Real-world treatment patterns and outcomes based on RAS/BRAF status in metastatic colorectal cancer-Analysis of the Prospective Dutch Colorectal Cancer cohort

Abstract

The treatment landscape for metastatic colorectal cancer (mCRC) has evolved into a continuum of care with an essential role for biomarkers and molecular subgroups. Treatment guidelines are primarily based on trial results; however, populations and outcomes differ from clinical practice. To support the interpretation of trial results and to assist in tailored patient counseling, we evaluated real-world treatment patterns and outcomes according to RAS/BRAF status. We included all patients diagnosed with BRAF^V600E-mutated mCRC in 2015-2020, participating in the Prospective Dutch Colorectal Cancer cohort study, plus a 1:2 random selection of patients with RAS-mutated and double wild-type mCRC. We evaluated differences in administered lines of treatment (LOTs), local treatment, attrition rates, treatment duration, progression-free survival (PFS) and overall survival (OS). 178 BRAF^V600E-mutated, 221 RAS-mutated, and 174 double wild-type patients were included. Of BRAF^V600E-mutated patients, 26% received ≥3 LOTs, compared to 42% and 47% of the RAS-mutated and double wild-type patients, respectively (p = .002). Local treatment was performed in 25% of BRAF^V600E-mutated, 43% of RAS-mutated, and 49% of double wild-type patients (p < .001). Median OS from diagnosis was 15.4, 24.1, and 32.6 months, respectively (p < .001) and loss of prognostic value of RAS/BRAF was observed from the 3rd LOT onwards (p = .17 and p = .54). This paper provides a comprehensive overview of the treatment landscape of mCRC per RAS/BRAF status in daily clinical practice. The observed substantial treatment heterogeneity within and between molecular subgroups underlines the importance of collecting real-world data to address post-trial knowledge gaps and to optimize individualized counseling for all mCRC patients.

Keywords: BRAF; RAS; metastatic colorectal cancer; real‐world; treatment patterns.

FIGURE 1

Flow diagram study population.

FIGURE 2

Alluvial plot with individual patient journeys. (A) Alluvial plot. (B) Table showing the numbers of patients presented in the alluvial plot.

FIGURE 3

Stacked funnel plots presenting the 1st to the 6th line of systemic treatment, including attrition rates stratified by RAS/BRAF mutational status. (A) Treatments—double wild‐type. (B) Treatments—RAS mutated. (C) Treatments—BRAFV600E mutated.

FIGURE 4

Stacked barplots demonstrating individual patients' time on and off treatment per RAS/BRAF subgroup. Each bar represents a unique patient. (Patients with missing data on systemic treatment line start‐ and/or stop dates were excluded from this analysis (n = 28).) The complete bar depicts the time in months from the diagnosis of metastatic colorectal cancer to death or the last moment of follow‐up. (A) Patients with double wild‐type mCRC. (B) Patients with RAS‐mutant mCRC. (C) Patients with BRAFV600E‐mutant mCRC.

FIGURE 5

Median OS and median PFS from the 1st to 6th systemic line of treatment by RAS/BRAF status. (A) Kaplan Meier curve—median OS from mCRC diagnosis. (B) Kaplan Meier curve—median PFS from 1st line systemic treatment. (C) mOS from initiation of the 1st (mOS1) to the 6th line (mOS) of treatment. (The vertical lines attached to each dot represent the 95% confidence intervals of the median OS and median PFS calculated using the Kaplan‐Meier method. The x‐axes do not represent time in graph (C) and (D).) (D) mPFS 1st (mPFS1) to 6th (mPFS6) line of treatment. (The vertical lines attached to each dot represent the 95% confidence intervals of the median OS and median PFS calculated using the Kaplan‐Meier method. The x‐axes do not represent time in graph (C) and (D).)