Mutational Analysis Advances Precision Medicine in Head and Neck Cancer

Ezra Cohen, MD

Several prespecified biomarkers were found to be associated with progression-free survival (PFS) and overall survival (OS) in patients with metastatic head and neck squamous cell carcinoma (HNSCC) who received second-line afatinib (Gilotrif), according to an analysis of the phase III LUX-Head and Neck 1 (LHN1) study, which compared afatinib with methotrexate.

The analysis looked at p16, EGFR amplification, PTEN, and HER3 expression in tumor samples from 268 of the 483 patients included in the LHN1 study, to determine if these biomarkers were associated with clinical outcomes.

In the overall study population, afatinib significantly improved PFS (HR, 0.80; P = .030).

The biomarker analysis showed that afatinib improved PFS over methotrexate for patients who were p16-negative (HR, 0.70), EGFR amplified (HR, 0.53), HER3-low (HR, 0.57) and PTEN-high (HR, 0.55,). Likewise, OS was improved with afatinib over methotrexate for patients who were EGFR amplified (HR, 0.76) and for patients who were PTEN-high (HR, 0.67).

OncLive: Can you discuss the goals of your analysis?

The mutational analysis was presented at the 2016 Multidisciplinary Head and Neck Cancer Symposium by lead study author Ezra Cohen, MD, associate director, Moores Cancer Center, professor of Medicine, University of California, San Diego Health System. OncLive spoke with Cohen to learn more about the results of the analysis and the next steps in understanding biomarkers for afatinib in patients with head and neck cancer.Cohen: LUX-Head and Neck 1 was a trial of afatinib versus methotrexate in patients who had recurrent metastatic head and neck cancer who had failed a first-line platinum regimen. In this study, patients were randomized in a 2:1 fashion to receive either afatinib or methotrexate. The primary endpoint of the trial was PFS, and that was met with a hazard ratio of about .8.

We used this study to look at specific tumor markers and test the hypotheses related to afatinib’s mechanism of action. One of them was whether EGFR amplification predicted a better outcome in afatinib versus methotrexate.

Other major hypotheses were surrounding the PI3 kinase pathway, with the thought that activation of the PI3 kinase would actually predict for resistance to afatinib.

With those things in mind, we had a list of pre-determined biomarkers that we were going to investigate. When we had the primary results of the study, we knew that we were going to go back and look at these biomarkers, and ask the simple question, “Are there patients who we could identify—from biomarker work—that will do better with afatinib versus the standard of care?”

What were the most significant findings?

In addition to that, we also looked at a very important biomarker: HPV status. We used p16 to look at that with the hypothesis that there may be HPV-negative patients who may actually do better with an EGFR inhibitor such as methotrexate.First off, let’s talk about p16. Strikingly, it looked like it was in fact the p16-negative patients who appeared to benefit from afatinib. There were almost no responses in the p16-negative cohort treated with methotrexate. Methotrexate was ineffective in that patient population, while we did see responses in approximately 14% of patients who were treated with afatinib.

Moreover, when we looked at EGFR amplification, we saw what we expected. The EGFR-amplified patients also appeared to benefit from afatinib over methotrexate. Lastly, we took p10—which is a tumor suppressor along the PI3 kinase pathway—and HER3, as indicators of activation of the PI3 kinase pathway.

By looking at those markers, we saw what we expected. If PTEN was high, those patients benefited from afatinib. If HER3 was low, indicating an intact PI3 kinase pathway, those patients benefited. Therefore, our hypothesis appears to be real. It certainly gives us ideas on where to take a drug like afatinib, if not afatinib itself, forward.

What was most striking was when we started to combine these biomarkers—if we looked at the p16-negative cohort patients who were EGFR amplified— we saw a 20% response rate, with a more than doubling of PFS with afatinib versus methotrexate.

Are there next steps planned for this research?

By the same token, if we looked at p16-negative patients who have been treated with an EGFR inhibitor, their response rate was almost 30%. This, again, gives us the idea that perhaps we can begin to think about selecting patients who might be the best candidates for afatinib and other EGFR inhibitors.Unfortunately, we have almost exhausted the tissue we have access to. However, we are now exploring a more extensive analysis, primarily one that is RNA- or DNA-based, to get more of a global view.

We are investigating a pathway analysis, which will hopefully support what we are seeing preliminary with p10 and HER3. Those are ongoing and, hopefully, we will have more data at the end of this year or early next year. Furthermore, there is actually another major trial still ongoing with afatinib in the locally advanced setting. This is the phase III LUX-Head and Neck 2, which uses adjuvant afatinib versus placebo in patients with no evidence of disease after curative intent therapy.

We are hopefully going to be able to learn a lot from those patients and those biomarkers, and we can apply some of the same biomarker analysis to that study. In the curative setting, we may be able to identify a group of patients who may especially benefit from afatinib.