Experts Examine Potential Biomarkers for Immunotherapy in Melanoma

Douglas B. Johnson, MD, MSCI

The emergence of immunotherapy in the melanoma treatment landscape has shifted the management of the disease; however, no strategies are in place to appropriately stratify patients for these potentially life-saving agents. Thus, the need to identify novel biomarkers that are predictive of response to checkpoint blockade has become increasingly imperative to maximize the effectiveness of this approach and avoid any unnecessary toxicity.

“Substantial unmet needs [exist with regard to biomarkers] at this point,” said Douglas B. Johnson, MD, MSCI, an assistant professor of medicine of hematology/oncology with Vanderbilt Institute, in an interview with OncLive. “We still lack a convincing biomarker that can tell us when to use an anti–PD-1 monotherapy, when to use a combination, whether either [regimen] is going to work, whether we should think about something different like a BRAF/MEK inhibitor [combination].... As far as the biomarker space goes, [we shouldn't ask,] ‘What are the unmet needs?’ They're all unmet needs at this point.”

Several prospective biomarkers have been examined as potential candidates, such as tumor-infiltrating lymphocytes, PD-L1 expression, and tumor mutational burden. In a collective effort to better examine potential candidates to fill this need, several research efforts are underway.

Body Composition

Immune checkpoint inhibitors (ICIs) have resulted in more durable responses and higher long-term survival rates in patients with metastatic melanoma. Following the success observed with this approach, investigators have shifted their focus to identifying biomarkers that will be indicative of therapeutic benefit, including patient factors that are adjustable like obesity.¹

Several recent studies have observed a correlation between obesity and response to ICIs in metastatic disease; 1 such study reported an improvement in overall survival (OS) and progression-free survival (PFS) in patients who were obese and received immunotherapy versus those with an average body mass index (BMI). However, other studies have not been able to establish such an association. As such, the mechanism of this potential relationship is not well understood and in need of further study.

“Several studies have suggested that patients who are obese or overweigh actually have better outcomes when treated with immune checkpoint inhibitors,” said Johnson, who was a senior author on the study. “This is a bit paradoxical because we normally think of obesity as being a more negative prognostic factor. However, several studies—one led by investigators from the University of Texas MD Anderson Cancer Center—have suggested that obese patients treated with a PD-1 monotherapy might have better outcomes.”

In 1 study, investigators reviewed an electronic medical record to identify a total of 349 patients with advanced melanoma who had previously been treated with either an anti–PD-1/PD-L1 monotherapy or a combination comprised of ipilimumab (Yervoy) and nivolumab (Opdivo). Of these patients, 290 had pretreatment scans that were available for further assessment. Three scans were excluded for excess artifact leaving a total of 287 patients who were included on the study.

For patients who had a BMI of less than 25, sarcopenia was defined as having a skeletal muscle index (SMI) of less than 43 cm/m2 for men and less than 41 cm/m2 for women. Additionally, for those with a BMI 26 or greater, sarcopenia was defined as less than 53 cm/m2 for men and less than 41 cm/m2 for women. Investigators defined sarcopenic obesity as those who met the same criteria for sarcopenia with a BMI of 30 or greater.

Skeletal muscle density (SMD) was also taken into consideration. For patients who had a BMI of less than 20 to 24.9, low SMD was determined to be less than 41 Houndsfield units (HU). Additionally, for patients who had a BMI greater than 25, low SMD was defined as less than 33 HU.

Pembrolizumab (Keytruda) was the most common ICI received (64.8%) followed by ipilimumab/nivolumab (21.6%). The median BMI was 28.9, which was noted to be a bit higher than the national average. More than half, or 53.7%, of patients were sarcopenic at baseline. Results from univariable or multivariable analyses did not demonstrate a significant association between BMI and response, toxicities, PFS, or OS.

Investigators also set out to determine whether different classes of obesity would correlate with more positive outcomes. No significant differences in response, toxicity, or PFS when looking at the different obesity classes; however, a substantial difference in OS was observed in patients with class III obesity, or a BMI of greater than 40, versus those with class I obesity (HR 2.4; 95% CI, 1.1-4.9; P =.03). Additionally, associations were not found to differ between the monotherapy and combination therapy cohorts.

“Surprisingly, in our cohort, we did not see any association between response and elevated BMI or obesity status,” Johnson said. “We did see, [however], that patients who had sarcopenic obesity—very little muscle, but a lot of fat—seemed to do the worst...We [also] saw that patients who tended to do the best had an intermediate BMI phenotype; their intermediate adiposity and higher muscle mass, perhaps, might be [representative of] the group that some of the other [studies] have picked out as doing particularly well [with ICIs].”

When examining the relationship between skeletal muscle gauge (SMG) and total adipose tissue index (TATI), investigators found no significant associations between BMI and response, PFS, or OS in the full group. However, notable differences in PFS and OS were noted between the group with the best outcomes (high SMG, mid TATI) and the worst outcomes (low SMG, high TATI). Thus, it was determined that patients with low SMG and high TATI had significantly worse survival outcomes with this approach (P =.02 and P=.02, respectively).

Although the relationship with outcomes and BMI were not observed, a trend toward worse outcomes in patients with higher adiposity and lower muscle quantity and quality was indicated. Even so, the trends observed were modest. As such, the investigators concluded that body composition could have some predictive value for ICI response, but it will likely not have a critical role in informing clinical decision-making.

“While we did see some interesting associations, from a clinical utility standpoint, I don't think we found anything that is going to change clinical practice,” concluded Johnson.

Tumor Mutational Burden and Circulating Tumor DNA

While immune checkpoint inhibitors have been shown to significantly improve outcomes in metastatic melanoma, approximately 40% to 50% of patients don't respond. Moreover, 60% of patients experience severe adverse effects (AEs) with this treatment, such as immune-mediated colitis, hepatitis, pneumonitis, or endocrinological diseases. However, investigators believe that tumor mutational burden (TMB) may be a potential predictor for response and circulating tumor DNA (ctDNA) may be a predictor for early response and prognosis.

“We know TMB for PD-1 response is important, [hence] the recent FDA approval for pembrolizumab [in this space],” Omid Hamid, MD, director of the Melanoma Center and Phase I Immuno-Oncology Program at The Angeles Clinic and Research Institute, said in an interview with OncLive. “It is an indicator of a [patient] who is going to have a higher response rate [to immunotherapy], but the question really becomes: Is it enough to tell us not to give a patient treatment? That's not clear at this time.”

In a recently published study, investigators examined patients with metastatic disease who had received systemic treatment from January 2018 on and had tumor tissue that was available for sequencing.² A total of 35 patients with melanoma who were receiving ipilimumab and nivolumab treatment were included in the analysis. A tumor panel of 710 tumor-associated genes was applied to all patients; this was followed by liquid biopsies that were done every 3-4 weeks.

Fifty-seven percent of the patients had cutaneous (n = 20), 1% had occult (n = 6), 11% had uveal (n = 4), 9% had acral (n = 3), and 6% had mucosal (n = 2) disease. The majority, or 63%, of patients were given ipilimumab and nivolumab as a first-line systemic treatment. Additionally, 29% (n = 10) of patients had previously been treated with a targeted therapy, while 9% (n = 3) had prior PD-1 antibodies.

Moreover, 49% of patients (n = 17) had completed 4 cycles of the immunotherapy combination, 23% (n = 8) received 3 cycles, 20% (n = 7) had 2 cycles, and 3 patients only had 1 cycle of treatment. One patient experienced rapid disease progression and died, 1 had severe myocarditis and diabetes mellitus and proceeded to receive single-agent nivolumab, and 1 patient refused further treatment.

The median TMB was 4.7 Mut/Mb (IQR 2–17), with patients being categorized into 1 of 3 groups: 37% were classified as TMB-low (n = 11; less than 3.3 Mut/Mb), 43% were TMB-intermediate (n = 13; 3.3–23.1 Mut/Mb), and 20% were TMB-high (n = 6; greater than 23.1 Mut/Mb). Cell-free DNA (cfDNA) was collected from 34 patients, with a median time of 23 days between baseline and first follow-up samples.

Results showed that the mean TMB was 43.2 in responders and 4.81 in non-responders. Moreover, median TMB was found to be significantly higher in patients who responded to the combination immunotherapy, with complete responders showing an even higher TMB value. Moreover, response to treatment appeared to significantly correlate to TMB, which was classified in 3 categories, including TMB-high (greater than 23.1 Mut/Mb), -intermediate (3.3–23.1 Mut/Mb) and -low less than 3.3 Mut/Mb). Trends of increased cfDNA concentration were noted in patients with progressive disease. Moreover, a more than 50% decrease in cfDNA concentration at first follow-up was significantly linked with response to the combination.

Although these findings are promising, Hamid cautions that TMB still requires further understanding before it can be used regularly in a clinical setting. “Would you remove the [possibility] of a patient with metastatic melanoma from receiving an immune checkpoint inhibitor based on TMB?” questioned Hamid. “Probably not, [because] the patient doesn't have many other options. It is something that is helping us understand how to make decisions. [It’s role] will [become clear] in the future as we understand [more about] how to utilize TMB with ctDNA.”

Additional results from the study showed that an increase in ctDNA copies were observed almost exclusively in patients with progressive disease at first follow-up. It was also found that remaining or undetectable ctDNA was shown to be more common in responders at first follow-up.

Results from the monovariate OS analysis per Kaplan-Meier showed that increasing cfDNA of greater than 50% or detectable ctDNA at first follow-up were found to negatively impact OS, while high TMB appeared to trend toward extended survival. However, TMB-low patients experienced improved survival if they did not experience an increase or detection of ctDNA at first follow-up.

ctDNA could additionally be an indicator for relapse in patients with melanoma, according to Hamid. "We know that the majority of melanomas have ctDNA, not just for the 50% who have BRAF mutations, but for other mutations as well. We know it's a good marker for risk of relapse in the adjuvant setting,” Hamid explained. “If [a patient has] ctDNA, you get your sentinel node [biopsy, but] if everything goes down to 0, you have a lower risk of relapse than a patient with persistent ctDNA. It has also been used as an indicator for response. There was an abstract [examining] patients who had pseudoprogression and then ctDNA was utilized to help understand what that means.”

microRNA

Differential expression profiles of microRNAs (miRNAs) have been reported for several tumor types, including melanoma. Investigators believe that these short regulatory RNAs may serve as an epigenetic marker for identifying new diagnostic and/or prognostic molecular markers; they are also thought to drive tumorigenesis and possibly represent new therapeutic targets.³

"miRNAs have been reported in a ton of cancers,” said Hamid. “They're regulators of gene expression, so they can be correlated to genes that can be expressed in malignancies. Their quantification can help [patients with melanoma] in the same way. Their potential biomarkers can help to diagnose early legions and early metastatic disease...In addition, they can also help to understand therapeutic targets. If these miRNAs are associated with certain pathways that we can target, then we can understand how to utilize them as therapeutic targets.”

Recently, circulating miRNA has gained interest in the field of melanoma, particularly as a potential marker for immunotherapy efficacy, in addition to classic circulating markers, such as lactate dehydrogenase (LDH), S100 calcium-binding protein B (S100B), and ctDNA. Due to the their structural stability, miRNAs have the potential to become robust noninvasive biomarkers.

Hamid additionally explained how miRNAs even have the potential to detect mutations earlier than current methods. “ctDNAs and miRNAs can help us to understand whether there is a target,” Hamid noted. “If there's circulating BRAF DNA, with a blood test, you may have an earlier way of knowing whether a tumor has that mutation and if that mutation [can be targeted].”

With the advent of immunotherapy agents in melanoma, finding miRNA efficacy markers has become an important focus of several research efforts. This need is heightened by the fact that the accumulation of myeloid-derived suppressor cells (MDSCs) that have a CD14+HLA-DR– phenotype can impede the efficacy of an immunotherapy agent.

Results from an additional recent study revealed that a panel of miRNAs was linked with MDSCs, immune checkpoint inhibitor resistance, and was even found to have involvement in the transformation of monocytes in MDSCs. miRNAs were found to increase amounts within blood CD14-positive cells, as well as in circulating plasma, and tumor tissues. It was also found to have a link with MDSC infiltrates. Following RNA extraction from circulating monocytes, extracellular vesicles, melanoma cells, and melanoma specimens, investigators were able to evaluate miRNA expression levels through a real-time quantitative polymerase-chain-reaction analysis.

“As we get more and more savy with understanding miRNAs and what happens with the evolution of the tumor, then we might be able to understand newer targets and how to introduce either combinatorial therapeutics or even a whole new therapeutic landscape for patients,” Hamid concluded.

The investigators of the review concluded that miRNAs could gain a stronger footing in melanoma, as new efficacy markers that could help to stratify patients based on immunotherapy benefit could be of great clinical importance in the future.

References

1. Young AC, Quach HT, Davis EJ, et al. Impact of body composition on outcomes from anti-PD1 +/− anti-CTLA-4 treatment in melanoma. J ImmunoTher Cancer. 2020;8(2):e000821. doi:10.1136/jitc-2020-000821
2. Forschner A, Battke F, Hadaschik, et al. Tumor mutation burden and circulating tumor DNA in combined CTLA-4 and PD-1 antibody therapy in metastatic melanoma – results of a prospective biomarker study. J Immunother Cancer. 2019;7(1):180. doi:10.1186/s40425-019-0659-0
3. Neagu M, Constantin C, Cretoiu SM, et al. miRNAs in the diagnosis and prognosis of skin cancer. Front Cell Dev Biol. 2020;71(8). doi:10.3389/fcell.2020.00071