(P 436) OUTCOME PROGNOSTICATORS IN ADOLESCENTS AND ADULTS WITH BONE SARCOMAS IN LOW-MIDDLE-INCOME COUNTRIES

Objective: Treating patients with bone sarcomas, including Ewing's Sarcoma (ES) and osteosarcoma (OGS) in Low-Middle Income Countries (LMIC) has multiple unique challenges. Social inhibitions and lack of awareness leads to high tumor burden and higher incidence of metastases at presentation. Inadvertent prior treatment by primary care physicians, high dropout rates and poor adherence compromises outcomes. We have aimed to identify prognostic factors including those specific to LMIC.

Methods: All adolescent and adult OGS and ES cases treated at Tata Memorial Centre, Mumbai, with the OGS-12 and Ewing's family of tumors-2001 [EFT-2001]) protocol from November 2011 to January 2021, and January 2013 to November 2018 respectively were prospectively analyzed. The OGS-12 protocol consists of doublets of doxorubicin, ifosfamide and cisplatin given sequentially for eight cycles in both neoadjuvant (NACT) and adjuvant (ACT) settings. The EFT-2001 protocol consists of a 12-month course of ifosfamide plus etoposide and vincristine, doxorubicin plus cyclophosphamide. Disease factors were symptom duration, lactate dehydrogenase (LDH), serum alkaline phosphatase (SAP), nutritional parameters, tumor size, presence and sites of metastases, pathological fracture and neurovascular bundle involvement. Treatment characteristics were treatment duration, protocol completion, prior treatment, febrile neutropenia (FN) and dose modifications. These were recorded from electronic medical records (EMR). The primary outcome was event-free survival (EFS). Secondary outcomes were overall survival (OS) and histological necrosis. Effect of covariates on histological response was estimated by logistic regression analysis and on survival using Cox proportional hazards analysis. Significant factors on univariate analysis were tested in multivariate analysis.

Results: =There were 1063 adolescents and adults aged >15 years with bone sarcomas, including 828 OGS cases and 235 ES cases. Males constituted the majority, 505 (62%) of OGS cases and 162 (69%) of ES cases respectively. The median age of presentation was 19 (range: 15-67) years. 256 (31%) OGS patients and 52 (22%) ES patients had metastatic disease. Most had high tumor burden and poor nutritional status, as described in Table 1. The median treatment duration was 40 (range: 36-61) weeks for OGS and 71 (range: 61-87) weeks for ES. 753 (91%) of OGS and 181 (77%) of ES cases completed the treatment protocol. 98 (12%) of OGS and 29 (9%) of ES cases had received prior treatment. Good histological necrosis (≥90%) was seen in 435 (55%) of 791 evaluable OGS patients. ≥90% necrosis was observed in 106 (48%) and ≥99% in 68 (31%) patients of 221 evaluable ES patients. 439 (53%) OGS patients and 103 (44%) of ES cases had FN, with 182 (22%) and 44 (19%) requiring dose modification, respectively. patients. At a median follow-up of 64.3 (range: 5-81 months), the median EFS ad OS of the OGS cohort was not reached, mean EFS and OS were 39 (95% CI: 5-61) months and 66 (95% CI: 8-91) months respectively. On univariable analysis, for both ES and OGS separately, the presence of baseline metastatic disease; tumor diameter >10cm; higher baseline SAP and LDH, absence of FN and failure to complete protocol and treatment duration >40 weeks were predictive of inferior EFS, as shown in Table 2. For OGS, baseline pathological fracture, necrosis < 90% and treatment duration >40 weeks were additional predictive factors. For ES, necrosis < 99% and treatment duration >72 weeks were predictive. However, on multivariable analysis, only the presence of baseline metastases (HR: 4.1, p< 0.001); pathological fracture (HR 1.3, p=0.04), poor histological necrosis (HR 1.9, p=0.01), failure to complete protocol (HR 1.2, p=0.04), and prior treatment (HR 1.7, p=0.02) independently predicted inferior EFS. These factors were also predictive of inferior OS, as shown in Table 3.

Conclusion: This is the largest prospective analysis of bone sarcomas in India, with generalizable results for LMIC. Metastatic disease for both OGS and ES and pathological fracture for OGS were disease characteristics resulting in worse outcomes. Importantly, indicators of poor treatment compliance such as failure to complete the planned protocol and social challenges such as prior inadvertent treatment were also independently predictive of inferior outcomes, Future directions include development of a predictive model for prognosticators in ES and OGS that will be of great importance in improving outcomes in LMIC.