What is the Likelihood That Tumor Endoprostheses Will Experience a Second Complication After First Revision in Patients With Primary Malignant Bone Tumors And What Are Potential Risk Factors?

backgroundEndoprosthetic reconstruction of massive bone defects has become the reconstruction method of choice after limb-sparing resection of primary malignant tumors of the long bones. Given the improved survival rates of patients with extremity bone sarcomas, an increasing number of patients survive but have prosthetic complications over time. Several studies have reported on the outcome of first endoprosthetic complications. However, no comprehensive data, to our knowledge, are available on the likelihood of an additional complication and the associated risk factors, despite the impact of this issue on the affected patients. QUESTIONS/PURPOSES: (1) What are the types and timing of complications and the implant survivorship free from revision after the first complication? (2) Does survivorship free from repeat revision for a second complication differ by anatomic sites? (3) Is the type of first complication associated with the risk or the type of a second complication? (4) Are patient-, tumor-, and treatment-related factors associated with a higher likelihood of repeat revision?

methodsBetween 1993 and 2015, 817 patients underwent megaprosthetic reconstruction after resection of a tumor in the long bones with a single design of a megaprosthetic system. No other prosthetic system was used during the study period. Of those, 75% (616 of 817) had a bone sarcoma. Seventeen patients (3%) had a follow-up of less than 6 months, 4.5% (27 of 599) died with the implant intact before 6 months and 43% (260 of 599 patients) underwent revision. Forty-three percent of patients (260 of 599) experienced a first prosthetic complication during the follow-up period. Ten percent of patients (26 of 260) underwent amputation after the first complication and were excluded from further analysis. Second complications were classified using the classification of Henderson et al. to categorize surgical results. Briefly, this system categorizes complications as wound dehiscence (Type 1); aseptic loosening (Type 2); implant fractures or breakage and periprosthetic fracture (Type 3); infection (Type 4); and tumor progression (Type 5). Implant survival curves were calculated with the Kaplan-Meier method and compared using the log-rank test. Hazard ratios (HR) were estimated with their respective 95% CIs in multivariate Cox regression models.

resultsA second complication occurred in 49% of patients (115 of 234) after a median of 17 months (interquartile range [IQR] 5 to 48) after the surgery for the first complication. The time to complication did not differ between the first (median 16 months; IQR 5 to 57) and second complication (median 17 months; IQR 5 to 48; p = 0.976). The implant survivorship free from revision surgery for a second complication was 69% (95% CI 63 to 76) at 2 years and 46% (95% CI 38 to 53) at 5 years. The most common mode of second complication was infection 39% (45 of 115), followed by structural complications with 35% (40 of 115). Total bone and total knee reconstructions had a reduced survivorship free from revision surgery for a second complication at 5 years (HR 2.072 [95% CI 1.066 to 3.856]; p = 0.031) compared with single joint replacements. With the numbers we had, we could not show a difference between the survivorship free of revision for a second complication based on the type of the first complication (HR 0.74 [95% CI 0.215 to 2.546]; p = 0.535). We did not detect an association between total reconstruction length, patient BMI, and patient age and survivorship free from revision for a second complication. Patients had a higher risk of second complications after postoperative radiotherapy (HR 1.849 [95% CI 1.092 to 3.132]; p = 0.022) but not after preoperative radiotherapy (HR 1.174 [95% CI 0.505 to 2.728]; p = 0.709). Patients with diabetes at the time of initial surgery had a reduced survivorship free from revision for a second complication (HR 4.868 [95% CI 1.497 to 15.823]; p = 0.009).

conclusionsPatients who undergo revision to treat a first megaprosthetic complication must be counseled regarding the high risk of future complications. With second complications occurring relatively soon after the first revision, regular orthopaedic follow-up visits are advised. Preoperative rather than postoperative radiotherapy should be performed when possible. Future studies should evaluate the effectiveness of different approaches in treating complications considering implant survivorship free of revision for a second complication. LEVEL OF EVIDENCE: Level III, therapeutic study.