Search for Predictive Biomarkers of PD-1 Pathway Blockade

John Powderly, MD

PD-L1 expression in tumors is a candidate molecular marker warranting further investigation as a means to select patients for immunotherapy with an anti PD-1 antibody, according to the authors of a 2012 article by Topalian et al¹published inThe New England Journal of Medicine, that validated the safety, activity, and immune correlates of the anti-PD-1 antibody nivolumab in several forms of cancer.

In the study, patients with non-small cell lung cancer (NSCLC), melanoma, or kidney cancer whose tumors were PD-L1-positive and who were treated with nivolumab had a 36% objective response rate (ORR). No PD-L1—negative tumors responded to the treatment, a finding suggesting that PD-L1 expression on the surface of tumor cells in pretreatment tumor specimens may be associated with ORR, and may be driven by constitutive oncogenic pathways.

Dr. Rizvi on PD-1 and PD-L1 Inhibitors in Lung Cancer

Rizvi is an associate attending physician at Memorial Sloan-Kettering Cancer Center.

However, another explanation for tumor cell expression of PD-L1, the authors suggested, relates to an “adaptive immune resistance in response to an endogenous antitumor immune response,” which may remain in check unless it is unleashed through blockade of the PD-1/PD-L1 pathway.¹

Basic questions such as this remain to be answered. But the search for biomarkers of clinical response proceeds in parallel with safety and efficacy investigations of PD-1 pathway inhibitor therapy in advanced solid tumors, specifically melanoma, NSCLC, and renal cell cancer.

Table. Advanced Melanoma Treated With Nivolumab, by PD-L1 Status

Endpoint

PD-L1 Status

N Event/N Subject (%)

Median, months (95% CI)

Melanoma

OS

+

-

8/16(50%)

10/18 (56%)

21.1 (9.4, -)

12.5% (8.2,-)

PFS

+

-

9/16 (56%)

12/18 (67%)

9.1 (1.8,-)

2.0 (1.8,9.3)

ORR

+

-

7/16 (44%)

3/18 (17%)

ORR = objective response rate; OS = overall response; PFS = progression-free survival.

A few of the biomarker investigations that were presented at the 2013 ASCO annual meeting point to more precise matching of patient to therapy:

• Patients with advanced melanoma who were PD-L1-positive as measured by a novel automated PD-L1 assay had a higher ORR and longer progressionfree survival (PFS) and overall survival (OS) when treated with nivolumab than did patients who were PD-L1 expression- negative (Table), suggesting that tumor PD-L1 status can be predictive of nivolumab activity in advanced solid tumors. In this phase I study, Grosso et al²describe evaluating tumor surface PD-L1 expression by immunohistochemistry (IHC) using an automated assay based on a sensitive and specific PD-L1 monoclonal antibody (28-8), distinct from 5H1.
• Topalian et al1 first examined immune correlates of anti-PD-1 antibody in cancer and suggested a correlation between pretreatment tumor PD-L1 expression in melanoma, NSCLC, and kidney cancer and response to therapy. At the time of their report presented at ASCO 2013, the investigators presented data for melanoma patients only. (The analysis of the link between PD-L1 expression with ORR, PFS, and OS in NSCLC was ongoing at the time of the ASCO report.)

• In patients with locally advanced or metastatic solid tumors, PD-L1 tumor expression, and a T-cell gene signature correlated with response to MPDL3280A, a human monoclonal antibody, in a study of biomarkers and associations with clinical activity of PD-L1 blockade.³In this companion investigation to a phase I study with MPDL3280A,⁴both reported at ASCO 2013, tumor samples that had been acquired from patients before and during treatment were analyzed by IHC (in 112 patients) and by means of a Genentech immunochip capable of measuring approximately 90 immune-related genes (in 96 patients), thus enabling a characterization of the tumor microenvironment. In addition, investigators undertook serial measurements of blood-based biomarkers and circulating immune cell subsets from an additional 23 patients for whom paired baseline and on-treatment samples were available.
• The researchers found that elevated baseline PD-L1 expression by IHC was positively associated with response to MPDL3280A. They also observed a coordinated expression of PD-L1 and CD8+ T-cells in tumor samples. In addition, a T-cell signature, including CD8, IFNg, and Granzyme-A, was associated with MPDL3280A response.
• Adaptive PD-L1 upregulation was evident from increasing PD-L1 expression and a Th1-dominant immune filtrate in responding tumors (In nonresponders, investigators noted minimal CD8+ T-cell infiltration and no activation of T-cells.)
• In a subpopulation of patients, according to investigator John Powderly, MD, from Carolina BioOncology, therapy with MPDL3280A led to T-cell reactivation and evidence for restored antitumor immunity—a rationale, along with the totality of evidence, for use of MPDL3280A as monotherapy in select patients with advanced or metastatic solid tumors.
• “Our data suggest that PD-L1 IHC status in the microenvironment may be associated with antitumor response. And patients with higher baseline expression of cytotoxic Th-1 T-cell markers, as per the immunochip MRA assay, appear to respond favorably if they have that Th1 signature at baseline,” said Powderly in summary remarks.
• A study evaluating the genetic subtype of metastatic melanoma, in particularNRAS-mutant melanoma, found that genetic subtype plays a role in predicting clinical benefit from immune- based therapy inBRAFwild-type metastatic melanoma.⁵Patients withNRAS-mutant multiple metastatic melanoma, a distinct subtype of the disease that is associated with poor prognoses, received increased clinical benefit from immune-based therapy versus patients withBRAF/NRASwild-type metastatic melanoma. There are currently no effective small-molecule inhibitors that specifically target NRAS, such as are available to treat BRAF-mutant metastatic melanoma. For this analysis, there were 173 patients with metastatic melanoma who underwent clinical genotyping for NRAS or BRAF mutation and were found to be BRAF wild-type, and were treated with either interleukin-2, ipilimumab, or anti-PD-1 (nivolumab) therapy. Patients withNRASmutation had improved clinical outcomes in terms of response rate (32% vs 18%;P=.042) and clinical benefit rate (49% vs 30%;P=.012).NRASmutation best predicted benefit (vs wild-type) for anti-PD-1/PDL1 therapy (response rate, 78% vs 19%;P=.002; clinical benefit rate, 78% vs 29%;P=.013). No significant differences were observed for interleukin-2 treatment.

References

1. Topalian SL, Hodi FS, Brahmer JR, et al. Safety, activity, and immune correlates of anti-PD-1 antibody in cancer.N Engl J Med. 2012;366:2443- 2454.
2. Grosso J, Horak CE, Inzunza D, et al. Association of tumor PD-L1 expression and immune biomarkers with clinical activity in patients (pts) with advanced solid tumors treated with nivolumab (anti-PD-1:BMS-936558; ONO-4538).J Clin Oncol. 2013;31(suppl; abstr 3016).
3. Powderly JD, Koeppen H, Hodi FS, et al. Biomarkers and associations with the clinical activity of PD-L1 blockade in a MPDL3280A study.J Clin Oncol. 2013;31(suppl; abstr 3001).
4. Herbst RS, Gordon MS, Fine GD, et al. A study of MPDL3280A, an engineered PD-L1 antibody in patients with locally advanced or metastatic tumors.J Clin Oncol. 2013;31(suppl; abstr 3000).
5. Johnson DB, Lovly CM, Flavin M, et al. NRAS mutation: a potential biomarker of clinical response to immune-based therapies in metastatic melanoma.J Clin Oncol. 2013;31(suppl; abstr 9019).