Portal de Periódicos da CAPES

Impact of primary tumour location on survival in patients with metastatic colorectal cancer receiving selective internal radiation therapy and chemotherapy as first-line therapy

2017; Elsevier BV; Volume: 28; Linguagem: Inglês

10.1093/annonc/mdx302.005

1569-8041

Guy van Hazel, Volker Heinemann, Navesh Sharma, Julien Taı̈eb, Jens Ricke, Marc Peeters, Michael Findlay, Peter Gibbs,

Radiomics and Machine Learning in Medical Imaging

Background: Primary tumour location of metastatic colorectal cancer (mCRC)is emerging as a major prognostic factor and predictor of response to some treatments, with right-sided primary tumours (RSP) having inferior prognosis and response. The impact of primary tumour side on outcomes after selective internal radiation therapy (SIRT) of patients with mCRC liver metastases has not previously been examined, but a survival benefit in patients with RSP was recently reported in an exploratory analysis of the FOXFIRE studies (Sharma RA et al. 2017 ASCO Annual Meeting; J Clin Oncol 2017;35(Suppl):Abs 3507). Therefore, we analysed data from the SIRFLOX (SF) and FOXFIRE-Global (FFG) trial cohorts to assess the impact of primary tumour location on survival and other outcomes. Methods: The SF and FFG studies were designed with similar protocols. Patients with previously untreated liver-only or liver-dominant mCRC were randomised (1:1) to receive mFOLFOX6 chemotherapy alone or with a single administration of SIRT using yttrium-90 resin (Y-90) microspheres (SIR-Spheres; Sirtex, North Sydney). Patients could receive bevacizumab at investigator discretion. Primary tumour location was captured prospectively in the case report form. RSP was defined as any primary tumour that was proximal to the splenic flexure; left-sided primary tumour (LSP) included a primary tumour at the splenic flexure, the more distal colon and the rectum. Progression-free survival (PFS) and overall survival (OS) data were examined with a focus on the influence of tumour side on outcomes. Results: Of the 739 patients recruited (SF n = 530; FFG n = 209), 179 (24.2%) had a RSP and 540 (73.1%) had a LSP. In the combined analysis of all 739 patients, the addition of SIRT to mFOLFOX6 chemotherapy had no impact on median PFS or OS [11.1 vs. 10.6 months (p = 0.22); 24.3 vs. 24.6 months (p = 0.84) with SIRT or without SIRT, respectively]. For patients with a RSP, OS was significantly improved with the addition of SIRT [22.0 vs. 17.1 months, respectively; p = 0.007; Hazard Ratio (HR): 0.64 (95% CI: 0.46-0.89)], but not for patients with a LSP [24.6 vs. 25.6 months; p = 0.279; HR: 1.12 (0.92-1.36)]. The location by treatment interaction was highly significant [Chi-square: 9.49; p = 0.002; HR: 0.548 (0.37-0.80)]. For patients with a RSP, the trend was for improved PFS with the addition of SIRT (10.8 vs. 8.7 months; p = 0.053; HR: 0.73 (0.53-1.01)], but not for patients with a LSP [11.4 vs. 10.8 months; p = 0.426; HR: 0.93 (0.78-1.11)]. The incidence of grade ≥3 adverse events did not differ for RSP and LSP (p > 0.05). Conclusions: In the SF and FFG studies, the addition of SIRT using Y-90 resin microspheres to first-line mFOLFOX6 chemotherapy was associated with a statistically and clinically significant gain in OS for patients with a RSP but not in patients with a LSP. These findings are being validated in the remaining FOXFIRE trial cohort. These data add to the growing literature describing primary tumour location-based differences in mCRC outcomes with some treatments, and may support a side-based approach to treatment selection including SIRT. The biological drivers of these differences remain to be defined.