As PD-1 inhibitors march toward earlier-stage head and neck cancers, researchers consider ways to improve the precision of immunotherapy through better patient selection and combinations for this accessible tumor type.
With Phase III trials ongoing of leading PD-1/L1 inhibitors in first-line treatment of recurrent or metastatic cancers of the head and neck, new research already is moving the future target to the neoadjuvant setting, prior to surgery and chemoradiation, with curative intent.
Squamous cell carcinomas of the head and neck (SCCHN or HNSCC) include tumors that start in the mouth, nose and throat and salivary glands. Typically, they are associated with use of tobacco and alcohol, but increasingly they are caused by human papillomavirus, or HPV. The Epstein-Barr virus also is associated with this tumor type.
SCHHN cancers caused by HPV have been on the rise, driving up overall incidence per year in the US from 39,700 in 2005 to 59,500 in 2015. Incidence has been rising, but standard therapies have been failing a large number of patients, noted Antonio Jimeno, University of Colorado, during a session at the American Society of Clinical Oncology (ASCO) annual meeting on June 6.
For those who are HPV-positive, the prognosis is more favorable compared to the negative type, but cure rates are far from complete, Jimeno said.
PD-1 checkpoint inhibitors represent a new treatment option for recurrent metastatic SCCHN. Bristol-Myers Squibb Co.'s Opdivo (nivolumab) was approved by FDA in October 2016 for recurrent metastatic SCCHN after the failure of platinum-based chemotherapy, a filing supported by the Phase III CheckMate 141 study, where the drug demonstrated superior overall survival compared to investigator's choice of methotrexate, docetaxel or Eli Lilly & Co.'s Erbitux (cetuximab). (Also see "Keeping Track: Mylan Submits Biosimilar Herceptin; FDA Clears Opdivo For Head And Neck Cancer" - Scrip, 11 Nov, 2016.) Opdivo was associated with a statistically significant 30% reduction in the risk of death in the study.
Merck & Co. Inc.'s competing PD-1 inhibitor Keytruda (pembrolizumab) received accelerated approval from FDA for the same indication in August 2016. Merck's filing was supported by a cohort with this tumor type in the single arm, Phase Ib KEYNOTE-012 study. In that trial, the drug demonstrated an objective response rate (ORR) of 16%, including a complete response rate of 5%. (Also see "Keeping Track: Opana ER Abuse Deterrence Claim Withdrawn; Third Neulasta Biosimilar Submitted; Keytruda Cleared In Head And Neck Cancer" - Scrip, 12 Aug, 2016.)
The data so far suggest that the PD-1 inhibitors are similar in terms of efficacy and there is not enough information to compare toxicity, University of California, San Diego's Ezra Cohen commented in an interview. There is clearly a big need for better therapy in recurrent and metastatic head and neck cancer and the PD-1 agents have been embraced and are used readily within the indication outlined in labeling, he said. A picture is starting to emerge of who gets the most out of PD-1 immunotherapy – those who are HPV-negative benefit, but those who are HPV-positive are a bit more likely to respond, Cohen told Scrip.
The market potential for this indication clearly is a lot smaller than other tumor types, like non-small cell lung cancer (NSCLC). Morningstar Research estimates PD-1/L1 inhibitors will have sales in head and neck cancer of $1bn out of a total of $28bn for all indications in 2021. But collectively, the smaller indications are worth a lot – about half of sales in 2021 will come from non-NSCLC tumor types, Morningstar estimates.
Phase III activity in the space is competitive. Keytruda and Opdivo are both in Phase III for first-line recurrent or metastatic SCCHN (no prior chemotherapy) and AstraZeneca PLC is running a Phase III study of its PD-L1 inhibitor Imfinzi (durvalumab) with its CTLA-4 inhibitor tremelimumab in a similar population (see table below). Merck KGAA and partner Pfizer Inc. have their PD-L1 inhibitor Bavencio (avelumab) in Phase III for use in a much earlier setting with chemoradiation in locally advanced SCCHN patients who have had no prior therapy for advanced disease, with curative intent.
Single Keytruda Dose Goes Far
Data presented at the ASCO meeting, held June 2-6 in Chicago, highlighted the potential for treating the disease in the neoadjuvant setting, prior to surgery and chemoradiation.
A Phase II trial sponsored by the University of Washington and the Dana-Farber Cancer Institute looked at Keytruda as a neoadjuvant treatment in patients with surgically resectable HPV-negative locally advanced SCCHN (mostly Stage IV) with high-risk pathological features. The typical rate of recurrence for these high-risk patients after undergoing post-operative adjuvant chemoradiation therapy (PO-ACRT) is high at 35%.
All patients received a single 200 mg dose of Keytruda prior to surgery, but surgical treatment was based on the tumor size prior to PD-1 treatment. Most of the participants had a history of heavy smoking.
In cases with extracapsular extension and positive margins after surgery, patients received PO-ACRT and Keytruda as a maintenance treatment. The goal was to reduce the recurrence rate from 35% to 15%.
Reporting results at ASCO, the Dana-Farber Cancer Institute's Ravindra Uppaluri noted that 10 of 24 (42%) patients had a pathological response in tumors or lymph nodes after the single pre-surgical dose of Keytruda – noting that this was a "finding that we were surprised to see." Six (25%) had a major treatment effect. The presence of CD8 T-cells, CD8/PD-1 positive cells and CD4 T-cells all correlated with response.
The number of patients with positive margins after surgery was much lower than what would be expected based on historical rates – 42% versus 71%. No delay was caused to surgery and treatment was well-tolerated, Upppaluri said.
Researchers concluded that using Keytruda in this setting has the potential to reduce the extent of surgery, risk of relapse and the need for PO-ACRT.
Reviewing the data at ASCO, University of Colorado's Jimeno described the potential for taming tumors and down-staging disease prior to surgery as "phenomenal." One risk, however, with PD-1 agents is that tumor flare can occur, though this is relatively rare in head and neck cancer.
He added that this is as "close as it gets in this setting to proof-of-concept" and that he is looking forward to seeing full pre- and post-surgical analyses in the future.
In a significant proportion of patients, the immune system was able to mount a response that appeared to be effective and it's possible that continuing treatment with immunotherapy longer – for two, three or more doses – would have had a robust clinical response, perhaps a complete response, Cohen commented to Scrip.
Jimeno also noted that there are some other investigator-led studies investigating immunotherapy in early-stage settings. David Neskey of the Medical University of South Carolina is leading a Phase II study (NCT03021993) of Opdivo as a neoadjuvant pre-surgical therapy in 19 patients with locally advanced oral cavity cancer that is due to complete in April 2019. The Netherlands Cancer Institute is exploring the combination of Yervoy and Opdivo as a neoadjuvant therapy prior to surgery in advanced or recurrent head and neck carcinoma in the IMCISION study (NCT03003637), which is due to complete in August 2019. Academic studies are being done with support of leading PD-1/L1 sponsors.
Bristol is also leading a Phase I/II, open label neoadjuvant study – CheckMate 358 – in virus-associated head and neck tumors. The company notes that developing treatment options in early treatment settings is a "key focus" and that neoadjuvant research is ongoing in a number of indications, including head and neck cancer and lung cancer.
Another small study presented at the ASCO meeting evaluated Keytruda in 27 patients with Stage III/IVb HSNCC with any HPV status and who were eligible for cisplatin-based chemotherapy. Patients received Keytruda at 200 mg 4-7 days before cisplatin based chemoradiation therapy and then along with chemo and following chemo.
Safety was the primary endpoint and investigators evaluated efficacy as the complete response rate at day 150 on imaging or during salvage surgery.
Almost 80% of patients were able to complete the dosing of Keytruda, Sanford Health researchers reported. The most common Grade 3 events were dysphagia, mucositis and dermatitis. There were also some Grade 4 adverse events – neutropenia in one patient and hypophosphatemia in one patient. There was one death, caused by anemia, but this was not deemed to be related to drug treatment. Three patients (11%) discontinued treatment due to adverse events.
Discussing results at the ASCO meeting, Jimeno noted that the upper GI side effects – dysphagia and mucositis – were substantial, but noted that chemo was given weekly, as opposed to every three weeks, in the trial, and that the hematological toxicity was par for the course. Dose intensity for chemoradiation therapy was on par with what they would aim to achieve as the standard of care, without Keytruda.
Investigators reported that 21 patients (78%) had a complete response (CR) at day 150. In those who were HPV-positive, the CR was 85% and it was 57% in those who were negative. Because this was a single arm study, it is not possible to tease out the added efficacy for including Keytruda, but at the very least the PD-1 inhibitor did not appear to detract from expected efficacy. Expansion cohorts of the trial in HPV-positive and HPV-negative disease are now up and running.
Jimeno concluded that adding Keytruda was feasible and well-tolerated.
The study raises questions about the sequencing of immunotherapy, the discussant said. PD-1/L1 inhibitors could be given before chemoradiation therapy (CRT) for priming, during CRT or after as a maintenance therapy, which would all have different effects, Jimeno said.
"The jury is still out as to the timing in head and neck cancer and ongoing studies will hopefully help us assess this," he noted.
Moving forward in the fast-paced field will require a better understanding of what is going on mechanistically with immunotherapy in cancer, better patient selection and effective use of combinations, Cohen said during the ASCO head and neck cancer session.
An analysis of Keytruda in the KEYNOTE-012 study presented at the ASCO meeting suggested that mutational load and the gene expression profile (GEP), which represents genes associated with predicted outcomes to PD-1 inhibitors, are helpful in identifying who will respond to treatment with PD-1 inhibitors.
Researchers reported that mutation load and GEP were significantly associated with response in the patients whose head and neck cancer was viral negative, that is not associated with a virus. The two markers were equally predictive of response and also independent predictors. Those who were viral negative and had both markers experienced the best response. In those who were viral positive, mutational load was not significantly associated with response, but GEP was.
Experience to date shows durable responses – going on for years – in some patients. This raises hope that immunotherapy alone in the curative-intent setting may cure some patients, Cohen said.
Understanding of the role of different targets in the immune system is growing with the release of clinical studies of IO/IO combinations.
Data for Incyte Corp.'s IDO inhibitor epacadostat with PD-1 inhibitors in a range of tumor types, including head and neck cancer, were presented at the ASCO meeting. (Also see "Merck & Co. Strengthening Its Strong Position In Immuno-Oncology" - Scrip, 13 Jun, 2017.)
Cohen sees IDO/PD-1 as the most interesting combination to emerge so far in head and neck cancer.
Tumors may escape the attention of the immune system by upregulating the enzyme IDO. Inhibiting IDO removes another brake on the immune system that may be complementary to inhibition of PD-1. (Also see "Scrip's Rough Guide To IDO" - Scrip, 18 May, 2017.)
In a cohort of 38 patients with head and neck cancer in the Phase I/II ECHO-202 study, the ORR was 34% overall, 39% in those who had 1-2 prior lines of therapy and 14% in patients who had three or more prior therapies. The combination was effective regardless of PD-L1 expression and whether the cancer was associated with the HPV virus, though results were stronger for HPV+ patients (46% ORR versus 29% ORR).
Investigators reported that responses were rapid, deep and durable – 10 of 13 were ongoing for a median of 18.4 weeks.
The rate of Grade 3 and 4 adverse events, which included fatigue, diarrhea and increased lipase, was 18%.
Investigators concluded that the Grade 3 and 4 event rate, numbers of discontinuations and adverse events of special interest for the combination were similar to what would be expected with Keytruda monotherapy.
Cohen commented that it is very encouraging that in addition to the response rate and depth of response, epacadostat did not seem to appreciably increase immune-related adverse events.
"I am sure as more data emerge they will form a greater picture. Remember, we are talking relatively of a small number of patients, but at least for now there isn't a danger signal that this combination cannot be taken forward," Cohen said.
Phase III Trials Of Checkpoint Inhibitors In SCCHN
||Start Date, Completion Date
|EAGLE (n=720): AstraZeneca's PD-L1 inhibitor Imfinzi with the investigational CTLA-4 inhibitor tremelimumab.
||Phase III randomized open label study of Imfinzi with or without tremelimumab vs. standard of care in recurrent or metastatic (R/M) SCHHN after failure of platinum-based chemo.
||Overall survival (OS)
||September 2015, February 2018
|KEYNOTE-048 (n=825): Merck's PD-1 inhibitor Keytruda
||Phase III randomized open label study first-line R/M SCHHN. Keytruda vs. Keytruda plus platinum-based chemo plus fluorouracil (5-FU) vs. active comparator (Erbitux plus platinum-based chemo plus 5-FU).
||Progression-free survival (PFS)
||March 2015, March 2018
|KESTREL (n=823): AstraZeneca's Imfinzi and tremelimumab
|Phase III randomized open label study of Imfinzi with or without tremelimumab vs. active comparator, the EXTREME regimen (Erbitux, 5-FU, cisplatin/carboplatin) in first-line R/M SCCHN.
|October 2015, March 2018
|CheckMate 651 (n=490): Bristol's PD-1 inhibitor Opdivo with its CTLA-4 inhibitor Yervoy (ipilimumab)
|Randomized open label study of Opdivo with Yervoy vs. EXTREME regimen in first-line R/M SCCHN.
|August 2016, January 2019
| JAVELIN Head and Neck 100 (n=640): Pfizer/Merck KGaA's Bavencio
|| Randomized double blind Phase III trial of Bavencio vs. placebo followed by maintenance therapy for up to 12 months in first-line treatment of locally advanced SCCHN.
|| November 2016, April 2021