Toxicities From Lenvatinib/Pembrolizumab Associated With Treatment Response in Gynecologic Malignancies

The number and severity of toxicities experienced by patients with gynecologic malignancies who received treatment with lenvatinib (Lenvima) and pembrolizumab (Keytruda) were shown to be associated with response, highlighting the importance of early intervention and management, according to data from a retrospective, single-center study presented at the 2024 Society of Gynecologic Oncology Winter Meeting.

In an evaluation of 88 patients treated with at least one cycle of lenvatinib and pembrolizumab from 2016 to 2023, investigators found that patients were more likely to respond to treatment if they experienced treatment-related toxicities. Notably, 79% of patients who experienced a complete or partial response experienced at least 3 toxicities. Moreover, these patients were also found to experience more intense toxicities at a grade of 3 or higher (32%) vs those with stable (24%) or progressive (13%) disease (P =.032).

“When we looked at clinical outcomes, the overall response rate [ORR] was significantly associated with [the] number of treatment-related toxicities, as patients with more toxicity were more likely to have a response to treatment,” Paulina J. Haight, MD, lead study author, said in an interview with OncLive®.

In the interview, Haight further discussed findings from the single-institution analysis, elucidating the need for further research into the association between the occurrence of toxicities with response to treatment with lenvatinib and pembrolizumab. Haight is an obstetrician-gynecologist at the Ohio State College of Medicine in Columbus.

OncLive: What is known about predictors of toxicity in gynecologic malignancies? How did this provide headway for your research?

Haight: Immunotherapy has become a very important part of the treatment paradigm for gynecologic malignancies, particularly for patients [with] endometrial cancer, as well as patients [with] cervical cancer and sometimes ovarian cancer, as well. There’s a lot of newer data that has shown a significant improvement in survival when we either utilize immunotherapy alone or add it in combination to chemotherapy or other targeted therapies. However, immunotherapy comes with a unique toxicity profile that results as the body’s anti-tumor response.

Unfortunately, adverse effects [AEs] are common with immunotherapy, as up to two thirds of patients can experience treatment-related toxicity, with approximately 15% of those being severe at grade 3 to 4. On average, approximately 5% of patients ultimately discontinue treatment related to toxicity. There’s been some interesting data that has been published recently in the realm of melanoma and lung cancer that has reported toxicity associated with immunotherapy is related to response rate and survival.

There’s a thought that not only the number of toxicities that a patient may experience, but also the severity of toxicity that a patient experiences, is associated with improved response rates and prolonged survival. Therefore, if toxicity may potentially be a biomarker for response to treatment, it’s going to be important that we do a good job of identifying that early and managing those toxicities so that these patients who experience them, can remain on treatment because we know that they’re the ones that have a higher likelihood of responding and doing well with that treatment.

There’s not a lot of data that looks at toxicity and its association with response in terms of immunotherapy that’s used in combination with other therapies. We were particularly interested in looking at patients who’ve received lenvatinib and pembrolizumab.

The rationale behind our study was to look at all our patients with gynecologic malignancies who have received treatment with lenvatinib and pembrolizumab, looking at their toxicity profile and correlating that with response.

What methods were used to conduct the study?

This was a single-institution review and is a retrospective study. We looked at all our patients with a gynecologic malignancy who had received at least one cycle of lenvatinib and pembrolizumab between the years 2016 and 2023. We ultimately ended up including 88 patients in our study.

Could you expand on how the toxicities for each patient were identified?

For the patients that we included in the study, we went through records from every treatment visit, and looked at their toxicity profile. If a grade of toxicity was not reported, the common terminology criteria for AEs was retroactively applied. We looked at the association between both the number of toxicities a patient experienced and the severity/the grade of toxicity.

We also looked at associations between not only demographics and clinical pathologic information, but clinical outcomes as well, including response and duration of response [DOR].

What were some of the key results that were identified from this investigation?

We included 88 patients in the study. Most of them had endometrial cancer–88%–and 11% had ovarian or primary peritoneal cancer. Most of the patients had either serous [32%], endometrioid [26%], or carcinosarcoma [17%] histology and 87% of tumors were mismatch repair proficient. We did have a 7% rate of mismatch repair–deficient tumors and 6% were unknown.

We grouped patients according to the number of toxicities that they experienced; 15% of patients experienced no toxicity; 30% of the cohort experienced 1 to 2 toxicities; 32% experienced 3 to 4 toxicities; and 24% experienced 5 or more toxicities.

The first thing that we did was look at patient demographics and clinical pathologic information. We wanted to see if there was any sort of an association between the number of toxicities or the severity of toxicities and those features. We didn’t find that toxicity was associated with demographic or clinical pathologic information other than the neutrophil to lymphocyte ratio. That’s something that has been studied and is still being studied in our patients with cancer in terms of whether this is a predictive biomarker for response. There’s been data from our institution previously that’s reported that a neutrophil to lymphocyte ratio of less than 6 has been a good prognostic marker, but not necessarily associated with response rates to immunotherapy.

We did find that the neutrophil to lymphocyte ratio being low, or less than 6, was associated with a higher number of toxicities experienced by patients.

Notably, the patients who had no toxicity had a response rate of 11%. Patients who had 5 or more toxicities had a response rate of 62%. The patients who did respond to treatment tended to experience more intense toxicity. Thirty-two percent of patients who responded had a grade 3 or 4 toxicity compared with those with stable disease, as only 24% had a grade 3 to 4 toxicity, and those with progressive disease had a 13% rate of grade 3 to 4 toxicity.

Still looking at those patients who responded, the DOR was significantly longer for patients who experienced more toxicity. For those who had 5 or more toxicities, the DOR was 18 months vs a median of 9 months for those who had fewer than 5 toxicities.

The last important thing to note is that patients who were on treatment for a longer period or received more cycles of treatment did experience a higher amount of toxicity. The median time to toxicity was 21 days across all patients and that indicates that toxicity is not just a result of longer duration on treatment.

Do any of these findings come as a surprise?

I’m not necessarily surprised by the results, especially based on some of the data that has been published in other solid tumor types like melanoma and lung cancer. I do think that this certainly adds to the body of literature and adds evidence that toxicity does seem to be an important predictive biomarker of response for patients.

It’s important that we be aware of that and keep an eye out for those patients who experience toxicity so that we can counsel them appropriately and identify them early, as well as manage them well, so that we can keep patients on treatment for longer.

Although this evaluation was specific to lenvatinib and pembrolizumab, could the results be extrapolated to other checkpoint inhibitors in gynecologic malignancies?

It’s very likely that these results will be able to be extrapolated to different forms of immunotherapy based on the mechanism of action of a lot of the immunotherapies that we utilize. That being said, I believe that we need to do a bit more work looking at alternative immunotherapy regimens, [as well as] immunotherapy as monotherapy for the treatment of patients with gynecologic malignancies, as that has not previously been reported.

What are the next steps for this research and what would you like your colleagues to take away from this investigation?

We decided to first look at patients with lenvatinib and pembrolizumab. Our next step is going to be expanding this cohort to all patients who have received immunotherapy within our division. That would include patients who have received pembrolizumab [Keytruda] as monotherapy, or in combination with other therapies or cytotoxic chemotherapy, as well as other immunotherapy regimens, including agents like durvalumab [Imfinzi] and dostarlimab-gxly [Jemperli].

One of the biggest takeaways from the study is that as gynecologic oncologists are utilizing immunotherapy a lot more frequently, our awareness and understanding of toxicities associated with these regimens [continues to be] important because the patients who experience significant toxicity are the ones that are more likely to achieve significant benefit from these regimens.

Reference

Haight PJ, Brown MD, Espelien BM, et al. Predictors of toxicity and response to treatment with lenvatinib and pembrolizumab in patients with gynecologic malignancy. Presented at: Society of Gynecologic Oncology Winter Meeting; January 25-27, 2024; Olympic Valley, CA. Poster 509234.