The American Journal of Orthopedics

ABSTRACT

Nonunion after a proximal tibia fracture is often associated with poor bone stock, (previous) infection, and compromised soft tissues. These conditions make revision internal fixation with double plating difficult. Combining a plate and contralateral 2-pin external fixator, coined composite fixation, can provide an alternative means of obtaining stability without further compromising soft tissues.

Three patients with a proximal tibia nonunion precluding standard internal fixation with double plating were treated with composite fixation. All 3 patients achieved union with deformity correction at a mean of 5.2 months (range, 5-5.5 months). The average range of motion (ROM) arc was 100° (range, 100°-115°) and postoperative ROM returned to pre-injury levels.

Composite fixation can be a helpful adjunct in the treatment of this challenging problem.

Continue to: Operative management of a proximal tibial nonunion...

Operative management of a proximal tibial nonunion is challenging, compromised by limited bone stock, pre-existing hardware, stiffness, poor soft tissue conditions, and infection. The goals of treatment include bone union, re-establishment of both joint stability and lower extremity alignment, restoration of an anatomic articular surface, and recovery of function.¹ Currently, various treatment options such as plate fixation, bone grafting, intramedullary nailing, external fixation, functional bracing, or a combination of these are available.^1-8 Rigid internal fixation is the gold standard for most nonunions. However, sometimes local soft tissues or bone quality preclude standard internal fixation. Bolhofner⁹ described the combination of a single plate and an external fixator on the contralateral side for the management of extra-articular proximal tibial fractures with compromised soft tissues, and the technique known as composite fixation was coined. The external fixator on 1 side and the plate on the other, generate a balanced, stable environment while limiting the use of foreign hardware, thereby avoiding both additional soft-tissue damage and periosteal stripping.^9-11 In this technical article, we describe the indication, technique, and outcomes of 3 patients with proximal tibial nonunions, who were successfully treated with composite fixation.

MATERIALS AND METHODS 

PATIENTS

Between January 2014 and July 2016, 3 patients each with a proximal tibial nonunion that developed after a bicondylar tibial plateau fracture (Schatzker type VI) were treated with composite fixation (Table). The 3 patients were female with an average age of 61 years (range, 60-62 years), and a body mass index of 23.7 kg/m² (range, 19.0-31.9 kg/m²). All 3 patients had sustained a tibial plateau fracture that was primarily treated with open reduction and internal fixation. Two of them had a diagnosis of rheumatoid arthritis and were being treated with methotrexate and Humira (adalimumab) (case 1), and with methotrexate, prednisolone, and etanercept (case 3). The etanercept was discontinued after discussion with the treating rheumatologist when a deep infection developed. Two patients (cases 1 and 2) were referred to us because of their nonunions. All 3 patients developed extra-articular nonunions with compromised bone stock. Two patients had developed deep infections during treatment of their plateau fractures; 1 of these patients underwent a medial gastrocnemius flap for wound coverage (case 1). The second patient (case 3) with a deep infection underwent partial hardware removal, a Masquelet salvage procedure, and revision plate fixation. However, the infection recurred. The hardware was removed, and 2 débridements with conversion to a hybrid external fixator with thin wire fixation were done. Due to her longstanding rheumatoid arthritis, the patient had bilateral valgus knee malalignment causing the ring fixator to strike her contralateral knee when she walked. The period from the initial tibial plateau fracture to our composite fixation averaged 11.3 months (range, 11-12 months). Indications for the use of the composite fixation comprised previously infected soft tissue on the lateral side and inability to walk with a hybrid thin wire fixator because of valgus knees (case 3), a medial gastrocnemius flap (case 2), and poor bone quality (case 1). Follow-up consisted of clinical examination, Timed Up and Go (TUG) test that is a standardized test for mobility, and radiographic evaluation at routine appointments up to 1 year or until healed.¹² At the last follow-up visit, patients filled out the International Knee Documentation Committee (IKDC) subjective knee form.¹³

SURGICAL TECHNIQUE

A fellowship-trained orthopedic trauma surgeon treated all patients. Patients were placed on a radiolucent operating table after general or regional anesthesia. Previous incisions were used. Two patients had a midline incision; the third had both a posteromedial and an anterolateral incision. Five deep tissue cultures were taken after which antibiotics were given intravenously. All unstable or failed hardware was removed. Aggressive débridement of the nonunion was performed. After débridement, multiple holes were drilled with a 2.0 mm drill bit until blood was seen to egress from both sides of the medullary canal. Malalignment of the proximal tibia was corrected and checked fluoroscopically. Fixation was done with an anatomic locking plate (LCP Proximal Tibia Plate 3.5; DePuy Synthes) with a mixture of locking and non-locking screws. In 2 patients, a tricortical graft from the posterior iliac crest was positioned in the defect. Additional autologous bone graft and demineralized bone matrix was added around the nonunion. Although locking screws were used, the fixation did not appear to be strong enough to resist the varus (cases 1 and 2), or the valgus (case 3) deforming forces. Additional fixation was thus needed. However, the contralateral soft tissues were compromised in case 2 (medial gastrocnemius flap), and case 3 (a previously infected area with very tenuous skin laterally), whereas the bone was considered to be of insufficient quality in case 1. The opposite side of the nonunion was stabilized using composite fixation with a 2-pin external fixator to circumvent the need for additional plate fixation. In 2 patients, the plate was placed laterally, and the external fixator medially. In the third patient, the plate was positioned medially, and the external fixator laterally. The plate was always placed first. The external fixator was placed last. Using fluoroscopy, we ensured that the fixator pins would not interfere with the screws. The pins were predrilled and positioned perpendicular to the tibia through small stab incisions. We prefer hydroxyapatite-coated pins (6-mm diameter, XCaliber Bone Screws; Pro-Motion Medical) to increase their holding power in the often osteopenic bone. Postoperative management consisted of toe-touch weight-bearing for 6 weeks and progressed to full weight-bearing at 3 months. Radiographs were taken on postoperative day 1, at 6 weeks, and at 12 weeks until healed. No continuous passive motion was used postoperatively. Antibiotics were continued until cultures were negative. No specific pin care was used. We advised patients to shower daily with the external fixator in place, once the wounds have healed.

Continue to: RESULTS... 

RESULTS

On average, patients were hospitalized for 5 days (range, 3-7 days). There were no postoperative complications. None of the patients developed a clinically significant pin site infection. There were no re-operations during follow-up. All patients achieved union at a mean of 5.2 months (range, 5-5.5 months) (Figure 1).

Deformity correction was achieved in all 3 patients. The average range of motion (ROM) arc was 100° (range, 100°-115°). None of the patients had an extension deficit. TUG test was <8 seconds in all patients. The IKDC knee score averaged 52 (range, 41-66). Of note is that 2 patients already had compromised knee function before the fracture because of rheumatoid arthritis. The Ahlbäck classification of osteoarthritis showed grade 1 in cases 1 and 3, and grade 2 in case 2.¹⁴ Postoperative ROM of the knee returned to pre-injury levels in all patients (Figure 2). The 2-pin external fixator was removed at 9 weeks on average (range, 6-12 weeks) postoperatively in the outpatient clinic. At the last follow-up appointment at an average of 10.3 months (range, 9-12 months), all wounds had healed without infection. All patients had a normal neurovascular examination.

DISCUSSION

Nonunion after a proximal tibial fracture is rare.⁴ In cases when nonunions do develop, they most often pertain to the extra-articular component with the plateau component healed. Surgical exposure for débridements, hardware removal, bone grafting, and revision of fixation carries the risk of wound breakdown, necrosis, and infection. The alternative strategy of composite fixation (a plate combined with a contralateral 2-pin external fixator) to limit additional soft tissue compromise was already described in proximal tibial fractures by Bolhofner.⁹ He treated 41 extra-articular proximal tibial fractures using this composite fixation technique and attained successful results with an average time to union of 12.1 weeks. There was only 1 malunion, 2 wound infections, and 3 delayed unions.

In our practice, we have extrapolated this idea to an extra-articular nonunion that developed after a tibial plateau fracture. With the use of an external fixator, we provided sufficient mechanical stability of the nonunion without unnecessarily compromising previously infected or tenuous soft tissues, a muscle flap, or further devascularizing poor bone. Limitations of this study include the retrospective data and small sample size prone to bias. However, all patients received the same treatment protocol from 1 orthopedic trauma surgeon, follow-up intervals were similar, and data were acquired consistently.

Meanwhile, we have used this technique in a fourth patient with a septic nonunion of a tibial plateau fracture. All 4 patients in whom we have used this method so far have healed successfully.

CONCLUSION

This technique respects both the demand for minimal soft tissue damage and a maximal stable environment without notable perioperative and postoperative complications. It also offers an alternative option for the treatment of a proximal tibial nonunion that is not amenable to invasive revision dual plate fixation. As such, it can be a useful addition to the existing armamentarium of the treating surgeon.

ABSTRACT

Nonunion after a proximal tibia fracture is often associated with poor bone stock, (previous) infection, and compromised soft tissues. These conditions make revision internal fixation with double plating difficult. Combining a plate and contralateral 2-pin external fixator, coined composite fixation, can provide an alternative means of obtaining stability without further compromising soft tissues.

Three patients with a proximal tibia nonunion precluding standard internal fixation with double plating were treated with composite fixation. All 3 patients achieved union with deformity correction at a mean of 5.2 months (range, 5-5.5 months). The average range of motion (ROM) arc was 100° (range, 100°-115°) and postoperative ROM returned to pre-injury levels.

Composite fixation can be a helpful adjunct in the treatment of this challenging problem.

Continue to: Operative management of a proximal tibial nonunion...

Operative management of a proximal tibial nonunion is challenging, compromised by limited bone stock, pre-existing hardware, stiffness, poor soft tissue conditions, and infection. The goals of treatment include bone union, re-establishment of both joint stability and lower extremity alignment, restoration of an anatomic articular surface, and recovery of function.¹ Currently, various treatment options such as plate fixation, bone grafting, intramedullary nailing, external fixation, functional bracing, or a combination of these are available.^1-8 Rigid internal fixation is the gold standard for most nonunions. However, sometimes local soft tissues or bone quality preclude standard internal fixation. Bolhofner⁹ described the combination of a single plate and an external fixator on the contralateral side for the management of extra-articular proximal tibial fractures with compromised soft tissues, and the technique known as composite fixation was coined. The external fixator on 1 side and the plate on the other, generate a balanced, stable environment while limiting the use of foreign hardware, thereby avoiding both additional soft-tissue damage and periosteal stripping.^9-11 In this technical article, we describe the indication, technique, and outcomes of 3 patients with proximal tibial nonunions, who were successfully treated with composite fixation.

MATERIALS AND METHODS 

PATIENTS

Between January 2014 and July 2016, 3 patients each with a proximal tibial nonunion that developed after a bicondylar tibial plateau fracture (Schatzker type VI) were treated with composite fixation (Table). The 3 patients were female with an average age of 61 years (range, 60-62 years), and a body mass index of 23.7 kg/m² (range, 19.0-31.9 kg/m²). All 3 patients had sustained a tibial plateau fracture that was primarily treated with open reduction and internal fixation. Two of them had a diagnosis of rheumatoid arthritis and were being treated with methotrexate and Humira (adalimumab) (case 1), and with methotrexate, prednisolone, and etanercept (case 3). The etanercept was discontinued after discussion with the treating rheumatologist when a deep infection developed. Two patients (cases 1 and 2) were referred to us because of their nonunions. All 3 patients developed extra-articular nonunions with compromised bone stock. Two patients had developed deep infections during treatment of their plateau fractures; 1 of these patients underwent a medial gastrocnemius flap for wound coverage (case 1). The second patient (case 3) with a deep infection underwent partial hardware removal, a Masquelet salvage procedure, and revision plate fixation. However, the infection recurred. The hardware was removed, and 2 débridements with conversion to a hybrid external fixator with thin wire fixation were done. Due to her longstanding rheumatoid arthritis, the patient had bilateral valgus knee malalignment causing the ring fixator to strike her contralateral knee when she walked. The period from the initial tibial plateau fracture to our composite fixation averaged 11.3 months (range, 11-12 months). Indications for the use of the composite fixation comprised previously infected soft tissue on the lateral side and inability to walk with a hybrid thin wire fixator because of valgus knees (case 3), a medial gastrocnemius flap (case 2), and poor bone quality (case 1). Follow-up consisted of clinical examination, Timed Up and Go (TUG) test that is a standardized test for mobility, and radiographic evaluation at routine appointments up to 1 year or until healed.¹² At the last follow-up visit, patients filled out the International Knee Documentation Committee (IKDC) subjective knee form.¹³

SURGICAL TECHNIQUE

A fellowship-trained orthopedic trauma surgeon treated all patients. Patients were placed on a radiolucent operating table after general or regional anesthesia. Previous incisions were used. Two patients had a midline incision; the third had both a posteromedial and an anterolateral incision. Five deep tissue cultures were taken after which antibiotics were given intravenously. All unstable or failed hardware was removed. Aggressive débridement of the nonunion was performed. After débridement, multiple holes were drilled with a 2.0 mm drill bit until blood was seen to egress from both sides of the medullary canal. Malalignment of the proximal tibia was corrected and checked fluoroscopically. Fixation was done with an anatomic locking plate (LCP Proximal Tibia Plate 3.5; DePuy Synthes) with a mixture of locking and non-locking screws. In 2 patients, a tricortical graft from the posterior iliac crest was positioned in the defect. Additional autologous bone graft and demineralized bone matrix was added around the nonunion. Although locking screws were used, the fixation did not appear to be strong enough to resist the varus (cases 1 and 2), or the valgus (case 3) deforming forces. Additional fixation was thus needed. However, the contralateral soft tissues were compromised in case 2 (medial gastrocnemius flap), and case 3 (a previously infected area with very tenuous skin laterally), whereas the bone was considered to be of insufficient quality in case 1. The opposite side of the nonunion was stabilized using composite fixation with a 2-pin external fixator to circumvent the need for additional plate fixation. In 2 patients, the plate was placed laterally, and the external fixator medially. In the third patient, the plate was positioned medially, and the external fixator laterally. The plate was always placed first. The external fixator was placed last. Using fluoroscopy, we ensured that the fixator pins would not interfere with the screws. The pins were predrilled and positioned perpendicular to the tibia through small stab incisions. We prefer hydroxyapatite-coated pins (6-mm diameter, XCaliber Bone Screws; Pro-Motion Medical) to increase their holding power in the often osteopenic bone. Postoperative management consisted of toe-touch weight-bearing for 6 weeks and progressed to full weight-bearing at 3 months. Radiographs were taken on postoperative day 1, at 6 weeks, and at 12 weeks until healed. No continuous passive motion was used postoperatively. Antibiotics were continued until cultures were negative. No specific pin care was used. We advised patients to shower daily with the external fixator in place, once the wounds have healed.

Continue to: RESULTS... 

RESULTS

On average, patients were hospitalized for 5 days (range, 3-7 days). There were no postoperative complications. None of the patients developed a clinically significant pin site infection. There were no re-operations during follow-up. All patients achieved union at a mean of 5.2 months (range, 5-5.5 months) (Figure 1).

Deformity correction was achieved in all 3 patients. The average range of motion (ROM) arc was 100° (range, 100°-115°). None of the patients had an extension deficit. TUG test was <8 seconds in all patients. The IKDC knee score averaged 52 (range, 41-66). Of note is that 2 patients already had compromised knee function before the fracture because of rheumatoid arthritis. The Ahlbäck classification of osteoarthritis showed grade 1 in cases 1 and 3, and grade 2 in case 2.¹⁴ Postoperative ROM of the knee returned to pre-injury levels in all patients (Figure 2). The 2-pin external fixator was removed at 9 weeks on average (range, 6-12 weeks) postoperatively in the outpatient clinic. At the last follow-up appointment at an average of 10.3 months (range, 9-12 months), all wounds had healed without infection. All patients had a normal neurovascular examination.

DISCUSSION

Nonunion after a proximal tibial fracture is rare.⁴ In cases when nonunions do develop, they most often pertain to the extra-articular component with the plateau component healed. Surgical exposure for débridements, hardware removal, bone grafting, and revision of fixation carries the risk of wound breakdown, necrosis, and infection. The alternative strategy of composite fixation (a plate combined with a contralateral 2-pin external fixator) to limit additional soft tissue compromise was already described in proximal tibial fractures by Bolhofner.⁹ He treated 41 extra-articular proximal tibial fractures using this composite fixation technique and attained successful results with an average time to union of 12.1 weeks. There was only 1 malunion, 2 wound infections, and 3 delayed unions.

In our practice, we have extrapolated this idea to an extra-articular nonunion that developed after a tibial plateau fracture. With the use of an external fixator, we provided sufficient mechanical stability of the nonunion without unnecessarily compromising previously infected or tenuous soft tissues, a muscle flap, or further devascularizing poor bone. Limitations of this study include the retrospective data and small sample size prone to bias. However, all patients received the same treatment protocol from 1 orthopedic trauma surgeon, follow-up intervals were similar, and data were acquired consistently.

Meanwhile, we have used this technique in a fourth patient with a septic nonunion of a tibial plateau fracture. All 4 patients in whom we have used this method so far have healed successfully.

CONCLUSION

This technique respects both the demand for minimal soft tissue damage and a maximal stable environment without notable perioperative and postoperative complications. It also offers an alternative option for the treatment of a proximal tibial nonunion that is not amenable to invasive revision dual plate fixation. As such, it can be a useful addition to the existing armamentarium of the treating surgeon.

ABSTRACT

Nonunion after a proximal tibia fracture is often associated with poor bone stock, (previous) infection, and compromised soft tissues. These conditions make revision internal fixation with double plating difficult. Combining a plate and contralateral 2-pin external fixator, coined composite fixation, can provide an alternative means of obtaining stability without further compromising soft tissues.

Three patients with a proximal tibia nonunion precluding standard internal fixation with double plating were treated with composite fixation. All 3 patients achieved union with deformity correction at a mean of 5.2 months (range, 5-5.5 months). The average range of motion (ROM) arc was 100° (range, 100°-115°) and postoperative ROM returned to pre-injury levels.

Composite fixation can be a helpful adjunct in the treatment of this challenging problem.

Continue to: Operative management of a proximal tibial nonunion...

Operative management of a proximal tibial nonunion is challenging, compromised by limited bone stock, pre-existing hardware, stiffness, poor soft tissue conditions, and infection. The goals of treatment include bone union, re-establishment of both joint stability and lower extremity alignment, restoration of an anatomic articular surface, and recovery of function.¹ Currently, various treatment options such as plate fixation, bone grafting, intramedullary nailing, external fixation, functional bracing, or a combination of these are available.^1-8 Rigid internal fixation is the gold standard for most nonunions. However, sometimes local soft tissues or bone quality preclude standard internal fixation. Bolhofner⁹ described the combination of a single plate and an external fixator on the contralateral side for the management of extra-articular proximal tibial fractures with compromised soft tissues, and the technique known as composite fixation was coined. The external fixator on 1 side and the plate on the other, generate a balanced, stable environment while limiting the use of foreign hardware, thereby avoiding both additional soft-tissue damage and periosteal stripping.^9-11 In this technical article, we describe the indication, technique, and outcomes of 3 patients with proximal tibial nonunions, who were successfully treated with composite fixation.

MATERIALS AND METHODS 

PATIENTS

Between January 2014 and July 2016, 3 patients each with a proximal tibial nonunion that developed after a bicondylar tibial plateau fracture (Schatzker type VI) were treated with composite fixation (Table). The 3 patients were female with an average age of 61 years (range, 60-62 years), and a body mass index of 23.7 kg/m² (range, 19.0-31.9 kg/m²). All 3 patients had sustained a tibial plateau fracture that was primarily treated with open reduction and internal fixation. Two of them had a diagnosis of rheumatoid arthritis and were being treated with methotrexate and Humira (adalimumab) (case 1), and with methotrexate, prednisolone, and etanercept (case 3). The etanercept was discontinued after discussion with the treating rheumatologist when a deep infection developed. Two patients (cases 1 and 2) were referred to us because of their nonunions. All 3 patients developed extra-articular nonunions with compromised bone stock. Two patients had developed deep infections during treatment of their plateau fractures; 1 of these patients underwent a medial gastrocnemius flap for wound coverage (case 1). The second patient (case 3) with a deep infection underwent partial hardware removal, a Masquelet salvage procedure, and revision plate fixation. However, the infection recurred. The hardware was removed, and 2 débridements with conversion to a hybrid external fixator with thin wire fixation were done. Due to her longstanding rheumatoid arthritis, the patient had bilateral valgus knee malalignment causing the ring fixator to strike her contralateral knee when she walked. The period from the initial tibial plateau fracture to our composite fixation averaged 11.3 months (range, 11-12 months). Indications for the use of the composite fixation comprised previously infected soft tissue on the lateral side and inability to walk with a hybrid thin wire fixator because of valgus knees (case 3), a medial gastrocnemius flap (case 2), and poor bone quality (case 1). Follow-up consisted of clinical examination, Timed Up and Go (TUG) test that is a standardized test for mobility, and radiographic evaluation at routine appointments up to 1 year or until healed.¹² At the last follow-up visit, patients filled out the International Knee Documentation Committee (IKDC) subjective knee form.¹³

SURGICAL TECHNIQUE

A fellowship-trained orthopedic trauma surgeon treated all patients. Patients were placed on a radiolucent operating table after general or regional anesthesia. Previous incisions were used. Two patients had a midline incision; the third had both a posteromedial and an anterolateral incision. Five deep tissue cultures were taken after which antibiotics were given intravenously. All unstable or failed hardware was removed. Aggressive débridement of the nonunion was performed. After débridement, multiple holes were drilled with a 2.0 mm drill bit until blood was seen to egress from both sides of the medullary canal. Malalignment of the proximal tibia was corrected and checked fluoroscopically. Fixation was done with an anatomic locking plate (LCP Proximal Tibia Plate 3.5; DePuy Synthes) with a mixture of locking and non-locking screws. In 2 patients, a tricortical graft from the posterior iliac crest was positioned in the defect. Additional autologous bone graft and demineralized bone matrix was added around the nonunion. Although locking screws were used, the fixation did not appear to be strong enough to resist the varus (cases 1 and 2), or the valgus (case 3) deforming forces. Additional fixation was thus needed. However, the contralateral soft tissues were compromised in case 2 (medial gastrocnemius flap), and case 3 (a previously infected area with very tenuous skin laterally), whereas the bone was considered to be of insufficient quality in case 1. The opposite side of the nonunion was stabilized using composite fixation with a 2-pin external fixator to circumvent the need for additional plate fixation. In 2 patients, the plate was placed laterally, and the external fixator medially. In the third patient, the plate was positioned medially, and the external fixator laterally. The plate was always placed first. The external fixator was placed last. Using fluoroscopy, we ensured that the fixator pins would not interfere with the screws. The pins were predrilled and positioned perpendicular to the tibia through small stab incisions. We prefer hydroxyapatite-coated pins (6-mm diameter, XCaliber Bone Screws; Pro-Motion Medical) to increase their holding power in the often osteopenic bone. Postoperative management consisted of toe-touch weight-bearing for 6 weeks and progressed to full weight-bearing at 3 months. Radiographs were taken on postoperative day 1, at 6 weeks, and at 12 weeks until healed. No continuous passive motion was used postoperatively. Antibiotics were continued until cultures were negative. No specific pin care was used. We advised patients to shower daily with the external fixator in place, once the wounds have healed.

Continue to: RESULTS... 

RESULTS

On average, patients were hospitalized for 5 days (range, 3-7 days). There were no postoperative complications. None of the patients developed a clinically significant pin site infection. There were no re-operations during follow-up. All patients achieved union at a mean of 5.2 months (range, 5-5.5 months) (Figure 1).

Deformity correction was achieved in all 3 patients. The average range of motion (ROM) arc was 100° (range, 100°-115°). None of the patients had an extension deficit. TUG test was <8 seconds in all patients. The IKDC knee score averaged 52 (range, 41-66). Of note is that 2 patients already had compromised knee function before the fracture because of rheumatoid arthritis. The Ahlbäck classification of osteoarthritis showed grade 1 in cases 1 and 3, and grade 2 in case 2.¹⁴ Postoperative ROM of the knee returned to pre-injury levels in all patients (Figure 2). The 2-pin external fixator was removed at 9 weeks on average (range, 6-12 weeks) postoperatively in the outpatient clinic. At the last follow-up appointment at an average of 10.3 months (range, 9-12 months), all wounds had healed without infection. All patients had a normal neurovascular examination.

DISCUSSION

Nonunion after a proximal tibial fracture is rare.⁴ In cases when nonunions do develop, they most often pertain to the extra-articular component with the plateau component healed. Surgical exposure for débridements, hardware removal, bone grafting, and revision of fixation carries the risk of wound breakdown, necrosis, and infection. The alternative strategy of composite fixation (a plate combined with a contralateral 2-pin external fixator) to limit additional soft tissue compromise was already described in proximal tibial fractures by Bolhofner.⁹ He treated 41 extra-articular proximal tibial fractures using this composite fixation technique and attained successful results with an average time to union of 12.1 weeks. There was only 1 malunion, 2 wound infections, and 3 delayed unions.

In our practice, we have extrapolated this idea to an extra-articular nonunion that developed after a tibial plateau fracture. With the use of an external fixator, we provided sufficient mechanical stability of the nonunion without unnecessarily compromising previously infected or tenuous soft tissues, a muscle flap, or further devascularizing poor bone. Limitations of this study include the retrospective data and small sample size prone to bias. However, all patients received the same treatment protocol from 1 orthopedic trauma surgeon, follow-up intervals were similar, and data were acquired consistently.

Meanwhile, we have used this technique in a fourth patient with a septic nonunion of a tibial plateau fracture. All 4 patients in whom we have used this method so far have healed successfully.

CONCLUSION

This technique respects both the demand for minimal soft tissue damage and a maximal stable environment without notable perioperative and postoperative complications. It also offers an alternative option for the treatment of a proximal tibial nonunion that is not amenable to invasive revision dual plate fixation. As such, it can be a useful addition to the existing armamentarium of the treating surgeon.

ABSTRACT

The incidence of acetabular fractures and associated in-hospital complication rates in the United States are poorly defined. Studies evaluating predictors of outcome for isolated acetabular fractures are weakly generalizable due to small sample sizes or the inclusion of all types of pelvic fractures. This study sought to analyze trends in acetabular fractures and associated complications in the US using the largest and most recent national dataset available.

The National Hospital Discharge Survey was queried to identify all patients admitted to US hospitals with acetabular fractures between 1990 and 2010. A representative cohort of 497,389 patients was identified, and multivariable logistic regression was used to identify independent predictors of mortality, adverse events, requirement of blood transfusion, and operative treatment with open reduction and internal fixation (ORIF).

Between 1990 and 2010, the population-adjusted incidence of acetabular fractures increased from 7.8 to 9.5/100,000 capita (P < .001). Mortality declined from 5.9% to 0.4% (P < .001), paralleling an increase in the proportion of patients treated with ORIF (12.6%-20.4%, P < .001), which was the variable associated with the lowest odds of mortality. Surgical intervention was associated with higher odds of adverse events and a requirement for blood transfusion. The average in-hospital length of stay decreased from 17.0 days to 10.3 days (P < .001).

This study provides the largest and most comprehensive epidemiologic analysis of acetabular fractures in the US. Knowledge of the increasing incidence of acetabular fractures and prognostic factors associated with poor outcomes may improve outcomes.

Continue to: Acetabular fractures are major injuries...

Acetabular fractures are major injuries frequently associated with life-altering sequelae¹ and a significant resulting cost to society.² Acetabular fractures are most often the result of a high-energy trauma^3-5 or fall from a height.^5,6 Functional outcomes and the prevention of post-traumatic arthritis have been shown to depend upon the accuracy of operative reduction.^7-9 However, literature on the epidemiology of acetabular fractures is largely limited to European countries,^1,10 and their incidence in the United States is more poorly defined.¹¹ Published mortality rates in the existing literature vary widely from 2% to 45%,^12-14 and few studies have identified the risk factors associated with in-hospital complications.¹⁵ While age, gender, and high-velocity mechanisms have been linked to increased mortality and complications,^14-16 the evidence for these associations is poorly generalizable due to the inclusion of all pelvic fractures in these studies. Some reports suggest that advances in surgical management have improved survival and functional outcome,^15,17 but these are based upon small cohorts. Knowledge of the incidence and patterns of disease burden are crucial for the allocation of limited healthcare resources.

ABSTRACT

The incidence of acetabular fractures and associated in-hospital complication rates in the United States are poorly defined. Studies evaluating predictors of outcome for isolated acetabular fractures are weakly generalizable due to small sample sizes or the inclusion of all types of pelvic fractures. This study sought to analyze trends in acetabular fractures and associated complications in the US using the largest and most recent national dataset available.

The National Hospital Discharge Survey was queried to identify all patients admitted to US hospitals with acetabular fractures between 1990 and 2010. A representative cohort of 497,389 patients was identified, and multivariable logistic regression was used to identify independent predictors of mortality, adverse events, requirement of blood transfusion, and operative treatment with open reduction and internal fixation (ORIF).

Between 1990 and 2010, the population-adjusted incidence of acetabular fractures increased from 7.8 to 9.5/100,000 capita (P < .001). Mortality declined from 5.9% to 0.4% (P < .001), paralleling an increase in the proportion of patients treated with ORIF (12.6%-20.4%, P < .001), which was the variable associated with the lowest odds of mortality. Surgical intervention was associated with higher odds of adverse events and a requirement for blood transfusion. The average in-hospital length of stay decreased from 17.0 days to 10.3 days (P < .001).

This study provides the largest and most comprehensive epidemiologic analysis of acetabular fractures in the US. Knowledge of the increasing incidence of acetabular fractures and prognostic factors associated with poor outcomes may improve outcomes.

Continue to: Acetabular fractures are major injuries...

Acetabular fractures are major injuries frequently associated with life-altering sequelae¹ and a significant resulting cost to society.² Acetabular fractures are most often the result of a high-energy trauma^3-5 or fall from a height.^5,6 Functional outcomes and the prevention of post-traumatic arthritis have been shown to depend upon the accuracy of operative reduction.^7-9 However, literature on the epidemiology of acetabular fractures is largely limited to European countries,^1,10 and their incidence in the United States is more poorly defined.¹¹ Published mortality rates in the existing literature vary widely from 2% to 45%,^12-14 and few studies have identified the risk factors associated with in-hospital complications.¹⁵ While age, gender, and high-velocity mechanisms have been linked to increased mortality and complications,^14-16 the evidence for these associations is poorly generalizable due to the inclusion of all pelvic fractures in these studies. Some reports suggest that advances in surgical management have improved survival and functional outcome,^15,17 but these are based upon small cohorts. Knowledge of the incidence and patterns of disease burden are crucial for the allocation of limited healthcare resources.

ABSTRACT

The incidence of acetabular fractures and associated in-hospital complication rates in the United States are poorly defined. Studies evaluating predictors of outcome for isolated acetabular fractures are weakly generalizable due to small sample sizes or the inclusion of all types of pelvic fractures. This study sought to analyze trends in acetabular fractures and associated complications in the US using the largest and most recent national dataset available.

The National Hospital Discharge Survey was queried to identify all patients admitted to US hospitals with acetabular fractures between 1990 and 2010. A representative cohort of 497,389 patients was identified, and multivariable logistic regression was used to identify independent predictors of mortality, adverse events, requirement of blood transfusion, and operative treatment with open reduction and internal fixation (ORIF).

Between 1990 and 2010, the population-adjusted incidence of acetabular fractures increased from 7.8 to 9.5/100,000 capita (P < .001). Mortality declined from 5.9% to 0.4% (P < .001), paralleling an increase in the proportion of patients treated with ORIF (12.6%-20.4%, P < .001), which was the variable associated with the lowest odds of mortality. Surgical intervention was associated with higher odds of adverse events and a requirement for blood transfusion. The average in-hospital length of stay decreased from 17.0 days to 10.3 days (P < .001).

This study provides the largest and most comprehensive epidemiologic analysis of acetabular fractures in the US. Knowledge of the increasing incidence of acetabular fractures and prognostic factors associated with poor outcomes may improve outcomes.

Continue to: Acetabular fractures are major injuries...

Acetabular fractures are major injuries frequently associated with life-altering sequelae¹ and a significant resulting cost to society.² Acetabular fractures are most often the result of a high-energy trauma^3-5 or fall from a height.^5,6 Functional outcomes and the prevention of post-traumatic arthritis have been shown to depend upon the accuracy of operative reduction.^7-9 However, literature on the epidemiology of acetabular fractures is largely limited to European countries,^1,10 and their incidence in the United States is more poorly defined.¹¹ Published mortality rates in the existing literature vary widely from 2% to 45%,^12-14 and few studies have identified the risk factors associated with in-hospital complications.¹⁵ While age, gender, and high-velocity mechanisms have been linked to increased mortality and complications,^14-16 the evidence for these associations is poorly generalizable due to the inclusion of all pelvic fractures in these studies. Some reports suggest that advances in surgical management have improved survival and functional outcome,^15,17 but these are based upon small cohorts. Knowledge of the incidence and patterns of disease burden are crucial for the allocation of limited healthcare resources.

ABSTRACT

A 61-year-old woman with a periprosthetic knee joint infection caused by Mycobacterium abscessus was successfully treated with surgical débridement, multidrug antimicrobial therapy, and staged reimplantation. To the authors’ knowledge, this represents the first report of successfully treating this organism after knee arthroplasty.

M. abscessus knee infections are rare, and there are no specific guidelines to inform treatment or successful treatment regimens for periprosthetic knee infections. Medical management alone was not successful in this case and hence cannot be recommended. Using a collaborative multidisciplinary approach, including surgical débridement, staged reimplantation, and multidrug antimicrobials, successful eradication of the periprosthetic joint infection caused by M. abscessus was achieved.  

Continue to: Total knee arthroplasty...

Total knee arthroplasty (TKA) procedures are projected to increase by more than 6-fold by 2030, with concurrent increases in revision TKA for infection projected.¹ Infection after TKA remains one of the most serious complications of the procedure, occurring in <2% of primary TKAs.² The majority of prosthetic joint infections (PJIs) are caused by staphylococci and streptococci.³ Although infection and treatment of PJIs by mycobacterial species have been described, there are presently no established treatment guidelines for mycobacterial PJIs.^4,5

Given the scarcity of clinical experience in dealing with these organisms, and the predicted increasing incidence of revision knee arthroplasty due to infection, we describe an unusual case of a PJI caused by Mycobacterium abscessus (M. abscessus), which was successfully treated using a combination of antimicrobial therapy and staged reconstruction. The patient provided written informed consent for print and electronic publication of this case report.

BACKGROUND

Mycobacteria are common environmental organisms that can survive harsh conditions, including low pH and extreme temperatures. They form biofilms and may be difficult to eradicate in cases of infection.⁶M. abscessus has proven to be difficult to eradicate due to limited antimicrobial susceptibility, lack of bactericidal options, and the variable presence of the erm gene, which yields inducible resistance to macrolides.⁷ Post-procedural outbreaks due to mycobacteria have been reported, often attributed to contaminated multiuse instruments, inadequate sterilization of tap water, multiuse vials, or improper skin preparation.^6,8-13

CASE REPORT 

A 61-year-old woman was referred with a 3-year history of progressive left knee pain and swelling. Before 8 months, she had undergone knee arthroscopy and had been treated with multiple steroid and hyaluronic acid injections, as well as ultrasound-guided aspiration of a Baker’s cyst (Figures 1A, 1B).

She elected to proceed with TKA 1 month after her last steroid injection. There was no preoperative concern for native joint infection. At the time of arthroplasty, clear joint fluid was encountered, and a deep tissue culture was taken (Figures 2A-2C).

Routine screening cultures for acid-fast bacilli (AFB) returned positive 9 days after the index arthroplasty, with subsequent identification of a nontuberculous mycobacterium (NTM), M. abscessus, subspecies massiliense. Sensitivity tests revealed susceptibility to amikacin, cefoxitin, and tigecycline (Table 1). The isolate was found to have inducible macrolide resistance by erm gene testing.

Table 1. Initial Mycobacterium abscessus massiliense Susceptibilities

^aAt 3 days; erm gene detected at 7 days.

Given no prior surgical suspicion for infection and the uncertain significance of the culture result, treatment options were debated. Medical management was selected based on the presumption that if infection was present, it was a native joint infection in which surgical débridement had already been undertaken at the time of primary arthroplasty. Similar reports for the treatment of M. tuberculosis infection in the knee have been reported with some success.^14,15 Short-interval reassessment was planned. Antimicrobial therapy was selected based on susceptibility data and clinical experience and consisted of intravenous (IV) cefoxitin, oral clarithromycin, and thrice-weekly intravenous amikacin. Over the ensuing weeks, she developed fevers, knee swelling, and persistent elevation of erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). With known potential of this organism for biofilm formation in other areas of the body and positive repeat cultures of the knee joint fluid, confirming the offending organism, a deep and resistant infection of the implant could not be excluded. Therefore, in an attempt to give the patient the best opportunity for clinical cure, the patient subsequently underwent a 2-stage antibiotic spacer explantation and exchange (Figures 3A, 3B). Moderate caseous material was present throughout the knee joint and the subcutaneous tissues. All bone was débrided, and complete synovectomy was undertaken, along with the removal of all implants. The antibiotic concentrations within the spacer were selected by guidance from the Infectious Disease and Pharmacy based on minimal inhibitory concentrations, with 3 packages of cement (40 g each) utilized and a total of 10 g of amikacin and 24 g of cefoxitin contained within the spacer. The patient continued systemic administration of amikacin, cefoxitin, and clarithromycin.

Continue to: One month postoperatively...

One month postoperatively, her constitutional symptoms, including fevers and night sweats, abated and inflammatory markers (ESR and CRP) had normalized. There were no clinical signs of infection. Amikacin was discontinued due to a 10-dB change on audiologic screening (4-6 kHz range), and tigecycline was substituted. Ultimately, she underwent 15 weeks of antimycobacterial therapy, 10 of which were after the explantation.

Eight weeks after cessation of her antibiotics, she underwent open biopsy. Multiple operative tissue samples showed negative results in pathology and culture tests.

Replantation was performed 14 weeks after stopping antimicrobials and 24 weeks after her explantation. The bone appeared healthy without evidence of osteomyelitis. A constrained reconstruction was secured with tobramycin-impregnated cement. One small island of necrotizing granuloma was observed within the bony cortex on histologic review; the granulomata appeared active with scattered neutrophils along with histiocytes and lymphocytes. AFB stains were negative. Intraoperative cultures, including mycobacterial cultures, were negative.

Based on the histologic evidence that infection may have persisted, and given the high stakes, antimicrobial treatment was reinitiated. Amikacin was again stopped after 3 weeks due to the development of tinnitus; tigecycline was substituted to complete the fourth and final week, at which point all antibiotics were discontinued. The patient was followed up uneventfully for 4 years (Figures 4A-4D and 5A-5C) with normal ESR and CRP. She continues to be ambulatory without assistive devices and walks an average of 30 miles per week without pain or constitutional symptoms.

Continue to: DISCUSSION...

DISCUSSION

Diagnosis of acute infection after TKA remains challenging, as some degree of pain, swelling, and even postoperative fevers may be common in noninfected TKA patients. Synovial white blood cell count and differential as well as alpha-defensin levels have been cited as predictive factors of infection.^16,17 Deep tissue and synovial fluid cultures offer the advantage of both identification and antimicrobial sensitivity testing of the offending organism. In this case, culture of the knee joint fluid at the time of TKA led to the unexpected finding of M. abscessus infection.

Preventable outbreaks due to M. abscessus have been reported and attributed to contaminated multiuse instruments, inadequate sterilization of tap water, multiuse vials, and improper skin preparation.^11-13 Rarely, M. abscessus has been reported as the cause of PJI. When an unusual organism is encountered after native joint instrumentation, an investigation should be undertaken to identify the source of contamination, with the assistance of infection control practitioners and/or the US Food and Drug Administration reporting. Reporting and investigation was undertaken in this case, though no suspect source could be identified.

Although there were no signs of infection prior to the TKA, there is an ongoing debate as to whether intra-articular corticosteroid injections increase the risk of PJIs, and if so, what the optimal amount of time to wait between procedures is. Although several earlier studies have been underpowered to answer these questions,¹⁸ this patient underwent TKA 1 month following the corticosteroid injection. Recent meta-analyses have shown no definitive evidence to indicate that this increased her risk of PJI.^19,20

Continue to: Treatments for mycobacterial infections...

Treatments for mycobacterial infections have been described with variable efficacy,^21,22 and only 2 cases of successfully treated PJIs have been reported after infection with M. abscessus. Both these cases were described in total hip arthroplasties,^23,24 and to the authors’ knowledge, this report represents the first described successfully treated case after TKA. Staged reconstruction remains a standard treatment for invasive organisms chronically infecting prosthetic joint implants, with reimplantation pending joint sterility and improvement in inflammatory markers.³ Previous successful reports of treating M. abscessus describe either resection arthroplasty²¹ or staged reconstruction.^23,24 The authors reported variable multidrug antimicrobial regimens, as summarized in Table 2, as guidelines for the treatment of mycobacterial PJI are currently not available.

CONCLUSION

This case report represents an episode of iatrogenic septic arthritis caused by Mycobacteria of the native knee after previous history of instrumentation, corticosteroid, and hyaluronic acid injections, with an overall indolent clinical course until subsequent arthroplasty. There were several important lessons learned, which are as follows: 1) Multidrug combination with antimicrobial therapy combined with aggressive surgical débridement and staged reimplantation permitted successful eradication of TKA PJI caused by M. abscessus in this patient. 2) Initial medical management alone was not successful and cannot be recommended for the treatment of M. abscessus in the setting of PJI. 3) Delaying the surgical débridement and the reconstructive course for a trial of medical management contributed to the ultimate requirement of a tibial tubercle osteotomy for an ankylosed knee at replantation. In this case, we initially had a low index of suspicion for deep infection, contributing to delayed surgical débridement. Ideally, a high degree of clinical suspicion should be maintained for joint infection in the presence of positive culture isolates of M. abscessus, as it may have a delayed clinical presentation of the typical features of PJI (fevers, swelling, erythema, etc). In such cases, the authors recommend consideration of early surgical débridement. 4) Medical management of TKA PJI is not without risks. Careful monitoring of patient side effects during antimicrobial administration remains paramount, as this patient did sustain a degree of hearing loss associated with prolonged medical therapy. 5) In complicated PJIs involving rare and intrinsically resistant organisms, a collaborative multidisciplinary approach, including specialists in orthopedic surgery, infectious disease, microbiology, pharmacy, and pathology, may be the preferred path to clinical cure.

ABSTRACT

A 61-year-old woman with a periprosthetic knee joint infection caused by Mycobacterium abscessus was successfully treated with surgical débridement, multidrug antimicrobial therapy, and staged reimplantation. To the authors’ knowledge, this represents the first report of successfully treating this organism after knee arthroplasty.

M. abscessus knee infections are rare, and there are no specific guidelines to inform treatment or successful treatment regimens for periprosthetic knee infections. Medical management alone was not successful in this case and hence cannot be recommended. Using a collaborative multidisciplinary approach, including surgical débridement, staged reimplantation, and multidrug antimicrobials, successful eradication of the periprosthetic joint infection caused by M. abscessus was achieved.  

Continue to: Total knee arthroplasty...

Total knee arthroplasty (TKA) procedures are projected to increase by more than 6-fold by 2030, with concurrent increases in revision TKA for infection projected.¹ Infection after TKA remains one of the most serious complications of the procedure, occurring in <2% of primary TKAs.² The majority of prosthetic joint infections (PJIs) are caused by staphylococci and streptococci.³ Although infection and treatment of PJIs by mycobacterial species have been described, there are presently no established treatment guidelines for mycobacterial PJIs.^4,5

Given the scarcity of clinical experience in dealing with these organisms, and the predicted increasing incidence of revision knee arthroplasty due to infection, we describe an unusual case of a PJI caused by Mycobacterium abscessus (M. abscessus), which was successfully treated using a combination of antimicrobial therapy and staged reconstruction. The patient provided written informed consent for print and electronic publication of this case report.

BACKGROUND

Mycobacteria are common environmental organisms that can survive harsh conditions, including low pH and extreme temperatures. They form biofilms and may be difficult to eradicate in cases of infection.⁶M. abscessus has proven to be difficult to eradicate due to limited antimicrobial susceptibility, lack of bactericidal options, and the variable presence of the erm gene, which yields inducible resistance to macrolides.⁷ Post-procedural outbreaks due to mycobacteria have been reported, often attributed to contaminated multiuse instruments, inadequate sterilization of tap water, multiuse vials, or improper skin preparation.^6,8-13

CASE REPORT 

A 61-year-old woman was referred with a 3-year history of progressive left knee pain and swelling. Before 8 months, she had undergone knee arthroscopy and had been treated with multiple steroid and hyaluronic acid injections, as well as ultrasound-guided aspiration of a Baker’s cyst (Figures 1A, 1B).

She elected to proceed with TKA 1 month after her last steroid injection. There was no preoperative concern for native joint infection. At the time of arthroplasty, clear joint fluid was encountered, and a deep tissue culture was taken (Figures 2A-2C).

Routine screening cultures for acid-fast bacilli (AFB) returned positive 9 days after the index arthroplasty, with subsequent identification of a nontuberculous mycobacterium (NTM), M. abscessus, subspecies massiliense. Sensitivity tests revealed susceptibility to amikacin, cefoxitin, and tigecycline (Table 1). The isolate was found to have inducible macrolide resistance by erm gene testing.

Table 1. Initial Mycobacterium abscessus massiliense Susceptibilities

^aAt 3 days; erm gene detected at 7 days.

Given no prior surgical suspicion for infection and the uncertain significance of the culture result, treatment options were debated. Medical management was selected based on the presumption that if infection was present, it was a native joint infection in which surgical débridement had already been undertaken at the time of primary arthroplasty. Similar reports for the treatment of M. tuberculosis infection in the knee have been reported with some success.^14,15 Short-interval reassessment was planned. Antimicrobial therapy was selected based on susceptibility data and clinical experience and consisted of intravenous (IV) cefoxitin, oral clarithromycin, and thrice-weekly intravenous amikacin. Over the ensuing weeks, she developed fevers, knee swelling, and persistent elevation of erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). With known potential of this organism for biofilm formation in other areas of the body and positive repeat cultures of the knee joint fluid, confirming the offending organism, a deep and resistant infection of the implant could not be excluded. Therefore, in an attempt to give the patient the best opportunity for clinical cure, the patient subsequently underwent a 2-stage antibiotic spacer explantation and exchange (Figures 3A, 3B). Moderate caseous material was present throughout the knee joint and the subcutaneous tissues. All bone was débrided, and complete synovectomy was undertaken, along with the removal of all implants. The antibiotic concentrations within the spacer were selected by guidance from the Infectious Disease and Pharmacy based on minimal inhibitory concentrations, with 3 packages of cement (40 g each) utilized and a total of 10 g of amikacin and 24 g of cefoxitin contained within the spacer. The patient continued systemic administration of amikacin, cefoxitin, and clarithromycin.

Continue to: One month postoperatively...

One month postoperatively, her constitutional symptoms, including fevers and night sweats, abated and inflammatory markers (ESR and CRP) had normalized. There were no clinical signs of infection. Amikacin was discontinued due to a 10-dB change on audiologic screening (4-6 kHz range), and tigecycline was substituted. Ultimately, she underwent 15 weeks of antimycobacterial therapy, 10 of which were after the explantation.

Eight weeks after cessation of her antibiotics, she underwent open biopsy. Multiple operative tissue samples showed negative results in pathology and culture tests.

Replantation was performed 14 weeks after stopping antimicrobials and 24 weeks after her explantation. The bone appeared healthy without evidence of osteomyelitis. A constrained reconstruction was secured with tobramycin-impregnated cement. One small island of necrotizing granuloma was observed within the bony cortex on histologic review; the granulomata appeared active with scattered neutrophils along with histiocytes and lymphocytes. AFB stains were negative. Intraoperative cultures, including mycobacterial cultures, were negative.

Based on the histologic evidence that infection may have persisted, and given the high stakes, antimicrobial treatment was reinitiated. Amikacin was again stopped after 3 weeks due to the development of tinnitus; tigecycline was substituted to complete the fourth and final week, at which point all antibiotics were discontinued. The patient was followed up uneventfully for 4 years (Figures 4A-4D and 5A-5C) with normal ESR and CRP. She continues to be ambulatory without assistive devices and walks an average of 30 miles per week without pain or constitutional symptoms.

Continue to: DISCUSSION...

DISCUSSION

Diagnosis of acute infection after TKA remains challenging, as some degree of pain, swelling, and even postoperative fevers may be common in noninfected TKA patients. Synovial white blood cell count and differential as well as alpha-defensin levels have been cited as predictive factors of infection.^16,17 Deep tissue and synovial fluid cultures offer the advantage of both identification and antimicrobial sensitivity testing of the offending organism. In this case, culture of the knee joint fluid at the time of TKA led to the unexpected finding of M. abscessus infection.

Preventable outbreaks due to M. abscessus have been reported and attributed to contaminated multiuse instruments, inadequate sterilization of tap water, multiuse vials, and improper skin preparation.^11-13 Rarely, M. abscessus has been reported as the cause of PJI. When an unusual organism is encountered after native joint instrumentation, an investigation should be undertaken to identify the source of contamination, with the assistance of infection control practitioners and/or the US Food and Drug Administration reporting. Reporting and investigation was undertaken in this case, though no suspect source could be identified.

Although there were no signs of infection prior to the TKA, there is an ongoing debate as to whether intra-articular corticosteroid injections increase the risk of PJIs, and if so, what the optimal amount of time to wait between procedures is. Although several earlier studies have been underpowered to answer these questions,¹⁸ this patient underwent TKA 1 month following the corticosteroid injection. Recent meta-analyses have shown no definitive evidence to indicate that this increased her risk of PJI.^19,20

Continue to: Treatments for mycobacterial infections...

Treatments for mycobacterial infections have been described with variable efficacy,^21,22 and only 2 cases of successfully treated PJIs have been reported after infection with M. abscessus. Both these cases were described in total hip arthroplasties,^23,24 and to the authors’ knowledge, this report represents the first described successfully treated case after TKA. Staged reconstruction remains a standard treatment for invasive organisms chronically infecting prosthetic joint implants, with reimplantation pending joint sterility and improvement in inflammatory markers.³ Previous successful reports of treating M. abscessus describe either resection arthroplasty²¹ or staged reconstruction.^23,24 The authors reported variable multidrug antimicrobial regimens, as summarized in Table 2, as guidelines for the treatment of mycobacterial PJI are currently not available.

CONCLUSION

This case report represents an episode of iatrogenic septic arthritis caused by Mycobacteria of the native knee after previous history of instrumentation, corticosteroid, and hyaluronic acid injections, with an overall indolent clinical course until subsequent arthroplasty. There were several important lessons learned, which are as follows: 1) Multidrug combination with antimicrobial therapy combined with aggressive surgical débridement and staged reimplantation permitted successful eradication of TKA PJI caused by M. abscessus in this patient. 2) Initial medical management alone was not successful and cannot be recommended for the treatment of M. abscessus in the setting of PJI. 3) Delaying the surgical débridement and the reconstructive course for a trial of medical management contributed to the ultimate requirement of a tibial tubercle osteotomy for an ankylosed knee at replantation. In this case, we initially had a low index of suspicion for deep infection, contributing to delayed surgical débridement. Ideally, a high degree of clinical suspicion should be maintained for joint infection in the presence of positive culture isolates of M. abscessus, as it may have a delayed clinical presentation of the typical features of PJI (fevers, swelling, erythema, etc). In such cases, the authors recommend consideration of early surgical débridement. 4) Medical management of TKA PJI is not without risks. Careful monitoring of patient side effects during antimicrobial administration remains paramount, as this patient did sustain a degree of hearing loss associated with prolonged medical therapy. 5) In complicated PJIs involving rare and intrinsically resistant organisms, a collaborative multidisciplinary approach, including specialists in orthopedic surgery, infectious disease, microbiology, pharmacy, and pathology, may be the preferred path to clinical cure.

ABSTRACT

A 61-year-old woman with a periprosthetic knee joint infection caused by Mycobacterium abscessus was successfully treated with surgical débridement, multidrug antimicrobial therapy, and staged reimplantation. To the authors’ knowledge, this represents the first report of successfully treating this organism after knee arthroplasty.

M. abscessus knee infections are rare, and there are no specific guidelines to inform treatment or successful treatment regimens for periprosthetic knee infections. Medical management alone was not successful in this case and hence cannot be recommended. Using a collaborative multidisciplinary approach, including surgical débridement, staged reimplantation, and multidrug antimicrobials, successful eradication of the periprosthetic joint infection caused by M. abscessus was achieved.  

Continue to: Total knee arthroplasty...

Total knee arthroplasty (TKA) procedures are projected to increase by more than 6-fold by 2030, with concurrent increases in revision TKA for infection projected.¹ Infection after TKA remains one of the most serious complications of the procedure, occurring in <2% of primary TKAs.² The majority of prosthetic joint infections (PJIs) are caused by staphylococci and streptococci.³ Although infection and treatment of PJIs by mycobacterial species have been described, there are presently no established treatment guidelines for mycobacterial PJIs.^4,5

Given the scarcity of clinical experience in dealing with these organisms, and the predicted increasing incidence of revision knee arthroplasty due to infection, we describe an unusual case of a PJI caused by Mycobacterium abscessus (M. abscessus), which was successfully treated using a combination of antimicrobial therapy and staged reconstruction. The patient provided written informed consent for print and electronic publication of this case report.

BACKGROUND

Mycobacteria are common environmental organisms that can survive harsh conditions, including low pH and extreme temperatures. They form biofilms and may be difficult to eradicate in cases of infection.⁶M. abscessus has proven to be difficult to eradicate due to limited antimicrobial susceptibility, lack of bactericidal options, and the variable presence of the erm gene, which yields inducible resistance to macrolides.⁷ Post-procedural outbreaks due to mycobacteria have been reported, often attributed to contaminated multiuse instruments, inadequate sterilization of tap water, multiuse vials, or improper skin preparation.^6,8-13

CASE REPORT 

A 61-year-old woman was referred with a 3-year history of progressive left knee pain and swelling. Before 8 months, she had undergone knee arthroscopy and had been treated with multiple steroid and hyaluronic acid injections, as well as ultrasound-guided aspiration of a Baker’s cyst (Figures 1A, 1B).

She elected to proceed with TKA 1 month after her last steroid injection. There was no preoperative concern for native joint infection. At the time of arthroplasty, clear joint fluid was encountered, and a deep tissue culture was taken (Figures 2A-2C).

Routine screening cultures for acid-fast bacilli (AFB) returned positive 9 days after the index arthroplasty, with subsequent identification of a nontuberculous mycobacterium (NTM), M. abscessus, subspecies massiliense. Sensitivity tests revealed susceptibility to amikacin, cefoxitin, and tigecycline (Table 1). The isolate was found to have inducible macrolide resistance by erm gene testing.

Table 1. Initial Mycobacterium abscessus massiliense Susceptibilities

^aAt 3 days; erm gene detected at 7 days.

Given no prior surgical suspicion for infection and the uncertain significance of the culture result, treatment options were debated. Medical management was selected based on the presumption that if infection was present, it was a native joint infection in which surgical débridement had already been undertaken at the time of primary arthroplasty. Similar reports for the treatment of M. tuberculosis infection in the knee have been reported with some success.^14,15 Short-interval reassessment was planned. Antimicrobial therapy was selected based on susceptibility data and clinical experience and consisted of intravenous (IV) cefoxitin, oral clarithromycin, and thrice-weekly intravenous amikacin. Over the ensuing weeks, she developed fevers, knee swelling, and persistent elevation of erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP). With known potential of this organism for biofilm formation in other areas of the body and positive repeat cultures of the knee joint fluid, confirming the offending organism, a deep and resistant infection of the implant could not be excluded. Therefore, in an attempt to give the patient the best opportunity for clinical cure, the patient subsequently underwent a 2-stage antibiotic spacer explantation and exchange (Figures 3A, 3B). Moderate caseous material was present throughout the knee joint and the subcutaneous tissues. All bone was débrided, and complete synovectomy was undertaken, along with the removal of all implants. The antibiotic concentrations within the spacer were selected by guidance from the Infectious Disease and Pharmacy based on minimal inhibitory concentrations, with 3 packages of cement (40 g each) utilized and a total of 10 g of amikacin and 24 g of cefoxitin contained within the spacer. The patient continued systemic administration of amikacin, cefoxitin, and clarithromycin.

Continue to: One month postoperatively...

One month postoperatively, her constitutional symptoms, including fevers and night sweats, abated and inflammatory markers (ESR and CRP) had normalized. There were no clinical signs of infection. Amikacin was discontinued due to a 10-dB change on audiologic screening (4-6 kHz range), and tigecycline was substituted. Ultimately, she underwent 15 weeks of antimycobacterial therapy, 10 of which were after the explantation.

Eight weeks after cessation of her antibiotics, she underwent open biopsy. Multiple operative tissue samples showed negative results in pathology and culture tests.

Replantation was performed 14 weeks after stopping antimicrobials and 24 weeks after her explantation. The bone appeared healthy without evidence of osteomyelitis. A constrained reconstruction was secured with tobramycin-impregnated cement. One small island of necrotizing granuloma was observed within the bony cortex on histologic review; the granulomata appeared active with scattered neutrophils along with histiocytes and lymphocytes. AFB stains were negative. Intraoperative cultures, including mycobacterial cultures, were negative.

Based on the histologic evidence that infection may have persisted, and given the high stakes, antimicrobial treatment was reinitiated. Amikacin was again stopped after 3 weeks due to the development of tinnitus; tigecycline was substituted to complete the fourth and final week, at which point all antibiotics were discontinued. The patient was followed up uneventfully for 4 years (Figures 4A-4D and 5A-5C) with normal ESR and CRP. She continues to be ambulatory without assistive devices and walks an average of 30 miles per week without pain or constitutional symptoms.

Continue to: DISCUSSION...

DISCUSSION

Diagnosis of acute infection after TKA remains challenging, as some degree of pain, swelling, and even postoperative fevers may be common in noninfected TKA patients. Synovial white blood cell count and differential as well as alpha-defensin levels have been cited as predictive factors of infection.^16,17 Deep tissue and synovial fluid cultures offer the advantage of both identification and antimicrobial sensitivity testing of the offending organism. In this case, culture of the knee joint fluid at the time of TKA led to the unexpected finding of M. abscessus infection.

Preventable outbreaks due to M. abscessus have been reported and attributed to contaminated multiuse instruments, inadequate sterilization of tap water, multiuse vials, and improper skin preparation.^11-13 Rarely, M. abscessus has been reported as the cause of PJI. When an unusual organism is encountered after native joint instrumentation, an investigation should be undertaken to identify the source of contamination, with the assistance of infection control practitioners and/or the US Food and Drug Administration reporting. Reporting and investigation was undertaken in this case, though no suspect source could be identified.

Although there were no signs of infection prior to the TKA, there is an ongoing debate as to whether intra-articular corticosteroid injections increase the risk of PJIs, and if so, what the optimal amount of time to wait between procedures is. Although several earlier studies have been underpowered to answer these questions,¹⁸ this patient underwent TKA 1 month following the corticosteroid injection. Recent meta-analyses have shown no definitive evidence to indicate that this increased her risk of PJI.^19,20

Continue to: Treatments for mycobacterial infections...

Treatments for mycobacterial infections have been described with variable efficacy,^21,22 and only 2 cases of successfully treated PJIs have been reported after infection with M. abscessus. Both these cases were described in total hip arthroplasties,^23,24 and to the authors’ knowledge, this report represents the first described successfully treated case after TKA. Staged reconstruction remains a standard treatment for invasive organisms chronically infecting prosthetic joint implants, with reimplantation pending joint sterility and improvement in inflammatory markers.³ Previous successful reports of treating M. abscessus describe either resection arthroplasty²¹ or staged reconstruction.^23,24 The authors reported variable multidrug antimicrobial regimens, as summarized in Table 2, as guidelines for the treatment of mycobacterial PJI are currently not available.

CONCLUSION

This case report represents an episode of iatrogenic septic arthritis caused by Mycobacteria of the native knee after previous history of instrumentation, corticosteroid, and hyaluronic acid injections, with an overall indolent clinical course until subsequent arthroplasty. There were several important lessons learned, which are as follows: 1) Multidrug combination with antimicrobial therapy combined with aggressive surgical débridement and staged reimplantation permitted successful eradication of TKA PJI caused by M. abscessus in this patient. 2) Initial medical management alone was not successful and cannot be recommended for the treatment of M. abscessus in the setting of PJI. 3) Delaying the surgical débridement and the reconstructive course for a trial of medical management contributed to the ultimate requirement of a tibial tubercle osteotomy for an ankylosed knee at replantation. In this case, we initially had a low index of suspicion for deep infection, contributing to delayed surgical débridement. Ideally, a high degree of clinical suspicion should be maintained for joint infection in the presence of positive culture isolates of M. abscessus, as it may have a delayed clinical presentation of the typical features of PJI (fevers, swelling, erythema, etc). In such cases, the authors recommend consideration of early surgical débridement. 4) Medical management of TKA PJI is not without risks. Careful monitoring of patient side effects during antimicrobial administration remains paramount, as this patient did sustain a degree of hearing loss associated with prolonged medical therapy. 5) In complicated PJIs involving rare and intrinsically resistant organisms, a collaborative multidisciplinary approach, including specialists in orthopedic surgery, infectious disease, microbiology, pharmacy, and pathology, may be the preferred path to clinical cure.

Reverse total shoulder arthroplasty (RTSA) is a common treatment for rotator cuff tear arthropathy. We performed a systematic review of all the RTSA literature to answer if we are treating the same patients with RTSA, across the world.

A systematic review was registered with PROSPERO, the international prospective register of systematic reviews, and performed with Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines using 3 publicly available free databases. Therapeutic clinical outcome investigations reporting RTSA outcomes with levels of evidence I to IV were eligible for inclusion. All study, subject, and surgical technique demographics were analyzed and compared between continents. Statistical comparisons were conducted using linear regression, analysis of variance (ANOVA), Fisher's exact test, and Pearson's chi-square test.

There were 103 studies included in the analysis (8973 patients; 62% female; mean age, 70.9 ± 6.7 years; mean length of follow-up, 34.3 ± 19.3 months) that had a low Modified Coleman Methodology Score (MCMS) (mean, 36.9 ± 8.7: poor). Most patients (60.8%) underwent RTSA for a diagnosis of rotator cuff arthropathy, whereas 1% underwent RTSA for fracture; indications varied by continent. There were no consistent reports of preopeartive or postoperative scores from studies in any region. Studies from North America reported significantly higher postoperative external rotation (34.1° ± 13.3° vs 19.3° ± 8.9°) (P < .001) and a greater change in flexion (69.0° ± 24.5° vs 56.3° ± 11.3°) (P = .004) compared with studies from Europe. North America had the greatest total number of publications followed by Europe. The total yearly number of publications increased each year (P < .001), whereas the MCMS decreased each year (P = .037).

The quantity, but not the quality of RTSA studies is increasing. Indications for RTSA varied by continent, although most patients underwent RTSA for rotator cuff arthropathy. The majority of patients undergoing RTSA are female over the age of 60 years for a diagnosis of rotator cuff arthropathy with pseudoparalysis.

Continue to: Reverse total shoulder arthroplasty...

Reverse total shoulder arthroplasty (RTSA) is a common procedure with indications including rotator cuff tear arthropathy, proximal humerus fractures, and others.^1,2 Studies have shown excellent, reliable, short- and mid-term outcomes in patients treated with RTSA for various indications.^3-5 Al-Hadithy and colleagues⁶ reviewed 41 patients who underwent RTSA for pseudoparalysis secondary to rotator cuff tear arthropathy and, at a mean follow-up of 5 years, found significant improvements in range of motion (ROM) as well as age-adjusted Constant and Oxford Outcome scores. Similarly, Ross and colleagues⁷ evaluated outcomes of RTSA in 28 patients in whom RTSA was performed for 3- or 4-part proximal humerus fractures, and found both good clinical and radiographic outcomes with no revision surgeries at a mean follow-up of 54.9 months. RTSA is performed across the world, with specific implant designs, specifically humeral head inclination, but is more common in some areas when compared with others.^3,8,9

The number of RTSAs performed has steadily increased over the past 20 years, with recent estimates of approximately 20,000 RTSAs performed in the United States in 2011.^10,11 However, there is little information about the similarities and differences between those patients undergoing RTSA in various parts of the world regarding surgical indications, patient demographics, and outcomes. The purpose of this study is to perform a systematic review and meta-analysis of the RTSA body of literature to both identify and compare characteristics of studies published (level of evidence, whether a conflict of interest existed), patients analyzed (age, gender), and surgical indications performed across both continents and countries. Essentially, the study aims to answer the question, "Across the world, are we treating the same patients?" The authors hypothesized that there would be no significant differences in RTSA publications, subjects, and indications based on both the continent and country of publication.

METHODS

A systematic review was conducted according to PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines using a PRISMA checklist.¹² A systematic review registration was performed using PROSPERO, the international prospective register of systematic reviews (registration number CRD42014010578).¹³Two reviewers independently conducted the search on March 25, 2014, using the following databases: Medline, Cochrane Central Register of Controlled Trials, SportDiscus, and CINAHL. The electronic search citation algorithm utilized was: (((((reverse[Title/Abstract]) AND shoulder[Title/Abstract]) AND arthroplasty[Title/Abstract]) NOT arthroscopic[Title/Abstract]) NOT cadaver[Title/Abstract]) NOT biomechanical[Title/Abstract]. English language Level I to IV evidence (2011 update by the Oxford Centre for Evidence-Based Medicine¹⁴) clinical studies were eligible. Medical conference abstracts were ineligible for inclusion. All references within included studies were cross-referenced for inclusion if missed by the initial search with any additionally located studies screened for inclusion. Duplicate subject publications within separate unique studies were not reported twice, but rather the study with longer duration follow-up or, if follow-up was equal, the study with the greater number of patients was included. Level V evidence reviews, letters to the editor, basic science, biomechanical and cadaver studies, total shoulder arthroplasty (TSA) papers, arthroscopic shoulder surgery papers, imaging, surgical techniques, and classification studies were excluded.

A total of 255 studies were identified, and, after implementation of the exclusion criteria, 103 studies were included in the final analysis (Figure 1). Subjects of interest in this systematic review underwent RTSA for one of many indications including rotator cuff tear arthropathy, osteoarthritis, rheumatoid arthritis, posttraumatic arthritis, instability, revision from a previous RTSA for instability, infection, acute proximal humerus fracture, revision from a prior proximal humerus fracture, revision from a prior hemiarthroplasty, revision from a prior TSA, osteonecrosis, pseudoparalysis, tumor, and a locked shoulder dislocation. There was no minimum follow-up or rehabilitation requirement. Study and subject demographic parameters analyzed included year of publication, years of subject enrollment, presence of study financial conflict of interest, number of subjects and shoulders, gender, age, body mass index, diagnoses treated, and surgical positioning. Clinical outcome scores sought were the DASH (Disability of the Arm, Shoulder, and Hand), SPADI (Shoulder Pain And Disability Index), Absolute Constant, ASES (American Shoulder and Elbow Score), KSS (Korean Shoulder Score), SST-12 (Simple Shoulder Test), SF-12 (12-item Short Form), SF-36 (36-item Short Form), SSV (Subjective Shoulder Value), EQ-5D (EuroQol-5 Dimension), SANE (Single Assessment Numeric Evaluation), Rowe Score for Instability, Oxford Instability Score, UCLA (University of California, Los Angeles) activity score, Penn Shoulder Score, and VAS (visual analog scale). In addition, ROM (forward elevation, abduction, external rotation, internal rotation) was analyzed. Radiographs and magnetic resonance imaging data were extracted when available. The methodological quality of the study was evaluated using the MCMS (Modified Coleman Methodology Score).¹⁵

STATISTICAL ANALYSIS

First, the number of publications per year, level of evidence, and Modified Coleman Methodology Score were tested for association with the calendar year using linear regression. Second, demographic data were tested for association with the continent using Pearson’s chi-square test or ANOVA. Third, indications were tested for association with the continent using Fisher’s exact test. Finally, clinical outcome scores and ROM were tested for association with the continent using ANOVA. Statistical significance was extracted from studies when available. Statistical significance was defined as P < .05.

Continue to: RESULTS...

 

 

RESULTS

There were 103 studies included in the analysis (Figure 1). A total of 8973 patients were included, 62% of whom were female with a mean age of 70.9 ± 6.7 years (Table 1). The average follow-up was 34.3 ± 19.3 months. North America had the overall greatest total number of publications on RTSA, followed by Europe (Figure 2). The total yearly number of publications increased by a mean of 1.95 publications each year (P < .001). There was no association between the mean level of evidence with the year of publication (P = .296) (Figure 3). Overall, the rating of studies was poor for the MCMS (mean 36.9 ± 8.7). The MCMS decreased each year by a mean of 0.76 points (P = .037) (Figure 4).

Table 1. Demographic Data by Continent

	North America	Europe	Asia	Australia	Total	P-value
Number of studies	52	43	4	4	103	-
Number of subjects	6158	2609	51	155	8973	-
Level of evidence						0.693
II	5 (10%)	3 (7%)	0 (0%)	0 (0%)	8 (8%)
III	10 (19%)	4 (9%)	0 (0%)	1 (25%)	15 (15%)
IV	37 (71%)	36 (84%)	4 (100%)	3 (75%)	80 (78%)
Mean MCMS	34.6 ± 8.4	40.2 ± 8.0	32.5 12.4	34.5 ± 6.6	36.9 ± 8.7	0.010
Institutional collaboration						1.000
Multi-center	7 (14%)	6 (14%)	0 (0%)	0 (0%)	13 (13%)
Single-center	45 (86%)	37 (86%)	4 (100%)	4 (100%)	90 (87%)
Financial conflict of interest						0.005
Present	28 (54%)	15 (35%)	0 (0%)	0 (0%)	43 (42%)
Not present	19 (37%)	16 (37%)	4 (100%)	4 (100%)	43 (42%)
Not reported	5 (10%)	12 (28%)	0 (0%)	0 (0%)	17 (17%)
Sex						N/A
Male	2157 (38%)	1026 (39%)	13 (25%)	61 (39%)	3257 (38%)
Female	3520 (62%)	1622 (61%)	38 (75%)	94 (61%)	5274 (62%)
Mean age (years)	71.3 ± 5.6	70.1 ± 7.9	68.1 ± 5.3	76.9 ± 3.0	70.9 ± 6.7	0.191
Minimum age (mean across studies)	56.9 ± 12.8	52.8 ± 15.7	62.8 ± 6.2	68.0 ± 12.1	55.6 ± 14.3	0.160
Maximum age (mean across studies)	82.1 ± 8.6	83.0 ± 5.5	73.0 ± 9.4	85.0 ± 7.9	82.2 ± 7.6	0.079
Mean length of follow-up (months)	26.5 ± 13.7	43.1 ± 21.7	29.4 ± 7.9	34.2 ± 16.6	34.3 ± 19.3	<0.001
Prosthesis type						N/A
Cemented	988 (89%)	969 (72%)	0 (0%)	8 (16%)	1965 (78%)
Press fit	120 (11%)	379 (28%)	0 (0%)	41 (84%)	540 (22%)

Abbreviations: MCMS, Modified Coleman Methodology Score; N/A, not available.

 

In studies that reported press-fit vs cemented prostheses, the highest percentage of press-fit prostheses compared with cemented prostheses was seen in Australia (84% press-fit), whereas the highest percentage of cemented prostheses was seen in North America (89% cemented). A higher percentage of studies from North America had a financial conflict of interest (COI) than did those from other countries (54% had a COI).

Continue to: Rotator cuff tear arthropathy...

 

 

Rotator cuff tear arthropathy was the most common indication for RTSA overall in 5459 patients, followed by pseudoparalysis in 1352 patients (Tables 2 and 3). While studies in North America reported rotator cuff tear arthropathy as the indication for RTSA in 4418 (75.8%) patients, and pseudoparalysis as the next most common indication in 535 (9.2%) patients, studies from Europe reported rotator cuff tear arthropathy as the indication in 895 (33.5%) patients, and pseudoparalysis as the indication in 795 (29.7%) patients. Studies from Asia also had a relatively even split between rotator cuff tear arthropathy and pseudoparalysis (45.3% vs 37.8%), whereas those from Australia were mostly rotator cuff tear arthropathy (77.7%).

Table 2. Number (Percent) of Studies With Each Indication by Continent

	North America	Europe	Asia	Australia	Total	P-value
Rotator cuff arthropathy	29 (56%)	19 (44%)	3 (75%)	3 (75%)	54 (52%)	0.390
Osteoarthritis	4 (8%)	10 (23%)	1 (25%)	1 (25%)	16 (16%)	0.072
Rheumatoid arthritis	9 (17%)	10 (23%)	0 (0%)	2 (50%)	21 (20%)	0.278
Post-traumatic arthritis	3 (6%)	5 (12%)	0 (0%)	1 (25%)	9 (9%)	0.358
Instability	6 (12%)	3 (7%)	0 (0%)	1 (25%)	10 (10%)	0.450
Revision of previous RTSA for instability	5 (10%)	1 (2%)	0 (0%)	1 (25%)	7 (7%)	0.192
Infection	4 (8%)	1 (2%)	1 (25%)	0 (0%)	6 (6%)	0.207
Unclassified acute proximal humerus fracture	9 (17%)	5 (12%)	1 (25%)	1 (25%)	16  (16%)	0.443
Acute 2-part proximal humerus fracture	0 (0%)	0 (0%)	0 (0%)	0 (0%)	0 (0%)	N/A
Acute 3-part proximal humerus fracture	2 (4%)	0 (0%)	0 (0%)	0 (0%)	2 (2%)	0.574
Acute 4-part proximal humerus fracture	5 (10%)	0 (0%)	0 (0%)	0 (0%)	5 (5%)	0.183
Acute 3- or 4-part proximal humerus fracture	6 (12%)	2 (5%)	0 (0%)	0 (0%)	8 (8%)	0.635
Revised from previous nonop proximal humerus fracture	7 (13%)	3 (7%)	0 (0%)	0 (0%)	10 (10%)	0.787
Revised from ORIF	1 (2%)	1 (2%)	0 (0%)	0 (0%)	2 (2%)	1.000
Revised from CRPP	0 (0%)	1 (2%)	0 (0%)	0 (0%)	1 (1%)	0.495
Revised from hemi	8 (15%)	4 (9%)	0 (0%)	1 (25%)	13 (13%)	0.528
Revised from TSA	15 (29%)	11 (26%)	0 (0%)	2 (50%)	28 (27%)	0.492
Osteonecrosis	4 (8%)	2 (5%)	1 (25%)	0 (0%)	7 (7%)	0.401
Pseudoparalysis irreparable tear without arthritis	20 (38%)	18 (42%)	2 (50%)	1 (25%)	41 (40%)	0.919
Bone tumors	0 (0%)	4 (9.3%)	0 (0%)	0 (0%)	4 (4%)	0.120
Locked shoulder dislocation	0 (0%)	0 (0%)	1 (25%)	0 (0%)	1 (1%)	0.078

Abbreviations: CRPP, closed reduction and percutaneous pinning; ORIF, open reduction internal fixation; RTSA, reverse total shoulder arthroplasty; TSA, total shoulder arthroplasty.

 

Table 3. Number of Patients With Each Indication as Reported by Individual Studies by Continent

	North America	Europe	Asia	Australia	Total
Rotator cuff arthropathy	4418	895	24	122	5459
Osteoarthritis	90	251	1	14	356
Rheumatoid arthritis	59	87	0	2	148
Post-traumatic arthritis	62	136	0	1	199
Instability	23	15	0	1	39
Revision of previous RTSA for instability	29	2	0	1	32
Infection	28	11	2	0	41
Unclassified acute proximal humerus fracture	42	30	4	8	84
Acute 3-part proximal humerus fracture	60	0	0	0	6
Acute 4-part proximal humerus fracture	42	0	0	0	42
Acute 3- or 4-part proximal humerus fracture	92	46	0	0	138
Revised from previous nonop proximal humerus fracture	43	53	0	0	96
Revised from ORIF	3	9	0	0	12
Revised from CRPP	0	3	0	0	3
Revised from hemi	105	51	0	1	157
Revised from TSA	192	246	0	5	443
Osteonecrosis	9	6	1	0	16
Pseudoparalysis irreparable tear without arthritis	535	795	20	2	1352
Bone tumors	0	38	0	0	38
Locked shoulder dislocation	0	0	1	0	1

Abbreviations: CRPP, closed reduction and percutaneous pinning; ORIF, open reduction internal fixation; RTSA, reverse total shoulder arthroplasty; TSA, total shoulder arthroplasty.

 

The ASES, SST-12, and VAS scores were the most frequently reported outcome scores in studies from North America, whereas the Absolute Constant score was the most common score reported in studies from Europe (Table 4). Studies from North America reported significantly higher postoperative external rotation (34.1° ± 13.3° vs 19.3° ± 8.9°) (P < .001) and a greater change in flexion (69.0° ± 24.5° vs 56.3° +/- 11.3°) (P = .004) compared with studies from Europe (Table 5).

Table 4. Outcomes by Continent

Metric (number of studies)	North America	Europe	Asia	Australia	P-value
DASH	1	2	0	0
Preoperative	54.0	62.0 ± 8.5	-	-	0.582
Postoperative	24.0	32.0 ± 2.8	-	-	0.260
Change	-30.0	-30.0 ± 11.3	-	-	1.000
SPADI	2	0	0	0
Preoperative	80.0 ± 4.2	-	-	-	N/A
Postoperative	34.8 ± 1.1	-	-	-	N/A
Change	-45.3 ± 3.2	-	-	-	N/A
Absolute constant	2	27	0	1
Preopeartive	33.0 ± 0.0	28.2 ± 7.1	-	20.0	0.329
Postoperative	54.5 ± 7.8	62.9 ± 9.0	-	65.0	0.432
Change	+21.5 ± 7.8	+34.7 ± 8.0	-	+45.0	0.044
ASES	13	0	2	0
Preoperative	33.2 ± 5.4	-	32.5 ± 3.5	-	0.867
Postoperative	73.9 ± 6.8	-	75.7 ± 10.8	-	0.752
Change	+40.7 ± 6.5	-	+43.2 ± 14.4	-	0.670
UCLA	3	2	1	0
Preoperative	10.1 ± 3.4	11.2 ± 5.7	12.0	-	0.925
Postoperative	24.5 ± 3.1	24.3 ± 3.7	24.0	-	0.991
Change	+14.4 ± 1.6	+13.1 ± 2.0	+12.0	-	0.524
KSS	0	0	2	0
Preopeartive	-	-	38.2 ± 1.1	-	N/A
Postoperative	-	-	72.3 ± 6.0	-	N/A
Change	-	-	+34.1 ± 7.1	-	N/A
SST-12	12	1	0	0
Preoperative	1.9 ± 0.8	1.2	-	-	N/A
Postoperative	7.1 ± 1.5	5.6	-	-	N/A
Change	+5.3 ± 1.2	+4.4	-	-	N/A
SF-12	1	0	0	0
Preoperative	34.5	-	-	-	N/A
Postoperative	38.5	-	-	-	N/A
Change	+4.0	-	-	-	N/A
SSV	0	5	0	0
Preopeartive	-	22.0 ± 7.4	-	-	N/A
Postoperative	-	63.4 ± 7.9	-	-	N/A
Change	-	+41.4 ± 2.1	-	-	N/A
EQ-5D	0	2	0	0
Preoperative	-	0.5 ± 0.2	-	-	N/A
Postoperative	-	0.8 ± 0.1	-	-	N/A
Change	-	+0.3 ± 0.1	-	-	N/A
OOS	1	0	0	0
Preoperative	24.7	-	-	-	N/A
Postoperative	14.9	-	-	-	N/A
Change	-9.9	-	-	-	N/A
Rowe	0	1	0	0
Preoperative	-	50.2	-	-	N/A
Postoperative	-	82.1	-	-	N/A
Change	-	31.9	-	-	N/A
Oxford	0	2	0	0
Preoperative	-	119.9 ± 138.8	-	-	N/A
Postoperative	-	39.9 ± 3.3	-	-	N/A
Change	-	-80.6 ± 142.2	-	-	N/A
Penn	1	0	0	0
Preoperative	24.9	-	-	-	N/A
Postoperative	66.4	-	-	-	N/A
Change	+41.5	-	-	-	N/A
VAS	10	1	1	1
Preoperative	6.6 ± 0.8	7.0	8.4	7.0	N/A
Postoperative	2.0 ± 0.7	1.0	0.8	0.8	N/A
Change	-4.6 ± 0.8	-6.0	-7.6	-6.2	N/A
SF-36 physical	2	0	0	0
Preoperative	32.7 ± 1.2	-	-	-	N/A
Postoperative	39.6 ± 4.0	-	-	-	N/A
Change	+7.0 ± 2.8	-	-	-	N/A
SF-36 mental	2	0	0	0
Preoperative	43.6 ± 2.8	-	-	-	N/A
Postoperative	48.1 ± 1.0	-	-	-	N/A
Change	+4.5 ± 1.8	-	-	-	N/A

Abbreviations: ASES, American Shoulder and Elbow Surgeon score; DASH, Disability of the Arm, Shoulder, and Hand; EQ-5D, EuroQol-5 Dimension; KSS, Korean Shoulder Scoring system; N/A, not available; OOS, Orthopaedic Outcome Score; SF, short form; SPADI, Shoulder Pain and Disability Index; SST, Simple Shoulder Test; SSV, Subjective Shoulder Value; UCLA, University of California, Los Angeles; VAS, visual analog scale.

 

Table 5. Shoulder Range of Motion, by Continent

Metric (number of studies)	North America	Europe	Asia	Australia	P-value
Flexion	18	22	1	1
Preoperative	57.6 ± 17.9	65.5 ± 17.2	91.0	30.0	0.060
Postoperative	126.6 ± 14.4	121.8 ± 19.0	133.0	150.0	0.360
Change	+69.0 ± 24.5	+56.3 ± 11.3	+42.0	120.0	0.004
Abduction	11	12	1	0
Preoperative	53.7 ± 25.0	52.0 ± 19.0	88.0	-	0.311
Postoperative	109.3 ± 15.1	105.4 ± 19.8	131.0	-	0.386
Change	55.5 ± 25.5	53.3 ± 8.3	43.0	-	0.804
External rotation	17	19	0	0
Preoperative	19.4 ± 9.9	11.2 ± 6.1	-	-	0.005
Postoperative	34.1 ± 13.3	19.3 ± 8.9	-	-	<0.001
Change	+14.7 ± 13.2	+8.1 ± 8.5	-	-	0.079

Continue to: DISCUSSION...

 

 

DISCUSSION

RTSA is a common procedure performed in many different areas of the world for a variety of indications. The study hypotheses were partially confirmed, as there were no significant differences seen in the characteristics of the studies published and patients analyzed; although, the majority of studies from North America reported rotator cuff tear arthropathy as the primary indication for RTSA, whereas studies from Europe were split between rotator cuff tear arthropathy and pseudoparalysis as the primary indication. Hence, based on the current literature the study proved that we are treating the same patients. Despite this finding, we may be treating them for different reasons with an RTSA.

RTSA has become a standard procedure in the United States, with >20,000 RTSAs performed in 2011.¹⁰ This number will continue to increase as it has over the past 20 years given the aging population in the United States, as well as the expanding indications for RTSA.¹¹ Indications of RTSA have become broad, although the main indication remains as rotator cuff tear arthropathy (>60% of all patients included in this study), and pseudoparalysis (>15% of all patients included in this study). Results for RTSA for rotator cuff tear arthropathy and pseudoparalysis have been encouraging.^16,17 Frankle and colleagues¹⁶ evaluated 60 patients who underwent RTSA for rotator cuff tear arthropathy at a minimum of 2 years follow-up (average, 33 months). The authors found significant improvements in all measured clinical outcome variables (P < .0001) (ASES, mean function score, mean pain score, and VAS) as well as ROM, specifically forward flexion increased from 55° to 105.1°, and abduction increased from 41.4° to 101.8°. Similarly, Werner and colleagues¹⁷ evaluated 58 consecutive patients who underwent RTSA for pseudoparalysis secondary to irreparable rotator cuff dysfunction at a mean follow-up of 38 months. Overall, significant improvements (P < .0001) were seen in the SSV score, relative Constant score, and Constant score for pain, active anterior elevation (42° to 100° following RTSA), and active abduction (43° to 90° following RTSA).

It is essential to understand the similarities and differences between patients undergoing RTSA in different parts of the world so the literature from various countries can be compared between regions, and conclusions extrapolated to the correct patients. For example, an interesting finding in this study is that the majority of patients in North America have their prosthesis cemented whereas the majority of patients in Australia have their prosthesis press-fit. While the patients each continent is treating are not significantly different (mostly older women), the difference in surgical technique could have implications in long- or short-term functional outcomes. Prior studies have shown no difference in axial micromotion between cemented and press-fit humeral components, but the clinical implications surrounding this are not well defined.¹⁸ Small series comparing cementless to cemented humeral prosthesis in RTSA have found no significant differences in clinical outcomes or postoperative ROM, but larger series are necessary to validate these outcomes.¹⁹ However, studies have shown lower rates of postoperative infections in patients who receive antibiotic-loaded cement compared with those who receive plain bone cement following RTSA.²⁰

Similarly, as the vast majority of patients in North America had an RTSA for rotator cuff arthropathy (75.8%) whereas those from Europe had RTSA almost equally for rotator cuff arthropathy (33.5%) and pseudoparalysis (29.7%), one must ensure similar patient populations before attempting to extrapolate results of a study from a different country to patients in other areas. Fortunately, the clinical results following RTSA for either indication have been good.^6,21,22

One final point to consider is the cost effectiveness of the implant. Recent evidence has shown that RTSA is associated with a higher risk for in-hospital death, multiple perioperative complications, prolonged hospital stay, and increased hospital cost when compared with TSA.²³ This data may be biased as the patient selection for RTSA varies from that of TSA, but it is a point that must be considered. Other studies have shown that an RTSA is a cost-effective treatment option for treating patients with rotator cuff tear arthropathy, and is a more cost-effective option in treating rotator cuff tear arthropathy than hemiarthroplasty.^24,25 Similarly, RTSA offers a more cost-effective treatment option with better outcomes for patients with acute proximal humerus fractures when compared with open reduction internal fixation and hemiarthroplasty.²⁶ However, TSA is a more cost-effective treatment option than RTSA for patients with glenohumeral osteoarthritis.²⁷ With changing reimbursement in healthcare, surgeons must scrutinize not only anticipated outcomes with specific implants but the cost effectiveness of these implants as well. Further cost analysis studies are necessary to determine the ideal candidate for an RTSA.

LIMITATIONS

Despite its extensive review of the literature, this study had several limitations. While 2 independent authors searched for studies, it is possible that some studies were missed during the search process, introducing possible selection bias. No abstracts or unpublished works were included which could have introduced publication bias. Several studies did not report all variables the authors examined, and this could have skewed some of the results since the reporting of additional variables could have altered the data to show significant differences in some measured variables. As outcome measures for various pathologies were not compared, conclusions cannot be drawn on the best treatment option for various indications. As case reports were included, this could have lowered both the MCMS as well as the average in studies reporting outcomes. Furthermore, given the overall poor quality of the underlying data available for this study, the validity/generalizability of the results could be limited as the level of evidence of this systematic review is only as high as the studies it includes. There are subtle differences between rotator cuff arthropathy and pseudoparalysis, and some studies may have classified patients differently than others, causing differences in indications. Finally, as the primary goal of this study was to report on demographics, no evaluation of concomitant pathology at the time of surgery or rehabilitation protocols was performed.

CONCLUSION

The quantity, but not the quality of RTSA studies is increasing. Indications for RTSA varied by continent although most patients underwent RTSA for rotator cuff arthropathy. The majority of patients undergoing RTSA are female over the age of 60 years for a diagnosis of rotator cuff arthropathy with pseudoparalysis.

This paper will be judged for the Resident Writer’s Award.