New Standard of Care? KEYNOTE-689 Results in LA-HNSCC

Ravindra Uppaluri, MD, PhD

Pembrolizumab (Keytruda) in combination with standard of care (SOC) resulted in a statistically significant improvement in event-free survival (EFS) when used for the treatment of newly diagnosed, resectable, locally advanced head and neck squamous cell carcinoma (LA-HNSCC), according to findings from the KEYNOTE-689 (NCT03765918) trial presented at the 2025 AACR Annual Meeting.

Historically, patients with LA-HNSCC have been treated with either upfront surgery followed by adjuvant therapy or definitive (chemo)radiotherapy. However, outcomes have been suboptimal, and patients suffer from adverse effect toxicities.

KEYNOTE-689 was a randomized trial comparing the SOC with the integration of pembrolizumab, given as neoadjuvant and adjuvant. The trial demonstrated a statistically significant improvement in EFS with the addition of pembrolizumab, with the median EFS being 51.8 months in the pembrolizumab arm vs 30.4 months in the SOC arm.

Additionally, the study showed significant pathological responses in the primary tumor and lymph nodes after neoadjuvant pembrolizumab. While overall survival was not statistically significant at the first interim analysis, follow-up is ongoing. Moreover, the adverse event profile of pembrolizumab and SOC was, in general, similar to SOC alone, and generally well managed.

These results suggest that integrating pembrolizumab into the treatment paradigm for resectable LA-HNSCC represents a significant advancement over the current SOC.

“The current standard of care was established over 20 years ago which, despite multimodality treatment, resulted in suboptimal outcomes. The KEYNOTE-689 results represent a dramatic improvement over the current standard of care. I think it is an exciting time for patients, and all providers who take care of patients with head and neck cancer,” Ravindra Uppaluri, MD, PhD, told Targeted Oncology^TM, in an interview.

In the interview, Uppaluri, chief of head and neck surgery at the Dana-Farber Brigham Cancer Center, and chief of the otolaryngology group at the Brigham and Women's Hospital, further discussed these data from the KEYNOTE-689 presented at this year’s AACR meeting.

Targeted Oncology^TM: Can you provide some background regarding locally advanced head and neck cancer?

Uppaluri: Newly diagnosed patients with stage III or IV locally advanced head and neck cancer are either treated with upfront surgery, followed by adjuvant therapy, or, nonsurgically, with definitive radiation, without or with chemotherapy. The adjuvant treatment after surgery is dictated by the surgical pathology findings from the resection specimens taken at the time of tumor surgery.

Patients who are considered low risk, those who have primary tumor clear margins and/or have tumor-involved lymph nodes without extranodal extension, are treated with adjuvant radiation after surgery. But if a patient has high-risk features, such as positive margins or extranodal extension, then those patients receive adjuvant chemoradiation after finishing surgery. It's a long process, and this paradigm has been in place for nearly 2 decades now, with the last studies done in the early 2000s. Given this static nature, we have hit a roadblock of improving patient outcomes, and remember these treatments are quite toxic.

It has been an exciting time in oncology, in general, because of immune checkpoint blockade therapies, and this has resulted in a new standard in recurrent/metastatic head and neck cancer. However, combining immune checkpoint blockade with chemoradiation in locally advanced head and neck cancer in the upfront setting has been negative in several phase 3 clinical trials.

With this context and preclinical findings, Douglas Adkins, MD, who is a medical oncologist at [Washington University in St. Louis] and I started early phase studies looking at integrating immune checkpoint inhibitors into surgical management of head and neck cancer patients. Our findings, and also [those] from other studies, were favorable and provided the impetus for the development of KEYNOTE-689.

What did the trial design of KEYNOTE-689 entail?

The trial was a randomized trial with over 700 patients with newly diagnosed locally advanced stage III/IV head and neck cancer, who were planned for surgery and then randomized to either standard of care or integrating pembrolizumab, neoadjuvant—that is, before surgery—and after surgery, along with standard of care, and comparing those 2 arms. It is a really novel trial design in head and neck cancer compared with all the other trials that have been done because of this perioperative component.

I want to highlight that it really requires multidisciplinary teams to make this happen. It is not just single-specialty folks saying we are going to treat cancer patients in a given way. It requires a multidisciplinary group of people to come together to implement this kind of workflow. The primary end point was event-free survival, which is defined as any disease progression or recurrence or death after randomization in the trial.

There were some secondary end points also. The neoadjuvant approach will alter the tumor microenvironment because it will activate the immune system. Studying the immune changes in the tumor, which is called pathologic response, was a secondary end point. An additional secondary end point was overall survival. It is clear from other studies that event-free survival is linked to overall survival, and this is being studied as a secondary end point.

What key findings were presented at AACR?

The key findings that were presented included the endpoints as described above and additional associations that were observed. The statistical design used a sequential strategy, where patient populations were analyzed based on the tumor combined positive score [CPS].

CPS is a commonly used biomarker for describing PD-L1 expression of tumors and the tumor environment, and a CPS of 10 or greater was the first analysis. For the CPS 10 or greater population, EFS was significantly improved in the pembrolizumab with standard of care versus the standard of care arm.

Sequential analysis showed that EFS improved for participants with a CPS of 1 or greater and for all participants in the pembrolizumab with standard of care group compared with the standard of care. For all participants, the median event-free survival across all patients was 51.8 months in the pembrolizumab arm vs 30.4 months after a median of 38.3 months of follow-up. This is highly significant in terms of event-free survival in these patients and was really the most exciting part of the study.

How do these results compare to previous standard-of-care treatments for this patient population?

The previous standard of care was established over 20 years ago, and despite this multimodality treatment, the outcomes were not optimal. KEYNOTE-689 findings dramatically improve over the standard of care. I think it is an exciting time for patients, providers, and all who take care of patients with head and neck cancer.

What findings from the secondary end points of the study were presented?

As I mentioned, there were 2 key secondary end points. One was the major pathological response. We expected that by giving an immune checkpoint blockade before surgery, we would observe pathological responses.

In a blinded, independent pathologic review, neoadjuvant pembrolizumab significantly induced pathologic responses in the primary tumor and lymph nodes compared with the standard of care.

Another secondary end point was overall survival, which on this first interim analysis was not significant. We do see a separation of curves favoring the pembrolizumab plus standard of care arm but at this point, longer follow-up is needed to look at these patients.

Were there any significant adverse events observed in the study, and how were they managed?

In terms of adverse events, both arms of the study had similar general rates. The most common adverse effects were related to the chemoradiation. As expected, in the pembrolizumab plus standard of care arm, there were more immune-mediated events, but there were really no new adverse effects that were observed, and these were generally well managed.

How do you see the findings from the study influencing the development of future trials or treatment strategies in this space?

I think that this trial is exciting for our patients and providers together and sets forth what we think will be a new standard of care for these patients. In terms of future studies, I think that seeing the real-world application is going to be the next exciting phase: that is how this approach is going to be integrated into the current standard of care. This will be something important to observe and also support patients and providers on, because it is a change to our current workflow.

The second aspect of what KEYNOTE-689 has done is to open the window into thinking about these approaches in the surgical setting. We now know this is a safe approach, and the majority of patients in both arms actually went on to surgery. There were really no concerns in that regard. Given that, I think there will be newer investigations looking at additional therapeutics that could be integrated into this type of clinical trial design.

In your opinion, what are the most important unanswered questions or areas of future research?

One key limitation to the study—note, this was designed nearly 7 years ago now—was that we had components of delivering pembrolizumab, both before surgery and after surgery. So which component actually led to the better clinical outcomes is unclear. Maybe giving the drug both before surgery and after surgery is important or one component is more important. I think that is a key question that needs to be answered in the future. It will require a lot more patients and a lot more studies, but I do think that's an important key question.

A second area which I think will be explored is integration of additional checkpoint therapies or other therapeutics combined with pembrolizumab, and we would expect the field to move towards these types of studies.

What else is important for an oncologist to know about this study and these findings?

I think there are 2 points I would like to make. First, in general, there's some hesitation by some providers including surgeons, regarding this idea that immunotherapy is delivered before surgery, and that this may lead to tumor progression or unresectable tumors. But KEYNOTE-689 findings are really reassuring in that sense. Nearly 90% of patients in both arms went on to surgery, and in the pembrolizumab arm, there was only 1 tumor found to be unresectable in nearly 300 patients. I think that kind of data, in terms of the safety of this approach, is very reassuring.

The other thing I would emphasize is this idea that—and I think it has been talked about many times—but locally advanced head and neck cancer care really requires multidisciplinary integration. The idea that only one specialty sees a patient and manages that patient is not ideal for locally advanced head and neck cancer. Moving forward, the KEYNOTE 689 new standard of care will clearly be beneficial for patients and really requires care team integration and discussion upfront in getting these therapeutics to patients and getting them to surgery.

Other than this study, are there any other research or trials in the space that have caught your eye?

One of the other studies in this field that is coming up is an adjuvant immunotherapy study from the GORTEC group in Europe, where they have tested adjuvant immunotherapy in a trial called NIVOPOSTOP [NCT03576417], and that study is to be presented at [the 2025 American Society of Clinical Oncology (ASCO) Annual Meeting]. We are all excited to see the results of that study. These studies address this unmet need in these surgically treated patients, and we are all excited about the next era for patients with head and neck cancer with these approaches integrating immunotherapy.