Background
Adjuvant platinum-doublet chemotherapy following surgical resection of non-small cell lung cancer (NSCLC) reduces risk of relapse, however not all patients can complete this treatment. This project aims to explore the frequency and determinants of early cessation of adjuvant chemotherapy in the Western Australian population and possible impacts on clinical outcomes.
Methods
This retrospective study analysed patients who received at least one cycle of adjuvant chemotherapy following NSCLC resection within WA public hospitals between years 2015 and 2022. We identified patients who received 4 cycles versus ≤3 cycles of chemotherapy and disease-free survival (DFS) was compared using Kaplan-Meier and Cox proportional hazards analysis, adjusting for clinicopathological variables including age, sex, stage, histopathology and comorbidities.
Results
Among 129 patients, median age was 67 years old and 50% were men. Pathological tumour stage of the cohort was IIA (13%), IIB (46%), IIIA (36%) and IIIB (4%). 74% had adenocarcinoma, with 13% and 47% harbouring EGFR and KRAS mutation(s), respectively. Median time from surgical resection to chemotherapy commencement was 51 days. 86% of patients received cisplatin, with vinorelbine (63%), pemetrexed (16%) and etoposide (11%) being common partner agents. 75% of patients received 4 cycles, and 10%, 12% and 3% received 3, 2 and 1 cycles respectively. Primary reasons for treatment discontinuation were patient preference and treatment toxicity. There was no significant difference in the median DFS between patients receiving 4 cycles or <3 cycles of chemotherapy (27.8 versus 22.3 months, p=0.67). Amongst the variables investigated, only age was a predictor of non-completion of adjuvant chemotherapy.
Conclusion
In this Western Australian population, most patients were able to complete adjuvant chemotherapy, with the most common reasons for treatment discontinuation being patient preference and treatment toxicity. There was no difference in DFS between the two groups, however statistical power is limited by the small sample size.