Factors Associated with Local Recurrence in Surgically Treated Osteosarcomas

By: Prof. Shah Alam Khan and Dr. Rishiram Poudel


Musculoskeletal oncology in the developing world is still in its incipient stages of development. The biologic aggressiveness of osteosarcomas with local recurrence in surgical cases is a disastrous event in resource-challenged nations. The surgical options for reconstruction in local recurrences are limited and amputation remains the treatment of choice. Moreover, life expectancy is considerably diminished, since most cases are associated with distant metastases.

Local control of the disease, therefore, remains a critical goal in the management of osteosarcoma. We conducted a study to determine the prognostic factors related to local recurrence, as their identification is critical to the development of a newer, more rational risk-adapted treatment strategy.

Retrospective analysis
The study involved 95 patients who had been surgically treated for osteosarcoma at our institute in New Delhi, India, from 2004 through 2010. Patients who had received chemotherapy outside our hospital, had metastasis at the time of presentation, died within 3 months after surgery, and had incomplete clinicopathologic/radiologic data were excluded.

Patients were divided into the following two groups: those with no local recurrence (NLR, mean age = 17 years, 69 percent male) and those with local recurrence (LR, mean age = 16 years, 60 percent male). Data on demographics, site of tumor, previous biopsy reports, type of surgery (limb salvage or amputation) and presence of metastases at presentation were obtained from hospital records. Radiographic images were studied to look for the presence of pathologic fracture, vicinity of neurovascular bundle, and to estimate tumor volume. The tumor volume at presentation was calculated using MR images using the formula [Volume = (π/6) × length × width × depth]. Pathology reports were reviewed to note the histologic subtype, chemotherapy-induced necrosis, and surgical margins.

The most common location of osteosarcoma in both groups was the distal femur. In both the groups, the more common histologic varieties were conventional (osteoblastic) osteosarcoma followed by the chondroblastic type (P = 0.395).

Patients undergoing surgery within 4 weeks of completion of neoadjuvant chemotherapy were considered to have surgery on time. Any associated distant metastasis during the disease course and time to local recurrence from surgery in the LR group was noted in the follow-up records. Prognostic significance of the abovementioned factors were compared between the two groups. Independent t-test & Wilcoxon rank-sum tests were applied to look for any significant changes between mean/median values between the two groups. Chi-square and Fisher’s exact tests were applied to look for any significant variations. Comparison of median survival time between the two groups was done using Kaplan-Meier survival curve and log rank test. Multivariate logistic regression model was used for factors that were significant on univariate analysis. 

At a mean follow-up of 2.8 years (range: 4 months to 81 months), local recurrence had developed in 15 patients (local recurrence rate = 15.7 percent). Although all 15 local recurrences were initially treated with limb salvage procedures, limb salvage surgery did not significantly increase the risk of local recurrence (P = 0.66).

Sixty percent of the patients in the LR group had undergone a previous biopsy procedure outside our institute, compared to 33.7 percent of patients in the NLR group (P = 0.05). The mean tumor volume in the NLR group was 406.74±771.67 cc and 195.77±226.8 cc in the LR group (P = 1.403). Sixty percent of patients in the NLR group had chemotherapy-induced necrosis < 90 percent, compared to 73 percent of patients in the LR group (P = 0.396). No patient in the LR group was operated “on time” (as defined in our protocol). The mean delay in the NLR group was 4.16±4.81 weeks, compared to 9.46±6.5 weeks in the LR group (P = 0.0002). Two patients (13.33 percent) in the LR group and 7 patients (8.75 percent) in the NLR group had inadequate surgical margins (P = 0.63).

All eight pathologic fractures occurred in the NLR group. In the NLR group, 68.75 percent had the neurovascular bundle (NVB) free of tumor, 18.75 percent had NVB displacement by the tumor, and NVB was encased in 12.50 percent; in the LR group, 60 percent had NVB free of tumor, 33.33 percent had NVB displaced by the tumor, and NVB was encased in 6.67 percent (P = 0.467). Thus, in univariate analysis, biopsy procedure done outside our hospital and a delay in surgery after completion of neoadjuvant chemotherapy were prognostic factors related to development of local recurrence in osteosarcoma.

On multivariate analysis in osteosarcoma groups, delay in surgery maintained its significance as a prognostic factor for development of local recurrence, whereas significance was reduced for previous biopsy done outside our hospital. Each week delay in surgery after completion of neoadjuvant chemotherapy increased the risk of local recurrence by 1.14 times (P = 0.01; SE = 0.60; 95% CI [1.032 and 1.269]). Of the 15 locally recurrent cases of osteosarcomas, 12 patients (80 percent) recurred within 12 months of the primary surgery. Of the 15 local recurrences in osteosarcoma, local relapse was followed by distant metastasis in 60 percent, metastasis was followed by local relapse in 20 percent. Isolated local recurrence occurred in 20 percent of the cases. 

At 2.8 year mean follow-up, 60 percent of patients in the NLR group survived, compared to 20 percent in the LR group. The median survival time for the patients in the NLR group was 46.78 months, compared to 22.18 months in the LR group (P = 0.0041).

Delay in surgery after completion of neoadjuvant chemotherapy emerged as a significant prognostic factor for development of local recurrence. This is a common problem in developing nations like India, where delay in surgery beyond the expected time of recovery from complications of chemotherapy can be attributed to a patient’s financial difficulties, delay in the release of funds through various welfare programs, or the family’s lack of an alternative caretaker. In addition, because very few centers offer a multidisciplinary approach to musculoskeletal malignancies, these centers have high patient volumes and long waiting lists for surgery. Our study highlights the need for high priority and prompt attention when treating osteosarcoma patients, particularly in resource-challenged environments.         

Dr. Shah Alam Khan is professor of orthopaedics, and Dr. Rishiram Poudel is assistant professor of orthopaedics, at the All India Institute of Medical Sciences in New Delhi, India.