Concept illustration of skin cancer.png
At the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting, significant attention was drawn to the evolving role of immunotherapy in cutaneous squamous cell carcinoma (CSCC), a nonmelanoma skin cancer affecting a growing number of patients. Among the headline studies was MK-3475-630/KEYNOTE-630 trial (NCT03833167), a phase 3 trial evaluating the efficacy of adjuvant pembrolizumab (Keytruda) following surgery and radiation in patients with high-risk CSCC.1,2
The trial’s primary end point was event-free survival (EFS), defined as the time to recurrence or death. KEYNOTE-630 did not meet this primary end point at its first interim analysis. However, a notable reduction in recurrence was observed, including a 2-year recurrence rate in the placebo group of 28% vs 14% in the pembrolizumab group. Additionally, there were 22 deaths with pembrolizumab vs 9 deaths with placebo.
Importantly, there were no grade 5 toxicities, indicating that the deaths were not directly attributed to treatment. The cause remains unclear but may relate to trial timing, COVID-19 complications, or other unknown factors.
In an interview with Targeted Oncology, Shlomo Koyfman, MD, a radiation oncologist at Cleveland Clinic, provided a comprehensive overview of the trial, its results, and its implications for future standards of care.
Targeted Oncology: Can you summarize this trial, the design, and the high-risk patient population that it focused on?
Shlomo Koyfman, MD: Most cutaneous squamous cell cancer skin cancers are super curable with just surgery, but there’s a small amount, maybe 10% of them, that have a higher risk of developing bad outcomes, either a recurrence or death. The highest risk population is those patients with very advanced disease, big T4 tumors, nodal disease, or a multiply recurrent T2 or T3 tumor with a lot of perineural invasion, or it is large, or it is deep, or things like that.
This trial, the KEYNOTE-630 trial, included all high-risk patients, all very high-risk patients with either big T4 disease, big nodal disease with multiple nodes or extraneural extension, or patients with T3 tumors with multiple risk factors. Essentially the standard of care for those patients is surgery and postoperative radiation. What this trial did was randomize patients after they completed surgery and postop[erative radiotherapy], randomized them to adjuvant pembrolizumab vs placebo, and that was done for 9 cycles, and the goal was to reduce recurrence and cancer related death and to improve the event-free survival.
What were the key efficacy outcomes, and how significant was the benefit with adjuvant pembrolizumab?
The study did not meet its primary end point. The primary end point was an approval of event-free survival with particular confidence intervals and statistical significance. The study did not meet that end point, unfortunately. However, when you look at the actual data, the recurrence rate was considerably lowered. The overall recurrence rate for patients in 2 years went from 28% for the placebo arm down to about 14% for the pembrolizumab arm. It reduced recurrence by 50%.
However, patients in the pembrolizumab arm had an increased risk of death, an increased episode of death. There were more deaths in the patients in the immunotherapy arm, which was unexpected. Now, what is weird about it is there were no grade 5 toxicities in the study, which means there were no actual deaths related to pembrolizumab. There was an increased number of deaths in the pembrolizumab arm. We do not know why that is, but because there were more deaths, the primary end point, which was event-free survival, included any recurrence or death as an end point. Even though the recurrences were much better and much lower, 50% lower with the immunotherapy, the death rate was higher. There were 22 deaths in the pembrolizumab arm, vs only 9 deaths in that arm. That essentially made it so that the trial did not meet its end point of improvement, event-free survival, and the trial was actually closed.
How do these results compare with previous standards of care or observation following surgery or radiation in this population?
If you look at the recurrence rates, it compares very favorably, and the curves separate with a statistically significant difference. It didn’t meet the prespecified P value. But certainly, the recurrences have improved. However, overall, the event-free survival was not improved, and therefore it was overall a negative trial. Still, patients in the pembrolizumab arm, at least from a recurrence standpoint, did better.
It is important to note that at the same ASCO session, presented at the same oral presentation, there will be presented the C-POST data. The C-POST was a similar trial with similar eligibility criteria designed by Regeneron using cemiplimab [Libtayo], and it was similarly structured—after surgery and radiation, adjuvant cemiplimab vs placebo. That trial was positive. There was a substantial reduction in events. The event-free survival was improved, and that did meet statistical significance.
Looking at these 2 trials, the cemiplimab C-POST trial being positive, and the KEYNOTE-630 trial not meeting its primary event free survival end point but significantly reducing recurrence, I believe that adjuvant immunotherapy compares favorably to the standard surgery and radiation, and, hopefully, there will be an FDA approval for cemiplimab in this disease.
Looking at both of those trials, are there any remaining questions about longer-term benefit or patient selection?
There are some unanswered questions. First of all, in the KEYNOTE-630 trial, why were there those unexplained deaths in the pembrolizumab? We don’t know. The trial did occur during [the] COVID [pandemic], and that complicated things. It was an international trial during COVID, which was complex. Similarly, in the KEYNOTE-630 trial, you were able to go on trial within 16 weeks of radiation, and that’s a long time—4 months between radiation and starting immunotherapy. If you look at the placebo arm, patients who did enroll, there were quite a number of people who actually recurred within the first few months. There are people that rapidly recur, and there were quite a number of people that, when we screened them, within that first 4 month period, they did not meet eligibility because they had already recurred. One of the challenges and the criticisms of the KEYNOTE-630 trial is that it was too long—4 months was too long a window to allow people to get the adjuvant immunotherapy. It is possible that if you are trying to give adjuvant immunotherapy to these patients, getting it in within the first couple of months is important. That is one unanswered question.
The other is, will this translate to an ultimate overall survival benefit? As of yet, there’s no overall survival benefit in the KEYNOTE-630 trial. I do not know about the C-POST trial—we have to see. It may translate to an overall survival benefit. However, in the KEYNOTE-630 trial, there was a crossover arm, so a large number of patients in the placebo arm who recurred ended up getting PD-1 therapy—mostly pembrolizumab, some cemiplimab—and of course, many of them had terrific responses. That is why the overall survival may not show, because a lot of these patients ended up getting immunotherapy anyway, because it is such an effective therapy. Overall, I do believe that certainly the aggregate of these 2 trials is that it is an effective therapy, it dramatically reduces recurrence, and in the right patient population, where we are not worried about excessive morbidity, I think my hope is that it is ingrained in our future standards of care.
What are the broader takeaways here about the evolving role of immunotherapy in nonmelanoma skin cancers?
There is a lot of excitement about neoadjuvant therapy for skin cancer. A very exciting trial done with cemiplimab showed exciting results with neoadjuvant [therapy]. In 79 patients, there was a 50% pathologic complete response rate. A confirmatory, large phase 3 trial is currently underway and open—it’s the NRG Oncology HN014 trial led by Neil Gross, MD, FACS. It is an exciting trial. We will be opening it along with many other centers in the US and Australia, and that will look at neoadjuvant immunotherapy vs the standard surgery-first approach. We are all optimistic that that will potentially lead to benefit and a new standard of care for patients who do get neoadjuvant immunotherapy.
I think the role of adjuvant immunotherapy is totally up in question, but for patients who get surgery first and adjuvant radiation, my hope is that—looking at the C-POST trial and the KEYNOTE-630 trial—that adjuvant immunotherapy for surgery-first patients becomes embedded in our future standards of care.