Identifying responder/non-responder status prior to anti-PD L1 or neo adjuvant/ palliative chemotherapy in Lung and Head and Neck cancers using a novel bioinformatic approach to analyze biomarker data.

Introduction and background:

Cancer is a group of genetic diseases that result from changes in the genome of cells, leading them to grow uncontrollably. Early cancers are treated with surgery and may be followed with chemo or radiotherapy. More advanced stages are treated with chemotherapy or radiotherapy. The newest treatment modality over the last few years has been immunotherapy. This includes harnessing the immune system to get rid of cancer cells without affecting the normal cells in the organ involved or other organs. The immune system has been harnessed by biologics that inhibit pathways that keep the immune system repressed. Two such pathways have seen considerable activity and success, namely the PDL1 pathway and the CTLA4 pathway. The biology invoked is that the tumor represses the immune system actively and removal of such repression by inhibiting or blocking PDL1 or PD1 or CTLA4 has shown dramatic “cures” in some cancers with the derepressed immune system killing the cancer cells. With the immune system shown to be repressed in many cancers, clinical researchers have been exploring the deblocking of the immune system in several cancers. While the presence or overexpression of PD1 or PDL 1 targets in the tumor and tumor microenvironment were initially guiding therapy, the presence or even overexpression of these “targets” has not correlated with response to anti-PDL 1 therapies in some cancers.

Therapy with anti-PDL1 antibodies does not however guarantee response or “cures’’. The response rates to anti-PDL 1 therapies have ranged from 15 – 30% in most solid cancers and 45 – 60% in melanoma. In other words, there are many partial and non-responders to anti-PD L1 therapy. Secondly, anti-PD L1 therapy is not without side effects. Blocking of PD-1/PD-L1 immune checkpoint leads to the development of new toxicities by reactivation of the immune system, also known as immune-related (irAEs). These irAEs may affect multiple organ systems and tissues, with clinical manifestations of autoimmune-like/inflammatory side effects that may cause damage to the skin, lungs, gastrointestinal tract, liver, endocrine glands, and skeletal muscle. In addition, the most common treatment-related adverse events (TRAEs) include fatigue, fever/chillness, and infusion reactions. Furthermore, rare and serious TRAEs have been reported, including immune-related encephalitis, myasthenia gravis, acute renal failure/interstitial nephritis, and myocarditis. Thirdly, anti-PDL 1 therapies are not inexpensive. The cost of a single dose of anti-PDL1 antibody in India is INR . 2 lakh per month.  The average patient requires 2-4 doses before a response can be evaluated. The average response rate to anti-PDL1 therapy is about 20% even after confirmation of the presence of PDL1 by IHC and enhanced tumor mutational burden.

These factors necessitate the need for prediction of response to anti-PDL 1 therapies. At the TMH we have been using anti-PDL 1 therapies for lung and oral cancers for the past 6-7  years. We have been using anti-PDL1 therapy for lung (TNM stage III/ IV ) and oral cancers at the TNM stage.

Our collaborator, Dr. Raman Govindarajan of Samrud Foundation has developed a biomarker discovery and bioinformatics algorithm to predict response to anti-PDL 1 therapy and has successfully demonstrated its utility in melanoma treated with anti-PD L1 antibodies. We wish to develop a similar method to predict response to anti-PDL 1 antibodies in lung and oral cancers

Aims/ Objectives:

The study will enroll lung and oral cancer patients slated to undergo anti-PD L1 therapy. Develop a biomarker identification method to differentiate patients who have a complete response from those having a partial or poor response to anti-PDL 1 antibody therapy with sampling done before the start of anti-PD L1 therapy.

Study methodology:

1. Using a discovery set of 30 patients each of NSCLC and OSCC, we will attempt to develop a method to predict response to anti-PDL 1/ neoadjuvant/ palliative chemo therapy. The outcome of therapy of these patients will be shared with the Samrud Foundation as the results of response to therapy are obtained.

 

2. Using a validation set of 150 patients each of NSCLC and OSCC, we will attempt to confirm/ validate/ improve the method to predict response to anti-PDL 1/ palliative/ neoadjuvant chemo therapy in lung and oral cancer patients. The validation set is requested at 150 patients to comfortably obtain all three response groups, namely partial, complete, and non-responders. This is a prospective study. We will use the discovery set to identify markers for each response group. Thereafter, we will attempt to use FFPE samples (from our past patients, whose outcomes we have documented) to evaluate if the validation can be done on retrospective samples. If this is not successful for technical reasons, we will proceed with the prospective design and accrue the above numbers prospectively.

 

3. Sampling of tumor and blood will be done before the start of therapy after ethical approval and patient informed consent. Blood will be sampled at the beginning of each cycle of therapy. This is to monitor response to therapy.

 

4. Anti-PD L1 antibody therapy will be administered once every 21 days and as per standard practice  until disease progression . This is the standard dosing regimen that we have employed in the past and is the basis of our past data.Response of tumor to therapy will be monitored clinically and by serial CT/MRI done at 8-week intervals.

 

5. Samrud Foundation will analyze tumors and blood for biomarkers using their proprietary method (which involves the determination of tumor-specific mutations and subtraction of germline mutations to arrive at true somatic mutations after which a proprietary algorithm determines mutations that can segregate various response phenotypes – partial, complete and non-responders) and algorithm development utilizing the discovery set samples and outcome of therapy information. Samrud Foundation will also carry out the biomarker analysis for the validation set and will predict the “response to therapy” in the patients belonging to the validation set. The validation set will be blinded to Samrud and revealed after prediction using the algorithm has been arrived at to assess the accuracy of the prediction method.

 

6. Validation of the algorithm will be done in a single-blinded fashion for the validation set. TMH will not share the outcome of therapy information and will await the prediction of response to therapy by Samrud before comparing the outcome results with the prediction results.

Risk & Benefit:

There is no change to the therapy planned as a matter of normal patient care. Additional biopsy if required entails a risk of pain and bleeding and standard precautions will be taken to minimize this risk and deal with it as part of the standard of care.

 Since this is an exploratory study, the participants will not be directly benefited. The results of the study may benefit patients if the prediction algorithm is accurate and reproducible. Such an approach may also be extended to other cancers and therapies which may benefit patients in the future in terms of cost, avoidance of side effects, and emotional distress.

 Risk & Benefit Assessment:

There is no risk  to the patients. It is likely that the markers identified during this study can identify high risk patients in future and also aid development of diagnostics and therapies for resistant/ recurrent and metastatic cancer.