Neoadjuvant Cemiplimab Elicits Responses and Surgical Benefit in HHI-Naive BCC of the Head and Neck

Hedgehog inhibitor (HHI) therapy has offered a systemic alternative to a potentially morbid surgery for patients with locally advanced basal cell carcinoma (BCC) of the head and neck, but tolerability issues frequently cut treatment short before patients get the full benefit. Investigators are now testing whether moving immunotherapy earlier in the treatment course, before HHI exposure, can change that calculus. In a phase 2 study (NCT05929664), neoadjuvant cemiplimab-rwlc (Libtayo) given to HHI-naive patients (n = 35) produced an objective response rate (ORR) of 60.1% (95% CI, 42.1%-76.1%) and allowed 42.9% (95% CI, 26.3%-60.6%) of patients to avoid sacrificing an at-risk organ altogether, according to study authors Hannah Kenny, MD, and Joseph M. Curry, MD.^1,2

“The rationale was if we’re seeing this signal in the HHI-naive population, maybe we can investigate it earlier on in the treatment regimen to target a greater number of patients who could see benefit in response from cemiplimab with hopefully lower levels of adverse effects [AEs] and better tolerability overall,” Kenny said in an interview with OncLive®. Kenny is a fourth-year resident in the Otolaryngology program at Sidney Kimmel Medical College, Thomas Jefferson University in Philadelphia, Pennsylvania.

“The idea is to capture [patients] earlier in the treatment algorithm, so that you can hopefully provide them with more benefit by going straight for immunotherapy and also provide them with a therapy that’s generally very well tolerated,” Curry added. Curry is a clinician at Jefferson Health in Philadelphia, Pennsylvania.

In the interview, Kenny and Curry discussed the rationale for evaluating cemiplimab ahead of HHI therapy, the trial’s response-adaptive design, the efficacy and safety data observed, as well as the concept of surgical benefit rate as an emerging clinician-reported end point.

OncLive: Why did you choose to investigate cemiplimab in the HHI-naive setting?

Kenny: Locally advanced BCC is a difficult clinical disease [because] there’s oftentimes significant morbidity that comes along with local treatment, including surgery and radiation. HHI therapy has been used as a systemic therapy to help treat this disease, and it does have some rate of efficacy, usually between 42% and 60% in this population. However, there are often difficulties with tolerability and duration of treatment. There has been a slight response in patients who have undergone HHI therapy and have to discontinue for any reason, including resistance, and they have [approximately] a 29% response rate to cemiplimab after that resistance is noted with HHI therapy.

The trial uses a response-adaptive design, reassessing patients after two treatment cycles. What was the rationale for that, and how did you ultimately define the threshold between responders and non-responders?

Kenny: The idea was that by having this adaptive system or schema, you are utilizing more treatment or seeing if you can get a further depth of response for patients who are responding, while avoiding further progression or stagnant response in patients who don’t seem to be responding to the treatment. So, by following an adaptive protocol, it allows the treatment to be tailored to the patient and to hone surgical decision-making about when to proceed to surgery or another standard of care. It allowed some flexibility to tailor based on the patient’s individual response.

Curry: There are a lot of unanswered questions in this space with neoadjuvant therapy in cancers that are operable. The question is often, how long is it safe to treat them for, and what risks are you presenting them with? We know in, for example, head neck cancers that after six or eight weeks, they’re likely to [develop progression], so your real clinical window is in that six- or eight-week time point which would be usually two doses or two cycles of immunotherapy. We wanted to be sure we evaluated them there but also tried to optimize their benefit, especially in this different disease. Basal cell [cancers are] typically more slow growing, so you may have a bigger window. But if you evaluate them after two cycles and you haven’t seen progression, then you have an opportunity to potentially deepen their response by continuing, and that was one of the things we were interested in looking at.

What was the ORR, and how did pathologic responses compare with radiographic responses?

Kenny: The two primary end points of the study were ORR and disease control rate [DCR]. The ORR was 60.1%, and the DCR was 91.4%. There was a large percentage of patients who were able to achieve response or stable disease, and we were able to meet our prespecified primary end points with those levels of response. In terms of correlations with radiographic response and pathologic tumor response, those analyses are preliminary, but what we’ve seen in our preliminary analysis is that there is a correlation between RECIST response overall. We looked at clinical caliper measurements, in addition to radiographic response, and found that about 40% of patients who had RECIST response had either a major or complete pathologic response compared with 0% of non-responders. The average pathologic tumor response was 70% in the responder group vs 10% in the non-responder group based on RECIST criteria.

Curry: It’s an interesting population and a challenging population because not all basal cells are the same. Some of them are poorly defined or hard to see on imaging, despite being relatively large in terms of surface area. Some are just small, and we were particularly targeting patients with high-risk tumors that were in critical anatomic locations to assess the benefit. We got imaging on everyone, but we couldn’t always rely on the imaging to give us the most accurate measurements. We used RECIST caliper measurements to track it, and that helped us really correlate with response. The imaging alone isn’t always adequate to capture the significance of the disease in some of these patients.

What about the durability of the responses that were seen?

Curry: It’s a little early to tell, but we will follow these patients out. Generally, what we’ve seen is that patients who have followed our recommendations, meaning they completed the neoadjuvant therapy and then underwent surgical resection, are doing very well, and we’ll [show] that in the forthcoming publication. We do have, I believe, two patients that we’re aware of that have [developed] recurrences. Each of those patients were patients who elected not to proceed to surgery. They chose to go another route and opt out of the planned course of treatment.

In terms of the secondary end point of quality of life (QOL), what do we know so far about patients’ experiences going through neoadjuvant immunotherapy?

Curry: In general, patients tolerated this treatment very well. The AEs were in line with what you would expect, so from that perspective, the neoadjuvant course was tolerated well. We didn’t have a lot of patients who, for example, had to exit [the study] early due to AEs. If you look at this study compared with some other studies, [we had] a longer neoadjuvant period than a lot of the preoperative studies because of the duration of therapy. We also had a number of patients who did have some surgical benefit associated with the neoadjuvant therapy, meaning that we were able to de-escalate their surgery. Another open question is: Is it safe and is it possible [to de-escalate surgery]? There’s not a lot of great long-term data on that, so that is one of the things we’re studying.

In terms of QOL, overall, like the findings suggest, this is a favorable clinical algorithm to proceed with, but we’ll need to see long-term follow-up and outcomes. But for the short-term surgical outcomes, it does look favorable.

You mentioned surgical benefit. How was this measured and what is its significance in this context?

Kenny: Something that’s unique about the study was the surgical benefit rate. In order for patients to qualify for this treatment, they had to have an identified organ at risk, whether that be the nose, the eyelid, the orbit, the ear, or the lip, and the plan for surgery was specified prior to treatment based on what the patient was presenting with at that time. After they underwent treatment, they underwent another assessment for what their surgery was going to look like, and their actual surgery was compared with that pretreatment plan to determine whether they achieved a benefit or avoided sacrifice of that organ that was determined to be at risk prior to treatment. We saw that 42.9% of patients were able to avoid some form of organ sacrifice based on this treatment. That was a secondary end point, but a meaningful outcome for many of our patients.

Curry: This is one of the things we’re thinking about in terms of changing practice and making the future better for our patients. We have some evidence that we have a good starting point. I’d like to see us push that forward either through combination therapies or through other methods to augment [the responses we’re seeing]. We also really need to show that we can do that safely, so we need to follow those patients that had de-escalated surgery and make sure that we’re also measuring this in the best way possible. We came up with this concept of surgical benefit rate, and other trials have measured things like this, but we do need a formal end point. In a way, it’s not really a patient-reported outcome, but it’s a clinician-reported outcome, and it’s a patient-relevant outcome that we show that we are helping patients by de-escalating the therapy that they need.

Do you expect the FDA to weigh in on surgical benefit rate if that end point does become more widely accepted?

Curry: We would love to get there. That’s one of the things we’re talking about now. How can we promote this because it’s something that is relevant to surgeons. It’s relevant beyond the current study. It’s relevant beyond this disease. Every patient that has a cancer that is potentially operable would be interested in less surgery or less morbid surgery. A lot of interest is out there, especially with the favorable responses to immunotherapy in terms of eliminating surgery. But I just don’t think that that’s realistic in all cases. Instead, optimizing the tools that we have at hand by decreasing the morbidity and the AEs of a systemic therapy or a surgery are where we need to be to get patients back to the lives that they had before and optimize QOL.

What specific AEs were seen?

Kenny: Across the cohort of 35 patients, 35.3% of patients experienced any treatment-related AEs, the majority of which were grade 1/2. Most commonly we saw pruritus and fatigue. We did have one grade 3 myalgia that led to treatment discontinuation, but that was the only treatment discontinuation in the setting of an AE. Overall, the safety profile was tolerable and favorable compared with standard of care, and that was one of the benefits that we have seen so far with this treatment.

You’re also collecting pre- and post-treatment tumor specimens in peripheral blood. What changes in the immune microenvironment have you seen so far, and are there any baseline biomarkers emerging that could be potentially predictive of response?

Kenny: We haven’t run those analyses yet, but those analyses will be forthcoming. It’s part of our plan, and we plan to evaluate both pre and post treatment tumor specimens to analyze those correlations and hopefully better prognosticate response to determine who could be the best candidates for this therapy going forward.

Curry: The best way to handle some of these analyses is to batch them at the end to avoid batch effects, so we frequently will send them all out at once. We’re in the process of analyzing them currently, but some of the things that we look for, for example, are changes in the immune elements that are within the tumor, and then in the peripheral blood, also changes in their overall immune status. A number of studies and algorithms have been proposed to measure this type of thing, and we’ll be applying those as we go forward.

What are the next steps toward bringing this approach into the clinic for patients?

Curry: This approach has a lot of promise. It is well tolerated, so the question comes down to long-term benefit. It would be great to see a bigger trial where we could generate more data, maybe head-to-head, to see how we help patients. One question is, could you get a bigger response faster with HHIs? Their time to response is quick, but is that benefit durable in terms of generating an immune response and generating a longer-term benefit for the patient’s antitumor response? We believe there’s a subset of patients who would benefit from this, and we’d like to see this move forward into a trial that would hopefully change clinical practice, and we think there’s enough evidence here to generate further study.

Disclosures: Kenny had no relevant disclosures and Curry disclosed research funding from Regeneron.

References

1. Kenny HL, Mastrolonardo E, Linnenbach A, et al. A phase II study of neoadjuvant cemiplimab in hedgehog inhibitor (HHI)–naive basal cell carcinoma of the head and neck. J Clin Oncol. 2026;44(suppl 16):9515. doi:10.1200/JCO.2026.44.16_suppl.9515
2. Cemiplimab plus fianlimab for the treatment of locally advanced head and neck basal cell carcinoma before surgery. ClinicalTrials.gov. Updated June 10, 2026. Accessed July 15, 2026. https://clinicaltrials.gov/study/NCT05929664