Research reveals practice patterns in precision oncology
A new study of a cohort of Veterans Health Affairs (VHA) oncology patients illustrates complexity of clinician decision making around targeted therapies in a fast-moving field
The field of precision oncology is challenging, because the details of next-generation sequencing (NGS) and variant categorization are new for many physicians. The percentage of cancer patients who are likely to benefit from genome-targeted therapy increased seven-fold between 2006 and 2018; yet in one study 20% of oncologists in the US reported that they “never or rarely” use next-generation sequencing tests.1, 2A retrospective study of non-small cell lung cancer patients treated since January 2017 at five community oncology practices found that 55% of patients with mutations amenable to targeted agents did not receive these treatments.3
Researchers from the United States Veterans Health Affairs (VHA) published an important study in this year’s ASCO proceedings pertaining to practice variation in precision oncology.4 They conducted a retrospective evaluation of a cohort of VHA patients with non-small cell lung cancer whose tumors were sent for NGS between July 2015 and February 2019. The output of DNA sequencing was analyzed by artificial intelligence-based software. A report describing genomic findings for which treatments were available was provided to the treating physicians. “Actionable” findings were defined as those for which level 1 and 2A evidence was available and approved by the Veterans Health Affairs’ National Precision Oncology Program. Variants included ALK, BRAF, EGFR, ROS1, ERBB2, MET, and RET.
Of 1750 patients with non-small cell lung cancer whose tumors underwent NGS, there was level 1 or 2A evidence to support the use of a targeted agent for 115. Of these, 37 (32.2%) were not treated with targeted agents. It is noteworthy that drug availability and affordability were not operative barriers. In 32.4% of cases, the patient did not have metastatic disease. In 19% of cases, the provider did not feel that the patient was a candidate for targeted therapy due to performance status or potential drug toxicity. In 24.3% of cases, the physician did not comment on the NGS results, suggesting that the information was not factored into treatment decisions.
There are numerous possible explanations as to why nearly a quarter of eligible patients did not receive targeted agents in the study. Perhaps patient-specific considerations that were not documented in the clinical notes precluded treatment. Perhaps NGS reports did not reach the treating physicians. Physician knowledge deficit could also have led to sub-optimal decision-making in the care of at least some of these patients.
Although this study was carried out at VHA facilities, rather than being specific to the Veterans Administration, these data are likely representative of oncology care across the US and globally. The research highlights an opportunity to support oncology clinicians in their awareness and understanding of the fast-moving field of precision oncology. Support for oncology clinicians can improve through a combination of education, clinician-focused NGS reports, artificial intelligence-augmented point-of-care decision support, and quality data analytics.
- Marquart J et al, JAMA Oncol. 2018;4(8):1093-1098
- Freedman A et al, Representative Survey of Oncologists in the United States; JCO Precis Oncol. JCO Precis Oncol. © 2018 by American Society of Clinical Oncology Creative Commons
- Gierman HJ et al; Journal of Clinical Oncology v37, no. 15_suppl (May 20, 2019) 1585
- Vashistha et al, Barriers to Prescribing Targeted Therapies for NSCLC Patients with Highly-Actionable Gene Variants in the VA National Precision Oncology Program; 2020 ASCO Abstract #33947