In this study, we found EGFR mutation, ALK translocation, and ROS-1 fusion in a proportion of 33%, 4%, and 1%, respectively. Even in a context of pharmaceutical industry-sponsored tests, many patients (61%) did not have their molecular profile completely characterized. Moreover, at the treatment initiation, the full driver mutations panel results were not available in 51% of the patients. Considering those harboring targetable molecular alteration, 55% did not undego the targeted therapy upfront, as recommended.
Previous studies have investigated the access to biomarker testing and its rates over time in many countries, including Brazil8-10. However, to the best of our knowledge, the present study is the first to focus on the time spent in the molecular characterization and its impact on treatment choice in a real-world scenario of a middle income country.
According to The College of American Pathologists (CAP), it should take less than 14 days from the availability of a suitable sample to the report of its final results11. In our study, we showed a median of 12 days, which is under this recommendation. Besides, other retrospective studies showed that EGFR analysis lasts from 8 to 17 days in different countries12.
More recently, a Japanese publication demonstrated a median time of 11 days between the test ordering and its conclusion . The molecular test included EGFR, ALK, ROS1, and PD-L1. Among the patients harboring a targetable mutation, 93% underwent a directed therapy as the first-line13. These data contrasts with ours. In our analysis, only 52% of the patients had their full driver mutations panel results available at the time of the treatment initiation. It may be explained by logistical challenges that lead to the long interval between the diagnosis procedure and the testing request (median of 29 days).
Awaiting biomarker testing results may delay treatment decisions in patients with advanced NSCLC, which may directly impact their clinical outcomes14.
This barrier to the personalized medicine implementation might be overcomed through the incorporating of reflex NSCLC biomarker testing at the level of the pathologist. Several studies had addressed the role of reflex testing in reducing the time between molecular investigation and treatment initiation. Phung et al. reported a reduction in this interval from 52 to around 23 days in a single-center study15. Similarly, a Canadian group demonstrated a shorter interval to the optimal first-line systemic therapy (median, 36 days [IQR, 16 to 91 days] versus 24 days [IQR, 8 to 43 days], p=.036) with the reflex testing utilization16. Moreover, according to an institutional review, when EGFR/ALK results were available since the first consultation with the oncologist, nsNSCLC-patients had their time to treatment improved significantly (16 versus 29 days, p=.004)17.
Our study also revealed another concern regarding precision medicine incorporation among NSCLC patients in our setting. Although the multiple driver mutations already identified in nsNSCLC, the molecular panel was not complete in a majority of the patients. The six most important targetable mutations were investigated only in 39% of cases. Our data differs to the MYLUNG Cousortium, in which 49.0% of the patients were completely tested for these 5 biomarkers, with a tendency of improvement in this rate through the last years18.
Moreover, in our cohort, less than 4% of patients had their material evaluated through a NGS method, a technique recognized as fast and accurate. Thus, the use of NGS may also be time-sparing, which contributes to avoiding the initiation of the first-line treatment before the availability of all driver mutations testing results. Our study did not assessed ethnic differences in terms of access to NGS testing. A recent study showed that African Americans were less likely to undergo NGS testing when compared to those who are Caucasian (39.8% versus 50.1%, p<0.0001)19.
NGS testing may also improve clinical trial participation wich represents a very important pathway to access to innovative therapies. The same study previously cited showed that African American were also less likely to be treated in clinical trials (1.9% versus 3.9%) due to the lack of access to NGS testing19.
Our study has some limitations. It was retrospective, and performed in a single-center, not reflecting the sociodemographic and genetic diversity of our population. Besides, we considered only the patients whose biomarker tests were requested. Thus, a selection bias could have occurred, since in daily practice many oncologists still use clinical predictors before requesting the tests. Maybe, it could justify the higher rate of EGFR mutations in the studied population (33.3%), comparing to previous publications4.
Although the remarkable progress in NSCLC treatment until the mid-2010s, the landscape has been changing rapidly in the recent years. Thefore, the interval of patients recruitment would be considered too long to precisely assess the compliance to the targeted therapy, which is another limitation of this study.