RESUMO
BACKGROUND: The Dutch Committee for the Evaluation of Oncological Agents (cieBOM) assesses the clinical benefit of systemic anti-cancer treatments (SACTs). For SACTs tested in non-randomized trials (NRTs), cieBOM primarily utilizes response-related thresholds as assessment criteria. As sufficiency of NRT-based evidence for benefit assessments is questionable, this study investigated whether and how NRTs can be used to assess the clinical benefit of new SACTs initially appraised by cieBOM based on randomized controlled trials (RCTs). METHODS: Using the RCTs underpinning cieBOM recommendations issued between 2015 and 2017, we searched for matching NRTs and applied the NRT-related assessment criteria by cieBOM to them. We then compared the assessment outcomes to the respective RCT-based cieBOM recommendations. Further, we investigated how the assessments would change when applying different response-related thresholds and adding a progression-free survival (PFS) threshold. RESULTS: For 13 of the 37 eligible recommendations, a matching NRT was found. Two treatments were assessed positively and six negatively; five treatments were non-assessable. Two positive recommendations matched a positive NRT-based assessment; one matching negative assessment was found, and one treatment could not be assessed based on either trial results. Adding a > 6 months PFS threshold decreased the number of non-assessable NRTs (five to two). CONCLUSIONS: Limited publications and inconsistent data reporting hampered the viability of NRTs for clinical benefit assessments of SACTs beyond the scope of rare indications. Further, response-related assessment criteria alone might not fully grasp the clinical benefit of novel SACTs. NRT-based assessments should be considered with caution due to uncertainty of the trial results.
RESUMO
The predictive value of the extent of peri-operative lymph node (LN) sampling in relation to disease relapse in patients with pulmonary carcinoid (PC) is unknown. Furthermore, post-surgery follow-up recommendations rely on institutional retrospective studies with short follow-ups. We aimed to address these shortcomings by examining the relation between LN sampling and relapse in a population-based cohort with long-term follow-up. By combining the Dutch nationwide pathology and cancer registries, all patients with surgically resected PC (2003-2012) were included in this analysis (last update 2020). The extent of surgical LN dissection was scored for the number of LN samples, location (hilar/mediastinal), and completeness of resection according to European Society of Thoracic Surgeons (ESTS) guidelines. Relapse-free interval (RFI) was evaluated using Kaplan Meier and multivariate regression analysis. 662 patients were included. The median follow-up was 87.5 months. Relapse occurred in 10% of patients, mostly liver (51.8%) and locoregional sites (45%). The median RFI was 48.1 months (95% CI 36.8-59.4). Poor prognostic factors were atypical carcinoid, pN1/2, and R1/R2 resection. In 546 patients LN dissection data could be retrieved; at least one N2 LN was examined in 44% and completeness according to ESTS in merely 7%. In 477 cN0 patients, 5.9% had pN1 and 2.5% had pN2 disease. In conclusion, relapse occurred in 10% of PC patients with a median RFI of 48.1 months thereby underscoring the necessity of long-term follow-up. Extended mediastinal LN sampling was rarely performed but systematic nodal evaluation is recommended as it provides prognostic information on distant relapse.