility in clinical practice has not been established and the use of these markers has not yet been incorporated into clinical guidelines.

How is disease activity measured in patients with RA?

In 2010 an international expert consensus panel published treatment recommendations for RA that emphasized a T2T strategy of individualizing and escalating treatment to achieve the lowest disease activity or remission. In clinical practice, numerous tools are available to measure RA disease activity. Herein, we mention several that are most commonly used in clinical practice.

DAS28 combines single activity measures into an overall continuous measure of disease activity and has been endorsed by both the American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR). It includes a 28-swollen joint count (SJC), 28-tender joint count (TJC), erythrocyte sedimentation rate (ESR; can also be calculated using C-reactive protein [CRP]), and a patient global assessment (PtGA). The cut-offs used for DAS28 interpretation are as follows: remission (< 2.6), low (≥ 2.6 but ≤ 3.2), moderate (> 3.2 but ≤ 5.1), or high (> 5.1).¹⁵ Some of the difficulties in using DAS28 in daily clinical practice include the need for a lab value and the time needed to perform the joint counts. Note also that due to the inclusion of ESR, which is influenced by age and other factors, DAS28 may underestimate remission in the elderly.

Another measure of RA disease activity is the Simplified Disease Activity Index (SDAI), which includes 28 SJC, 28 TJC, PtGA, provider global assessment (PrGA), and CRP in mg/dL. The level of disease activity using the SDAI is interpreted as: remission (SDAI ≤ 3.3), low (≥ 3.4 but ≤ 11), moderate (> 11 but ≤ 26), or high (> 26). The advantage of the SDAI is that a calculator or computer is not required for calculations. Another measure, the Clinical Disease Activity Index (CDAI), includes a 28 SJC, 28 TJC, PtGA, and PrGA. Because a laboratory value is not needed to calculate the CDAI, it is well-suited for use in clinical practice. When using the CDAI, the level of disease activity can be defined as remission (CDAI ≤ 2.8), low (> 2.8 but ≤ 10), moderate (> 10 but ≤ 22), or high (> 22). Again, as with the SDAI, a calculator or computer is not needed for calculations.

What are the alternative treatment options after first biologic failure?

In patients who have failed treatment with an initial biologic, usually a TNFi, the treating rheumatologist has the following options (Figure), with the best treatment strategy being driven by individualized patient and disease-related factors (Table 1 and Table 2):

• TNFi dose escalation
• Trial of an alternate TNFi agent (the “cycling” strategy)
• Optimization of therapy conjoined with a conventional synthetic DMARD (csDMARD)
• Use of a non-TNF biologic or targeted synthetic DMARD (the “swapping” strategy)

jcom02604181_f1.JPG

If all the listed strategies fail, the next step can be the addition of short-term, low-dose glucocorticoid therapy.

jcom02604181_t1.JPG

TNFi Dose Escalation

The available data have demonstrated the safety, efficacy, and cost-effectiveness of dose escalation in patients with RA receiving infliximab.^16-18 The ATTRACT trial first demonstrated this, with greater clinical and radiographic improvements in those with higher trough serum concentrations, suggesting that doses higher than 3 mg/kg or more frequent than every 8 weeks may be needed for full response in some patients.¹⁹

jcom02604181_t2.JPG

There is a lack of studies in RA patients to determine the most effective dose escalation strategy. A study in patients with Crohn disease showed that intensification to 10 mg/kg every 8 weeks (dose doubling) was at least as effective as 5 mg/kg every 4 weeks (halving interval) at 12 months.¹⁶ Due to greater patient and administration convenience of dose-doubling, this strategy may be preferred.¹⁷ A starting dose of 10 mg/kg every 8 weeks is not routinely recommended due to an increased risk of serious infection; these adverse events were not found when the dose was gradually increased, as clinically indicated, starting at 3 mg/kg.^19,20 Further studies are needed to explore this approach in RA patients.

These results, however, have not been replicated with other TNFi agents. No significant clinical improvements were identified with etanercept 50 mg twice weekly,²¹ adalimumab 40 mg every week in the PREMIER trial,¹⁸ or certolizumab 400 mg every other week in an open-label extension phase of the RAPID 1 study.²² A Japanese study found significantly worse clinical outcomes with dose escalation of golimumab.²³ Conversely, 2 studies found clinical benefits after escalating the tocilizumab dose, the first a real-world review from the Consortium of Rheumatology Researchers of North America (CORRONA) registry using the intravenous formulation,²⁴ and the other the BREVACTA study utilizing subcutaneous tocilizumab.²⁵ No studies to date have been published on dose escalation of abatacept in patients with RA who respond poorly. Overall, previous studies support dose escalation in individuals being treated with infliximab to improve clinical outcomes, but additional studies are needed for other bDMARDs.

Per the ACR/EULAR^26,27 guidelines, all approved bDMARDs may be used without hierarchical positioning. However, after the failure of a TNFi agent, these guidelines do not provide specific advice about a preference between the “cycling” strategy (switching to an alternative TNFi) and “swapping” strategy (switching to a therapy with a different mode of action). Cycling might work for several reasons, including differences in the agents’ molecular structure, immunological mechanism of action, immunogenicity, and pharmacokinetics.^28-30 The cycling strategy is a well-established approach adopted by more than 94% of practicing rheumatologists, according to a national survey,³¹ and its efficacy is supported by trials and additional observational studies.^32-35

The greater clinical effectiveness of switching to infliximab compared with continuing with etanercept in patients with inadequate response to etanercept (n = 28) was suggested in the open-label OPPOSITE trial.³⁶ Data from the GO-AFTER trial³⁷ suggests that a greater proportion of patients with RA refractory to adalimumab, etanercept, or infliximab who were treated with golimumab achieved an ACR20 and ACR50 response compared with patients who received placebo, and this response persisted through 5 years.³⁸ More recently, certolizumab pegol and adalimumab were compared head-to-head in the EXXELERATE trial.³⁹ The results of this trial revealed the adequate efficacy of cycling to another TNFi after primary insufficient response to the first.

In studies from Finland and Sweden,^35,40 it has been observed that a better response is achieved in patients in whom TNF failure was initially due to secondary failure or intolerance rather than primary failure. A post-hoc analysis of the results of the GO-AFTER trial⁴¹ and from a few observational studies^35,40,42 revealed that switching from one TNFi to another, especially from a monoclonal antibody to a soluble receptor, was often more beneficial for RA patients than switching from a soluble receptor to a monoclonal antibody.

Optimization of Therapy Conjoined with csDMARDs

Methotrexate is one of the oldest and most effective csDMARDs available for the treatment of RA.⁴³ The 2016 EULAR guidelines recommend the addition of methotrexate and/or other csDMARDs to potentiate the effect of bDMARDs.²⁶ In the case of TNFi therapy, the observed synergistic effect between the monoclonal antibody and methotrexate may be explained by sustained suppression of ADA formation.⁴⁴ In the TEMPO,⁴⁵ PREMIER,¹⁸ and GO-BEFORE⁴⁶ trials, the addition of methotrexate led to improved clinical and radiological outcomes in patients treated with etanercept, adalimumab, and golimumab,⁴⁷ respectively. These findings were also demonstrated in several registries, where significant improvement in clinical response and retention rate of the TNFi agents was noted. Results have been replicated with non-TNFi bDMARDs, including abatacept^48,49 and rituximab.⁵⁰ Patients treated with interleukin (IL)-6 inhibitors in combination with methotrexate have shown significantly less radiographic progression compared to those treated with tocilizumab alone and those treated with monotherapy tocilizumab versus monotherapy methotrexate.^51,52 Results possibly favor the use of IL-6 inhibitors alone in those who cannot tolerate or have contraindications to methotrexate.

An open prospective study by Cohen et al added methotrexate to the treatment regimens of individuals on bDMARD monotherapy with a primary failure and found favorable changes in ACR20 and DAS28 scores at 3 and 12 months and therapeutic biological response (ESR, CRP) at 3 months.⁵³ Unlike monotherapy, in these situations methotrexate is known to be efficacious even at a lower dose, possibly at 7.5 mg to 10 mg per week. Some studies have shown that methotrexate administered parenterally may be more efficacious than when given orally.^54-58

In clinical trials and observational studies, leflunomide, sulfasalazine, and hydroxychloroquine have been used as alternate csDMARDs added to the treatment regimen.^59-62 There are, however, only 2 trials comparing the efficacy of methotrexate with that of other csDMARDs as concomitant treatment in patients with inadequate response to TNFi therapy. The RABBIT trial found a slight decrease in effectiveness with concomitant TNFi and leflunomide compared to TNFi/methotrexate, but overall each group had similar EULAR responses at 24 months.⁶³ A study by De Stefano et al found comparable ACR20 and DAS28 responses among individuals receiving TNFis with methotrexate or leflunomide.⁶¹

The “Swapping” Strategy

The efficacy of the swapping strategy has been shown in 3 randomized clinical trials demonstrating the superiority of abatacept, tocilizumab, and rituximab in the treatment of individuals with RA refractory to TNFis. Tocilizumab was studied in the RADIATE⁶⁴ trial, which involved 499 patients with inadequate response to 1 or more TNFi agents. The primary endpoint (24-week ACR20) was achieved by 50.0%, 30.4%, and 10.1% of patients in the 8 mg/kg, 4 mg/kg, and control groups, respectively (P < 0.001 for both tocilizumab groups versus placebo). The utility of abatacept as second-line therapy after initial TNF failure was evaluated in the ATTAIN⁶⁵ study. Participants with an inadequate response to etanercept or infliximab were randomly assigned to receive either abatacept or placebo. ACR50 response rates after 6 months of treatment were 20.3% with abatacept and 3.8% with placebo (P < 0.001). The SWITCH-RA study,⁶⁶ an observational study, compared rituximab to TNFis in 1112 participants with inadequate response to initial anti-TNF therapy. At 6 months, mean change in DAS28 was small but significantly greater for the rituximab group (–1.5 vs –1.1; P = 0.007). The difference in response rates was greatest among seropositive patients. These data suggest that rituximab has efficacy following TNFi failure, particularly for seropositive patients. Additionally, REFLEX⁶⁷ is the sole randomized controlled trial in patients with insufficient response to TNFis that showed significant prevention of radiographic progression at week 56 in patients on rituximab compared to placebo (mean change from baseline in total Genant-modified Sharp score, 1.00 vs 2.31, respectively; P = 0.005).

One study randomly assigned 399 patients with active RA who had inadequate response to prior TNFi therapy to tofacitinib⁶⁸ (5 mg twice daily or 10 mg twice daily) or placebo, both with methotrexate.⁶ After 3 months of treatment, ACR20 response rates (41.7% for 5 mg, 28.1% for 10 mg, 24.4% for placebo) and DAS28 remission rates (6.7% for 5 mg, 8.8% for 10 mg, 1.7% for placebo) were significantly greater among patients treated with tofacitinib compared to those treated with placebo. More recently, the RA-BEACON trial⁶⁹ demonstrated a consistent, beneficial treatment effect of baricitinib in patients with insufficient response to 1 or more TNFis. In this trial, 527 patients with an inadequate response to bDMARDs were randomly assigned to receive baricitinib 2 mg or 4 mg daily or placebo for 24 weeks. A higher proportion of patients receiving baricitinib 4 mg had an ACR20 response at week 12 compared with those treated with placebo (55% vs 27%, P < 0.001), and patients receiving the 4-mg dose had significant improvements from baseline in DAS28 and Health Assessment Questionnaire–Disability Index scores (P < 0.001 for both comparisons).

To Cycle or to Swap?

Several observational studies (SCQM-RA,⁷⁰ STURE,⁷¹ BSRBR,⁷² Favalli,⁴³ MIRAR,⁷³ SWITCH-RA,⁷⁴ ROC⁷²) have clearly demonstrated that the swapping strategy is favored over the cycling strategy. In the ROC study,⁷² patients were randomly assigned (based on physician discretion) to receive a non-TNF biologic or a TNFi. More patients in the non-TNF group than in the TNFi group showed low disease activity at week 24 (45% vs 28%; odds ratio [OR], 2.09; 95% confidence interval [CI], 1.27-3.43; P = 0.004) and at week 52 (41% vs 23%; OR, 2.26; 95% CI, 1.33-3.86; P = 0.003). The authors concluded that in patients having an insufficient response to TNFi therapy, a non-TNF biologic agent may be more effective than a second TNFi drug. Only a few studies^75-77 have demonstrated similar results between the 2 strategies. Overall, the available evidence seems to suggest the superiority of the swapping over the cycling strategy.

An important clinical pearl to keep in mind is that both swapping and cycling strategies might theoretically increase the risk of infection; however, limited evidence is reported in the literature. In a large retrospective analysis⁷⁸ of data on 4332 RA patients from a large US claims database, patients who had cycled between TNFi agents had a 30% to 40% increased risk of infection compared to patients treated with rituximab. Patients on infliximab had a 62% higher hazard of severe infections, and this has also been reported in an observational study.⁷⁹ In another study,⁷⁰ 41% of 201 patients with RA followed between 1999 and 2013 who swapped to abatacept/rituximab or tocilizumab developed adverse events, as compared to 59% of those who switched to a second TNFi.

Currently, there are no specific guidelines or biomarkers available to facilitate selection of specific treatment from among the classes of biologics. With the development of several new drugs and regulatory approval of baricitinib, physicians now have several biologic options to treat patients. A recent large time-trend study⁸⁰ deriving data from more than 200,000 patients with RA showed that etanercept remains the most frequently used agent for the treatment of RA; it also showed that the use of adalimumab and infliximab is decreasing, and that the use of newer agents, especially abatacept, golimumab, and certolizumab, has considerably risen in recent years. In this study, abatacept, rituximab, certolizumab, golimumab, tocilizumab, and tofacitinib accounted for 13.2%, 13.8%, 6.9%, 11.9%, and 7.5% switches from first TNFi therapy.

Jin et al⁸¹ studied factors associated with the choice of bDMARD for initial and subsequent use. They found that patients with commercial insurance had an 87% higher likelihood of initiating a bDMARD. In the Medicaid subgroup, African Americans had lower odds of initiating and switching bDMARDs than non-Hispanic whites. Prior use of steroids and nonbiologic DMARDs predicted both bDMARD initiation and subsequent switching. Etanercept, adalimumab, and infliximab were the most commonly used first- and second-line bDMARDS; patients on anakinra and golimumab were most likely to be switched to other bDMARDs.

Which treatment strategy is the most cost-effective?

Several studies have reported better treatment persistence rates among patients who are treated with the swapping strategy compared to the cycling strategy. In a retrospective analysis of claims data,⁸² the authors examined treatment persistence and health care costs in patients switching to biologics with a different mechanism of action or cycling to another TNFi. The mean cost was significantly lower among patients treated using the swapping strategy than among the TNFi cyclers, both for the total cost of care for RA and for the total cost of the targeted DMARDs in the first year after the change in therapy. The authors concluded that switching to a drug with a different mechanism of action is associated with higher treatment persistence and lower health care costs than TNFi cycling.

What about biosimilars?

Biosimilars are copies of already licensed biologics that are very similar to the biologics, but are made by different sponsors using independently derived cell lines and separately developed manufacturing processes.⁸³ Regarding biosimilar use, EULAR²⁶ states that biosimilar bDMARDs approved by the European Medicines Agency or US Food and Drug Administration have similar efficacy and safety as the originator bDMARDs, and recommends them as preferred agents if they are indeed appreciably cheaper than originator or other bDMARDs.

What are the novel treatment targets in RA?

New therapeutics for RA continue to be developed. One of the new agents is peficitinib (ASP015K), an oral, once-daily Janus kinase (Jak) inhibitor targeting Jak-1, Jak-2, and tyrosine kinase-2, with moderate selectivity for Jak-3. In a phase 2b trial, 100-mg and 150-mg doses of peficitinib achieved a statistically significant ACR20 response (48.3% and 56.3%) compared to placebo (29.4%) at 12 weeks.⁸⁴

Given the benefit of targeting TNF-α and IL-17 in RA, a novel molecule (ABT-122) that targets both human TNF and IL-17 has been developed. Two phase 1 studies⁸⁵ showed that dual neutralization of TNF and IL-17 with ABT-122 has characteristics acceptable for further exploration of therapeutic potential of this agent in TNF- and IL-17A–driven immune-mediated inflammatory diseases. Another novel drug is mavrilimumab, a human monoclonal antibody that targets granulocyte–macrophage colony-stimulating factor receptor α. A recent studyshowed that long-term treatment with mavrilimumab maintained response and was well-tolerated, with no increased incidence of treatment-emergent adverse events.⁸⁶

Namilumab (AMG203) is an immunoglobulin G1 monoclonal antibody that binds with high affinity to the GM-CSF ligand. In a phase 1b, randomized, double-blind study (PRIORA)⁸⁷ to assess namilumab in treating active, mild-to-moderate RA, significant improvement was seen in the DAS28-CRP score with namilumab (150 and 300 mg groups combined) compared with placebo at day 43 (P = 0.0117) and also 8 weeks after last dosing at day 99 (P = 0.0154). Adverse events were similar across different doses of namilumab and placebo, and included nasopharyngitis and exacerbation/worsening of RA. Another drug showing promise in RA is fosdagrocorat (PF-04171327), a potential dissociated agonist of the glucocorticoid receptor. A multicenter, double-blind, parallel-group, active- and placebo-controlled phase 2 study randomly assigned 86 patients to receive fosdagrocorat 10 mg, fosdagrocorat 25 mg, prednisone 5 mg, or placebo, all with stable background methotrexate therapy.⁸⁸ Both fosdagrocorat doses demonstrated efficacy in improving signs and symptoms in RA patients, with manageable adverse events.

Case Conclusion

There are several available treatment options for the case patient. Based on the PREMIER trial, solely increasing the dose of adalimumab is unlikely to provide a therapeutic benefit. Adding low-dose methotrexate (possibly via a parenteral route because of patient-reported gastrointestinal discomfort) might provide some synergistic and therapeutic effect. However, because of primary failure with TNFi therapy, she may benefit from the initiation of a biologic with a different mechanism of action (ie, swapping strategy). Therapeutic options include tocilizumab, abatacept, rituximab, and the Jak inhibitors (tofacitinib and baricitinib).

Summary

The optimal treatment of RA aims at achieving, and then maintaining, remission or a low disease activity. The choice of best treatment must be individualized to the patient, keeping in mind other factors, including comorbidities like fibromyalgia, history of diverticulitis (prior to use of tocilizumab), history of chronic obstructive pulmonary disease (prior to the use of abatacept), malignancy, and the presence of risk factors for infections (age, diabetes, chronic bronchitis). In a patient with inadequate response to initial biologic therapy, several options exist for the rheumatologist. Current evidence supports TNFi dose escalation for only infliximab; optimization of concurrent csDMARD or switching to a different csDMARD are other options. Cycling and swapping are other alternate approaches supported by many observational studies. While no head-to-head trials exist comparing the 2 strategies, data suggest superiority of the swapping strategy over the cycling approach. With the continuing development of novel therapeutics in RA, physicians have a growing list of treatment options to help their patients achieve disease remission.

Corresponding author: Namrata Singh, MD, 200 Hawkins Drive, Iowa City, IA 52242.

Financial disclosures: None.

From the University of Iowa Hospitals and Clinics, Iowa City, IA.

Abstract

• Objective: To discuss the variability in response to tumor necrosis factor inhibitors (TNFis) observed in patients with rheumatoid arthritis (RA) and discuss therapeutic options for patients who do not respond to initial TNFi therapy.
• Methods: Review of the literature.
• Results: Optimal treatment of RA aims at achieving and then maintaining remission or low disease activity. In a patient with an inadequate response to initial biologic therapy, several therapeutic options exist. Current evidence supports TNFi dose escalation for only infliximab; optimization of concurrent conventional synthetic disease-modifying antirheumatic drug (csDMARD) or switching to a different csDMARD are other options. Cycling (switching to an alternative TNFi) and swapping (switching to a therapy with a different mode of action) strategies are other alternate approaches supported by many observational studies. While no head-to-head trials exist directly comparing the 2 strategies, data suggest superiority of the swapping strategy over the cycling approach. Also, several studies have shown that switching to a drug with a different mechanism of action is associated with higher treatment persistence and lower health care costs than TNFi cycling.
• Conclusion: Physicians have a growing list of treatment options to help their patients with RA achieve disease remission. The choice of best treatment for a given patient needs to be individualized, keeping in mind other factors, including comorbidities.

Keywords: biologics; rheumatoid arthritis; swapping strategy; cycling strategy; TNF inhibitors.

Following the discovery of tumor necrosis factor (TNF) as a proinflammatory cytokine 30 years ago, the use of TNF antagonists has revolutionized the treatment of rheumatoid arthritis (RA). Although TNF inhibitors (TNFIs) are frequently used as a first-line biologic disease-modifying antirheumatic drug (bDMARD), they are not uniformly efficacious in achieving remission in all patients with RA. This article highlights the reasons for such variability in observed response and discusses therapeutic options for patients who do not respond to TNFi therapy.

Case Presentation

A 60-year-old woman is evaluated in the clinic for complaints of pain in her hands, morning stiffness lasting 2 hours, and swelling in her wrists, all of which have been ongoing for 3 months. Physical exam reveals evidence of active inflammation, with synovitis in her second, third, and fourth metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints bilaterally, swelling over both wrists, and a weak grip. Inflammatory markers are elevated, and rheumatoid factor and anti-cyclic citrullinated peptide (anti-CCP) are both positive at high titer. Radiographs reveal evidence of small erosions at the third and fourth MCPs and PIPs bilaterally and periarticular osteopenia. The patient is diagnosed with seropositive, erosive RA based on history, physical exam, laboratory studies, and imaging. She is started on 20 mg of prednisone for acute treatment of her symptoms along with methotrexate, and, initially, her symptoms are well controlled. A few months after starting treatment, she develops voluminous diarrhea that necessitates cessation of methotrexate. Leflunomide also causes similar symptoms. The combination of sulfasalazine and hydroxychloroquine does not adequately control her symptoms, and ongoing use of low-dose glucocorticoids is required to improve functionality in all joints. Using the treat-to-target (T2T) strategy, adalimumab is initiated. However, she continues to report persistent swelling and pain and still requests oral glucocorticoids to help decrease inflammation. The 28-joint Disease Activity Score (DAS28) is 4.8, suggestive of moderate disease activity.

Why are TNFi agents sometimes ineffective?

The introduction of monoclonal antibodies and fusion proteins to block TNF and other cytokines was a remarkable development in the treatment of RA that revolutionized patient care. Despite the efficacy of TNFis, clinical response to these agents is not universal and only some patients achieve complete remission. In targeting the eventual goal of remission or low disease activity in patients with RA, the concept of “TNF failure” becomes extremely relevant. These inadequate responses to anti-TNF therapy may be due to primary failures, or complete lack of clinical response after initiation of the bDMARD, and secondary failures, or the loss of initially achieved clinical response to therapy. Other reasons for discontinuation of a given TNFi include partial disease control and intolerance to the medication (possible injection-site or infusion reactions). Keystone and Kavanaugh¹ divided causes of failure of TNF agents into 2 broad categories: perceptual (related to natural variations in disease course like hormonal variation and physical and emotional stress) and pathophysiological failures (genetic variations, high body mass index, concomitant cigarette use).

Another important consideration in patients treated with a TNFi is the consequent formation of anti-drug antibodies (ADAs). TNFi agents are immunogenic and normally elicit an immune response. The appearance of such ADAs may reduce the bioavailability of free drug, resulting in a decreased clinical response,² or may lead to serious adverse effects.

Several studies have reported that the prevalence of primary failure, secondary failure, and intolerance to TNFis ranges from 30% to 40%.^3-6 Female sex,⁷ concurrent prednisone use,⁸ high disease activity scores,^6,8,9 and the absence of treatment with low-dose methotrexate^7,8 have all been shown to be negative predictors of bDMARD retention and response.¹⁰

Are there any factors that predict TNFi failure?

There are no specific parameters to accurately predict responses to TNFI therapy.¹¹ Several clinical and molecular biomarkers in synovium (initial TNF levels, macrophages, T cells)¹² and peripheral blood (serum myeloid-related protein 8 and 14 complex levels,¹³ prealbumin, platelet factor 4, and S100A12)¹⁴ have been described as predictors of clinical response to TNFis, but their utility in clinical practice has not been established and the use of these markers has not yet been incorporated into clinical guidelines.

How is disease activity measured in patients with RA?

In 2010 an international expert consensus panel published treatment recommendations for RA that emphasized a T2T strategy of individualizing and escalating treatment to achieve the lowest disease activity or remission. In clinical practice, numerous tools are available to measure RA disease activity. Herein, we mention several that are most commonly used in clinical practice.

DAS28 combines single activity measures into an overall continuous measure of disease activity and has been endorsed by both the American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR). It includes a 28-swollen joint count (SJC), 28-tender joint count (TJC), erythrocyte sedimentation rate (ESR; can also be calculated using C-reactive protein [CRP]), and a patient global assessment (PtGA). The cut-offs used for DAS28 interpretation are as follows: remission (< 2.6), low (≥ 2.6 but ≤ 3.2), moderate (> 3.2 but ≤ 5.1), or high (> 5.1).¹⁵ Some of the difficulties in using DAS28 in daily clinical practice include the need for a lab value and the time needed to perform the joint counts. Note also that due to the inclusion of ESR, which is influenced by age and other factors, DAS28 may underestimate remission in the elderly.

Another measure of RA disease activity is the Simplified Disease Activity Index (SDAI), which includes 28 SJC, 28 TJC, PtGA, provider global assessment (PrGA), and CRP in mg/dL. The level of disease activity using the SDAI is interpreted as: remission (SDAI ≤ 3.3), low (≥ 3.4 but ≤ 11), moderate (> 11 but ≤ 26), or high (> 26). The advantage of the SDAI is that a calculator or computer is not required for calculations. Another measure, the Clinical Disease Activity Index (CDAI), includes a 28 SJC, 28 TJC, PtGA, and PrGA. Because a laboratory value is not needed to calculate the CDAI, it is well-suited for use in clinical practice. When using the CDAI, the level of disease activity can be defined as remission (CDAI ≤ 2.8), low (> 2.8 but ≤ 10), moderate (> 10 but ≤ 22), or high (> 22). Again, as with the SDAI, a calculator or computer is not needed for calculations.

What are the alternative treatment options after first biologic failure?

In patients who have failed treatment with an initial biologic, usually a TNFi, the treating rheumatologist has the following options (Figure), with the best treatment strategy being driven by individualized patient and disease-related factors (Table 1 and Table 2):

• TNFi dose escalation
• Trial of an alternate TNFi agent (the “cycling” strategy)
• Optimization of therapy conjoined with a conventional synthetic DMARD (csDMARD)
• Use of a non-TNF biologic or targeted synthetic DMARD (the “swapping” strategy)

jcom02604181_f1.JPG

If all the listed strategies fail, the next step can be the addition of short-term, low-dose glucocorticoid therapy.

jcom02604181_t1.JPG

TNFi Dose Escalation

The available data have demonstrated the safety, efficacy, and cost-effectiveness of dose escalation in patients with RA receiving infliximab.^16-18 The ATTRACT trial first demonstrated this, with greater clinical and radiographic improvements in those with higher trough serum concentrations, suggesting that doses higher than 3 mg/kg or more frequent than every 8 weeks may be needed for full response in some patients.¹⁹

jcom02604181_t2.JPG

There is a lack of studies in RA patients to determine the most effective dose escalation strategy. A study in patients with Crohn disease showed that intensification to 10 mg/kg every 8 weeks (dose doubling) was at least as effective as 5 mg/kg every 4 weeks (halving interval) at 12 months.¹⁶ Due to greater patient and administration convenience of dose-doubling, this strategy may be preferred.¹⁷ A starting dose of 10 mg/kg every 8 weeks is not routinely recommended due to an increased risk of serious infection; these adverse events were not found when the dose was gradually increased, as clinically indicated, starting at 3 mg/kg.^19,20 Further studies are needed to explore this approach in RA patients.

These results, however, have not been replicated with other TNFi agents. No significant clinical improvements were identified with etanercept 50 mg twice weekly,²¹ adalimumab 40 mg every week in the PREMIER trial,¹⁸ or certolizumab 400 mg every other week in an open-label extension phase of the RAPID 1 study.²² A Japanese study found significantly worse clinical outcomes with dose escalation of golimumab.²³ Conversely, 2 studies found clinical benefits after escalating the tocilizumab dose, the first a real-world review from the Consortium of Rheumatology Researchers of North America (CORRONA) registry using the intravenous formulation,²⁴ and the other the BREVACTA study utilizing subcutaneous tocilizumab.²⁵ No studies to date have been published on dose escalation of abatacept in patients with RA who respond poorly. Overall, previous studies support dose escalation in individuals being treated with infliximab to improve clinical outcomes, but additional studies are needed for other bDMARDs.

Per the ACR/EULAR^26,27 guidelines, all approved bDMARDs may be used without hierarchical positioning. However, after the failure of a TNFi agent, these guidelines do not provide specific advice about a preference between the “cycling” strategy (switching to an alternative TNFi) and “swapping” strategy (switching to a therapy with a different mode of action). Cycling might work for several reasons, including differences in the agents’ molecular structure, immunological mechanism of action, immunogenicity, and pharmacokinetics.^28-30 The cycling strategy is a well-established approach adopted by more than 94% of practicing rheumatologists, according to a national survey,³¹ and its efficacy is supported by trials and additional observational studies.^32-35

The greater clinical effectiveness of switching to infliximab compared with continuing with etanercept in patients with inadequate response to etanercept (n = 28) was suggested in the open-label OPPOSITE trial.³⁶ Data from the GO-AFTER trial³⁷ suggests that a greater proportion of patients with RA refractory to adalimumab, etanercept, or infliximab who were treated with golimumab achieved an ACR20 and ACR50 response compared with patients who received placebo, and this response persisted through 5 years.³⁸ More recently, certolizumab pegol and adalimumab were compared head-to-head in the EXXELERATE trial.³⁹ The results of this trial revealed the adequate efficacy of cycling to another TNFi after primary insufficient response to the first.

In studies from Finland and Sweden,^35,40 it has been observed that a better response is achieved in patients in whom TNF failure was initially due to secondary failure or intolerance rather than primary failure. A post-hoc analysis of the results of the GO-AFTER trial⁴¹ and from a few observational studies^35,40,42 revealed that switching from one TNFi to another, especially from a monoclonal antibody to a soluble receptor, was often more beneficial for RA patients than switching from a soluble receptor to a monoclonal antibody.

Optimization of Therapy Conjoined with csDMARDs

Methotrexate is one of the oldest and most effective csDMARDs available for the treatment of RA.⁴³ The 2016 EULAR guidelines recommend the addition of methotrexate and/or other csDMARDs to potentiate the effect of bDMARDs.²⁶ In the case of TNFi therapy, the observed synergistic effect between the monoclonal antibody and methotrexate may be explained by sustained suppression of ADA formation.⁴⁴ In the TEMPO,⁴⁵ PREMIER,¹⁸ and GO-BEFORE⁴⁶ trials, the addition of methotrexate led to improved clinical and radiological outcomes in patients treated with etanercept, adalimumab, and golimumab,⁴⁷ respectively. These findings were also demonstrated in several registries, where significant improvement in clinical response and retention rate of the TNFi agents was noted. Results have been replicated with non-TNFi bDMARDs, including abatacept^48,49 and rituximab.⁵⁰ Patients treated with interleukin (IL)-6 inhibitors in combination with methotrexate have shown significantly less radiographic progression compared to those treated with tocilizumab alone and those treated with monotherapy tocilizumab versus monotherapy methotrexate.^51,52 Results possibly favor the use of IL-6 inhibitors alone in those who cannot tolerate or have contraindications to methotrexate.

An open prospective study by Cohen et al added methotrexate to the treatment regimens of individuals on bDMARD monotherapy with a primary failure and found favorable changes in ACR20 and DAS28 scores at 3 and 12 months and therapeutic biological response (ESR, CRP) at 3 months.⁵³ Unlike monotherapy, in these situations methotrexate is known to be efficacious even at a lower dose, possibly at 7.5 mg to 10 mg per week. Some studies have shown that methotrexate administered parenterally may be more efficacious than when given orally.^54-58

In clinical trials and observational studies, leflunomide, sulfasalazine, and hydroxychloroquine have been used as alternate csDMARDs added to the treatment regimen.^59-62 There are, however, only 2 trials comparing the efficacy of methotrexate with that of other csDMARDs as concomitant treatment in patients with inadequate response to TNFi therapy. The RABBIT trial found a slight decrease in effectiveness with concomitant TNFi and leflunomide compared to TNFi/methotrexate, but overall each group had similar EULAR responses at 24 months.⁶³ A study by De Stefano et al found comparable ACR20 and DAS28 responses among individuals receiving TNFis with methotrexate or leflunomide.⁶¹

The “Swapping” Strategy

The efficacy of the swapping strategy has been shown in 3 randomized clinical trials demonstrating the superiority of abatacept, tocilizumab, and rituximab in the treatment of individuals with RA refractory to TNFis. Tocilizumab was studied in the RADIATE⁶⁴ trial, which involved 499 patients with inadequate response to 1 or more TNFi agents. The primary endpoint (24-week ACR20) was achieved by 50.0%, 30.4%, and 10.1% of patients in the 8 mg/kg, 4 mg/kg, and control groups, respectively (P < 0.001 for both tocilizumab groups versus placebo). The utility of abatacept as second-line therapy after initial TNF failure was evaluated in the ATTAIN⁶⁵ study. Participants with an inadequate response to etanercept or infliximab were randomly assigned to receive either abatacept or placebo. ACR50 response rates after 6 months of treatment were 20.3% with abatacept and 3.8% with placebo (P < 0.001). The SWITCH-RA study,⁶⁶ an observational study, compared rituximab to TNFis in 1112 participants with inadequate response to initial anti-TNF therapy. At 6 months, mean change in DAS28 was small but significantly greater for the rituximab group (–1.5 vs –1.1; P = 0.007). The difference in response rates was greatest among seropositive patients. These data suggest that rituximab has efficacy following TNFi failure, particularly for seropositive patients. Additionally, REFLEX⁶⁷ is the sole randomized controlled trial in patients with insufficient response to TNFis that showed significant prevention of radiographic progression at week 56 in patients on rituximab compared to placebo (mean change from baseline in total Genant-modified Sharp score, 1.00 vs 2.31, respectively; P = 0.005).

One study randomly assigned 399 patients with active RA who had inadequate response to prior TNFi therapy to tofacitinib⁶⁸ (5 mg twice daily or 10 mg twice daily) or placebo, both with methotrexate.⁶ After 3 months of treatment, ACR20 response rates (41.7% for 5 mg, 28.1% for 10 mg, 24.4% for placebo) and DAS28 remission rates (6.7% for 5 mg, 8.8% for 10 mg, 1.7% for placebo) were significantly greater among patients treated with tofacitinib compared to those treated with placebo. More recently, the RA-BEACON trial⁶⁹ demonstrated a consistent, beneficial treatment effect of baricitinib in patients with insufficient response to 1 or more TNFis. In this trial, 527 patients with an inadequate response to bDMARDs were randomly assigned to receive baricitinib 2 mg or 4 mg daily or placebo for 24 weeks. A higher proportion of patients receiving baricitinib 4 mg had an ACR20 response at week 12 compared with those treated with placebo (55% vs 27%, P < 0.001), and patients receiving the 4-mg dose had significant improvements from baseline in DAS28 and Health Assessment Questionnaire–Disability Index scores (P < 0.001 for both comparisons).

To Cycle or to Swap?

Several observational studies (SCQM-RA,⁷⁰ STURE,⁷¹ BSRBR,⁷² Favalli,⁴³ MIRAR,⁷³ SWITCH-RA,⁷⁴ ROC⁷²) have clearly demonstrated that the swapping strategy is favored over the cycling strategy. In the ROC study,⁷² patients were randomly assigned (based on physician discretion) to receive a non-TNF biologic or a TNFi. More patients in the non-TNF group than in the TNFi group showed low disease activity at week 24 (45% vs 28%; odds ratio [OR], 2.09; 95% confidence interval [CI], 1.27-3.43; P = 0.004) and at week 52 (41% vs 23%; OR, 2.26; 95% CI, 1.33-3.86; P = 0.003). The authors concluded that in patients having an insufficient response to TNFi therapy, a non-TNF biologic agent may be more effective than a second TNFi drug. Only a few studies^75-77 have demonstrated similar results between the 2 strategies. Overall, the available evidence seems to suggest the superiority of the swapping over the cycling strategy.

An important clinical pearl to keep in mind is that both swapping and cycling strategies might theoretically increase the risk of infection; however, limited evidence is reported in the literature. In a large retrospective analysis⁷⁸ of data on 4332 RA patients from a large US claims database, patients who had cycled between TNFi agents had a 30% to 40% increased risk of infection compared to patients treated with rituximab. Patients on infliximab had a 62% higher hazard of severe infections, and this has also been reported in an observational study.⁷⁹ In another study,⁷⁰ 41% of 201 patients with RA followed between 1999 and 2013 who swapped to abatacept/rituximab or tocilizumab developed adverse events, as compared to 59% of those who switched to a second TNFi.

Currently, there are no specific guidelines or biomarkers available to facilitate selection of specific treatment from among the classes of biologics. With the development of several new drugs and regulatory approval of baricitinib, physicians now have several biologic options to treat patients. A recent large time-trend study⁸⁰ deriving data from more than 200,000 patients with RA showed that etanercept remains the most frequently used agent for the treatment of RA; it also showed that the use of adalimumab and infliximab is decreasing, and that the use of newer agents, especially abatacept, golimumab, and certolizumab, has considerably risen in recent years. In this study, abatacept, rituximab, certolizumab, golimumab, tocilizumab, and tofacitinib accounted for 13.2%, 13.8%, 6.9%, 11.9%, and 7.5% switches from first TNFi therapy.

Jin et al⁸¹ studied factors associated with the choice of bDMARD for initial and subsequent use. They found that patients with commercial insurance had an 87% higher likelihood of initiating a bDMARD. In the Medicaid subgroup, African Americans had lower odds of initiating and switching bDMARDs than non-Hispanic whites. Prior use of steroids and nonbiologic DMARDs predicted both bDMARD initiation and subsequent switching. Etanercept, adalimumab, and infliximab were the most commonly used first- and second-line bDMARDS; patients on anakinra and golimumab were most likely to be switched to other bDMARDs.

Which treatment strategy is the most cost-effective?

Several studies have reported better treatment persistence rates among patients who are treated with the swapping strategy compared to the cycling strategy. In a retrospective analysis of claims data,⁸² the authors examined treatment persistence and health care costs in patients switching to biologics with a different mechanism of action or cycling to another TNFi. The mean cost was significantly lower among patients treated using the swapping strategy than among the TNFi cyclers, both for the total cost of care for RA and for the total cost of the targeted DMARDs in the first year after the change in therapy. The authors concluded that switching to a drug with a different mechanism of action is associated with higher treatment persistence and lower health care costs than TNFi cycling.

What about biosimilars?

Biosimilars are copies of already licensed biologics that are very similar to the biologics, but are made by different sponsors using independently derived cell lines and separately developed manufacturing processes.⁸³ Regarding biosimilar use, EULAR²⁶ states that biosimilar bDMARDs approved by the European Medicines Agency or US Food and Drug Administration have similar efficacy and safety as the originator bDMARDs, and recommends them as preferred agents if they are indeed appreciably cheaper than originator or other bDMARDs.

What are the novel treatment targets in RA?

New therapeutics for RA continue to be developed. One of the new agents is peficitinib (ASP015K), an oral, once-daily Janus kinase (Jak) inhibitor targeting Jak-1, Jak-2, and tyrosine kinase-2, with moderate selectivity for Jak-3. In a phase 2b trial, 100-mg and 150-mg doses of peficitinib achieved a statistically significant ACR20 response (48.3% and 56.3%) compared to placebo (29.4%) at 12 weeks.⁸⁴

Given the benefit of targeting TNF-α and IL-17 in RA, a novel molecule (ABT-122) that targets both human TNF and IL-17 has been developed. Two phase 1 studies⁸⁵ showed that dual neutralization of TNF and IL-17 with ABT-122 has characteristics acceptable for further exploration of therapeutic potential of this agent in TNF- and IL-17A–driven immune-mediated inflammatory diseases. Another novel drug is mavrilimumab, a human monoclonal antibody that targets granulocyte–macrophage colony-stimulating factor receptor α. A recent studyshowed that long-term treatment with mavrilimumab maintained response and was well-tolerated, with no increased incidence of treatment-emergent adverse events.⁸⁶

Namilumab (AMG203) is an immunoglobulin G1 monoclonal antibody that binds with high affinity to the GM-CSF ligand. In a phase 1b, randomized, double-blind study (PRIORA)⁸⁷ to assess namilumab in treating active, mild-to-moderate RA, significant improvement was seen in the DAS28-CRP score with namilumab (150 and 300 mg groups combined) compared with placebo at day 43 (P = 0.0117) and also 8 weeks after last dosing at day 99 (P = 0.0154). Adverse events were similar across different doses of namilumab and placebo, and included nasopharyngitis and exacerbation/worsening of RA. Another drug showing promise in RA is fosdagrocorat (PF-04171327), a potential dissociated agonist of the glucocorticoid receptor. A multicenter, double-blind, parallel-group, active- and placebo-controlled phase 2 study randomly assigned 86 patients to receive fosdagrocorat 10 mg, fosdagrocorat 25 mg, prednisone 5 mg, or placebo, all with stable background methotrexate therapy.⁸⁸ Both fosdagrocorat doses demonstrated efficacy in improving signs and symptoms in RA patients, with manageable adverse events.

Case Conclusion

There are several available treatment options for the case patient. Based on the PREMIER trial, solely increasing the dose of adalimumab is unlikely to provide a therapeutic benefit. Adding low-dose methotrexate (possibly via a parenteral route because of patient-reported gastrointestinal discomfort) might provide some synergistic and therapeutic effect. However, because of primary failure with TNFi therapy, she may benefit from the initiation of a biologic with a different mechanism of action (ie, swapping strategy). Therapeutic options include tocilizumab, abatacept, rituximab, and the Jak inhibitors (tofacitinib and baricitinib).

Summary

The optimal treatment of RA aims at achieving, and then maintaining, remission or a low disease activity. The choice of best treatment must be individualized to the patient, keeping in mind other factors, including comorbidities like fibromyalgia, history of diverticulitis (prior to use of tocilizumab), history of chronic obstructive pulmonary disease (prior to the use of abatacept), malignancy, and the presence of risk factors for infections (age, diabetes, chronic bronchitis). In a patient with inadequate response to initial biologic therapy, several options exist for the rheumatologist. Current evidence supports TNFi dose escalation for only infliximab; optimization of concurrent csDMARD or switching to a different csDMARD are other options. Cycling and swapping are other alternate approaches supported by many observational studies. While no head-to-head trials exist comparing the 2 strategies, data suggest superiority of the swapping strategy over the cycling approach. With the continuing development of novel therapeutics in RA, physicians have a growing list of treatment options to help their patients achieve disease remission.

Corresponding author: Namrata Singh, MD, 200 Hawkins Drive, Iowa City, IA 52242.

Financial disclosures: None.

From the University of Iowa Hospitals and Clinics, Iowa City, IA.

Abstract

• Objective: To discuss the variability in response to tumor necrosis factor inhibitors (TNFis) observed in patients with rheumatoid arthritis (RA) and discuss therapeutic options for patients who do not respond to initial TNFi therapy.
• Methods: Review of the literature.
• Results: Optimal treatment of RA aims at achieving and then maintaining remission or low disease activity. In a patient with an inadequate response to initial biologic therapy, several therapeutic options exist. Current evidence supports TNFi dose escalation for only infliximab; optimization of concurrent conventional synthetic disease-modifying antirheumatic drug (csDMARD) or switching to a different csDMARD are other options. Cycling (switching to an alternative TNFi) and swapping (switching to a therapy with a different mode of action) strategies are other alternate approaches supported by many observational studies. While no head-to-head trials exist directly comparing the 2 strategies, data suggest superiority of the swapping strategy over the cycling approach. Also, several studies have shown that switching to a drug with a different mechanism of action is associated with higher treatment persistence and lower health care costs than TNFi cycling.
• Conclusion: Physicians have a growing list of treatment options to help their patients with RA achieve disease remission. The choice of best treatment for a given patient needs to be individualized, keeping in mind other factors, including comorbidities.

Keywords: biologics; rheumatoid arthritis; swapping strategy; cycling strategy; TNF inhibitors.

Following the discovery of tumor necrosis factor (TNF) as a proinflammatory cytokine 30 years ago, the use of TNF antagonists has revolutionized the treatment of rheumatoid arthritis (RA). Although TNF inhibitors (TNFIs) are frequently used as a first-line biologic disease-modifying antirheumatic drug (bDMARD), they are not uniformly efficacious in achieving remission in all patients with RA. This article highlights the reasons for such variability in observed response and discusses therapeutic options for patients who do not respond to TNFi therapy.

Case Presentation

A 60-year-old woman is evaluated in the clinic for complaints of pain in her hands, morning stiffness lasting 2 hours, and swelling in her wrists, all of which have been ongoing for 3 months. Physical exam reveals evidence of active inflammation, with synovitis in her second, third, and fourth metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints bilaterally, swelling over both wrists, and a weak grip. Inflammatory markers are elevated, and rheumatoid factor and anti-cyclic citrullinated peptide (anti-CCP) are both positive at high titer. Radiographs reveal evidence of small erosions at the third and fourth MCPs and PIPs bilaterally and periarticular osteopenia. The patient is diagnosed with seropositive, erosive RA based on history, physical exam, laboratory studies, and imaging. She is started on 20 mg of prednisone for acute treatment of her symptoms along with methotrexate, and, initially, her symptoms are well controlled. A few months after starting treatment, she develops voluminous diarrhea that necessitates cessation of methotrexate. Leflunomide also causes similar symptoms. The combination of sulfasalazine and hydroxychloroquine does not adequately control her symptoms, and ongoing use of low-dose glucocorticoids is required to improve functionality in all joints. Using the treat-to-target (T2T) strategy, adalimumab is initiated. However, she continues to report persistent swelling and pain and still requests oral glucocorticoids to help decrease inflammation. The 28-joint Disease Activity Score (DAS28) is 4.8, suggestive of moderate disease activity.

Why are TNFi agents sometimes ineffective?

The introduction of monoclonal antibodies and fusion proteins to block TNF and other cytokines was a remarkable development in the treatment of RA that revolutionized patient care. Despite the efficacy of TNFis, clinical response to these agents is not universal and only some patients achieve complete remission. In targeting the eventual goal of remission or low disease activity in patients with RA, the concept of “TNF failure” becomes extremely relevant. These inadequate responses to anti-TNF therapy may be due to primary failures, or complete lack of clinical response after initiation of the bDMARD, and secondary failures, or the loss of initially achieved clinical response to therapy. Other reasons for discontinuation of a given TNFi include partial disease control and intolerance to the medication (possible injection-site or infusion reactions). Keystone and Kavanaugh¹ divided causes of failure of TNF agents into 2 broad categories: perceptual (related to natural variations in disease course like hormonal variation and physical and emotional stress) and pathophysiological failures (genetic variations, high body mass index, concomitant cigarette use).

Another important consideration in patients treated with a TNFi is the consequent formation of anti-drug antibodies (ADAs). TNFi agents are immunogenic and normally elicit an immune response. The appearance of such ADAs may reduce the bioavailability of free drug, resulting in a decreased clinical response,² or may lead to serious adverse effects.

Several studies have reported that the prevalence of primary failure, secondary failure, and intolerance to TNFis ranges from 30% to 40%.^3-6 Female sex,⁷ concurrent prednisone use,⁸ high disease activity scores,^6,8,9 and the absence of treatment with low-dose methotrexate^7,8 have all been shown to be negative predictors of bDMARD retention and response.¹⁰

Are there any factors that predict TNFi failure?

There are no specific parameters to accurately predict responses to TNFI therapy.¹¹ Several clinical and molecular biomarkers in synovium (initial TNF levels, macrophages, T cells)¹² and peripheral blood (serum myeloid-related protein 8 and 14 complex levels,¹³ prealbumin, platelet factor 4, and S100A12)¹⁴ have been described as predictors of clinical response to TNFis, but their utility in clinical practice has not been established and the use of these markers has not yet been incorporated into clinical guidelines.

How is disease activity measured in patients with RA?

In 2010 an international expert consensus panel published treatment recommendations for RA that emphasized a T2T strategy of individualizing and escalating treatment to achieve the lowest disease activity or remission. In clinical practice, numerous tools are available to measure RA disease activity. Herein, we mention several that are most commonly used in clinical practice.

DAS28 combines single activity measures into an overall continuous measure of disease activity and has been endorsed by both the American College of Rheumatology (ACR) and European League Against Rheumatism (EULAR). It includes a 28-swollen joint count (SJC), 28-tender joint count (TJC), erythrocyte sedimentation rate (ESR; can also be calculated using C-reactive protein [CRP]), and a patient global assessment (PtGA). The cut-offs used for DAS28 interpretation are as follows: remission (< 2.6), low (≥ 2.6 but ≤ 3.2), moderate (> 3.2 but ≤ 5.1), or high (> 5.1).¹⁵ Some of the difficulties in using DAS28 in daily clinical practice include the need for a lab value and the time needed to perform the joint counts. Note also that due to the inclusion of ESR, which is influenced by age and other factors, DAS28 may underestimate remission in the elderly.

Another measure of RA disease activity is the Simplified Disease Activity Index (SDAI), which includes 28 SJC, 28 TJC, PtGA, provider global assessment (PrGA), and CRP in mg/dL. The level of disease activity using the SDAI is interpreted as: remission (SDAI ≤ 3.3), low (≥ 3.4 but ≤ 11), moderate (> 11 but ≤ 26), or high (> 26). The advantage of the SDAI is that a calculator or computer is not required for calculations. Another measure, the Clinical Disease Activity Index (CDAI), includes a 28 SJC, 28 TJC, PtGA, and PrGA. Because a laboratory value is not needed to calculate the CDAI, it is well-suited for use in clinical practice. When using the CDAI, the level of disease activity can be defined as remission (CDAI ≤ 2.8), low (> 2.8 but ≤ 10), moderate (> 10 but ≤ 22), or high (> 22). Again, as with the SDAI, a calculator or computer is not needed for calculations.

What are the alternative treatment options after first biologic failure?

In patients who have failed treatment with an initial biologic, usually a TNFi, the treating rheumatologist has the following options (Figure), with the best treatment strategy being driven by individualized patient and disease-related factors (Table 1 and Table 2):

• TNFi dose escalation
• Trial of an alternate TNFi agent (the “cycling” strategy)
• Optimization of therapy conjoined with a conventional synthetic DMARD (csDMARD)
• Use of a non-TNF biologic or targeted synthetic DMARD (the “swapping” strategy)

jcom02604181_f1.JPG

If all the listed strategies fail, the next step can be the addition of short-term, low-dose glucocorticoid therapy.

jcom02604181_t1.JPG

TNFi Dose Escalation

The available data have demonstrated the safety, efficacy, and cost-effectiveness of dose escalation in patients with RA receiving infliximab.^16-18 The ATTRACT trial first demonstrated this, with greater clinical and radiographic improvements in those with higher trough serum concentrations, suggesting that doses higher than 3 mg/kg or more frequent than every 8 weeks may be needed for full response in some patients.¹⁹

jcom02604181_t2.JPG

There is a lack of studies in RA patients to determine the most effective dose escalation strategy. A study in patients with Crohn disease showed that intensification to 10 mg/kg every 8 weeks (dose doubling) was at least as effective as 5 mg/kg every 4 weeks (halving interval) at 12 months.¹⁶ Due to greater patient and administration convenience of dose-doubling, this strategy may be preferred.¹⁷ A starting dose of 10 mg/kg every 8 weeks is not routinely recommended due to an increased risk of serious infection; these adverse events were not found when the dose was gradually increased, as clinically indicated, starting at 3 mg/kg.^19,20 Further studies are needed to explore this approach in RA patients.

These results, however, have not been replicated with other TNFi agents. No significant clinical improvements were identified with etanercept 50 mg twice weekly,²¹ adalimumab 40 mg every week in the PREMIER trial,¹⁸ or certolizumab 400 mg every other week in an open-label extension phase of the RAPID 1 study.²² A Japanese study found significantly worse clinical outcomes with dose escalation of golimumab.²³ Conversely, 2 studies found clinical benefits after escalating the tocilizumab dose, the first a real-world review from the Consortium of Rheumatology Researchers of North America (CORRONA) registry using the intravenous formulation,²⁴ and the other the BREVACTA study utilizing subcutaneous tocilizumab.²⁵ No studies to date have been published on dose escalation of abatacept in patients with RA who respond poorly. Overall, previous studies support dose escalation in individuals being treated with infliximab to improve clinical outcomes, but additional studies are needed for other bDMARDs.

Per the ACR/EULAR^26,27 guidelines, all approved bDMARDs may be used without hierarchical positioning. However, after the failure of a TNFi agent, these guidelines do not provide specific advice about a preference between the “cycling” strategy (switching to an alternative TNFi) and “swapping” strategy (switching to a therapy with a different mode of action). Cycling might work for several reasons, including differences in the agents’ molecular structure, immunological mechanism of action, immunogenicity, and pharmacokinetics.^28-30 The cycling strategy is a well-established approach adopted by more than 94% of practicing rheumatologists, according to a national survey,³¹ and its efficacy is supported by trials and additional observational studies.^32-35

The greater clinical effectiveness of switching to infliximab compared with continuing with etanercept in patients with inadequate response to etanercept (n = 28) was suggested in the open-label OPPOSITE trial.³⁶ Data from the GO-AFTER trial³⁷ suggests that a greater proportion of patients with RA refractory to adalimumab, etanercept, or infliximab who were treated with golimumab achieved an ACR20 and ACR50 response compared with patients who received placebo, and this response persisted through 5 years.³⁸ More recently, certolizumab pegol and adalimumab were compared head-to-head in the EXXELERATE trial.³⁹ The results of this trial revealed the adequate efficacy of cycling to another TNFi after primary insufficient response to the first.

In studies from Finland and Sweden,^35,40 it has been observed that a better response is achieved in patients in whom TNF failure was initially due to secondary failure or intolerance rather than primary failure. A post-hoc analysis of the results of the GO-AFTER trial⁴¹ and from a few observational studies^35,40,42 revealed that switching from one TNFi to another, especially from a monoclonal antibody to a soluble receptor, was often more beneficial for RA patients than switching from a soluble receptor to a monoclonal antibody.

Optimization of Therapy Conjoined with csDMARDs

Methotrexate is one of the oldest and most effective csDMARDs available for the treatment of RA.⁴³ The 2016 EULAR guidelines recommend the addition of methotrexate and/or other csDMARDs to potentiate the effect of bDMARDs.²⁶ In the case of TNFi therapy, the observed synergistic effect between the monoclonal antibody and methotrexate may be explained by sustained suppression of ADA formation.⁴⁴ In the TEMPO,⁴⁵ PREMIER,¹⁸ and GO-BEFORE⁴⁶ trials, the addition of methotrexate led to improved clinical and radiological outcomes in patients treated with etanercept, adalimumab, and golimumab,⁴⁷ respectively. These findings were also demonstrated in several registries, where significant improvement in clinical response and retention rate of the TNFi agents was noted. Results have been replicated with non-TNFi bDMARDs, including abatacept^48,49 and rituximab.⁵⁰ Patients treated with interleukin (IL)-6 inhibitors in combination with methotrexate have shown significantly less radiographic progression compared to those treated with tocilizumab alone and those treated with monotherapy tocilizumab versus monotherapy methotrexate.^51,52 Results possibly favor the use of IL-6 inhibitors alone in those who cannot tolerate or have contraindications to methotrexate.

An open prospective study by Cohen et al added methotrexate to the treatment regimens of individuals on bDMARD monotherapy with a primary failure and found favorable changes in ACR20 and DAS28 scores at 3 and 12 months and therapeutic biological response (ESR, CRP) at 3 months.⁵³ Unlike monotherapy, in these situations methotrexate is known to be efficacious even at a lower dose, possibly at 7.5 mg to 10 mg per week. Some studies have shown that methotrexate administered parenterally may be more efficacious than when given orally.^54-58

In clinical trials and observational studies, leflunomide, sulfasalazine, and hydroxychloroquine have been used as alternate csDMARDs added to the treatment regimen.^59-62 There are, however, only 2 trials comparing the efficacy of methotrexate with that of other csDMARDs as concomitant treatment in patients with inadequate response to TNFi therapy. The RABBIT trial found a slight decrease in effectiveness with concomitant TNFi and leflunomide compared to TNFi/methotrexate, but overall each group had similar EULAR responses at 24 months.⁶³ A study by De Stefano et al found comparable ACR20 and DAS28 responses among individuals receiving TNFis with methotrexate or leflunomide.⁶¹

The “Swapping” Strategy

The efficacy of the swapping strategy has been shown in 3 randomized clinical trials demonstrating the superiority of abatacept, tocilizumab, and rituximab in the treatment of individuals with RA refractory to TNFis. Tocilizumab was studied in the RADIATE⁶⁴ trial, which involved 499 patients with inadequate response to 1 or more TNFi agents. The primary endpoint (24-week ACR20) was achieved by 50.0%, 30.4%, and 10.1% of patients in the 8 mg/kg, 4 mg/kg, and control groups, respectively (P < 0.001 for both tocilizumab groups versus placebo). The utility of abatacept as second-line therapy after initial TNF failure was evaluated in the ATTAIN⁶⁵ study. Participants with an inadequate response to etanercept or infliximab were randomly assigned to receive either abatacept or placebo. ACR50 response rates after 6 months of treatment were 20.3% with abatacept and 3.8% with placebo (P < 0.001). The SWITCH-RA study,⁶⁶ an observational study, compared rituximab to TNFis in 1112 participants with inadequate response to initial anti-TNF therapy. At 6 months, mean change in DAS28 was small but significantly greater for the rituximab group (–1.5 vs –1.1; P = 0.007). The difference in response rates was greatest among seropositive patients. These data suggest that rituximab has efficacy following TNFi failure, particularly for seropositive patients. Additionally, REFLEX⁶⁷ is the sole randomized controlled trial in patients with insufficient response to TNFis that showed significant prevention of radiographic progression at week 56 in patients on rituximab compared to placebo (mean change from baseline in total Genant-modified Sharp score, 1.00 vs 2.31, respectively; P = 0.005).

One study randomly assigned 399 patients with active RA who had inadequate response to prior TNFi therapy to tofacitinib⁶⁸ (5 mg twice daily or 10 mg twice daily) or placebo, both with methotrexate.⁶ After 3 months of treatment, ACR20 response rates (41.7% for 5 mg, 28.1% for 10 mg, 24.4% for placebo) and DAS28 remission rates (6.7% for 5 mg, 8.8% for 10 mg, 1.7% for placebo) were significantly greater among patients treated with tofacitinib compared to those treated with placebo. More recently, the RA-BEACON trial⁶⁹ demonstrated a consistent, beneficial treatment effect of baricitinib in patients with insufficient response to 1 or more TNFis. In this trial, 527 patients with an inadequate response to bDMARDs were randomly assigned to receive baricitinib 2 mg or 4 mg daily or placebo for 24 weeks. A higher proportion of patients receiving baricitinib 4 mg had an ACR20 response at week 12 compared with those treated with placebo (55% vs 27%, P < 0.001), and patients receiving the 4-mg dose had significant improvements from baseline in DAS28 and Health Assessment Questionnaire–Disability Index scores (P < 0.001 for both comparisons).

To Cycle or to Swap?

Several observational studies (SCQM-RA,⁷⁰ STURE,⁷¹ BSRBR,⁷² Favalli,⁴³ MIRAR,⁷³ SWITCH-RA,⁷⁴ ROC⁷²) have clearly demonstrated that the swapping strategy is favored over the cycling strategy. In the ROC study,⁷² patients were randomly assigned (based on physician discretion) to receive a non-TNF biologic or a TNFi. More patients in the non-TNF group than in the TNFi group showed low disease activity at week 24 (45% vs 28%; odds ratio [OR], 2.09; 95% confidence interval [CI], 1.27-3.43; P = 0.004) and at week 52 (41% vs 23%; OR, 2.26; 95% CI, 1.33-3.86; P = 0.003). The authors concluded that in patients having an insufficient response to TNFi therapy, a non-TNF biologic agent may be more effective than a second TNFi drug. Only a few studies^75-77 have demonstrated similar results between the 2 strategies. Overall, the available evidence seems to suggest the superiority of the swapping over the cycling strategy.

An important clinical pearl to keep in mind is that both swapping and cycling strategies might theoretically increase the risk of infection; however, limited evidence is reported in the literature. In a large retrospective analysis⁷⁸ of data on 4332 RA patients from a large US claims database, patients who had cycled between TNFi agents had a 30% to 40% increased risk of infection compared to patients treated with rituximab. Patients on infliximab had a 62% higher hazard of severe infections, and this has also been reported in an observational study.⁷⁹ In another study,⁷⁰ 41% of 201 patients with RA followed between 1999 and 2013 who swapped to abatacept/rituximab or tocilizumab developed adverse events, as compared to 59% of those who switched to a second TNFi.

Currently, there are no specific guidelines or biomarkers available to facilitate selection of specific treatment from among the classes of biologics. With the development of several new drugs and regulatory approval of baricitinib, physicians now have several biologic options to treat patients. A recent large time-trend study⁸⁰ deriving data from more than 200,000 patients with RA showed that etanercept remains the most frequently used agent for the treatment of RA; it also showed that the use of adalimumab and infliximab is decreasing, and that the use of newer agents, especially abatacept, golimumab, and certolizumab, has considerably risen in recent years. In this study, abatacept, rituximab, certolizumab, golimumab, tocilizumab, and tofacitinib accounted for 13.2%, 13.8%, 6.9%, 11.9%, and 7.5% switches from first TNFi therapy.

Jin et al⁸¹ studied factors associated with the choice of bDMARD for initial and subsequent use. They found that patients with commercial insurance had an 87% higher likelihood of initiating a bDMARD. In the Medicaid subgroup, African Americans had lower odds of initiating and switching bDMARDs than non-Hispanic whites. Prior use of steroids and nonbiologic DMARDs predicted both bDMARD initiation and subsequent switching. Etanercept, adalimumab, and infliximab were the most commonly used first- and second-line bDMARDS; patients on anakinra and golimumab were most likely to be switched to other bDMARDs.

Which treatment strategy is the most cost-effective?

Several studies have reported better treatment persistence rates among patients who are treated with the swapping strategy compared to the cycling strategy. In a retrospective analysis of claims data,⁸² the authors examined treatment persistence and health care costs in patients switching to biologics with a different mechanism of action or cycling to another TNFi. The mean cost was significantly lower among patients treated using the swapping strategy than among the TNFi cyclers, both for the total cost of care for RA and for the total cost of the targeted DMARDs in the first year after the change in therapy. The authors concluded that switching to a drug with a different mechanism of action is associated with higher treatment persistence and lower health care costs than TNFi cycling.

What about biosimilars?

Biosimilars are copies of already licensed biologics that are very similar to the biologics, but are made by different sponsors using independently derived cell lines and separately developed manufacturing processes.⁸³ Regarding biosimilar use, EULAR²⁶ states that biosimilar bDMARDs approved by the European Medicines Agency or US Food and Drug Administration have similar efficacy and safety as the originator bDMARDs, and recommends them as preferred agents if they are indeed appreciably cheaper than originator or other bDMARDs.

What are the novel treatment targets in RA?

New therapeutics for RA continue to be developed. One of the new agents is peficitinib (ASP015K), an oral, once-daily Janus kinase (Jak) inhibitor targeting Jak-1, Jak-2, and tyrosine kinase-2, with moderate selectivity for Jak-3. In a phase 2b trial, 100-mg and 150-mg doses of peficitinib achieved a statistically significant ACR20 response (48.3% and 56.3%) compared to placebo (29.4%) at 12 weeks.⁸⁴

Given the benefit of targeting TNF-α and IL-17 in RA, a novel molecule (ABT-122) that targets both human TNF and IL-17 has been developed. Two phase 1 studies⁸⁵ showed that dual neutralization of TNF and IL-17 with ABT-122 has characteristics acceptable for further exploration of therapeutic potential of this agent in TNF- and IL-17A–driven immune-mediated inflammatory diseases. Another novel drug is mavrilimumab, a human monoclonal antibody that targets granulocyte–macrophage colony-stimulating factor receptor α. A recent studyshowed that long-term treatment with mavrilimumab maintained response and was well-tolerated, with no increased incidence of treatment-emergent adverse events.⁸⁶

Namilumab (AMG203) is an immunoglobulin G1 monoclonal antibody that binds with high affinity to the GM-CSF ligand. In a phase 1b, randomized, double-blind study (PRIORA)⁸⁷ to assess namilumab in treating active, mild-to-moderate RA, significant improvement was seen in the DAS28-CRP score with namilumab (150 and 300 mg groups combined) compared with placebo at day 43 (P = 0.0117) and also 8 weeks after last dosing at day 99 (P = 0.0154). Adverse events were similar across different doses of namilumab and placebo, and included nasopharyngitis and exacerbation/worsening of RA. Another drug showing promise in RA is fosdagrocorat (PF-04171327), a potential dissociated agonist of the glucocorticoid receptor. A multicenter, double-blind, parallel-group, active- and placebo-controlled phase 2 study randomly assigned 86 patients to receive fosdagrocorat 10 mg, fosdagrocorat 25 mg, prednisone 5 mg, or placebo, all with stable background methotrexate therapy.⁸⁸ Both fosdagrocorat doses demonstrated efficacy in improving signs and symptoms in RA patients, with manageable adverse events.

Case Conclusion

There are several available treatment options for the case patient. Based on the PREMIER trial, solely increasing the dose of adalimumab is unlikely to provide a therapeutic benefit. Adding low-dose methotrexate (possibly via a parenteral route because of patient-reported gastrointestinal discomfort) might provide some synergistic and therapeutic effect. However, because of primary failure with TNFi therapy, she may benefit from the initiation of a biologic with a different mechanism of action (ie, swapping strategy). Therapeutic options include tocilizumab, abatacept, rituximab, and the Jak inhibitors (tofacitinib and baricitinib).

Summary

The optimal treatment of RA aims at achieving, and then maintaining, remission or a low disease activity. The choice of best treatment must be individualized to the patient, keeping in mind other factors, including comorbidities like fibromyalgia, history of diverticulitis (prior to use of tocilizumab), history of chronic obstructive pulmonary disease (prior to the use of abatacept), malignancy, and the presence of risk factors for infections (age, diabetes, chronic bronchitis). In a patient with inadequate response to initial biologic therapy, several options exist for the rheumatologist. Current evidence supports TNFi dose escalation for only infliximab; optimization of concurrent csDMARD or switching to a different csDMARD are other options. Cycling and swapping are other alternate approaches supported by many observational studies. While no head-to-head trials exist comparing the 2 strategies, data suggest superiority of the swapping strategy over the cycling approach. With the continuing development of novel therapeutics in RA, physicians have a growing list of treatment options to help their patients achieve disease remission.

Corresponding author: Namrata Singh, MD, 200 Hawkins Drive, Iowa City, IA 52242.

Financial disclosures: None.