User login
CHICAGO – , according to Priyanka Iyer, MD, an endocrinology fellow at MD Anderson Cancer Center, Houston.
“As long as we get glycemic control, they can continue,” she said at the annual meeting of the Endocrine Society.
Diabetes is a known side effect of immune checkpoint inhibitors (ICIs) but it’s rare, occurring in maybe 0.17% of patients, and its natural history and risk factors are unknown.
ICIs are fairly new agents, and as their use expands beyond clinical trials, “we anticipate seeing larger numbers of cases. Patients should be educated about the symptoms of uncontrolled blood sugars while on ICIs,” and endocrinologists “have to get involved and recognize this entity sooner,” Dr. Iyer said.
In short, her team found that ICI-mediated diabetes can occur in patients with or without preexisting diabetes, and that most patients have evidence of beta-cell failure, likely T-cell mediated destruction due to immune activation. In all but one case, patients remained on insulin at a median follow-up of 44 weeks, even after stopping ICIs. For most, ICI-mediated diabetes likely means lifelong insulin.
They were all on the programmed cell death protein (PD-1) inhibitors nivolumab (Opdivo) or pembrolizumab (Keytruda), or the PD-1 ligand (PD-L) inhibitor durvalumab (Imfinzi). The agents are used for a range of cancers, including renal cell, melanoma, and Hodgkin lymphoma. There were no diabetes cases in patients on single-agent ipilimumab (Yervoy) or tremelimumab, which target cytotoxic T-lymphocyte associated antigen-4 and are used for melanoma and mesothelioma.
Median time to diabetes presentation after the start of ICI treatment was 12.3 weeks but ranged from 1 to 67.2 weeks. Half of the cases presented in diabetic ketoacidosis (DKA). Patients had upward trending hyperglycemia and most had diabetes symptoms for a while before diagnosis. They presented with a blood glucose above 250 mg/dL, and more than half above 500 mg/dL. Median hemoglobin A1c at presentation was 8%, but ranged up to 12.5%.
Every patient required insulin, including the six that discontinued ICIs after developing diabetes. Diabetes resolved in just one patient at 10.2 months; she presented with DKA.
There were no obvious predisposing factors. None of the patients had histories of type 1 diabetes or other autoimmune disease. Five patients had well-controlled type 2 diabetes prior to ICI initiation; four had prediabetes. Some had family members with type 2 diabetes, but not type 1. Four had prior ICI exposure. Just three patients were on concomitant steroids.
A few patients also developed thyroid or pituitary dysfunction, which are more common side effects of ICIs.
The median age at diabetes presentation was 61 years and ranged from 32 to 82 years. The majority of patients were men, which reflects MD Anderson demographics, not a predisposing risk factor, Dr. Iyer said.
Melanoma was the most common cancer, followed by renal cell and prostate; patients had stage 2-4 disease. About half the subjects were on single agent anti-PD-1 treatment, about a third on anti-PD-1 combination treatment, and the rest on anti-PD-L1 combination therapy. C-peptide levels were below 0.9 ng/mL at diabetes diagnosis in most of the patients. Eleven of the 20 tested (55%) were positive for the pancreatic islet cell antibody GAD65.
The investigators had no disclosures. A funding source was not reported.
SOURCE: Iyer PC et al. Abstract OR05-5.
CHICAGO – , according to Priyanka Iyer, MD, an endocrinology fellow at MD Anderson Cancer Center, Houston.
“As long as we get glycemic control, they can continue,” she said at the annual meeting of the Endocrine Society.
Diabetes is a known side effect of immune checkpoint inhibitors (ICIs) but it’s rare, occurring in maybe 0.17% of patients, and its natural history and risk factors are unknown.
ICIs are fairly new agents, and as their use expands beyond clinical trials, “we anticipate seeing larger numbers of cases. Patients should be educated about the symptoms of uncontrolled blood sugars while on ICIs,” and endocrinologists “have to get involved and recognize this entity sooner,” Dr. Iyer said.
In short, her team found that ICI-mediated diabetes can occur in patients with or without preexisting diabetes, and that most patients have evidence of beta-cell failure, likely T-cell mediated destruction due to immune activation. In all but one case, patients remained on insulin at a median follow-up of 44 weeks, even after stopping ICIs. For most, ICI-mediated diabetes likely means lifelong insulin.
They were all on the programmed cell death protein (PD-1) inhibitors nivolumab (Opdivo) or pembrolizumab (Keytruda), or the PD-1 ligand (PD-L) inhibitor durvalumab (Imfinzi). The agents are used for a range of cancers, including renal cell, melanoma, and Hodgkin lymphoma. There were no diabetes cases in patients on single-agent ipilimumab (Yervoy) or tremelimumab, which target cytotoxic T-lymphocyte associated antigen-4 and are used for melanoma and mesothelioma.
Median time to diabetes presentation after the start of ICI treatment was 12.3 weeks but ranged from 1 to 67.2 weeks. Half of the cases presented in diabetic ketoacidosis (DKA). Patients had upward trending hyperglycemia and most had diabetes symptoms for a while before diagnosis. They presented with a blood glucose above 250 mg/dL, and more than half above 500 mg/dL. Median hemoglobin A1c at presentation was 8%, but ranged up to 12.5%.
Every patient required insulin, including the six that discontinued ICIs after developing diabetes. Diabetes resolved in just one patient at 10.2 months; she presented with DKA.
There were no obvious predisposing factors. None of the patients had histories of type 1 diabetes or other autoimmune disease. Five patients had well-controlled type 2 diabetes prior to ICI initiation; four had prediabetes. Some had family members with type 2 diabetes, but not type 1. Four had prior ICI exposure. Just three patients were on concomitant steroids.
A few patients also developed thyroid or pituitary dysfunction, which are more common side effects of ICIs.
The median age at diabetes presentation was 61 years and ranged from 32 to 82 years. The majority of patients were men, which reflects MD Anderson demographics, not a predisposing risk factor, Dr. Iyer said.
Melanoma was the most common cancer, followed by renal cell and prostate; patients had stage 2-4 disease. About half the subjects were on single agent anti-PD-1 treatment, about a third on anti-PD-1 combination treatment, and the rest on anti-PD-L1 combination therapy. C-peptide levels were below 0.9 ng/mL at diabetes diagnosis in most of the patients. Eleven of the 20 tested (55%) were positive for the pancreatic islet cell antibody GAD65.
The investigators had no disclosures. A funding source was not reported.
SOURCE: Iyer PC et al. Abstract OR05-5.
CHICAGO – , according to Priyanka Iyer, MD, an endocrinology fellow at MD Anderson Cancer Center, Houston.
“As long as we get glycemic control, they can continue,” she said at the annual meeting of the Endocrine Society.
Diabetes is a known side effect of immune checkpoint inhibitors (ICIs) but it’s rare, occurring in maybe 0.17% of patients, and its natural history and risk factors are unknown.
ICIs are fairly new agents, and as their use expands beyond clinical trials, “we anticipate seeing larger numbers of cases. Patients should be educated about the symptoms of uncontrolled blood sugars while on ICIs,” and endocrinologists “have to get involved and recognize this entity sooner,” Dr. Iyer said.
In short, her team found that ICI-mediated diabetes can occur in patients with or without preexisting diabetes, and that most patients have evidence of beta-cell failure, likely T-cell mediated destruction due to immune activation. In all but one case, patients remained on insulin at a median follow-up of 44 weeks, even after stopping ICIs. For most, ICI-mediated diabetes likely means lifelong insulin.
They were all on the programmed cell death protein (PD-1) inhibitors nivolumab (Opdivo) or pembrolizumab (Keytruda), or the PD-1 ligand (PD-L) inhibitor durvalumab (Imfinzi). The agents are used for a range of cancers, including renal cell, melanoma, and Hodgkin lymphoma. There were no diabetes cases in patients on single-agent ipilimumab (Yervoy) or tremelimumab, which target cytotoxic T-lymphocyte associated antigen-4 and are used for melanoma and mesothelioma.
Median time to diabetes presentation after the start of ICI treatment was 12.3 weeks but ranged from 1 to 67.2 weeks. Half of the cases presented in diabetic ketoacidosis (DKA). Patients had upward trending hyperglycemia and most had diabetes symptoms for a while before diagnosis. They presented with a blood glucose above 250 mg/dL, and more than half above 500 mg/dL. Median hemoglobin A1c at presentation was 8%, but ranged up to 12.5%.
Every patient required insulin, including the six that discontinued ICIs after developing diabetes. Diabetes resolved in just one patient at 10.2 months; she presented with DKA.
There were no obvious predisposing factors. None of the patients had histories of type 1 diabetes or other autoimmune disease. Five patients had well-controlled type 2 diabetes prior to ICI initiation; four had prediabetes. Some had family members with type 2 diabetes, but not type 1. Four had prior ICI exposure. Just three patients were on concomitant steroids.
A few patients also developed thyroid or pituitary dysfunction, which are more common side effects of ICIs.
The median age at diabetes presentation was 61 years and ranged from 32 to 82 years. The majority of patients were men, which reflects MD Anderson demographics, not a predisposing risk factor, Dr. Iyer said.
Melanoma was the most common cancer, followed by renal cell and prostate; patients had stage 2-4 disease. About half the subjects were on single agent anti-PD-1 treatment, about a third on anti-PD-1 combination treatment, and the rest on anti-PD-L1 combination therapy. C-peptide levels were below 0.9 ng/mL at diabetes diagnosis in most of the patients. Eleven of the 20 tested (55%) were positive for the pancreatic islet cell antibody GAD65.
The investigators had no disclosures. A funding source was not reported.
SOURCE: Iyer PC et al. Abstract OR05-5.
REPORTING FROM ENDO 2018
Key clinical point: Be on the lookout for new-onset diabetes when patients start immune checkpoint inhibitors.
Major finding: In all but one case, patients remained on insulin at a median follow-up of 44 weeks, even after stopping ICIs.
Study details: Review of 24 cases.
Disclosures: The investigators had no disclosures. A funding source was not reported.
Source: Iyer PC et al. Abstract OR05-5.