Precision medicine with TKIs and immunotherapies is revolutionizing the treatment of patients with NSCLC. Paul A. Bunn, Jr, MD, head of the Division of Medical Oncology at the University of Colorado, discussed the current translational advances with these agents in a discussion held at this yearâ€™s International Lung Cancer Congress.
Paul A. Bunn, Jr, MD
Precision medicine with TKIs and immunotherapies is revolutionizing the treatment of patients with non-small cell lung cancer (NSCLC). Paul A. Bunn, Jr, MD, head of the Division of Medical Oncology at the University of Colorado, discussed the current translational advances with these agents in a discussion held at this year’s International Lung Cancer Congress.
Mutational testing for TKI therapy has moved well beyond testing for onlyEGFRand ALKmutations. “The Lung Cancer Mutation Consortium study published last year showed that there are a number of drivers other thanEGFRand ALKand that if you have a driver and you have a driver treatment, the median survival was between 3 and 4 years,” said Bunn (JAMA. 2014;311:1998-2006). This is well beyond the 1-year median survival commonly projected for stage-IV lung cancer patients.
Bunn also noted that upfront TKI therapy is not only warranted in patients with EGFR and ALK mutations, but also in those with other driver mutations, as well. For example, he noted that response rates of 60% and 67%-80%, respectively, have been observed with early use of TKIs in patients withROS1and METmutations.
Despite success with TKIs in NSCLC, resistance mechanisms eventually emerge. In patients receiving anti-EGFR TKIs, “half or more of the progressions are because of the main gatekeeper mutation,T790M,” said Bunn.
Bunn believes that, for the most part, these mutations are present at diagnosis and not subsequently developed. “We can argue whether theT790Mmutation was there at the time of diagnosismy belief is that in close to 100% it was, but it was in a very low number and you couldn’t detect it. But when you blunt the growth of all the others, even though it grows slowly, it will outgrow the others.”
In testing forT790M, Bunn suggested that, based on data presented at ASCO 2015 (abstract 8001), use of less-invasive plasma genotyping appears to be as effective as tumor tissue genotyping. “The sensitivity of plasma testing is not as good as tissue testing, but the specificity is 99% or 100%.”
The two primary treatments that have emerged for patients withT790Mmutations are the third-generation TKIs, rociletinib (CO-1686) and AZD9291. Bunn presented a comparison across trials of monotherapy with these agents (ASCO 2014, abstracts 8009 and 8010) versus a combination of afatinib (Gilotrif) and cetuximab (Erbitux) in patients with NSCLC.
InT790M-positive patients, response rates with AZD9291, rociletinib, and afatinib/cetuximab were 65%, 58%, and 32%, respectively. However, inT790M-negative patients, response rates were similar, at 22%, 28%, and 28%, respectively. Bunn also noted that inT790M-negative patients who responded to the TKIs, the response duration was much shorter compared with responses in patients who harbored the mutation.
Rolling submission of a new drug application (NDA) for rociletinib has been initiated for patients with EGFRT790M-positive metastatic NSCLC following prior treatment with an EGFR-targeted therapy, according to the drug’s developer, Clovis Oncology. AstraZeneca, the company developing AZD9291, is expected to submit an NDA for the drug within the next few months.
If both drugs receive regulatory approval, Bunn said there are no head-to-head efficacy data to support the use of one drug over the other. “Whether the response rate and PFS is different between the two drugs is not known.”
Toxicity differences have emerged between the drugs that will likely be a key factor in treatment selection. “The primary reason for the toxicity differences is that there’s a metabolite of CO-1686 that inhibits IGF1R.” Bunn said that two toxicities that arise in CO-1686 due to the metabolite are hyperglycemia and elevation of the QTc.
Rates of hyperglycemia in NSCLC patients receiving CO-1686 have ranged from about 22% to 53%, depending on the study, Bunn noted. “The majority of those patients will require an oral antihyperglycemic drug.”
Elevation of the QTc requires extensive assessment and monitoring of other drugs that the patient is receiving that also interfere with QTc.
There are also some toxicities, such as rash, that are higher with AZD9291 than CO-1686, Bunn said.
In the end, Bunn indicated that with the currently available data, either drug is an acceptable option in this setting. “In my prediction, patients who progress on first-lineEGFRTKIs are going to get a plasma test to see if they have T790Mand if it’s positive, they should get one of these drugs, and if it’s negative, they should get a biopsy.”
“It’s a little hard to talk about translational advances without talking about immunotherapy,” said Bunn.
In March 2015, the FDA approved nivolumab (Opdivo) in the second-line setting for patients with advanced squamous NSCLC. Data presented at ASCO 2015 showed an overall survival benefit in the second-line setting for patients with nonsquamous NSCLC, as well.
PD-L1 status emerged as an efficacy biomarker in the pivotal nonsquamous trial, but not in the pivotal squamous study. Regardless, Bunn said that the NCCN NSCLC guidelines recommend second-line nivolumab for squamous as well as nonsquamous patients and, “Insurance companies are paying for nivolumab irrespective of PD-L1 expression and irrespective of histology.”
The PD-1 inhibitor pembrolizumab (Keytruda) is currently being reviewed by the FDA as a second-line treatment for patients with advanced NSCLC across all histologies, with a final decision scheduled to be made by October 2, 2015. In the pivotal phase I trial on which the pembrolizumab submission was based, PD-L1 status was shown to be a biomarker for efficacy.
While PD-L1 is the primary biomarker that has emerged with PD-1 agents, there is much debate over its interpretation and clinical value. As researchers continue to explore PD-L1 and search for other biomarkers, Bunn said, “At the moment, one could argue that the best biomarker is the number of mutations that an individual has. It does appear that the more mutations your tumor has, the more likely it is to respond.”