Multi-platform molecular profiling increases the median overall survival of those who receive treatment with clinical benefit by 422 days from time of profiling
This observational study aimed to primarily measure outcomes and determine the clinical utility of Caris Molecular Intelligence (CMI) multi-platform molecular profiling (MMP), as well as assess how it impacted physicians’ treatment decisions. 1,180 cases of solid cancers were enrolled in the Caris registry between 2009 and March 2015. Registry data included baseline clinical information at time of enrolment, as well as information collected at nine-month intervals post-enrolment. Specific testing was performed on tumour biopsies from all patients. This testing involved a combination of sequencing (Sanger, NGS or pyrosequencing), protein expression (IHC), gene amplification (CISH or FISH) and/or RNA fragment analysis.1
Of the 1180 patients, 153 were lost to follow up and a further 21 were excluded based on reported histology. The remaining patients (n=1027) were divided into two cohorts: a matched cohort (defined as having received at least one treatment associated with potential benefit and no treatment associated with a lack of benefit at any time post diagnosis, n=534) and an unmatched cohort (all patients who received at least one treatment associated with a lack of benefit at any time post diagnosis, n=493). Selected biomarker expression was shown to vary between the cohorts. The matched cohort had significantly more potential sensitivity to platinum agents (ERCC1 loss), gemcitabine (RRM1 loss), 5-FU based antimetabolites (TS loss), taxanes (TUBB3 loss), androgen deprivation therapy (AR), and hormone therapy (ER and PR). The matched cohort had less BRCP expression, and the unmatched cohort had more actionable mutations (KRAS, BRAF, PIK3CA).1
Survival analysis showed that patients in the matched cohort had a significant increase in median overall survival (OS) of 422 days from the time of profiling compared with the unmatched cohort (median OS 1068 vs 646 days). Median OS from diagnosis, including treatments given prior to MMP, also demonstrated a survival benefit in the matched cohort of 398 days (978 vs 580 days). The study also confirmed the predictive value of certain biomarkers. ERCC1-negative subjects had a longer median OS survival than ERCC1-positive subjects if they received platinum agent(s) at any time after diagnosis (median OS 1038 vs 705 days) or after profiling (median OS 1075 vs 685 days). Similar outcomes were also observed for other biomarkers, including RRM1 loss with gemcitabine.1
The CMI report influenced the treatment decision in 53% of cases (n=629), with 97% of these changed decisions (n=611) resulting in a treatment with potential benefit being administered, and 46% (n=292) resulting in the avoidance of treatment associated with potential lack of benefit. Patients in the matched cohort were administered less treatments overall after profiling (median 3.2 vs 4.2 therapies).1
- Median overall survival was shown to increase by 422 days (1068 vs 646 days) in patients who only received treatments associated with clinical benefit after multi-platform molecular profiling.
- The study also confirmed the predictive value of certain biomarkers, such as ERCC1-loss and platinum agent(s), as well as RRM1 loss and gemcitabine.
- The CMI report influenced physicians’ treatment decision in 53% of cases (n=629/1180).
- Patients in the matched cohort of the study received less treatments after profiling compared with the unmatched cohort (3.2 vs 4.2 therapies).
- Spetzler D, Xiao N, Burnett K, et al. Multi-platform molecular profiling of 1,180 patients increase median overall survival and influences treatment decision in 53% of cases. Eur J Cancer. 2015; 51(Suppl 3):S44. [PDF]