Oesophageal cancer, encompassing adenocarcinoma (OAC) and squamous cell carcinoma (OSCC), is a significant global health concern. For locally advanced OSCC, definitive chemoradiotherapy (dCRT) is a standard treatment, while for OAC, it's used when surgery is not feasible. The systemic component of dCRT significantly impacts outcomes, yet systemic failure remains high (30-50%). While cisplatin and capecitabine are commonly used, carboplatin and paclitaxel offer potentially lower toxicity with comparable efficacy. The lack of biomarkers to guide chemotherapy selection in dCRT prompted this study. Previous research suggests that 18F-Fluorodeoxyglucose positron emission tomography (PET)-CT can predict response to neoadjuvant chemotherapy, leading to the hypothesis that early PET-CT assessment could guide optimal concurrent chemotherapy choices in dCRT for OAC and OSCC. This substudy within the SCOPE2 trial aimed to evaluate this hypothesis, leveraging a pre-existing framework for radiotherapy dose escalation.

Existing evidence supports the importance of the systemic component in dCRT for oesophageal cancer, although outcomes remain poor despite various attempts at improvement. Studies like RTOG 85-01 highlighted the benefit of concurrent chemotherapy, but toxicity remains a concern. SCOPE1 showed higher cisplatin dose intensity predicted improved survival, but concurrent chemotherapy was the main source of toxicity. While carboplatin and paclitaxel offer a potentially less toxic alternative, no reliable biomarkers currently guide chemotherapy doublet selection. In the neoadjuvant setting, PET-CT has shown promise in predicting response and guiding treatment personalization, however, its utility in the context of dCRT was unknown. This trial aimed to address this knowledge gap by investigating the use of early PET-CT assessment to direct the choice of systemic therapy during dCRT.

This was a multi-center, randomized, open-label, phase II substudy of the SCOPE2 trial. Patients with histologically confirmed oesophageal carcinoma (OAC or OSCC) eligible for dCRT were recruited. Patients received one cycle of induction cisplatin/capecitabine, followed by a day 14 PET-CT scan. Non-responders (less than 35% reduction in SUVmax from baseline) were randomly assigned (1:1) to either continue cisplatin/capecitabine or switch to carboplatin/paclitaxel for a further induction cycle and concurrent with radiotherapy. Responders continued cisplatin/capecitabine. All patients were also randomized to standard (50Gy) or high (60Gy) dose radiotherapy. The primary endpoint was treatment failure-free survival (TFFS) at week 24 in non-responders. Secondary endpoints included acute toxicity, overall survival, and progression-free survival (PFS). Statistical analysis included chi-square tests, logistic regression, and Kaplan-Meier curves. The trial was stratified by recruiting hospital, reason for nonsurgical therapy, and stage.

The PET-CT substudy was stopped early due to futility and potential harm. 103 patients participated, with 63 (61.2%) being non-responders. In the OSCC cohort, 24-week TFFS was significantly better in the cisplatin/capecitabine arm (92.6% vs 68.0%, p=0.028). Overall survival also favored cisplatin/capecitabine (42.5 vs 20.4 months, adjusted HR 0.36, p=0.018). There was a non-significant trend toward better survival in non-responders receiving cisplatin/capecitabine compared to responders. In the OAC cohort, the sample size was too small to draw meaningful conclusions. Neither baseline SUVmax nor percentage change in SUVmax correlated with 24-week TFFS. Treatment compliance was high in both arms, with similar rates of grade 3/4 toxicity.

The study's findings demonstrate that in OSCC, early metabolic response assessment via PET-CT at day 14 of induction chemotherapy does not predict outcome, and switching to carboplatin/paclitaxel in non-responders leads to inferior results. The lack of benefit in switching therapy based on early PET-CT contrasts with some neoadjuvant studies in OAC, which is likely due to factors such as the differences in inherent radiosensitivity, propensity for systemic relapse, or possibly an interaction between cisplatin/capecitabine and higher-dose radiotherapy. Further investigation is needed to explore the complex relationship between metabolic response, treatment choice, and outcome in both OAC and OSCC. The small sample size in the OAC cohort limits definitive conclusions for this subtype.

This study concludes that early PET-CT-guided selection of systemic therapy based on metabolic response is not recommended for patients with OSCC receiving dCRT. Switching from cisplatin/capecitabine to carboplatin/paclitaxel in non-responders resulted in worse outcomes. Further research is needed to optimize the systemic component of dCRT and identify predictive biomarkers for treatment personalization. The inconclusive OAC results highlight the need for larger trials to investigate this patient subgroup.

The early termination of the trial due to futility and potential harm limits the statistical power and generalizability of the findings. The relatively small sample size, particularly in the OAC cohort, hampers definitive conclusions for this subtype. The lack of central review of PET-CT scans may have introduced some variability. Selection bias might have occurred due to limited participation of centers in the PET-CT substudy. The study protocol allowed for CT scan results alone to determine the primary endpoint when endoscopy was not feasible which introduces potential bias although this was uncommon.