Impact of a national mesothelioma network on outcomes including survival.

INTRODUCTION: Equitable delivery of high-quality care is challenging in mesothelioma. Previous UK audits report variable practice and outcomes. The Scottish Mesothelioma Network was launched in April 2019, funding lead clinicians and clinical nurse specialists, a weekly national multidisciplinary team meeting and service improvement driven by quality performance indicators. We report outcomes from an embedded impact assessment. RESEARCH DESIGN AND METHODS: A multicentre ambispective cohort study was performed, including all cases diagnosed from April 2017 to April 2022. Baseline data, treatment and survival outcomes were collected prospectively in all network cases (April 2019 to April 2022) and West of Scotland (WoS) pre-network cases (April 2017 to March 2019). Data were retrieved retrospectively for pre-network non-WoS cases via cancer networks, supplemented by Public Health Scotland registry data. Overall survival (OS) was compared between pre-network and network cohorts using restricted mean survival time (RMST). RMST differences were integrated with baseline features and treatment utilisation, including ipilimumab-nivolumab and surgery. Treatment effects were modelled by propensity score matching. RESULTS: 659 consecutive patients were included (pre-network n=273; network n=386). Cohort demographics were well-balanced. In the network era, data quality improved and diagnostic pathways were not delayed. Adjusted RMST was 4.13 months longer in non-epithelioid network cases (p<0.001), but did not differ in epithelioid cases. Ipilimumab-nivolumab was used more commonly in network cases (47/279 (16.8%) vs 4/154 (2.6%) performance status (PS) 0-1 pre-network), but similar RMST extension was observed in untreated patients (+4.45 months, p=0.01). RMST differences were associated with improved PS at diagnosis, but not with stage, ipilimumab-nivolumab or surgery (in 17/659 (2.6%)). CONCLUSION: Establishment of a national mesothelioma network was associated with improved OS in non-epithelioid patients. Causal inferences cannot be drawn using the methods used, but these data support wider development of specialist mesothelioma networks.