Serum Immunoregulatory Proteins As Predictors of Overall Survival of Metastatic Melanoma Patients Treated with Ipilimumab Yoshinobu Koguchi1, Helena M

Cancer Microenvironment and Immunology Research

Serum Immunoregulatory Proteins as Predictors of Overall Survival of Metastatic Melanoma Patients Treated with Ipilimumab Yoshinobu Koguchi1, Helena M. Hoen1, Shelly A. Bambina1, Michael D. Rynning2, Richard K. Fuerstenberg2, Brendan D. Curti1, Walter J. Urba1, Christina Milburn3, Frances Rena Bahjat3, Alan J. Korman3, and Keith S. Bahjat1

Abstract

Treatment with ipilimumab improves overall survival (OS) [log10 CXCL11: HR, 1.88; 95% confidence interval (CI), 1.14– in patients with metastatic melanoma. Because ipilimumab 3.12; P ¼ 0.014; and log10 sMICA quadratic effect P ¼ 0.066; targets T lymphocytes and not the tumor itself, efficacy may sMICA ( 247 vs. 247): HR, 1.75; 95% CI, 1.02–3.01]. Mul- be uniquely sensitive to immunomodulatory factors present at tivariate analysis of an independent ipilimumab-treated cohort the time of treatment. We analyzed serum from patients with confirmed the association between log10 CXCL11 and OS (HR, metastatic melanoma (247 of 273, 90.4%) randomly assigned 3.18; 95% CI, 1.13–8.95; P ¼ 0.029), whereas sMICA was less to receive ipilimumab or gp100 peptide vaccine. We quantified strongly associated with OS [log10 sMICA quadratic effect P ¼ candidate biomarkers at baseline and assessed the association 0.16; sMICA (247 vs. 247): HR, 1.48; 95% CI, 0.67–3.27]. of each using multivariate analyses. Results were confirmed in High baseline CXCL11 and sMICA were associated with poor an independent cohort of similar patients (48 of 52, 92.3%) OS in patients with metastatic melanoma after ipilimumab treated with ipilimumab. After controlling for baseline covari- treatment but not vaccine treatment. Thus, pretreatment ates, elevated chemokine (C-X-C motif) ligand 11 (CXCL11) CXCL11 and sMICA may represent predictors of survival benefit and soluble MHC class I polypeptide–related chain A (sMICA) after ipilimumab treatment as well as therapeutic targets. Cancer were associated with poor OS in ipilimumab-treated patients Res; 75(23); 5084–92. 2015 AACR.

Introduction received ipilimumab monotherapy achieved a 28.4% four-year OS rate (6). Despite the success of ipilimumab, the majority of the Ipilimumab is a human monoclonal antibody targeting CTLA- patients on this study died as a consequence of melanoma. With 4 (cytotoxic T lymphocyte antigen-4). CTLA-4 is expressed on an increasing number of treatment options available and activated T cells, has structural similarities to the costimulatory increased use of targeted therapies, predictive biomarkers that molecule CD28, and binds to the same ligands as CD28 albeit identify those patients most likely to benefit from a specific with higher affinity. Binding of CTLA-4 to CD80/CD86 inhibits T- treatment are needed (7, 8). cell activation by limiting IL2 production and expression of the Unlike traditional cancer therapies, immunotherapeutics act IL2 receptor (CD25; ref. 1). Ipilimumab prevents CTLA-4 from primarily upon cells of the immune system. The requirement for binding its ligands, thus promoting activation of effector T cells the immune system as a third-party mediator of the drug's activity via prolonged CD28 signaling (2). In addition, anti-CTLA-4 suggests the balance of positive and negative regulators of the antibodies can deplete intratumoral regulatory T cells, subverting immune response at the time of therapy may be a critical deter- yet another mechanism of immunosuppression (3). minant of efficacy for any immunotherapy. Cytokines, chemo- Ipilimumab improved overall survival (OS) in patients with kines, and soluble receptors regulate the survival, activity, and metastatic melanoma in a randomized, double-blinded phase III location of immune effector cells and thus represent potential clinical trial (4, 5). Patients with metastatic melanoma who players in determining drug efficacy. Of particular interest are soluble factors involved in the recruitment and regulation of effector T cells representing the most readily measurable clinical 1Earle A. Chiles Research Institute, Providence Cancer Center, Port- biomarkers. 2 3 land, Oregon. R&D Systems, Minneapolis, Minnesota. Bristol-Myers To identify candidate soluble factor(s) predictive of improved Squibb, Redwood City, California. survival following ipilimumab treatment, we analyzed pretreat- Note: Supplementary data for this article are available at Cancer Research ment sera from treatment (ipilimumab) and "active control" Online (http://cancerres.aacrjournals.org/). (gp100 vaccine) patients from the pivotal phase III clinical trial Corresponding Author: Keith S. Bahjat, Earle A. Chiles Research Institute, of ipilimumab (4) for a variety of factors and correlated their Providence Cancer Center, 4805 NE Glisan Street, 2N83, Portland, OR 97213. levels with OS. A hypothesis-guided panel of candidate biomar- Phone: 503-215-7229; Fax: 503-215-6841; E-mail: [email protected] kers was selected, including biomarkers previously reported to doi: 10.1158/0008-5472.CAN-15-2303 associate with response to ipilimumab. Each analyte was assessed 2015 American Association for Cancer Research. in univariate and multivariate models for its correlation with OS.

5084 Cancer Res; 75(23) December 1, 2015

Correlative biomarkers identified in the initial screen were further phase III trial, differences within treatment group were of primary validated by testing sera from an independent cohort of ipilimu- interest and thus tested in separate models. Survival was defined mab-treated patients at our institution. as time from beginning of ipilimumab treatment to date of death, censoring at date of last follow-up. To calculate median follow-up time, deaths were censored. Patients and Methods Univariate survival analysis was performed for each treatment Clinical trials subgroup using Cox proportional hazards regression. Effects of Detailed information regarding the phase III clinical trial of CXCL11, sMICA, sMICB, sCD25, VEGF, absolute lymphocyte ipilimumab (NCT00094653) was reported elsewhere (4). Briefly, counts (ALC), tumor burden, and effect of LDH [ vs. > upper patients with metastatic melanoma having failed at least one prior limit of normal (ULN)] were shown. Models of quadratic effects therapy that may have included IL2, dacarbazine, and/or temo- to examine possible nonlinear effects and models of linear effects zolomide were enrolled excluding those with ocular melanoma. of continuous variables were tested. When the quadratic effect was þ All patients were HLA-A 0201 as the restricting element for the not significant or the linear effect was more strongly significant, gp100 peptides used. All ipilimumab-treated patients received main effects model results were reported. For sMICA and VEGF, ipilimumab alone at 3 mg/kg every 3 weeks for 4 treatments. In quadratic effect was significant in some models. To present an HR, the gp100 group, patients received two peptides (1 mg each), results are also reported for a categorized variable, with the cutoff injected subcutaneously as an emulsion with incomplete Freund's point determined as the quintile where a threshold effect was adjuvant (Montanide ISA-51). Peptide injections were given observed in the phase III trial. Kaplan–Meier plots used for immediately after 90-minute intravenous infusion of placebo. determining cutoff points are shown in Supplementary Fig. S1. Tumor burden was assessed by the treating physician as previ- In multivariate analyses, model results of the remaining variables ously described (4). (other than sMICA), however, are from the model containing the Serum samples were also obtained from patients treated on an continuous form of the variable with the quadratic effect. Con- expanded access program at the Earle A. Chiles Research Institute tinuous measures were approximately log normal and analyzed as (EACRI cohort). Detailed information regarding this Compas- log10-transformed. sionate Use Trial for Unresectable Melanoma with Ipilimumab is In multivariate analysis of OS, Cox proportional hazards available elsewhere (NCT00495066). All patients received ipili- regression was used to test effects of biomarker candidates on mumab alone (3 or 10 mg/kg every 3 weeks for 4 treatments) with survival after controlling for other biomarkers and baseline no exclusions for ocular primary melanomas or HLA type. patient characteristics. Only CXCL11 and sMICA were included, All patients provided written informed consent and all studies as they were significant in univariate survival models of the were carried out in accordance with the Declaration of Helsinki ipilimumab group but not the gp100 group. Covariates in models under good clinical practice and Institutional Review Board for both studies were age, gender, ECOG status, prior immuno- approval. therapy, LDH, and ALC. Tumor burden was also included in multivariate model for the phase III trial cohort, but not the EACRI Serum cytokine analysis cohort as these data were not captured. Serum was collected and stored at 80 C. Chemokine (C-C Analyses were performed using SAS 9.3 (SAS Institute Inc.). motif) ligand 2 (CCL2), CCL3, CCL4, CCL8, CCL18, CCL26, Forest plots were prepared using Forest Plot Viewer (9) and edited chemokine (C-X-C motif) ligand (CXCL9), CXCL10, CXCL11, using Adobe Illustrator. GraphPad Prism was used for depicting CXCL13, and VEGF were measured using a bead-based multi- some Kaplan–Meier plots. plexed immunoassay (R&D Systems). Soluble MHC class I polypeptide–related sequence A (sMICA), sMICB, soluble UL16-binding protein (sULBP)-1, sULBP-2, sULBP-3, and sULBP-4 were Results measured using a custom multiplex bead array (R&D Systems). Patient characteristics of phase III study Bead-based assays were analyzed using the Luminex-based Bio- Demographics for the phase III study were previously reported Plex system (BIO-RAD). Soluble CD25 (sCD25) and soluble (4). Briefly, 676 patients were enrolled with 137 selected to receive lymphocyte-activation gene 3 (sLAG-3) were measured by ELISA ipilimumab monotherapy (treatment group), 136 to receive (R&D Systems). Serum sHLA-G was measured by ELISA (Exbio gp100 monotherapy (control group), and 403 treated with the Vestec). Only serum cytokines having statistical significance in combination of these agents. Biomarker analysis was restricted to univariate analyses of OS were reported. the monotherapy groups. Baseline characteristics were similar between monotherapy groups (Table 1), except that a higher Statistical considerations proportion of patients received prior immunotherapy in the Differences in patient baseline characteristics between treat- gp100 alone group (P ¼ 0.036; Table 1, column D). Patients ment groups (ipilimumab vs. gp100) or trials (phase III vs. were followed for a median of 31 months (range, 27–43 months). compassionate use) were evaluated using a t test for age and c2 OS of the ipilimumab group was 45.6% at 12 months, 33.2% at tests for gender, Eastern Cooperative Oncology Group (ECOG) 18 months, and 23.5% at 24 months, with a median OS of 10.1 performance status, lactate dehydrogenase (LDH), prior IL2 ther- months [95% confidence interval (CI), 8.0–13.8]. OS of the apy, and prior immunotherapy. Differences in baseline serum gp100 group was 25.3% at 12 months, 16.3% at 18 months, and biomarkers between study and treatment groups were tested with 13.7% at 24 months with a median OS of 6.4 months (95% CI, Wilcoxon rank-sum tests because of skewed distributions. 5.5–8.7). Analysis of soluble immunomodulatory proteins was Analysis of OS was conducted in the phase III trial and sepa- performed on serum collected prior to treatment. Baseline rately in the confirmatory EACRI cohort due to the differences in CXCL11 concentrations were comparable between ipilimumab the patient populations, study design, and study protocol. In the (median, 38; range, 2–1,027 pg/mL) and gp100 (median, 39;

www.aacrjournals.org Cancer Res; 75(23) December 1, 2015 5085

Table 1. Baseline patient characteristics ABCDE Phase III trial cohort EACRI cohort Statistics Demographic or Ipilimumab gp100 monotherapy Ipilimumab P:A P:A clinical characteristic monotherapy (n ¼ 124) (n ¼ 123) monotherapy (n ¼ 48) vs. Ba vs. Ca Age, y 0.99 0.51 Median 57 57 60 Range 23–90 19–88 36–81 Male, % 61.3 54.5 60.4 0.28 0.92 ECOG, % 0.85d 0.22d 0 51.6 52.9 41.7 1 47.6 43.9 52.1 2 0.8 3.2 6.3 LDH > ULN, % 37.1 36.4 66.0 0.91 0.0007 ALC, 109/L 0.27 0.61 Median 1.3 1.2 1.3 Range 0.4–3.3 0.3–2.8 0.3–4.1 Prior IL2 therapy, % 23.4 25.2 60.4 0.74 <0.0001 Prior immunotherapy, %b 39.5 52.9 77.1 0.036 <0.0001 CXCL11, pg/mL 0.51 0.0003 Median 38 39 14 Range 2–1027 2–911 3–153 sMICA, pg/mL 0.99 <0.0001 Median 115 121 299 Range 13–1573 13–2074 12–2456 Median survival, moc 10.1 6.9 8.6 aP values shown from the t test for age, Wilcoxon log-rank test for CXCL11 and sMICA, and c2 tests for all other variables. bIncluding IL2. cMedian survivals times calculated from Kaplan–Meier estimates. dECOG 1–2 versus 0.

range, 2–911 pg/mL) groups. Similarly, baseline levels of sMICA Within the ipilimumab-treated group CXCL11 and LDH, but not were consistent between ipilimumab (median, 115; range, 13– tumor burden, ALC, age, sex, or ECOG score, were associated with 1,573 pg/mL) and gp100 (median, 121; range, 13–2,074 pg/mL) OS (log10 CXCL11 HR, 1.88; 95% CI; 1.14–3.12; P ¼ 0.014: LDH groups. HR, 2.99; 95% CI, 1.78–5.02; P < 0.0001; Fig. 2A). sMICA was also associated with OS [log10 sMICA quadratic effect P ¼ 0.0659; CXCL11, sMICA, and OS sMICA (247 vs. <247): HR, 1.75; 95% CI, 1.02–3.01 with P ¼ Univariate analysis of ipilimumab-treated patients showed that 0.0420] but less strongly than CXCL11 (Fig. 2A). In the gp100- a 10-fold increase in CXCL11 was associated with double the risk treated group, only LDH was independently associated with OS of death (HR, 2.08; 95% CI, 1.40–3.11; P ¼ 0.0003; Fig. 1), (LDH: HR, 2.24; 95% CI, 1.36–3.69; P ¼ 0.0016; Fig. 2B). These whereas CXCL11 was not associated with OS in the gp100 group multivariate results again suggest that CXCL11 and sMICA are (HR, 1.21; 95% CI, 0.87–1.68; P ¼ 0.2597). The effect of CXCL11 potential predictive biomarkers of OS in ipilimumab-treated on OS was significantly different for the ipilimumab group versus melanoma patients, whereas LDH represents a prognostic bio- the gp100 group (P ¼ 0.040). In the univariate analysis of log10 marker for patients with melanoma irrespective of treatment. sMICA, higher sMICA was associated with decreased survival in the ipilimumab group [log10 sMICA quadratic effect P < 0.0001; CXCL11 and sMICA in an independent ipilimumab-treated sMICA (247 vs. <247): HR, 3.46; 95% CI, 2.16–5.56 with P < cohort 0.0001] but not in the gp100 group (log10 sMICA HR, 0.91; 95% We analyzed sera from patients with melanoma (48 of 52, CI, 0.61–1.36; P ¼ 0.6373). Elevated sMICB, LDH, tumor burden, 92.3%) collected prior to treatment with ipilimumab in an and sCD25 were all associated with poorer survival regardless of expanded access program at our institution (EACRI cohort). treatment (Fig. 1). Elevated VEGF was also associated with When comparing patient characteristics between the ipilimu- decreased survival in both groups, although marginally so for mab-treated phase III trial cohort and the EACRI cohort, we found the ipilimumab-treated group (Fig. 1). Higher numbers of lym- more patients in the EACRI cohort with elevated LDH (P ¼ phocytes (ALC) at baseline were associated with better OS in both 0.0007), prior IL2 therapy (P < 0.0001), and prior immunother- treatment groups (Fig. 1). These univariate analyses suggest that apy (P < 0.0001; Table 1, column A, C, and E). This comparison CXCL11 and sMICA are potential predictors of OS in ipilimumab- suggests that the EACRI cohort included more patients with treated melanoma patients, whereas sMICB, sCD25, VEGF, LDH, advanced disease and poorer prognosis. This discrepancy may tumor burden, and ALC represent putative prognostic biomarkers. account for shorter median survival of the EACRI cohort (8.6 Multivariate analyses were also conducted focusing on CXCL11 months) relative to that of ipilimumab-treated phase III trial and sMICA, as these two biomarkers were identified in the cohort (10.1 months). In the EACRI cohort, median follow-up univariate analysis as correlating with ipilimumab but not gp100 was 39 months (range, 0.8–40 months). treatment. Models were used to test the independent effects of Univariate analyses showed that elevated pretreatment con- CXCL11 and sMICA after adjusting for each other and the cov- centrations of CXCL11, sCD25, and LDH were associated with an ariates LDH, ALC, tumor burden, age, sex, and ECOG status. increased risk of death (Fig. 3). Similar to phase III study findings,

5086 Cancer Res; 75(23) December 1, 2015 Cancer Research

Biomarker Treatment HR (95% CI) P

Log10 CXCL11 ipi 2.08 (1.40–3.11) 0.0003

Log10 CXCL11 gp100 1.21 (0.87–1.68) 0.2597

sMICA (≥247 vs. <247) ipi 3.46 (2.16–5.56)* <0.0001

Log10 sMICA gp100 0.91 (0.61–1.36) 0.6373

Log10 sMICB ipi 5.56 (3.12–9.91) <0.0001

Log10 sMICB gp100 2.86 (1.80–4.54) <0.0001

Log10 sCD25 ipi 3.24 (1.20–8.74) 0.0203

Log10 sCD25 gp100 7.75 (3.22–18.67) <0.0001

Log10 VEGF ipi 1.66 (0.97–2.84) 0.0630

VEGF (≥ 157 vs. <157) gp100 2.82 (1.06–3.87)§ <0.0001

LDH; > ULN vs. ≤ ULN ipi 3.37 (2.19–5.19) <0.0001

LDH; > ULN vs. ≤ ULN gp100 2.79 (1.88–4.16) <0.0001

Log10 ALC ipi 0.06 (0.02–0.22) <0.0001

Log10 ALC gp100 0.21 (0.09–0.51) 0.0006

Log10 tumor burden ipi 2.24 (1.56–3.21) <0.0001

Log10 tumor burden gp100 2.18 (1.48–3.21) <0.0001

0.01 0.1 1 10 High biomarker level better High biomarker level worse

Figure 1. Univariate analysis of biomarker effects on OS for patients from the phase III clinical trial. HR and CI for association with OS of patients treated with ipilimumab (ipi) or gp100. Cox proportional hazards regression was used for univariate analysis of biomarker effects on OS. HR is numerator versus denominator. Among 124 patients analyzed, 35 were censored in ipilimumab-treated group. Among 123 patients analyzed, 13 were censored in gp100-treated group. Missing data: LDH, ALC, tumor burden 1–2 missing in gp100 group; CXCL11, sCD25, VEGF 5 missing in gp100 group, 11 missing in ipilimumab group. , in quadratic effects 2 2 model of ipilimumab group, (log10 sMICA) P < 0.0001. x, in gp100 group, (log10 VEGF) P ¼ 0.0002.

a 10-fold increase in CXCL11 was associated with a 3.7-fold 0.0457) after controlling for each other, gender, age, ECOG status, increase in the risk of death (HR, 3.74; 95% CI, 1.71–8.22; P ¼ and prior immunotherapy (Fig. 4). sMICA may be associated with 0.0010). sMICA and VEGF effects were nonlinear, depicted by the OS in this cohort [log10 sMICA quadratic effect: P ¼ 0.1589; threshold effect as seen in the Kaplan–Meier plots of survival sMICA (247 vs. <247): HR, 1.48; 95% CI, 0.67–3.27 with P ¼ (Supplementary Fig. S2). Elevated sMICA was also associated with 0.3284; Fig. 4], a result due in part to adjusting for CXCL11 and increased risk of death [log10 sMICA quadratic effect: P ¼ 0.0244; LDH and somewhat small cohort size. Thus, we conﬁrmed the sMICA (247 vs. <247): HR, 2.06; 95% CI, 1.06–4.00 with P ¼ predictive association between CXCL11 and OS in ipilimumab- 0.0324; Fig. 3]. Elevated VEGF was associated with decreased treated melanoma patients but found a weaker association survival and sMICB and ALC were not associated with survival. between sMICA and OS in the EACRI cohort. We also conﬁrmed Multivariate analysis showed that CXCL11 and LDH were the association between LDH and OS in the EACRI cohort, associated with OS (log10 CXCL11: HR, 3.18; 95% CI, 1.13– compatible with the notion that LDH is a prognostic marker for 8.95; P ¼ 0.0288 and LDH: HR, 2.24; 95% CI, 1.02–4.95; P ¼ patients with metastatic melanoma.

www.aacrjournals.org Cancer Res; 75(23) December 1, 2015 5087

A Biomarker HR (95% CI) P

Log10 CXCL11 1.88 (1.14–3.12) 0.0141

sMICA (≥247 vs. <247): 1.75 (1.02–3.01)* 0.0420

LDH; > ULN vs. ≤ ULN 2.99 (1.78–5.02) <0.0001

0.0639 Log10 ALC 0.25 (0.06–1.08)

0.0865 Log10 tumor burden 1.48 (0.95–2.32)

Sex: male vs. female 1.10 (0.65–1.86) 0.7242

Age 0.99 (0.97–1.01) 0.1800

ECOG; 1-2 vs 0 1.43 (0.84–2.43) 0.1918

Prior immunotherapy: 0.83 (0.50–1.40) 0.4915 Figure 2. yes vs. no Multivariate analysis of biomarker effects on OS for patients from the phase III clinical trial. HR and CI for association of potential biomarker 0.1 1 10 with OS of patients treated with ipilimumab (A) or High biomarker level better High biomarker level worse gp100 (B). Cox proportional hazards regression B was used for multivariate analysis of biomarker Biomarker HR (95% CI) P effects on OS. Among the 113 total patients analyzed, 34 were censored in ipilimumab-treated group. Among total 115 patients analyzed, Log10 CXCL11 1.01 (0.70–1.44) 0.9741 13 were censored in gp100-treated group. ,in quadratic effects model of ipilimumab group, 2 (log10 sMICA) P ¼ 0.0659. 0.94 (0.60–1.47) Log10 sMICA 0.7731

LDH; > ULN vs. ≤ ULN 2.24 (1.36–3.69) 0.0016

Log10 ALC 0.46 (0.14–1.50) 0.1966

1.45 (0.93–2.26) Log10 tumor burden 0.1050

Sex: male vs. female 1.21 (0.79–1.84) 0.3731

Age 1.00 (0.98–1.02) 0.9419

ECOG; 1-2 vs 0 1.51 (0.99–2.32) 0.0569

Prior immunotherapy: 1.29 (0.80–2.07) 0.2956 yes vs. no

0.1 111 100 High biomarker level better High biomarker level worse

5088 Cancer Res; 75(23) December 1, 2015 Cancer Research

Biomarker HR (95% CI) P

3.74 (1.71–8.22) Log10 CXCL11 0.0010

sMICA 2.06 (1.06–4.00)* 0.0324 (≥247 vs. <247)

Figure 3. Log sMICB 1.31 (0.74–2.29) 0.3533 Univariate analysis of the ipilimumab- 10 treated EACRI cohort. HR and CI for association of potential biomarker with OS of patients treated with ipilimumab. Cox proportional hazards Log10 sCD25 6.68 (1.36–32.75) 0.0192 regression was used for univariate analysis of biomarker effects on OS. HR is numerator versus denominator. Among 48 total patients analyzed, 8 VEGF 2.36 (1.21–4.63)§ 0.0121 were censored. , in quadratic 2 (≥157 vs. <157) effects model, (log10 sMICA) 2 P ¼ 0.0244. x, (log10 VEGF) P ¼ 0.0200. LDH; 2.11 (1.04–4.28) 0.0376 > ULN vs. ≤ ULN)

Log10 ALC 0.58 (0.16–2.07) 0.3989

0.1 1 10 High biomarker level better High biomarker level worse

Kaplan–Meier survival curves benefit from ipilimumab. Because a minority of patients with To illustrate the effects of CXC11 and sMICA on OS in the melanoma benefit from ipilimumab, avoiding treatment of these phase III trial, Kaplan–Meier survival plots are shown (Fig. 5) refractory patients would reduce exposure to inefficient therapy, for the biomarker high or low groups on the basis of selected eliminate their risks for adverse effects and lower overall costs of cutoff points (the median of CXC11; 35 pg/mL, the 80th therapy. percentile of sMICA; 247 pg/mL, Fig. 5A and B). Because of Potential predictors of responsiveness to ipilimumab treatment the quadratic association of sMICA with survival, the 80th have been identified in previous reports. For instance, elevated percentile was the cutoff point chosen on the basis of the levels of several candidate biomarkers [e.g., C-reactive protein approximate threshold value seen in the sMICA quintile plot (CRP; refs. 10, 11), erythrocyte sedimentation rate (ESR; ref. 12), (Supplementary Fig. S1). Kaplan–Meier survival plots for the LDH (10–14), S100 protein (12), sCD25 (15), and VEGF (16)] EACRI cohort are also shown using the same cutoff points as were thought to associate with reduced benefit following ipili- used for the phase III study data plot (Fig. 5C and D). The mumab treatment. In contrast, increased baseline ALC were distribution of baseline CXCL11 was lower and the distribution associated with improved OS upon ipilimumab treatment (10– of sMICA levels was higher in the confirmatory cohort (Table 1, 14, 17). Interestingly, each of these candidate biomarkers were at column E). Nonetheless, both cutoff points successfully dichot- some time reported as prognostic biomarkers for melanoma (18, omize patients treated with ipilimumab into patients with poor 19). We were fortunate to have serum samples from the phase III or better OS. study comparing an ipilimumab-treated cohort with a control cohort allowing us to differentiate predictive versus prognostic biomarkers (20). We also used samples from a second indepen- Discussion dent cohort of ipilimumab-treated patients to validate our basic We found that high baseline serum CXCL11 and sMICA were findings, which together with multivariate analyses, diminished associated with poor OS in patients with metastatic melanoma the possibility of coincidental influence from other covariates. treated with ipilimumab but not in patients treated with a These analyses have demonstrated the predictive nature of "control" gp100 vaccine. This association was validated in an CXCL11 and sMICA for patients treated with ipilimumab, and independent cohort of ipilimumab-treated melanoma patients, the prognostic nature of sMICB, VEGF, sCD25, LDH, and ALC for strongly suggesting that measurement of pretreatment serum all patients with melanoma independent of treatment with CXCL11 and sMICA levels may identify patients most likely to ipilimumab.

www.aacrjournals.org Cancer Res; 75(23) December 1, 2015 5089

Biomarker HR (95% CI) P

Log10 CXCL11 3.18 (1.13–8.95) 0.0288

sMICA (≥247 vs. <247) 1.48 (0.67–3.27)* 0.3284

LDH; 2.24 (1.02–4.95) 0.0457 ≤ > ULN vs. ULN) Figure 4. Multivariate analysis of the Log ALC 0.67 (0.16–2.88) 0.5918 ipilimumab-treated EACRI cohort. HR 10 and CI for association of potential biomarker with OS of patients treated with ipilimumab. Cox proportional Sex: 0.97 (0.44–2.14) 0.9404 hazards regression was used for male vs. female multivariate analysis of biomarker effects on OS. HR is numerator versus denominator. Among Age 1.00 (0.97–1.03) 0.8977 47 total patients analyzed, 8 were censored. , in quadratic effects 2 model, (log10 sMICA) P ¼ 0.1589.

ECOG; 0.97 (0.44–2.16) 0.9407 1-2 vs. 0

Prior immunotherapy: 1.73 (0.70–4.30) 0.2382 yes vs. no

0.1 1 10 High biomarker level better High biomarker level worse

Our data revealed a strong association between elevated Elevated levels of sMICA and sMICB have been reported in serum CXCL11 protein and reduced OS upon ipilimumab several types of malignant diseases and implicated in cancer treatment in patients with metastatic melanoma. CXCL11, immunoevasion (32). While the cellular stress response in trans- along with CXCL9 and CXCL10, binds to CXCR3, a critical formed cells induces expression of membrane-bound MICA/ chemokine receptor for directional migration of TH1andcyto- MICB, the cleavage of these molecules produces soluble forms toxic T cells (21). In contrast to our findings with serum of MICA and MICB capable of inhibiting interactions between CXCL11, tissue expression of CXCL11 mRNA correlates with membrane-bound MICA/MICB and NKG2D, thus desensitizing T-cell infiltration into tumors and improved prognosis (22). the activation signal through NKG2D in effector T cells and Microarray analysis of mRNA expression in melanoma tissues natural killer (NK) cells (33). The cleavage of membrane-bound from ipilimumab-treated patients also associated baseline and MICA/MICB is promoted by the metalloprotease ADAM10 and is posttreatment tissue expression of CXCL11 with presence of T enhanced within hypoxic tumor environments (34). As hypoxia cells in tumor and favorable clinical responses (23). The dis- also promotes expression of immune inhibitory molecules (e.g., crepancy between these results and ours may be attributed to PD-L1 and LAG-3) and favors accumulation of regulatory sample source (tissue vs. serum) and/or assay targets (mRNA immune cells (35), sMICA may be a biomarker that reflects this vs. protein). Paired analysis of mRNAs and proteins in tissue immunosuppressive tumor environment. and serum might provide evidence to resolve this discrepancy. The greatest significance of these findings may ultimately be in As CXCL11 has distinct immunoregulatory functions, in con- the identification of CXCL11 and sMICA as immunotherapeutic trast to immunostimulatory functions mediated by CXCL9 and targets. Agents that inhibit the immunosuppressive activity of CXCL10 (24, 25), we prefer the following nonmutually exclu- CXCL11 and sMICA without preventing interaction of the recep- sive explanations of how CXCL11, even in the presence of tors (CXCR3 and NKG2D, respectively) with immunopotentiat- CXCL9 and CXCL10, may limit T-cell effector function as a ing ligands (CXCL9/CXCL10 and membrane-bound MICA, part of negative feedback loop: (i) disrupting chemokine gra- respectively) may have therapeutic activity in patients with cancer, dients for directional migration of T cells (26), (ii) preventing either alone or with other agents, including chemotherapeutics, CXCR3–CXCL9/10 interaction by promoting receptor internal- radiation, or other immunotherapeutics. To explore this idea, ization (24), (iii) suppressing T-cell responses through induc- future studies should address whether CXCL11 and sMICA direct- tion and/or recruitment of regulatory T cells (25, 27), and (iv) ly interfere with ipilimumab-enabled effector T cells or are merely promoting growth and metastasis of tumors expressing CXCR3 elevated as a result of the immunosuppressive environment found (28–31). in patients refractory to ipilimumab therapy.

5090 Cancer Res; 75(23) December 1, 2015 Cancer Research

Figure 5. Kaplan–Meier curves for OS according to pretreatment CXCL11 or sMICA status. Curves for OS obtained by applying selected cut points for CXCL11 (A and C) and sMICA (B and D) to the phase III trial cohort (A and B) or the EACRI cohort (C and D). Numbers of subjects at risk at each 10-month interval are listed below each graph. The difference in treatment effect [ipilimumab (ipi) or gp100] according to CXCL11 or sMICA concentration is represented in the top right of each graph (log-rank).

Disclosure of Potential Conﬂicts of Interest Administrative, technical, or material support (i.e., reporting or organizing Y. Koguchi, F.R. Bahjat, H.M. Hoen, and A.J. Korman have ownership interest data, constructing databases): H.M. Hoen, K.S. Bahjat in a pending patent. C. Milburn is an employee of Bristol-Myers Squibb. F.R. Study supervision: B.D. Curti, A.J. Korman, K.S. Bahjat Bahjat is the co-founder of NeurAlexo, VP Pharmalocalogy at Oncovir and reports receiving other commercial research support from Oncovir, Bristol Acknowledgments Myers Squibb and is also a consultant/advisory board member of Bristol Myers Squibb. A.J. Korman has ownership interest in Bristol-Myers Squibb. K.S. Bahjat The authors thank Gwen Kramer for assistance with sample preparation, reports receiving commercial research grant from Bristol-Myers Squibb. No Michael Gough, Marka Crittenden, and Will Redmond for helpful discussions, potential conﬂicts of interest were disclosed by the other authors. and the dedication and skill of the clinical research team at the Providence Cancer Center.

Authors' Contributions Conception and design: Y. Koguchi, C. Milburn, A.J. Korman, K.S. Bahjat Grant Support Development of methodology: C. Milburn, K.S. Bahjat These studies were supported by institutional funding from the Providence Acquisition of data (provided animals, acquired and managed patients, Portland Medical Foundation and a sponsored-research grant from Bristol- provided facilities, etc.): Y. Koguchi, S.A. Bambina, R.K. Fuerstenberg, Myers Squibb (BMS). B.D. Curti, A.J. Korman, K.S. Bahjat The costs of publication of this article were defrayed in part by the Analysis and interpretation of data (e.g., statistical analysis, biostatistics, payment of page charges. This article must therefore be hereby marked advertisement computational analysis): Y. Koguchi, H.M. Hoen, B.D. Curti, F.R. Bahjat, in accordance with 18 U.S.C. Section 1734 solely to indicate K.S. Bahjat this fact. Writing, review, and/or revision of the manuscript: Y. Koguchi, H.M. Hoen, M.D. Rynning, R.K. Fuerstenberg, B.D. Curti, W.J. Urba, C. Milburn, F.R. Bahjat, Received August 19, 2015; revised September 16, 2015; accepted September A.J. Korman, K.S. Bahjat 23, 2015; published online December 1, 2015.

www.aacrjournals.org Cancer Res; 75(23) December 1, 2015 5091

References 1. Wolchok JD, Hodi FS, Weber JS, Allison JP, Urba WJ, Robert C, et al. 17. Ku GY, Yuan J, Page DB, Schroeder SEA, Panageas KS, Carvajal RD, et al. Development of ipilimumab: a novel immunotherapeutic approach for Single-institution experience with ipilimumab in advanced melanoma the treatment of advanced melanoma. Ann N Y Acad Sci 2013;1291:1–13. patients in the compassionate use setting. Cancer 2010;116:1767–75. 2. Rudd CE, Taylor A, Schneider H. CD28 and CTLA-4 coreceptor expression 18. Mouawad R, Spano J-P, Khayat D. Old and new serological biomarkers in and signal transduction. Immunol Rev 2009;229:12–26. melanoma: where we are in 2009. Melanoma Res 2010;20:67–76. 3. Simpson TR, Li F, Montalvo-Ortiz W, Sepulveda MA, Bergerhoff K, Arce F, 19. Rochet NM, Kottschade LA, Grotz TE, Porrata LF, Markovic SN. The et al. Fc-dependent depletion of tumor-infiltrating regulatory T cells co- prognostic role of the preoperative absolute lymphocyte count and abso- defines the efficacy of anti-CTLA-4 therapy against melanoma. J Exp Med lute monocyte count in patients with resected advanced melanoma. Am J 2013;210:1695–710. Clin Oncol 2015;38:252–8. 4. Hodi FS, O'Day SJ, McDermott DF, Weber RW, Sosman JA, Haanen JB, et al. 20. Simon R. Clinical trials for predictive medicine: new challenges and Improved survival with ipilimumab in patients with metastatic melanoma. paradigms. Clin Trials 2010;7:516–24. N Engl J Med 2010;363:711–23. 21. Van Raemdonck K, Van den Steen PE, Liekens S, Van Damme J, Struyf S. 5. Page DB, Postow MA, Callahan MK, Allison JP, Wolchok JD. Immune CXCR3 ligands in disease and therapy. Cytokine Growth Factor Rev modulation in cancer with antibodies. Annu Rev Med 2014;65: 2014;26:311–27. 185–202. 22. Galon J, Angell HK, Bedognetti D, Marincola FM. The continuum of cancer 6. Wolchok JD, Weber JS, Maio M, Neyns B, Harmankaya K, Chin K, et al. immunosurveillance: prognostic, predictive, and mechanistic signatures. Four-year survival rates for patients with metastatic melanoma who Immunity 2013;39:11–26. received ipilimumab in phase II clinical trials. Ann Oncol 2013;24: 23. Ji R-R, Chasalow SD, Wang L, Hamid O, Schmidt H, Cogswell J, et al. An 2174–80. immune-active tumor microenvironment favors clinical response to ipi- 7. Sharma P, Allison JP. The future of immune checkpoint therapy. Science limumab. Cancer Immunol Immunother 2011;61:1019–31. 2015;348:56–61. 24. Sauty A, Colvin RA, Wagner L, Rochat S, Spertini F, Luster AD. CXCR3 8. Curti BD, Urba WJ. Clinical deployment of antibodies for treatment of internalization following T cell-endothelial cell contact: preferential role of melanoma. Mol Immunol 2015;67:18–27. IFN-inducible T cell chemoattractant (CXCL11). J Immunol 2001;167: 9. Boyles AL, Harris SF, Rooney AA, Thayer KA. Forest plot viewer: a new 7084–93. graphing tool. Epidemiology 2011;22:746–7. 25. Zohar Y, Wildbaum G, Novak R, Salzman AL, Thelen M, Alon R, et al. 10. Wilgenhof S, Four Du S, Vandenbroucke F, Everaert H, Salmon I, Lienard D, CXCL11-dependent induction of FOXP3-negative regulatory T cells sup- et al. Single-center experience with ipilimumab in an expanded access presses autoimmune encephalomyelitis. J Clin Invest 2014;124:2009–22. program for patients with pretreated advanced melanoma. J Immunother 26. Griffith JW, Sokol CL, Luster AD. Chemokines and chemokine receptors: 2013;36:215–22. positioning cells for host defense and immunity. Annu Rev Immunol 11. Simeone E, Gentilcore G, Giannarelli D, Grimaldi AM, Caraco C, Curvietto 2014;32:659–702. M, et al. Immunological and biological changes during ipilimumab 27. Koch MA, Tucker-Heard G, Perdue NR, Killebrew JR, Urdahl KB, Campbell treatment and their potential correlation with clinical response and sur- DJ. The transcription factor T-bet controls regulatory T cell homeostasis and vival in patients with advanced melanoma. Cancer Immunol Immunother function during type 1 inflammation. Nat Immunol 2009;10:595–602. 2014;63:675–83. 28. Billottet C, Quemener C, Bikfalvi A. CXCR3, a double-edged sword in 12. Kelderman S, Heemskerk B, van Tinteren H, van den Brom RRH, Hospers tumor progression and angiogenesis. Biochim Biophys Acta 2013;1836: GAP, van den Eertwegh AJM, et al. Lactate dehydrogenase as a selection 287–95. criterion for ipilimumab treatment in metastatic melanoma. Cancer 29. Robledo MM, Bartolome RA, Longo N, Rodríguez-Frade JM, Mellado M, Immunol Immunother 2014;63:449–58. Longo I, et al. Expression of functional chemokine receptors CXCR3 and 13. Di Giacomo AM, Danielli R, Calabro L, Bertocci E, Nannicini C, Giannarelli CXCR4 on human melanoma cells. J Biol Chem 2001;276:45098–105. D, et al. Ipilimumab experience in heavily pretreated patients with mel- 30. Monteagudo C, Martin JM, Jorda E, Llombart-Bosch A. CXCR3 chemokine anoma in an expanded access program at the University Hospital of Siena receptor immunoreactivity in primary cutaneous malignant melanoma: (Italy). Cancer Immunol Immunother 2010;60:467–77. correlation with clinicopathological prognostic factors. J Clin Pathol 14. Delyon J, Mateus C, Lefeuvre D, Lanoy E, Zitvogel L, Chaput N, et al. 2007;60:596–9. Experience in daily practice with ipilimumab for the treatment of 31. Longo-Imedio MI, Longo N, Trevino~ I, Lazaro P, Sanchez-Mateos P. patients with metastatic melanoma: an early increase in lymphocyte Clinical significance of CXCR3 and CXCR4 expression in primary mela- and eosinophil counts is associated with improved survival. Ann Oncol noma. Int J Cancer 2005;117:861–5. 2013;24:1697–703. 32. Le Bert N, Gasser S. Advances in NKG2D ligand recognition and responses 15. Hannani D, Vetizou M, Enot D, Rusakiewicz S, Chaput N, Klatzmann D, by NK cells. Immunol Cell Biol 2014;92:230–6. et al. Anticancer immunotherapy by CTLA-4 blockade: obligatory contri- 33. Raulet DH, Gasser S, Gowen BG, Deng W, Jung H. Regulation of ligands for bution of IL-2 receptors and negative prognostic impact of soluble CD25. the NKG2D activating receptor. Annu Rev Immunol 2013;31:413–41. Cell Res 2015;25:208–24. 34. Baginska J, Viry E, Paggetti J, Medves S, Berchem G, Moussay E, et al. The 16. Yuan J, Zhou J, Dong Z, Tandon S, Kuk D, Panageas KS, et al. Pretreatment critical role of the tumor microenvironment in shaping natural killer cell- serum VEGF is associated with clinical response and overall survival in mediated anti-tumor immunity. Front Immun 2013;4:490. advanced melanoma patients treated with ipilimumab. Cancer Immunol 35. Labiano S, Palazon A, Melero I. Immune response regulation in the tumor Res 2014;2:127–32. microenvironment by hypoxia. Semin Oncol 2015;42:378–86.

5092 Cancer Res; 75(23) December 1, 2015 Cancer Research

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2015 American Association for Cancer Research. Serum Immunoregulatory Proteins as Predictors of Overall Survival of Metastatic Melanoma Patients Treated with Ipilimumab

Yoshinobu Koguchi, Helena M. Hoen, Shelly A. Bambina, et al.

Cancer Res 2015;75:5084-5092.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/75/23/5084

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2015/10/14/0008-5472.CAN-15-2303.DC1

Cited articles This article cites 35 articles, 6 of which you can access for free at: http://cancerres.aacrjournals.org/content/75/23/5084.full#ref-list-1

Citing articles This article has been cited by 4 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/75/23/5084.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/75/23/5084. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.