Mind the gap: Precision medicine shortcomings in NSCLC
ASCO abstracts highlight new molecular targets in advanced NSCLC, but we're failing to execute on the best care available now.
This year’s ASCO Annual Meeting included several major successes that should translate to new molecular targets in advanced non-small cell lung cancer, but recent real-world evidence should lead us to take a hard look in the mirror and realize we’re failing to deliver on the promise of the treatments we already have.
There are four biomarkers in this setting — EGFR, ALK, ROS1 and BRAF V600E — that have not only strong evidence to support highly effective targeted therapies, but also FDA-approved biomarker-driven therapies that are indicated standards of care, as defined by National Comprehensive Cancer Network guidelines. For EGFR and ALK, at least, there also are studies that establish targeted therapies as overwhelmingly superior to nontargeted therapies as first-line treatment.
Three other biomarkers — MET exon 14 mutation or high-level amplification, HER2 and RET — are listed as emerging biomarkers with data supporting use of novel therapies.
‘Glimpse’ of promising new treatments
At ASCO, we saw new data that reveal comparably high objective response rates with both capmatinib (INC280, Incyte) and tepotinib (EMD 1214063, Merck) for patients with MET exon 14 mutation-positive advanced NSCLC.
Remarkably impressive data also were presented for patients with advanced NSCLC that harbors a RET rearrangement and who received the novel agent BLU-667 (Blueprint Medicines), adding another option along with the similarly profoundly active selpercatinib (LOXO-292, Loxo Oncology) in this setting. These agents all merit FDA approval and commercial use as soon as feasible.
Further behind but with tremendous potential future implications was an early report of AMG 510 (Amgen), a tyrosine kinase inhibitor of KRASG12C, which is seen in approximately 13% of patients with advanced NSCLC. Although only preliminary data in a phase 1 trial, five of 10 patients with advanced NSCLC that harbors this mutation demonstrated a partial response to AMG 510, offering a glimpse of a potential treatment approach that could deliver highly effective targeted therapy for a biomarker we have known about but that has been widely considered “undruggable.”
‘Woefully inadequate’ testing rates
But we must unfortunately temper our optimism, because we’ve also seen new data that clearly illustrate how we are falling behind in executing on the promise of precision medicine.
Several months ago, Singal and colleagues reported a comprehensive review of both molecular and clinical data in databases from Foundation Medicine and Flatiron Health, which included over 3,500 patients with stage III or stage IV NSCLC. They reported results that were not mentioned in the discussion or accompanying editorial, but that should have been the dismaying focus of the paper.
All patients had undergone next-generation sequencing that was reported back to the ordering physician, but only 64% of patients with an EGFR mutation (n = 405 of 630) and only 70% of patients with an ALK rearrangement (n = 75 of 107) ever received the appropriate targeted therapy for their driver mutation.
These are the most established, overwhelmingly effective biomarker-driven treatments we have in advanced NSCLC, and more than one-third of the patients with a demonstrated target not only didn’t receive what ideally should be first-line treatment, but they never received the therapy.
Results for patients with less common but still NCCN-specified biomarkers were far more dismal, with only 33% (n = 179 of 543) ever receiving the targeted therapy recommended in the guidelines.
To me, these results are inexplicably poor and beg the question of whether the data are accurate. Unfortunately, new results presented at ASCO are astonishingly similar.
Looking at electronic medical records and claims data from the large Integra Connect database, Gierman and colleagues examined results from 1,203 patients with stage IIIb or stage IV NSCLC treated since the beginning of 2017 at five cancer care practices.
First, researchers found that molecular testing rates were very disappointing, with the most frequent testing, for an EGFR mutation, topping out at 54% of patients, and rates dropping steadily from there.
Only 22% of patients had testing for all four standard-of-care biomarkers specified in the NCCN guidelines, and a mere 7% of patients had testing for all seven that are suggested in the guidelines as emerging targets. We may note that molecular testing is most clearly indicated for patients with stage IV nonsquamous NSCLC, but the testing rate was still woefully inadequate in this large sample.
But beyond the poor execution of molecular testing that should be a standard of care, the use of the information obtained recapitulated what was reported in the Foundation Medicine/Flatiron Health findings.
The medical records and claims data from the Integra Connect database revealed that only 45% of patients whose cancer harbored an alteration in EGFR, ALK, ROS1 or BRAF V600E had evidence of ever receiving the appropriate targeted therapy for their driver mutation. And, despite patients with an EGFR mutation or ALK rearrangement being excluded from trials of first-line immunotherapy, more than one-third of the patients identified as EGFR- or ALK-positive received an immune checkpoint inhibitor prior to indicated targeted therapy.
This was not simply because treatment was required before molecular testing results were back. The median time to results for tissue testing was only 14 days (and a median of only 4 days for blood-based testing), and the EGFR and ALK testing results were reported back before the start of immunotherapy in nearly two-thirds of these cases.
Delivering on promise
We can scratch our heads about whether these results are real, and we can invoke hand-waving explanations to explain away these egregiously awful results — yes, it’s possible that these databases weren’t capturing every patient treatment accurately, and maybe some patients declined or couldn’t afford the oral therapy that is otherwise the clear standard of care.
But there is no way for methodological issues to justify these shortcomings.
We must acknowledge that our hard-won new treatments are being undermined by our failure to execute precision medicine broadly.
We should devote far more time and attention to examining why we have these gaps, then ensuring that the appropriate testing and treatment based on the results are being practiced in the rank and file oncology practices of the world.
We cannot fix what we don’t measure. It is clearly time to focus more on delivering on the promise of our best treatments. Having remarkably effective targeted therapies helps nobody if we aren’t testing or treating properly once they are available.
Basch E and Schrag D. JAMA. 2019;doi:10.1001/jama.2019.4064.
Maemondo M, et al. N Engl J Med. 2010;doi:10.1056/NEJMoa0909530.
NCCN Clinical Practice Guidelines in Oncology: Non-Small Cell Lung Cancer, Version 5.2019. Available at: www.nccn.org/professionals/physician_gls/default.aspx#nscl. Accessed on June 20, 2019.
Oxnard G, et al. Abstract OA12.17. Presented at: IASLC World Conference on Lung Cancer; Sept. 23-26, 2018; Toronto.
Román M, et al. Mol Cancer. 2018;doi:10.1186/s12943-018-0789-x.
Singal G, et al. JAMA. 2019;doi:10.1001/jama.2019.3241.
Soria JC, et al. Lancet. 2017;doi:10.1016/S0140-6736(17)30123-X.
Yang JC, et al. Lancet Oncol. 2015;doi:10.1016/S1470-2045(14)71173-8.
The following were presented at ASCO Annual Meeting; May 31-June 4, 2019; Chicago:
Fakih M, et al. Abstract 3003.
Gainor JF, et al. Abstract 9008.
Gierman HJ, et al. Abstract 1585.
Paik PK, et al. Abstract 9005.
Wolf J, et al. Abstract 9004.
For more information:
Howard (Jack) West, MD, is associate clinical professor of medical oncology at City of Hope Comprehensive Cancer Center. He can be reached at email@example.com.
To contribute to this column or suggest topics, email Wafik S. El-Deiry, MD, PhD, FACP, at firstname.lastname@example.org.
Disclosure: West reports honoraria from and consultant roles with Ariad/Takeda, AstraZeneca, Boehringer Ingelheim, Genentech/Roche, Loxo Oncology, Pfizer and Spectrum, as well as speakers roles with Ariad/Takeda and AstraZeneca.