Exceptional Responder to Pembrolizumab Guides Immunotherapy Progress in Leiomyosarcoma

Suzanne George, MD, assistant professor of Medicine, Harvard Medical School, clinical director, Center for Sarcoma and Bone Oncology, Dana-Farber Cancer Institute

Suzanne George, MD

Researchers at Dana-Farber Cancer Institute identified a treatment-naïve patient with metastatic uterine leiomyosarcoma who had a complete tumor remission for more than 2 years on pembrolizumab (Keytruda).^1,2

A complete pathologic response to pembrolizumab was observed at all except 1 of the patient’s metastatic uterine leiomyosarcoma sites. To explore potential mechanisms of immunotherapy sensitivity and resistance, the researchers examined samples from the primary tumor, the sole treatment-resistant metastasis, and germline tissue.

In the resistant tumor, PD-1 cell infiltration significantly decreased (P = .039). Additionally, the resistant tumor uniquely harbored biallelic PTEN loss and had lower expression of 2 neoantigens that showed strong immunoreactivity with patient T cells in vitro.

OncLive: Please describe your research.

In an interview with OncLive, lead study author Suzanne George, MD, assistant professor of Medicine, Harvard Medical School, clinical director, Center for Sarcoma and Bone Oncology, Dana-Farber Cancer Institute, discussed the exceptional responder to pembrolizumab and the significance of these findings for further immunotherapy advances in leiomyosarcoma.George: The research that we recently published is a collaborative effort between the sarcoma and genomics team at Dana-Farber. We described a patient with an exceptional response to the PD-1 inhibitor pembrolizumab. This is a patient who received pembrolizumab as first-line therapy for metastatic recurrent uterine leiomyosarcoma. This patient had a dramatic response to the therapy in multiple sites of disease with the exception of a single site.

Is it unusual to have a resistant tumor left over that you can study?

We were ultimately able to resect that resistance site of disease. We gathered tissue samples from her baseline tumor, which had a dramatic response, and the resistant tumor to attempt to determine the factors for resistance. With differences between the tumors at those 2 time points that might help us understand why the tumor became resistant. It was quite a unique opportunity to have tissue sample from both the baseline and a resistant tumor. It also gave us the opportunity to study this more fully because we had the samples over time. Some of the information that we've learned from this experience is just highlighting the importance of biopsies and tissue acquisition in patients on treatments over time.

The specific work that we've done related to this case was looking at whole-exome sequencing, meaning looking at the gene profile of the tumor at baseline and compared to the resistant tumor. We also focused on changes of neoantigen expression, which we believe may be relevant in sensitivity and ability for responses to be seen in PD-1 antibodies.

Entering this research would you have expected a patient with this diagnosis to respond to PD-1 inhibitors?

The entire field of immunotherapy has dramatically changed the landscape of oncology in many areas but we know that there is a tremendous amount of work to be done to identify which patients are sensitive to these approaches and which are resistant. When we started, this patient began with pembrolizumab. We were unsure if sarcoma and specifically leiomyosarcoma would respond to this approach. This research occurred in the very early stages of using PD-1 antibodies in soft-tissue sarcoma and uterine leiomyosarcoma.

Have you learned anything from this exceptional, specific case that could be applied to the general patient population?

What we have since learned is that this case was an exceptional one. There have since been other studies that have been presented using pembrolizumab and nivolumab (Opdivo), that have shown almost no responses in leiomyosarcoma. Having that background further heightens the exceptional nature of this patient’s response. In my opinion, this work is very hypothesis generating. We've shown that the genomic landscape is modulated over time in leiomyosarcoma. We believe that the neoantigen expression is important in immunotherapy and is also modulated over time in exposure to these compounds.

Was the tumor responsive right from the start of treatment?

How many of the tumors were you able to sample?

It's been very exciting for the patient’s family and our team. It emphasizes our need to continue to learn and understand why this case has been so exceptional. Hopefully, we can take that knowledge and apply it to other cases.It was. During the first scan assessment, we could see a dramatic response in the bulk of the tumor in one part of the abdomen. It shrunk remarkably. There was a small lesion in another part of the abdomen that increased slightly. Over time, it was clear that as the bulk tumor continued to shrink to become almost non-visible on imaging, this small lesion increased slowly throughout treatment. The comparison of tissue that we used for the analysis of this study included the patient’s baseline tumor. At the time of surgery when the resistant tumor had grown to a point that, for clinical care, it needed to be removed, we then had the sample of the resistant tumor. The surgeon explored the abdomen in detail, attempting to enter areas where the responding tumors were. When those tumors were explored and reviewed by the pathologist, there was no viable tumor remaining. The only viable tumor that remained in the abdomen at the time of the resection was the resistant tumor.

How long did the patient receive pembrolizumab?

Can you explain why certain tumors responded to the treatment?

You measured PD-L1 and PD-L2?

Importantly, no intervening treatment had been administered between the initial diagnosis and initiation of pembrolizumab. This was first-line therapy for this patient.The patient was initially diagnosed and then began pembrolizumab for first-line therapy for recurrent disease within 6 months of initial diagnosis. The patient has remained on pembrolizumab for more than 2 years due to excellent disease control. It is a remarkable finding to have a patient with recurrent leiomyosarcoma continue to have disease control measured in years.The work has given some hints as to why 1 tumor was resistant. We saw differences from baseline to resistance that might suggest possible mechanisms of escape from this therapy. I think the question as to why she responded initially is still one that we need to understand better. That's right. We measured PD-L1, PD-L2, and PD-1 on the tumor specimens both at baseline and at the time of resistance. Interestingly, we saw that this tumor has a strong expression of PD-L2. We've seen that with other studies in leiomyosarcoma.

What are the next steps regarding this research?

This case at baseline also had an infiltration of PD-L1—positive cells. The tumor cells themselves did not have extensive PD-L1 positivity, but the infiltrated cells in the tumor stroma did have PD-L1 positivity. When one compared that to the resistant tumor, there was a significant decrease in the PD-L1 infiltrating cells in the resistant tumor. That was a clear difference between baseline and resistance. We have several additional upcoming studies looking at combinations of immunotherapy. We also will be biopsying and obtaining tissues at baseline and at the time of resistance or response from participants to help build on the knowledge that we have currently.

In addition, we have a data set of patients with leiomyosarcoma that also received immunotherapy who did not respond as well as this patient. We may be going back into the data set to see if we can find similarities between that resistant population and this single patient’s resistant lesion.

References

1. Miao D, Adeegbe D, Rodig SJ, et al. Response and oligoclonal resistance to pembrolizumab in uterine leiomyosarcoma: Genomic, neoantigen, and immunohistochemical evaluation. J Clin Oncol 34, 2016 (suppl; abstr 11043).
2. George S, Miao D, Demetri GD, et al. Loss of PTEN is associated with resistance to anti-PD-1 checkpoint blockade therapy in metastatic uterine leiomyosarcoma. Immunity. 2017;46(2):197-204. doi:10.1016/j.immuni.2017.02.001.