Meeting News

Strategies to detect immune-based biomarkers evolving for melanoma treatment

NEW YORK — Several blood- and tissue-based assays are in development to identify appropriate treatment options for patients with melanoma, according to a speaker at HemOnc Today Melanoma and Cutaneous Malignancies.

“There have been a number of new therapies since 2011, including immunotherapies and targeted therapies,” Ryan J. Sullivan, MD, assistant in medicine at Massachusetts General Hospital and assistant professor at Harvard Medical School, said during his presentation. “But when you have lots of options, it does bring up a challenge: Who do we treat with what and when? The development of blood- and tissue-based predictors of benefit to therapy is an unmet need.”

When considering patient selection for therapy, there are several models. One model involves blood or tissue genotyping of patients to identify BRAF, NRAS, NF1 or CKIT mutations. However, this approach is dependent on the availability of targeted therapies for specific genotypes.

“For only one of these do we have approved therapies that we can offer that are targeting their specific mutation,” Sullivan said. “For the rest of the categories, by and large, it’s a barren field of targeted therapy. Selection of immunotherapy can be used by default in patients who do not harbor BRAF mutations. It’s not a very empiric way of approaching the therapy.”

Ryan J. Sullivan
Ryan J. Sullivan

Thus, researchers have started to focus on doing a tissue-based analysis to identify patients who are and aren’t likely to respond to immunotherapy.

“The advantages of this are that it is using emerging technology and approaches to assay tissue and/or blood and, as opposed to the first model, the selection of immunotherapy is active,” Sullivan said. “However, it minimizes the potential selection of longer-term survivors of targeted therapy.”

Emerging predictive models of PD-1 responsiveness include those for mutational load, neoantigen load, antigen expression machinery, T-cell infiltration and PD-L1 expression.

Patients with melanoma have the highest mutational burden compared with other malignancies where immunotherapy works, such as lung cancer and head and neck cancer.

Higher-mutational burden has demonstrated a trend to predict responsiveness, but there does not appear to a specific cutoff to determine which patients will benefit.

In one study, patients with high mutational load had significantly longer median PFS (not reached vs. 89 days vs. 86 days; P < .001) and median OS (not reached vs. 300 days vs. 375 days;

P < 0.001) compared with patients with intermediate or low mutational burden.

“The reason we propose that mutational load matters is because if you have a high mutational burden, you’re more likely to have some of those mutations trigger your immune system to recognize them, and then allow your immune system to respond,” Sullivan said. “It’s not something important about high mutation load, other than you’re more likely to make neoantigens.”

Thus, higher neoantigen load also is associated with benefit from therapy. However, those antigens need to be expressed. This led researchers to observe that MHC II expression is associated with better outcomes. The right T cells need to then be present in the tumor to have a response. For instance, tumor responsiveness is associated with interferon-gamma signature and not with mesenchymal/EMT/angiogenesis–associated signature.

In order to live, the tumor’s response to this is expression of PD-L1, Sullivan said.

“There is an association with PD-L1 and response,” Sullivan said. “Still, 25% of patients without PD-L1 have a response, so it is not a perfect marker.”

Further, when used to determine whether patients should receive combination therapy, PD-L1 expression appeared to predict benefit in terms of PFS, but not in terms of response rate. The absolute increase in response rate based on PD-L1 expression was 16.2% for nivolumab (Opdivo, Bristol-Myers Squibb) monotherapy compared with 17.3% for the combination of nivolumab and ipilimumab (Yervoy, Bristol-Myers Squibb).

“Based on these data, I don’t recommend doing routine PD-L1 analysis to select patients for combination therapy,” Sullivan said.

In blood, models are in development to assess circulating tumor DNA or TCR sequencing, Sullivan added.

“There are been dramatic advances in therapeutics, and dramatic advances in technology that are helping us to begin selection strategies to better pick patients for therapy,” Sullivan said. – by Alexandra Todak

Reference:

Sullivan RJ, et al. Biomarkers: Current status and clinical application in melanoma. Presented at: HemOnc Today Melanoma and Cutaneous Malignancies; March 24-25, 2017; New York.

Disclosure: Sullivan reports advisory board/consultant roles with Amgen, Biodesix, Novartis and Takeda, as well as research funding paid to his institution from Amgen, BioMedValley, Lilly and Merck.

NEW YORK — Several blood- and tissue-based assays are in development to identify appropriate treatment options for patients with melanoma, according to a speaker at HemOnc Today Melanoma and Cutaneous Malignancies.

“There have been a number of new therapies since 2011, including immunotherapies and targeted therapies,” Ryan J. Sullivan, MD, assistant in medicine at Massachusetts General Hospital and assistant professor at Harvard Medical School, said during his presentation. “But when you have lots of options, it does bring up a challenge: Who do we treat with what and when? The development of blood- and tissue-based predictors of benefit to therapy is an unmet need.”

When considering patient selection for therapy, there are several models. One model involves blood or tissue genotyping of patients to identify BRAF, NRAS, NF1 or CKIT mutations. However, this approach is dependent on the availability of targeted therapies for specific genotypes.

“For only one of these do we have approved therapies that we can offer that are targeting their specific mutation,” Sullivan said. “For the rest of the categories, by and large, it’s a barren field of targeted therapy. Selection of immunotherapy can be used by default in patients who do not harbor BRAF mutations. It’s not a very empiric way of approaching the therapy.”

Ryan J. Sullivan
Ryan J. Sullivan

Thus, researchers have started to focus on doing a tissue-based analysis to identify patients who are and aren’t likely to respond to immunotherapy.

“The advantages of this are that it is using emerging technology and approaches to assay tissue and/or blood and, as opposed to the first model, the selection of immunotherapy is active,” Sullivan said. “However, it minimizes the potential selection of longer-term survivors of targeted therapy.”

Emerging predictive models of PD-1 responsiveness include those for mutational load, neoantigen load, antigen expression machinery, T-cell infiltration and PD-L1 expression.

Patients with melanoma have the highest mutational burden compared with other malignancies where immunotherapy works, such as lung cancer and head and neck cancer.

Higher-mutational burden has demonstrated a trend to predict responsiveness, but there does not appear to a specific cutoff to determine which patients will benefit.

In one study, patients with high mutational load had significantly longer median PFS (not reached vs. 89 days vs. 86 days; P < .001) and median OS (not reached vs. 300 days vs. 375 days;

P < 0.001) compared with patients with intermediate or low mutational burden.

“The reason we propose that mutational load matters is because if you have a high mutational burden, you’re more likely to have some of those mutations trigger your immune system to recognize them, and then allow your immune system to respond,” Sullivan said. “It’s not something important about high mutation load, other than you’re more likely to make neoantigens.”

Thus, higher neoantigen load also is associated with benefit from therapy. However, those antigens need to be expressed. This led researchers to observe that MHC II expression is associated with better outcomes. The right T cells need to then be present in the tumor to have a response. For instance, tumor responsiveness is associated with interferon-gamma signature and not with mesenchymal/EMT/angiogenesis–associated signature.

In order to live, the tumor’s response to this is expression of PD-L1, Sullivan said.

“There is an association with PD-L1 and response,” Sullivan said. “Still, 25% of patients without PD-L1 have a response, so it is not a perfect marker.”

Further, when used to determine whether patients should receive combination therapy, PD-L1 expression appeared to predict benefit in terms of PFS, but not in terms of response rate. The absolute increase in response rate based on PD-L1 expression was 16.2% for nivolumab (Opdivo, Bristol-Myers Squibb) monotherapy compared with 17.3% for the combination of nivolumab and ipilimumab (Yervoy, Bristol-Myers Squibb).

“Based on these data, I don’t recommend doing routine PD-L1 analysis to select patients for combination therapy,” Sullivan said.

In blood, models are in development to assess circulating tumor DNA or TCR sequencing, Sullivan added.

“There are been dramatic advances in therapeutics, and dramatic advances in technology that are helping us to begin selection strategies to better pick patients for therapy,” Sullivan said. – by Alexandra Todak

Reference:

Sullivan RJ, et al. Biomarkers: Current status and clinical application in melanoma. Presented at: HemOnc Today Melanoma and Cutaneous Malignancies; March 24-25, 2017; New York.

Disclosure: Sullivan reports advisory board/consultant roles with Amgen, Biodesix, Novartis and Takeda, as well as research funding paid to his institution from Amgen, BioMedValley, Lilly and Merck.

    See more from HemOnc Today New York