Window of Opportunity Trials Are Needed for Novel Tissue Biomarker Analysis in Head and Neck Cancers

Drug development is a difficult and ever-changing process, with new agents being developed by industry and academic partners. Still, the average time from investiga- tor new drug filing to final new drug application approval is 8.7 years because of the need for vetting in the preclinical and clinical trial setting.¹ Although the process is rigorous to provide approval at a faster pace, in general, the process still takes several years of clinical development before being approved for use in the clinical setting.^1,2 The traditional trajectory for early-phase cancer drug development is typically conducted in patients who are heavily pretreated, as novel investigational agents are used when no other stan- dard-of-care therapies exist.

Novel trial designs have been used, with close collaborations between clinicians, laboratory researchers, and industry partners, such as single-patient trials (N-of-1) trials,³ to provide patients with new therapeutic options when no other treatments exist. For example, N-of-1 trials were conducted in patients receiving larotrectinib (Vitrakvi) for TRK fusion–positive cancers, as well as LOXO-292 (now approved by the FDA as selpercatinib [Retevmo]) for patients with RET alterations to speed drug development and guide treatment decisions earlier.^4,5 In both single-patient clinical trials, clinicians worked closely with scientific researchers in industry to drive the research forward with rapid intra-patient, pharmacokinetic-guided dose escalations to provide clinical treatment options for these patients.

We propose that window of opportunity clinical trials can be used for a more rapid translation of basic science discovery into the clinic with pharmacodynamic and biomarker analysis in the common treatment-naïve tumor tissue.6 Since window studies are traditionally conducted before surgical resection, most patients undergoing the initial diagnostic/ research biopsy have not been exposed to systemic chemo- therapy or cancer-related treatment. The research specimen obtained at the time of surgical resection not only provides a significantly larger amount of tissue for biomarker analysis, but the timing of treatment before surgery offers
the purest evaluation of changes in the tumor after the window treatment.

A more complicated version of a window of opportunity clinical trial design can also occur in the metastatic setting by layering individual agents and using novel sequences of therapeutic drugs and treatments. Incorporating research biopsies at rationally selected intervals offers biomarker analysis with single, double, and possible triple therapy at each time point. This creates a simple but sophisticated layering of therapeutics with associated biopsies for mechanistic analysis. Clinical trials with a lead-in window of a single agent offer this window before a planned combination therapy. At our institution, we have used window of opportunity clinical trials for patients with head and neck cancer in two settings: as a treatment before surgical resection and in the metastatic setting where patients are undergoing palliative radiation therapy with systemic chemotherapy.

In the surgical setting, Wheeler’s preclinical data indicated that overexpression of the recep- tor tyrosine kinase AXL is tightly linked to cetuximab resistance.^7,8 In the previous window of opportunity trial, we employed two weeks of cetuximab (Erbitux) treatment before surgical resection to evaluate changes in Ki67 expression and gauge whether expression of AXL could serve as a predictive biomarker in cetuximab resistance (NCT03769311). Further preclinical data from our team suggested that tumors resistant to cetuximab had elevated AXL and AXL was responsible, in part, for the activation of the kinase c-Abl.⁹ Combination therapy of cetuximab plus imatinib (Gleevec) led to complete tumor regression without recurrence in animal models. To test this newly discovered combination in head and neck cancer we developed a surgical window of opportunity trial to test the combination by looking at Ki67 expression as a surrogate for proliferation (Figure 1).

In our clinical trial evaluating INCB081776, pembrolizumab (Keytruda), and palliative radi- ation therapy in patients with metastatic head and neck cancer (Figure 2), patients undergo a baseline biopsy followed by 14 days of treatment with INCB081776. A second research biopsy is obtained after completion of INCB081776 treatment, but before the addition of pembrolizumab monotherapy. Subsequently, patients
are treated with palliative radiation therapy for metastatic disease to alter the tumor microenvironment and stimulate the immune system. A third optional biopsy may be obtained for additional biomarker analysis. Patients will continue to receive combination therapy with INCB081776 plus pembrolizumab until disease progression or unacceptable toxicity. In this design, our team endeavored to provide a mechanism for obtaining tissue specimens to be analyzed in real time during the clinical trial, which would provide ongoing proof of concept. In addition, having biospecimens available for analysis during therapy can help refine and correlate toxicities and provide an early look at potential treatment changes.

We acknowledge that the traditional design of window trials may not provide an immediate effect for patients. Still, the data generated paves the way to refine treatment strategies for future cohorts and the next generation of clinical trials. As clinicians, we are obligated
to patients to be good stewards of research specimens provided to spur additional clinical research. We must engage with our scientific partners for close collaborations for real-time analysis to efficiently refine our scientific questions to speed the drug development process and ultimately improve the survival of patients with cancer. As part of the University of Wisconsin Head and Neck Specialized Program of Research Excellence grant, our work continues to improve upon this, focusing on horizontal and vertical collaborations to ensure that all cancer patients benefit.

Acknowledgments

This project was supported by the Specialized Program of Research Excellence program, through the National Institutes of Health (NIH) National Institute for Dental and Craniofacial Research and National Cancer Institute grants P50DE026787, P50CA278595, and R01CA262292. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH. Figures were created with Biorender.com