Purpose:

PD-1 checkpoint blockade immunotherapy induces long and durable response in patients with advanced melanoma. However, only a subset of patients with melanoma benefit from this approach. The mechanism triggering the innate resistance of anti–PD-1 therapy remains unclear.

Experimental Design: Whole-exome sequencing (WES) and RNA sequencing (RNA-Seq) analyses were performed in a training cohort (n = 31) using baseline tumor biopsies of patients with advanced melanoma treated with the anti–PD-1 antibody. Copy-number variations (CNVs) for the genes CDK4, CCND1, and CDKN2A were assayed using a TaqMan copy-number assay in a validation cohort (n = 85). The effect of CDK4/6 inhibitors combined with anti–PD-1 antibody monotherapy was evaluated in PD-1–humanized mouse (C57BL/6-hPD-1) and humanized immune system (HIS) patient-derived xenograft (PDX) models.

Results:

WES revealed several significant gene copy-number gains in the patients of no clinical benefit cohort, such as 12q14.1 loci, which harbor CDK4. The association between CDK4 gain and innate resistance to anti–PD-1 therapy was validated in 85 patients with melanoma (P < 0.05). RNA-Seq analysis of CDK4-normal cell lines and CDK4-normal tumors showed altered transcriptional output in TNFα signaling via NF-κB, inflammatory response, and IFNγ response gene set. In addition, CDK4/6 inhibitor (palbociclib) treatment increased PD-L1 protein levels and enhanced efficacy (P < 0.05) in the C57BL/6-hPD-1 melanoma cell and the HIS PDX model.

Conclusions:

In summary, we discovered that genetic aberrations in the CDK4 pathway are associated with innate resistance to anti–PD-1 therapy in patients with advanced melanoma. Moreover, our study provides a strong rationale for combining CDK4/6 inhibitors with anti–PD-1 antibody for the treatment of advanced melanomas.

Translational Relevance

Anti–PD-1 antibodies have received considerable attention as effective therapies for advanced melanoma. However, the majority of the patients failed to response to this approach, due to genetic aberrations in the key signaling pathways. Here, based on the whole-exome sequencing and RNA-sequencing profiling, we identified obvious CDK4, CCND1 gain, and CDKN2A loss in the patients with melanoma with resistance to anti–PD-1 immunotherapy. Using the B16 cells–implanted C57BL/6-Pdcd1tm1(hPDCD1)/Bcgen (C57BL/6-hPD-1) and anti–PD-1–resistant melanoma-derived humanized immune system patient-derived xenograft mouse model, we found that CDK4/6 inhibitors activated the immune system. Importantly, combination of CDK4/6 inhibitors and anti–PD-1 antibodies significantly suppressed the tumor growth in both mouse models. We conclude that CDK4/6 can be targeted in advanced melanoma to improve the efficacy of immunotherapy.

As an immune checkpoint blockade, anti–PD-1/PD-L1 antibodies have improved outcomes in many cancer types. This approach, which has been approved for the treatment of metastatic melanoma, non–small cell lung cancer (NSCLC), head and neck cancer, and urothelial carcinoma, may cause release of endogenous antitumor immunity, eliminate tumor cells, and mediate durable tumor regression (1). Previous trials have revealed that 35%–60% of patients with melanoma can achieve clinical response, based on RECIST, to anti–PD-1 mAb treatment (2, 3). Unfortunately, 40%–65% of patients have none or short-lived responses, and 43% of patients experience drug resistance within 3 years (3). On the basis of a phase II anti–PD-1 clinical trial that our center reported at the American Society of Clinical Oncology (ASCO) meeting, the overall objective response rate (ORR) of a cohort of predominantly acral and mucosal melanoma was 20.7 overall, which is lower than that of the Caucasian population. Within acral and mucosal subtypes, the response rate was clearly lower than that observed in cutaneous melanomas (14.3% and 0%, respectively; ref. 4). It is vital to understand the resistance mechanisms of immunotherapy and to explore new strategies to overcome treatment resistance in various melanoma populations.

To date, most studies of anti–PD-1 therapeutic resistance have mainly concentrated on immune factors, such as PD-L1 expression and tumor-infiltrating CD8+ T cells (5, 6). However, somatic genetic factors may play equally vital roles, and several studies have revealed that the mutational burden may also be associated with the response of immunotherapy (7, 8). Moreover, tumors with a high mutational burden, such as melanoma, NSCLC, and bladder cancer, have been correlated with superior clinical responses to immune checkpoint blockade compared with tumors with lower mutational rates (9, 10). This may be because a greater number of genetic variants increase the probability of stimulating tumor-specific antigens (neoantigens), leading to superior antitumor immune response during immunotherapy (11). There are significant differences in the incidence, gene spectrum, and clinical manifestations between Caucasian and non-Caucasian populations (12–14). For example, a whole-genome sequencing (WGS) study revealed that acral and mucosal melanomas, which are the predominant subtypes among non-Caucasians, are dominated by somatic structural variants, whereas the burden of point and structural mutations is not heavy compared with that of cutaneous melanoma (15). The major somatic variants include KIT, TERT, NRAS, EZH2, and CDK4 in non-Caucasians melanomas (16–20). Such distinct molecular differences may affect the immune microenvironment and responsiveness to immunotherapy.

Aberrant cell-cycle progression is a hallmark of cancer cells. Cyclin-dependent kinase 4 (CDK4), or its close homolog CDK6, forms an active complex with Cyclin D that phosphorylates and inhibits the Rb protein, promoting the G1–S-phase transition of the cell cycle (21). Commonly dysregulated in most cancers, this CDK4 pathway (p16Ink4A-Cyclin D-CDK4-Rb) is one of the drivers of melanoma (15, 22). Recent studies have shown that genetic alterations of CDK4 pathway components are frequently found in acral melanoma, with a higher prevalence than in cutaneous melanoma, suggesting that the CDK4 pathway may be an effective therapeutic target in melanoma in the non-Caucasian population (13, 15, 20). Furthermore, recent studies have demonstrated a correlation between the CDK4 pathway and tumor immunity. Zhang and colleagues found that cyclin D-CDK4 regulates PD-L1 protein abundance to affect cancer immune surveillance (23). Deng and colleagues reported that CDK4/6 inhibition enhances T-cell activation via derepression of nuclear factor of activated T cell (NFAT) family proteins to augment antitumor immunity (24). However, the above studies are still at laboratory research stage. Explorations of clinical data and preclinical studies are required.

In this study, we analyzed clinical samples in an effort to explore factors that may influence innate sensitivity or resistance to anti–PD-1 therapy and to screen druggable targets with the potential to be combined with anti–PD-1 antibody monotherapy for treatment of advanced melanoma.

Patients and tissue samples

In this study, we obtained 31 tumor biopsies, 116 blood samples, 24 normal control tissues, and 85 formalin-fixed, paraffin-embedded (FFPE) tumor tissues from baseline lesions of 116 patients with advanced melanoma who were treated with an anti–PD-1 antibody (JS001, Shanghai Junshi Biosciences, clinical trial ID: NCT02836795 and NCT03013101) at Peking University Cancer Hospital (Beijing, China) between March 2016 and December 2017. Informed consent for the use of materials in the study (including archived materials, as well as the establishment of cell lines and tumor models) was obtained from all patients enrolled in clinical trials. Informed written consent was obtained from each subject or each subject's guardian. Clinical data, including sex, age, tumor side, tumor thickness, metastasis status, and clinical efficacy, were collected. Clinical efficacy was evaluated per RECISTv1.1 by the independent review committee (IRC) every 8 weeks. Diagnosis was histopathologically confirmed for all patients. The flow chart describing the total population of the study is shown in Supplementary Fig. S1. This study was approved by the medical ethics committee of Peking University Cancer Hospital and Institute (Beijing, China) and was conducted according to Declaration of Helsinki Principles. The institutional review board number is 2016KT50.

WES

Genomic DNA was extracted from 31 matched tumor biopsies and blood samples using DNeasy Blood & Tissue Kit (Qiagen). Exome capture was performed using SureSelect Human All Exon V6 (Agilent Technologies) according to the manufacturer's instructions. The quantity of libraries was assessed using Qubit 2.0 Fluoromete. The library quality and size were measured using a 2100 Bioanalyzer high sensitivity DNA assay according to the manual. For Illumina sequencing, qualified libraries were applied for 2 × 150 bp paired-end sequencing using the Illumina HiSeq X-ten Platform (Illumina). All whole-exome capture, sequencing, and data analyses were performed at Shanghai Biotechnology Corporation. Bioinformatics analysis of WES results can be found in the Supplementary Materials and Methods.

RNA-Seq

Total RNA was isolated from 24 matched tumor biopsies and normal control tissues using RNeasy Mini Kit (Qiagen) following the manufacturer's instructions. Strand-specific libraries were prepared using the TruSeq Stranded Total RNA Sample Preparation Kit (Illumina) according to the manufacturer's instructions. The purified libraries were quantified using a Qubit 2.0 Fluorometer (Life Technologies) and validated with an Agilent 2100 Bioanalyzer (Agilent Technologies) to confirm the insert size and to calculate the mole concentration. Clustering was generated with cBot using the library diluted to 10 pmol/L and then sequenced using the Illumina HiSeq X-ten System (Illumina). Library construction and sequencing were performed at Shanghai Biotechnology Corporation. Data analysis for gene expression based on RNA-Seq is described in the Supplementary Materials and Methods.

QuantiGenePlex RNA assay (panel)

All experiments were performed at Shanghai Biotechnology Corporation following the user manual of QuantiGene Plex Assay (Panomics). See details in Supplementary Materials and Methods.

DNA preparation and TaqMan copy-number assays

The copy numbers of CDK4, CCND1, and CDKN2A were calculated using TaqMan Copy Number Assays (Applied Biosystems of Thermo Fisher Scientific); the Rnasep gene TaqMan probe was used as a control. See details in Supplementary Materials and Methods.

FISH

CDK4 gain, CCND1 gain, and CDKN2A loss were identified using array-based comparative genomic hybridization. Interphase FISH was performed using CDK4/CEN12q FISH Probe (FG0033, Abnova) with a CCND1/CEP11 amplification probe (Vysis) and a CDKN2A/CEP9 loss probe (Vysis). The imprint slides were covered with 10 mL of dual-probe hybridization mixture containing a pair of chromosome probes and subtelomere probes. The slides and probes were denatured at 75°C for 5 minutes and hybridized at 37°C for 16 hours. Post-hybridization washes were performed according to the manufacturer's protocol.

IHC

IHC analyses were performed using antibodies against CD4 (ab183685, Abcam), CD8 (ab217344, Abcam), B220 (14-0452-82, eBioscience), FOXP3 (14-5773-80, eBioscience), and PD-L1 (SP142 and M4422, Spring), as described previously (25). Staining intensity and percentage were independently scored by three pathologists. We used ImageJ software to quantify the IHC results.

Immunofluorescence

Immunofluorescence (IF) analyses were performed using antibodies against CD4 (ab196372, Abcam), CD8 (ab4055, Abcam), CD19 (ab134114, Abcam), and FOXP3 (ab20034, Abcam).

Immunology multiplex cytokine/chemokine profiling

A panel of 38 cytokines, chemokines, and growth factors in plasma was analyzed using MILLIPLEX MAP Human Cytokine/Chemokine Magnetic Bead Panel, Premixed 38 Plex, Immunology Multiplex Assay (HCYTMAG-60K-PX38, Merck). Measurements were performed using a Bio-Plex MAGPIX Multiplex Reader (Bio-Rad Laboratories).

NanoString-based gene expression profiling

Tumor RNA samples were subjected to nCounter PanCancer Immune Profiling Panel (NanoString Technologies) analysis consisting of 770 human genes. See details in Supplementary Materials and Methods.

Cell lines and primary cell culture

A2058 (catalog No. CRL-11147), SK-MEL-5 (catalog No. HTB-70), and WM-266-4 (catalog No. CRL-1676) cell lines were obtained from ATCC. B16-OVA cell lines were a gift from Shanghai Junshi Biosciences. AMC-3 (acral primary melanoma cell) cell lines were derived from a patient-derived xenograft (PDX) model. See details in Supplementary Materials and Methods.

Western blotting

All cells were harvested and lysed using the PhosphoSafe Extraction Reagent (Millipore). Supernatants were collected by centrifugation. Western blot analysis of protein complexes was performed using antibodies against the following: PD-L1 (catalog No. 13684, Cell Signaling Technology) and mouse PD-L1/B7-H1 (MAB90781-100, R&D Systems).

Animals

Humanized PD-1–knockin C57Bl/6 mice (C57BL/6-Pdcd1tm1(hPDCD1)/Bcgen and C57BL/6-hPD-1) were purchased from Biocytogen Gene Biotechnology Co., Ltd. The C57BL/6-hPD-1 mice were developed by knocking-in human PDCD1 protein coding region inserted into the ATG site of murine Pdcd1 gene, the expression of endogenous murine Pdcd1 was replaced by the expression of human full-length PDCD1 protein, and PD-1 expression was detected by flow cytometry (Supplementary Fig. S2). NOD/SCID/IL2 receptor γ chain null mice (NOD/SCID/IL2γnull, NSG) were obtained from the Vital River Laboratory Animal Technology Co., Ltd. and bred and raised under specific pathogen-free conditions. All animal care and experimental procedures were carried out in accordance with the Animal Care Ethics approved by the Medical Ethics Committee of the Beijing Cancer Hospital & Institute.

C57BL/6-hPD-1 mouse model and treatment

B16-OVA cells (1 × 105) were subcutaneously injected into 6-week-old C57BL/6-hPD-1 female mice. Six days later, the mice were daily treated with palbociclib (150 mg/kg body weight, by gastric gavage), the anti–PD-1 antibody (pembrolizumab 200 μg, every 3 days), or vehicle (sodium lactate buffer solution) for 12 days. The CDK4/6 inhibitor, palbociclib (PD0332991, catalog No. S1579) was purchased from Selleck Chemicals. Tumor size and mouse weight were measured every 3 days and tumor volume was calculated using the following formula: volume = length × width2/2. Samples of tumor and spleen were processed for FACS analysis or IHC.

Humanized immune system PDX model and treatment

NSG mice were sublethally irradiated with 200 cGy. Human peripheral blood monocular cells (PBMC) were isolated using Ficoll Hypaque centrifugation from a healthy donor and transplanted (1 × 107) into the irradiated NSG mice via tail vein injection.

One week after PBMC transplantation, the fragments of patient-derived melanoma tissues were subcutaneously inoculated into the PBMC-humanized NSG mice. When the tumor volume reached approximately 70 mm3 and hCD45+ cells constituted about 5% of the peripheral blood, the mice were treated with palbociclib (50 mg/kg, orally, daily dosing), the anti–PD-1 antibody (200 μg, i.p., every 3 days), or vehicle (sodium lactate buffer solution) for 21 days. Tumor size and mouse body weight were measured twice a week, and tumor volume was calculated using the following formula: volume = length × width2/2. Samples of tumor and spleen were processed for FACS analysis or IF. The Humanized immune system (HIS) PDX experiment was performed at Yicon Biomedical Technology Inc. The similar HIS PDX mouse model was described previously (26, 27).

FACS

FITC-, PE-, PE-Cyanine7-, APC-, APC-eFluor-, PerCP-, or PerCP-Cyanine5.5–conjugated anti-CD3, -CD4, -CD8, -CD19, -CD45, and -CD49b mAbs were used to stain lymphocytes (all from eBioscience).

Statistical analysis

The statistical evaluation was conducted with IBM SPSS statistical software (v20.0). t Test was used to analyze mean values for normally distributed continuous variables, whereas the Mann–Whitney U test was applied to compare mean values for abnormally distributed continuous variables. Correlations between the aberration status and clinical parameters were evaluated by χ2 test or Fisher exact test. For all statistical tests, P < 0.05 (two-tailed test) was considered statistically significant.

Patient cohort, checkpoint blockade treatment, and tumor biopsies

To explore differential genetic aberrations associated with the response of advanced melanoma to anti–PD-1 immune checkpoint blockade, we assembled a test cohort of 116 patients with advanced melanoma who were treated with this therapy. Among them, 31 patients were classified as a training cohort and 85 patients as a validation cohort. The clinical pathologic parameters of the patients are shown in Supplementary Table S1. Clinical efficacy was evaluated per RECIST 1.1 by the IRC every 8 weeks. Clinical benefit (CB) cohort included patients with a complete response (CR) or partial response (PR) according to RECIST 1.1 (i.e., tumor shrinkage > 30% from baseline) or stable disease (SD) if they had any objective reduction in tumor burden lasting at least 6 months. No clinical benefit (NCB) were defined as those experiencing progressive disease according to RECIST 1.1 or SD lasting ≤6 months and were discontinued from immunotherapy within 3 months (28, 29). As of November 30, 2018, among 116 evaluable patients with at least three IRC evaluations for clinical efficacy, one achieved CR, 21 PR, and 42 SD.

CDK4 pathway aberrations associate significantly with NCB to PD-1 blockade

To investigate genetic variation associated with innate resistance to anti–PD-1 therapy, genomic DNA extracted from tumors and matched peripheral blood cells of 31-sample cohort were subjected to WES, including 15 patients of CB cohort and 16 NCB. The mean sequencing depth in the training cohort was 338-fold for tumor and 135-fold for matched blood. At least 93.27% of exome-wide target bases were covered to a depth of more than 10 ×. The WES demonstrated that the overall nonsynonymous mutation burden was moderate in the training cohort with a median number of 90 (ranging from 6 to 3,068). The tumors of patients in the CB and NCB groups showed similar mutation burdens (Fig. 1A). The overall mutations and copy-number changes in selected known melanoma driver genes are shown in Fig. 1B and C. Frequencies of alteration in various key genes and signaling pathways between CB and NCB cohort are shown in Fig. 1D–F. Alterations in the cell-cycle pathway occurred more frequently in NCB cohort than CB cohort. The copy-number alterations affecting antigen-presentation machinery and HLA class I alleles were frequent in tumors of both groups, whereas mutations were relatively uncommon (Fig. 1G).

Figure 1.

Diversity of mutations and copy-number changes between patients in the CB and NCB cohorts. A, Mutation burden in the CB and NCB groups. B, Mutations in significantly mutated genes and selected known melanoma driver genes of CB and NCB groups. C, Copy-number changes in selected melanoma-associated genes of CB and NCB groups. D and E, Percentage of samples with protein-affecting aberrations in candidate driver genes, as grouped by pathway: copy-number amplification (copy number > 3, red), copy-number deletion (copy number < 2, blue), structural variants (yellow). F, Frequency of aberrations in pathways as a percentage of CB and NCB groups. SWI/SNF, SWItch/sucrose nonfermentable nucleosome remodeling complex. G, Alterations in HLA alleles and antigen presentation machinery in CB and NCB groups. H and I, According to GISTIC2.0 output, there were significant copy-number changes in the CB and NCB groups.

Figure 1.

Diversity of mutations and copy-number changes between patients in the CB and NCB cohorts. A, Mutation burden in the CB and NCB groups. B, Mutations in significantly mutated genes and selected known melanoma driver genes of CB and NCB groups. C, Copy-number changes in selected melanoma-associated genes of CB and NCB groups. D and E, Percentage of samples with protein-affecting aberrations in candidate driver genes, as grouped by pathway: copy-number amplification (copy number > 3, red), copy-number deletion (copy number < 2, blue), structural variants (yellow). F, Frequency of aberrations in pathways as a percentage of CB and NCB groups. SWI/SNF, SWItch/sucrose nonfermentable nucleosome remodeling complex. G, Alterations in HLA alleles and antigen presentation machinery in CB and NCB groups. H and I, According to GISTIC2.0 output, there were significant copy-number changes in the CB and NCB groups.

Close modal

According to GISTIC2.0 output of single cohort of samples, WES revealed several significant gene copy-number gains in the NCB group, such as 5p15.33, 7p22.3, 7q36.1, and 12q14.1 loci, which harbor CDK4, TERT, CARD11, and EZH2. In addition, several significant gene copy-number losses were observed in the NCB group, such as 9q21.3 and 10q23.31 loci, which harbor CDKN2A and PTEN (Fig. 1H and I).

Confirmation of CDK4 pathway aberration associated with innate resistance to anti–PD-1 therapy

To validate the findings from WES, we investigated CNVs in CDK4, CCND1, and CDKN2A in an independent cohort (n = 85). On the basis of data of 85 patients, we found a negative correlation between CDK4 copy-number gain and anti–PD-1 response, moreover, a positive correlation between CDK4 copy-normal and anti–PD-1 response (P = 0.013; Fig. 2A; Table 1). Although patients with CCND1 copy-number gain and CDKN2A copy-number loss tend to be nonresponsive to anti–PD-1 therapy, the difference was not statistically significant (P = 0.643, 0.318, respectively; Fig. 2A; Table 1). Among 32 patients with acral subtype melanoma, CCND1 copy-number gain was associated with a lack of response to anti–PD-1 therapy (P = 0.043); among 13 patients with unknown primary subtype melanoma, CDKN2A copy-number loss was associated with nonresponse to anti–PD-1 therapy (P = 0.021; Table 1). We also performed FISH to confirm the observed CDK4 gain, CCND1 gain, and CDKN2A loss (Fig. 2B).

Figure 2.

Correlation between CDK4 pathway aberrations and CB of anti–PD-1 therapy in melanoma. A, Genetic aberrations in the CDK4 pathway and response to anti–PD-1 therapies based on sample. CB, n = 33; NCB, n = 52. B, FISH analysis of CNVs in CDK4, CCND1, and CDKN2A. C, RNA expression levels of CDK4, CCND1, and CDKN2A in tumor biopsy specimens of NCB and CB groups according to data from the QuantiGenePlex RNA assay. CB, n = 22; NCB, n = 12. D, RNA expression levels of CDK4, CCND1, and CDKN2A in FFPE tumor tissue specimens from NCB and CB groups according to the QuantiGenePlex RNA assay (CB, n = 16; NCB, n = 26; *, P < 0.05).

Figure 2.

Correlation between CDK4 pathway aberrations and CB of anti–PD-1 therapy in melanoma. A, Genetic aberrations in the CDK4 pathway and response to anti–PD-1 therapies based on sample. CB, n = 33; NCB, n = 52. B, FISH analysis of CNVs in CDK4, CCND1, and CDKN2A. C, RNA expression levels of CDK4, CCND1, and CDKN2A in tumor biopsy specimens of NCB and CB groups according to data from the QuantiGenePlex RNA assay. CB, n = 22; NCB, n = 12. D, RNA expression levels of CDK4, CCND1, and CDKN2A in FFPE tumor tissue specimens from NCB and CB groups according to the QuantiGenePlex RNA assay (CB, n = 16; NCB, n = 26; *, P < 0.05).

Close modal
Table 1.

Correlation of CDK4 pathway aberrations with a CB of anti–PD-1 therapy in melanoma

CDK4 AberrationCCND1 AberrationCDKN2A Aberration
GainNormalPaGainNormalLossPaLossNormalPa
CB (%) 
 CB 16 (18.8) 17 (20) 0.013 14 (16.5) 8 (9.4) 7 (8.2) 0.643 18 (21.2) 15 (17.6) 0.318 
 NCB 39 (45.9) 13 (15.3)  30 (35.3) 11 (12.9) 15 (17.6)  34 (40) 18 (21.2)  
Acral subtype (%) 
 CB 6 (18.8) 3 (9.4) 0.398 3 (9.4) 1 (3.1) 5 (5.9) 0.043 6 (18.8) 3 (9.4) 0.638 
 NCB 18 (21.2) 5 (5.9)  17 (20) 4 (12.5) 2 (6.3)  15 (17.6) 8 (25)  
Mucosal subtype (%) 
 CB 1 (7.6) 1 (7.6) 0.154 1 (7.6) 0 (0) 1 (7.6) 0.538 0 (0) 2 (15.3) 0.192 
 NCB 11 (84.6) 0 (0)  3 (23) 3 (23) 5 (38.4)  7 (53.8) 4 (30.8)  
CSD subtype (%) 
 CB 3 (17.6) 4 (23.5) 0.646 4 (23.5) 2 (11.8) 1 (5.9) 0.581 5 (29.4) 2 (11.8) 0.354 
 NCB 4 (23.5) 6 (35.3)  5 (29.4) 4 (23.5) 1 (5.9)  5 (29.4) 5 (29.4)  
Non-CSD subtype (%) 
 CB 3 (30) 3 (30) 0.452 2 (20) 2 (20) 2 (20) 0.262 5 (50) 1 (10) 0.667 
 NCB 3 (30) 1 (10)  3 (30) 0 (0) 1 (10)  3 (30) 1 (10)  
Unknown primary subtype (%) 
 CB 3 (23.1) 6 (46.2) 0.217 4 (30.7) 3 (23.1) 2 (15.4) 0.657 2 (15.4) 7 (53.8) 0.021 
 NCB 3 (23.1) 1 (7.7)  2 (15.4) 0 (0) 2 (15.4)  4 (30.7) 0 (0)  
CDK4 AberrationCCND1 AberrationCDKN2A Aberration
GainNormalPaGainNormalLossPaLossNormalPa
CB (%) 
 CB 16 (18.8) 17 (20) 0.013 14 (16.5) 8 (9.4) 7 (8.2) 0.643 18 (21.2) 15 (17.6) 0.318 
 NCB 39 (45.9) 13 (15.3)  30 (35.3) 11 (12.9) 15 (17.6)  34 (40) 18 (21.2)  
Acral subtype (%) 
 CB 6 (18.8) 3 (9.4) 0.398 3 (9.4) 1 (3.1) 5 (5.9) 0.043 6 (18.8) 3 (9.4) 0.638 
 NCB 18 (21.2) 5 (5.9)  17 (20) 4 (12.5) 2 (6.3)  15 (17.6) 8 (25)  
Mucosal subtype (%) 
 CB 1 (7.6) 1 (7.6) 0.154 1 (7.6) 0 (0) 1 (7.6) 0.538 0 (0) 2 (15.3) 0.192 
 NCB 11 (84.6) 0 (0)  3 (23) 3 (23) 5 (38.4)  7 (53.8) 4 (30.8)  
CSD subtype (%) 
 CB 3 (17.6) 4 (23.5) 0.646 4 (23.5) 2 (11.8) 1 (5.9) 0.581 5 (29.4) 2 (11.8) 0.354 
 NCB 4 (23.5) 6 (35.3)  5 (29.4) 4 (23.5) 1 (5.9)  5 (29.4) 5 (29.4)  
Non-CSD subtype (%) 
 CB 3 (30) 3 (30) 0.452 2 (20) 2 (20) 2 (20) 0.262 5 (50) 1 (10) 0.667 
 NCB 3 (30) 1 (10)  3 (30) 0 (0) 1 (10)  3 (30) 1 (10)  
Unknown primary subtype (%) 
 CB 3 (23.1) 6 (46.2) 0.217 4 (30.7) 3 (23.1) 2 (15.4) 0.657 2 (15.4) 7 (53.8) 0.021 
 NCB 3 (23.1) 1 (7.7)  2 (15.4) 0 (0) 2 (15.4)  4 (30.7) 0 (0)  

Abbreviations: CSD, melanomas on skin with chronic sun-induced damage; non-CSD, melanoma on skin without sun-induced damage.

aP values of the χ2 test or Fisher exact test were used for evaluating the CB.

Furthermore, we performed QuantiGenePlex RNA analysis to investigate differential RNA expression levels of CDK4, CCND1, and CDKN2A between the CB and NCB groups. On the basis of 34 tumor biopsy specimens, the RNA expression levels of CDK4, CCND1, and CDKN2A in the NCB group were higher than those in the CB group (Fig. 2C; 0.28 vs. 0.09, P = 0.03; 1.27 vs. 0.32, P = 0.05; 0.033 vs. 0.018, P = 0.2, respectively). Similar results were found in 42 FFPE tumor tissue specimens (Fig. 2D). The tumor biopsies in Fig. 2C were mainly metastatic lesions, while most of the FFPE tumor tissue specimens in Fig. 2D were primary melanomas (Supplementary Table S2).

Correlation of CDK4 pathway aberrations with PD-L1 protein expression in patients with melanoma

Because PD-L1 expression has been demonstrated to partially predict the efficacy of PD-1 mAb, we investigate the association between CDK4 pathway aberrations with PD-L1 protein expression. We performed IHC to detect levels of PD-L1 protein expression, in which PD-L1 ≥ 5% was defined as positive staining (Fig. 3A). The overall PD-L1–positive rate was 23.3%. Patients with greater levels of PD-L1 expression by IHC had higher response rates (P = 0.024), with a response rate for PD-L1–positive versus -negative patients was 64.7% versus 33% (Supplementary Table S3). Moreover, patients with CDK4 gain showed a lower PD-L1–positive rate than those with a CDK4-normal status (20.8% vs. 28%; P = 0.492), although the difference was not significant (Supplementary Table S3).

Figure 3.

Correlation of CDK4 pathway aberrations with immune gene expression. A, Representative images of melanoma tumor cells with PD-L1 expression. B, Differential expression of genes between the cohort of CDK4-gain group and CDK4-normal group. C, Differential expression of genes likely to be associated with PD-1 mAb treatment was reported between the cohort of CDK4-gain group and CDK4-normal group. D, GSEA was performed on RNA-Seq from tumor biopsy samples from patients with CDK4-gain and CDK4-normal. Enrichment plots show differential expression of TNFα signaling via NF-κB, inflammatory response, and IFNγ response gene set in the CDK4-normal tumors. E, GSEA was performed on RNA-Seq from tumor cell line with CDK4-gain and CDK4-normal. Enrichment plots show differential expression of TNFα signaling via NF-κB, and IL6/JAK/STAT3 signaling pathway gene set in CDK4-normal cell line (SK-MEL-5). F, Heatmap representation of differentially expressed genes (P < 0.05) between CDK4-gain and CDK4-normal tumors based NanoString analysis. G, Differential expression of cytokines, chemokines, and growth factors in plasma between CDK4-gain and CDK4-normal tumors based on Immunology Multiplex cytokine/chemokine profiling. CDK4 normal, n = 38; CDK4 gain, n = 47. The two independent sample t test was used to determine significance.

Figure 3.

Correlation of CDK4 pathway aberrations with immune gene expression. A, Representative images of melanoma tumor cells with PD-L1 expression. B, Differential expression of genes between the cohort of CDK4-gain group and CDK4-normal group. C, Differential expression of genes likely to be associated with PD-1 mAb treatment was reported between the cohort of CDK4-gain group and CDK4-normal group. D, GSEA was performed on RNA-Seq from tumor biopsy samples from patients with CDK4-gain and CDK4-normal. Enrichment plots show differential expression of TNFα signaling via NF-κB, inflammatory response, and IFNγ response gene set in the CDK4-normal tumors. E, GSEA was performed on RNA-Seq from tumor cell line with CDK4-gain and CDK4-normal. Enrichment plots show differential expression of TNFα signaling via NF-κB, and IL6/JAK/STAT3 signaling pathway gene set in CDK4-normal cell line (SK-MEL-5). F, Heatmap representation of differentially expressed genes (P < 0.05) between CDK4-gain and CDK4-normal tumors based NanoString analysis. G, Differential expression of cytokines, chemokines, and growth factors in plasma between CDK4-gain and CDK4-normal tumors based on Immunology Multiplex cytokine/chemokine profiling. CDK4 normal, n = 38; CDK4 gain, n = 47. The two independent sample t test was used to determine significance.

Close modal

CDK4 gain status in melanoma influences immune gene expression

To further explore the consequences of CDK4 gene copy-number gain, the tumor biopsies of 24 patients treated with PD-1 mAb were subjected to RNA-Seq analysis. After performing differentially expressed analysis between CDK4 gain tumors and CDK4-normal tumors, and gene set enrichment analysis (GSEA) for differentially expressed genes, we found that three gene sets, including TNFα signaling via NF-ĸB, inflammatory response, and IFNγ response, were enriched in CDK4-normal tumors (P < 0.05; q < 0.25; Fig. 3B–D). As for cell lines, compared with the cell line with CDK4 gain (SK-MEL-5), the CDK4-normal cell line (A2058) highly expressed genes enriched in TNFα signaling via NF-κB and IL6/JAK/STAT3 signaling pathway (P < 0.05 but q > 0.25; Fig. 3E).

Furthermore, we performed NanoString-based gene expression profiling and Immunology Multiplex cytokine/chemokine profiling. Among 55 patients, NanoString-based gene expression profiling revealed high expression levels of IL17A, TNFRSF10C, CCL28, MICB, LICRB3, CREB1, NOTCH1, and IL6 among the tumor specimen of patients with CDK4 gain (Fig. 3F). Immunology Multiplex cytokine/chemokine profiling was performed to verify these results, and patients' tumors with CDK4 gain had higher levels of IL17A and IL6 in plasma than did patients with a CDK4-normal status (59.7 vs. 33.4, P = 0.016; 34.9 vs. 27.8, P = 0.038), as well as lower levels of eotaxin (260.0 vs. 339.2; P = 0.044; Fig. 3G).

Effects of combined CDK4/6 inhibitors and anti–PD-1 on the growth of melanoma in the C57BL/6-hPD-1 model

To further explore the therapeutic effects of CDK4 inhibitors in combination with PD-1, we performed in vivo experiments. C57BL/6-hPD-1 mice were subcutaneously injected with B16-OVA cells. Six days later, the mice were treated with either vehicle, palbociclib, the anti–PD-1 antibody, or a combination of palbociclib and the antibody for 12 days. According to tumor measurements over the treatment period, there were significant reductions in growth with the combined therapy of palbociclib and the anti–PD-1 antibody compared with monotherapy or the control (P < 0.05; Fig. 4A).

Figure 4.

Effects of combined CDK4/6 inhibitors and anti–PD-1 treatment on the growth of melanoma in the C57BL/6-hPD-1 model. A, Sensitivity of the C57BL/6-hPD-1 PDX model to combined PD-1 blockade and CDK4/6 inhibitors in vivo. n = 5 per group. B, Histology of melanoma with enhanced presence of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells in tumors of C57BL/6-hPD-1 PDX models treated with palbociclib or combined therapy. H&E, hematoxylin and eosin. C, The IHC results of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells were presented as mean ± SE. D, FACS analysis of immune cells from melanoma showed increased leukocytes in anti–PD-1 antibody, palbociclib, and combined treatment groups, with significant increases in B cells, T cells, Tc cells, Th cells, NK cells, and NKT cells in palbociclib alone–treated and combined treatment groups. Error bars indicate SD, and P values were determined by Student–Newman–Keuls (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant).

Figure 4.

Effects of combined CDK4/6 inhibitors and anti–PD-1 treatment on the growth of melanoma in the C57BL/6-hPD-1 model. A, Sensitivity of the C57BL/6-hPD-1 PDX model to combined PD-1 blockade and CDK4/6 inhibitors in vivo. n = 5 per group. B, Histology of melanoma with enhanced presence of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells in tumors of C57BL/6-hPD-1 PDX models treated with palbociclib or combined therapy. H&E, hematoxylin and eosin. C, The IHC results of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells were presented as mean ± SE. D, FACS analysis of immune cells from melanoma showed increased leukocytes in anti–PD-1 antibody, palbociclib, and combined treatment groups, with significant increases in B cells, T cells, Tc cells, Th cells, NK cells, and NKT cells in palbociclib alone–treated and combined treatment groups. Error bars indicate SD, and P values were determined by Student–Newman–Keuls (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant).

Close modal

Histology of the B16-OVA melanoma tumors revealed increased presence of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells in the palbociclib-alone group as well as in the group treated with the combination therapy, compared with PD-1 antibody–alone group or the control group (Fig. 4B). FACS analysis was performed to detect immune cells infiltrating into the tumors after therapy, with an increase in the leukocyte content in the palbociclib alone or combined treatment group. In addition, there was significant increase in B cells, T cells, cytotoxic T (Tc) cells, Th cells, NK (natural killer) cells, and NKT cells in the palbociclib alone–treated tumors and combined treatment tumors (Fig. 4C). However, there was no difference in the proportion of immune cells among the four treatment groups.

Effects of combined CDK4/6 inhibitors and anti–PD-1 treatment on the growth of melanoma in the HIS PDX model

The results for B16-OVA mouse tumors prompted us to assess how palbociclib affects tumor growth and antitumor immunity in a patient-derived melanoma model. Thus, we generated the HIS PDX model, in which tumors from patients with advanced melanoma were engrafted with the human PBMCs into irradiated NSG mice. This patient is in the NCB cohort, with CDK4 gain. Tumor growth curves for each treatment group demonstrated the improved efficacy of combining PD-1 blockade with the CDK4/6 inhibitor in this model, compared with the monotherapy or the control group (P < 0.05; Fig. 5A).

Figure 5.

Effects of combined CDK4/6 inhibitors and anti–PD-1 therapy on the growth of melanoma in the HIS PDX model. A, Sensitivity of the HIS PDX model to combined PD-1 blockade and CDK4/6 inhibitors in vivo. n = 5 per group. B, Histology of melanoma with enhanced presence of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ in tumors of HIS PDX models treated with palbociclib or combined therapy. H&E, hematoxylin and eosin. C, The IF results of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells were presented as mean ± SE. D, FACS analysis of immune cells from melanoma showed increased leukocytes in anti–PD-1 antibody, palbociclib, and combined treatment groups, with significant increases in immune cells, T cells, Tc cells, and Th cells in palbociclib alone–treated and combined treatment groups. Error bars indicate SD, and P values were determined by Student–Newman–Keuls (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant).

Figure 5.

Effects of combined CDK4/6 inhibitors and anti–PD-1 therapy on the growth of melanoma in the HIS PDX model. A, Sensitivity of the HIS PDX model to combined PD-1 blockade and CDK4/6 inhibitors in vivo. n = 5 per group. B, Histology of melanoma with enhanced presence of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ in tumors of HIS PDX models treated with palbociclib or combined therapy. H&E, hematoxylin and eosin. C, The IF results of B cells, CD4+ T cells, CD8+ T cells, and FoxP3+ cells were presented as mean ± SE. D, FACS analysis of immune cells from melanoma showed increased leukocytes in anti–PD-1 antibody, palbociclib, and combined treatment groups, with significant increases in immune cells, T cells, Tc cells, and Th cells in palbociclib alone–treated and combined treatment groups. Error bars indicate SD, and P values were determined by Student–Newman–Keuls (*, P < 0.05; **, P < 0.01; ***, P < 0.001; ns, not significant).

Close modal

The effects of combined treatment on B cells, Tc cells, Th cells, and regulatory T cells (Treg) in HIS PDX model tumors were greater than those of monotherapy (Fig. 5B). We performed immuno-phenotyping analysis by FACS in these tumors after therapy. Tumor-bearing mice treated with palbociclib and combined therapy exhibited an increase in immune cells, T cells, Tc cells, and Th cells of tumor (Fig. 5C).

CDK4/6 inhibitor treatment positively regulates PD-L1 protein expression and influences immune gene expression

The above in vivo experiments showed that CDK4 inhibitors can activate antitumor immunity and synergistically improve the therapeutic effect of PD-1 mAb. We conducted the following experiments to explore related mechanisms. SK-MEL-5 cell line (with CCND1 gain plus CDKN2A loss) and the primary acral melanoma line AMC-3 cell line (with CDK4 gain plus CDKN2A loss), which the two cell lines have been proven to be sensitive to palbociclib (20), were treated with palbociclib for 24 hours. Western blot assays showed that palbociclib treatment increased PD-L1 protein levels in the SK-MEL-5 cell line, AMC-3 cell line, C57BL/6-hPD-1, and HIS PDX murine models (Supplementary Fig. S3A).

RNA-Seq analysis revealed that palbociclib treatment enhanced immune gene expression in the samples of SK-MEL-5 cell line, AMC-3 cell line, and C57BL/6-hPD-1 murine model, including genes related to the IL6/JAK/STAT3 signaling pathway (P < 0.05 but q > 0.25; Supplementary Fig. S3B).

Anti–PD-1 therapy has revolutionized treatment for advanced melanoma and many other cancer types. However, only a minority of patients benefit from monotherapy, especially small, among the Asian melanoma population (4, 30, 31). According to data from a large-scale prospective anti–PD-1 clinical study published by our center at ASCO, the ORR of chronic sun-induced damage (CSD), non-CSD, mucosal, and acral melanomas was 35.3, 33.3, 0, and 14.3, respectively (4). Overall, the anti–PD-1 antibody appeared to be similarly efficacious for CSD and non-CSD in a Chinese population compared with a Caucasian population, with greater effects than the mucosal and acral subtypes (4). The lack of predictors of a response, mechanisms of therapeutic resistance, and effective combination therapies remains a challenge for treatment of patients with mucosal and acral subtypes. Here, we present the role of specific genetic aberrations in determining responses of advanced melanoma to anti–PD-1 therapy. We demonstrated that genetic aberrations in the CDK4 pathway were associated with the innate resistance to anti–PD-1 therapy in patients with advanced melanoma. Gene expression analysis showed that CDK4-normal tumors and cells exhibited enriched transcriptional output in TNFα signaling via NF-κB, inflammatory response, and IFNγ response gene set. Moreover, our in vivo study provided a rationale of combining CDK4/6 inhibitors with anti–PD-1 antibody for the treatment of advanced melanomas.

Our results showed a similar tumor mutation rate in the CB and NCB groups. Several studies have noted that the overall mutation load correlates with clinical responses to anti–PD-1 therapy in NSCLC, colon cancer, and melanoma (32–34), with a statistically significant difference in the median, and the range of mutation burden in tumors among CB significantly overlaps with the range in NCB (35). Nonetheless, there are still patients with a high mutation load with NCB (33). Moreover, previous research indicates high mutational load does not associate with tumor response to anti–PD-1 therapy but correlates with improved patient survival of metastatic melanoma. Therefore, additional genomic and immunologic features are likely to contribute to the response patterns of anti–PD-1 therapy.

We performed WES of 31 pretreated advanced melanoma tumors and found significant CDK4 gain and CDKN2A loss in the NCB group, and the association between CDK4 gain and innate resistance to anti–PD-1 therapy was validated in another melanoma cohort (n = 85), moreover, we also found that CB was associated with CDK4 normal in the validation cohort. We found CCND1 copy-number gain to be associated with a lack of response to anti–PD-1 therapy among 32 patients with acral subtype melanoma, and among 13 patients with unknown primary subtype melanoma, CDKN2A copy-number loss was associated with anti–PD-1 therapy nonresponse. Thus, the genetic aberrations in the CDK4 pathway are associated with innate resistance to PD-1 blockade. According to SNP array analyses of 46 melanoma cell lines and data from The Cancer Genome Atlas, Leonardelli and colleagues found that because of associated JAK2 allele losses, melanoma harboring allelic CDKN2A deletions may be more prone to develop resistance to immunotherapy. As drivers of somatic copy-number alterations, CDK4 amplifications have been verified in a series of cancers, such as melanoma (20), NSCLC (36), and urothelial carcinoma (37). In addition, PTEN loss is associated with the resistance to immunotherapy (38, 39). We found PTEN loss significantly occurred in the NCB group, but the correlation between PTEN loss and resistance to PD-1 blockade may require a large sample of cohorts for validation. Recently, several selective inhibitors of CDK4/6, such as palbociclib, ribociclib, and abemaciclib, have been approved by the FDA for the treatment of metastatic breast cancer and are in clinical trials for several other indications (40). Indeed, a series of preclinical experiments indicate the feasibility of using CDK4 as an antitumor target of melanoma (20, 41), and a recent clinical case report showed that two patients with metastatic melanoma with genetic aberrations in the CDK4 pathway achieved tumor control for over 6 months with palbociclib treatment (42).

To further explore the relevant mechanisms triggered by CDK4 gain, RNA-Seq was performed to investigate alterations induced by amplification of CDK4 pathway components, in which CDK4-normal cell lines and CDK4-normal tumors displayed altered transcriptional output in immune signaling pathways, such as TNFα signaling via NF-κB pathways, inflammatory response, and IFNγ response gene set, differential expression of these genes may impact CDK4-normal patients' response to anti–PD-1 therapy. Luoto and colleagues reported glioblastoma cases with focal amplification of CDK4 that presented negative adaptive immune responses that were associated with a lower macrophage and CD4+ T-cell component (43). In addition, recent studies have revealed essential roles of the noncanonical NF-κB pathway in regulating different aspects of immune function (44–46). For example, several members of the TNF receptor (TNFR) superfamily regulate the generation of immunosuppressive Tregs by controlling the development of medullary thymic epithelial cells (mTEC), which are known to mediate activation of both the canonical and noncanonical NF-κB pathways (47). Moreover, recent studies have shown that IFNγ is a key driver of PD-L1 expression, and the deficiency of the IFNγ signaling pathway has been linked with anti–PD-1 therapy resistance (48, 49). NanoString-based gene expression profiling of patients with C DK4 gain exhibited altered output in IL17A, TNFRSF10C, CCL28, MICB, LICRB3, CREB1, NOTCH1, and IL6. These molecules can provide clues for combination therapy. Furthermore, Immunology Multiplex cytokine/chemokine profiling confirmed higher levels of IL17A and IL6, as well as lower levels of Eotaxin in patients with CDK4 gain. NF-κB pathways modulate the effector function of differentiated T cells via Th17 cells, which secrete IL17A (50). Several studies have also demonstrated that NF-κB signaling plays a role in regulating the development of NKT cells, as indirectly mediated via mTEC regulation (51). Emerging evidences also show that CREB is able to regulate immune responses via inhibition of NF-κB activity. It regulates macrophage, T, and B lymphocytes (such as Th17) survival and induces transcription of immune-related genes (52, 53). Furthermore, IL17A and IL6 may affect expression of PD-L1. Ma and colleagues showed that targeting IL17A inhibits PD-L1 expression in tumor cells, decreases the percentage of Tregs in tumor-infiltrating lymphocytes, promotes IFNγ secretion by CD4+ and CD8+ T cells, and exhibits a synergistic antitumor effect with the anti–PD-1 antibody in estrogen receptor (ER)-negative breast cancer (54). Kim and colleagues also reported that PD-1 is overexpressed in IL17A-producing T cells in patients with psoriasis (54). Tsukamoto and colleagues found that IL6 blockade upregulated expression of PD-L1 on melanoma cells and that IL6 blockade promoted infiltration of IFNγ-producing CD4+ T cells into tumor tissues, exerting a synergistic antitumor effect with the anti–PD-1 antibody (55). A previous study also demonstrated that activated EGFR regulates PD-L1 expression via the IL6/JAK/STAT3 pathway in NSCLC (56).

Several studies have demonstrated that CDK4/6 inhibitors alter the tumor immune microenvironment, which may enhance the effects of immuno-oncology agents (57, 58). In fact, previous studies have shown that CDK4/6 inhibitors enhance immune cell infiltration and that secretion of chemokine (C-C motif) ligand 5 (CCL5) is upregulated in palbociclib-treated melanoma cells, possibly promoting T-cell infiltration (59). CDK4/6 inhibitors also increase tumor infiltration and activation of effector T cells via derepression of NFAT family proteins and their target genes (24), inactivating immunosuppressive cells such as myeloid cells and Tregs (60, 61), and enhancing tumor antigen presentation. Moreover, another study demonstrated that cyclin D-CDK4 promoted proteasome-mediated degradation of PD-L1 proteins via Cul3SPOP in melanoma (B16) and colon carcinoma (MC38) murine model, which is a critical factor for response to anti–PD-1 therapy (23). The research also found that CDK4/6 inhibition could augment the response to PD-1 blockade in colon carcinoma (MC38 and CT26) murine model (23). On the basis of promising preclinical data, three clinical trials are currently underway for FDA-approved CDK4/6 inhibitors in combination with anti–PD-1/PD-L1 antibodies and aromatase inhibitors in the treatment of ER+ breast cancer (NCT03147287, NCT02778685, and NCT03294694). In another clinical trial, the combination of CDK4/6 inhibitors with anti–PD-1/PD-L1 antibodies is currently being investigated for advanced solid tumors (NCT02791334), including melanoma.

To further explore whether inhibition of the CDK4 pathway can activate antitumor immunity, we performed a series of experiments in vitro and in vivo. Our data confirm cooperative antitumor effects of CDK4/6 inhibitor (palbociclib) and the anti–PD-1 antibody in mouse melanoma models. Furthermore, we performed a simulated combined dosing regimen of patients resistant to anti–PD-1 therapy and achieved effectiveness. We also demonstrated that palbociclib treatment increased PD-L1 protein levels and the quantity of B cells, T cells, and NK cells in a melanoma cell line and mouse model. Furthermore, RNA-Seq analysis revealed that palbociclib treatment enhanced immune gene expression in a melanoma cell line and mouse model, including genes related to the T-cell receptor signaling pathway, leukocyte transendothelial migration, and NK-cell–mediated cytotoxicity. On the basis of these findings, CDK4/6 inhibitors may synergize with anti–PD-1 immune checkpoint blockade in the treatment of advanced melanoma.

There are some limitations in our study as well as potential perspectives. Because of limited follow-up time, we did not have complete progression-free survival (PFS) and overall survival (OS) data, and we will continuously monitor the efficacy readouts of the patients (PFS and OS) for future analyses. In addition, the sample size was relatively small. As the PD-1 mAb was approved by the China Food and Drug Administration in 2018, we will include more patients for data analysis in future studies. Furthermore, we did not investigate whether various genetic aberrations in the CDK4 pathway in cancer cells may influence the response to anti–PD-1 treatment in melanoma mouse models. In the future, we will generate B16 stable cell lines ectopically expressing CDK4 and inject them into C57BL/6-hPD-1 mice and establish additional HIS PDX models with various genetic aberrations in the CDK4 pathway to assess their sensitivity to anti–PD-1 treatment.

Our study demonstrated that CDK4 pathway genetic aberrations can serve as genomic biomarkers for predicting the response to anti–PD-1 therapy in advanced melanoma. We also revealed that CDK4/6 inhibition enhances antitumor immunity and improves susceptibility to anti–PD-1 therapy in melanoma, providing an important rationale for developing precision combination therapy and carrying out clinical trials for advanced melanoma.

K.T. Flaherty reports receiving commercial research grants from Novartis and Sanofi, holds ownership interest (including patents) in Clovis Oncology, Strata Oncology, Vivid Biosciences, Checkmate Pharmaceuticals, X4 Pharmaceuticals, PIC Therapeutics, Fount Therapeutics, Shattuck Labs, Apricity, Oncoceutics, Fog Pharma, Tvardi, and xCures, and is a consultant/advisory board member for Clovis Oncology, Strata Oncology, Vivid Biosciences, Checkmate Pharmaceuticals, X4 Pharmaceuticals, PIC Therapeutics, Sanofi, Amgen, Asana Biosciences, Adaptimmune, Fount Therapeutics, Aeglea, Array Biopharma, Shattuck Labs, Tolero, Apricity, Oncoceutics, Fog Pharma, Neon, Tvardi, xCures, Monopteros, Novartis, Genentech, Bristol-Myers Squibb, Merck, Takeda, Verastem, Boston Biomedical, Pierre Fabre, Cell Medica, and Debiopharm. No potential conflicts of interest were disclosed by the other authors.

The Editor-in-Chief of Clinical Cancer Research is an author on this article. In keeping with AACR editorial policy, a senior member of the Clinical Cancer Research editorial team managed the consideration process for this submission and independently rendered the final decision concerning acceptability.

Conception and design: Y. Kong, B. Zheng, J. Guo

Development of methodology: Y. Kong, J. Yu, J. Guo

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Kong, J. Yu, J. Yan, Q. Guo, Z. Chi, B. Tang, J. Yu, T. Yin, Z. Cheng, X. Wu, H. Yu, J. Dai, X. Sheng, L. Si, C. Cui, X. Bai, L. Mao, B. Lian, X. Wang, X. Yan, S. Li, L. Zhou, J. Guo

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Kong, K.T. Flaherty, J. Guo

Writing, review, and/or revision of the manuscript: Y. Kong, J. Yu, B. Zheng, K.T. Flaherty, J. Guo

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): Y. Kong, J. Guo

Study supervision: Y. Kong, J. Guo

We would like to thank Yicon Biomedical Technology Inc. (Beijing, China) for humanized immune system patient-derived xenograft model experiments. We would like to thank American Journal Experts for English language editing. This work was supported by grants from Natural Science Foundation of China (81672696 and 81772912), Fostering Young Scholars of Peking University Health Science Center, Beijing Baiqianwan Talents Project, and Beijing Municipal Administration of Hospitals' Ascent Plan (DFL20181101), Clinical Medicine Plus X-Young Scholars Project (Peking University), the Fundamental Research Funds for the Central Universities.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Topalian
SL
,
Drake
CG
,
Pardoll
DM
. 
Immune checkpoint blockade: a common denominator approach to cancer therapy
.
Cancer Cell
2015
;
27
:
450
61
.
2.
Larkin
J
,
Chiarion-Sileni
V
,
Gonzalez
R
,
Grob
JJ
,
Cowey
CL
,
Lao
CD
, et al
Combined nivolumab and ipilimumab or monotherapy in untreated melanoma
.
N Engl J Med
2015
;
373
:
23
34
.
3.
Robert
C
,
Schachter
J
,
Long
GV
,
Arance
A
,
Grob
JJ
,
Mortier
L
, et al
Pembrolizumab versus ipilimumab in advanced melanoma
.
N Engl J Med
2015
;
372
:
2521
32
.
4.
Chi
Z
,
Tang
B
,
Sheng
X
,
Si
L
,
Cui
C
,
Kong
Y
, et al
A phase II study of JS001, a humanized PD-1 mAb, in patients with advanced melanoma in China
.
J Clin Oncol
2018
;
36
:
15s
(
suppl; abstr 9539
)
5.
Patel
SP
,
Kurzrock
R
. 
PD-L1 expression as a predictive biomarker in cancer immunotherapy
.
Mol Cancer Ther
2015
;
14
:
847
56
.
6.
Topalian
SL
,
Taube
JM
,
Anders
RA
,
Pardoll
DM
. 
Mechanism-driven biomarkers to guide immune checkpoint blockade in cancer therapy
.
Nat Rev Cancer
2016
;
16
:
275
87
.
7.
Axelrod
ML
,
Johnson
DB
,
Balko
JM
. 
Emerging biomarkers for cancer immunotherapy in melanoma
.
Semin Cancer Biol
2018
;
52
:
207
15
.
8.
Huang
AC
,
Postow
MA
,
Orlowski
RJ
,
Mick
R
,
Bengsch
B
,
Manne
S
, et al
T-cell invigoration to tumour burden ratio associated with anti–PD-1 response
.
Nature
2017
;
545
:
60
5
.
9.
Alexandrov
LB
,
Nik-Zainal
S
,
Wedge
DC
,
Aparicio
SA
,
Behjati
S
,
Biankin
AV
, et al
Signatures of mutational processes in human cancer
.
Nature
2013
;
500
:
415
21
.
10.
Postow
MA
,
Callahan
MK
,
Wolchok
JD
. 
Immune checkpoint blockade in cancer therapy
.
J Clin Oncol
2015
;
33
:
1974
82
.
11.
Chan
TA
,
Wolchok
JD
,
Snyder
A
. 
Genetic basis for clinical response to CTLA-4 blockade in melanoma
.
N Engl J Med
2015
;
373
:
1984
.
12.
Cui
C
,
Lian
B
,
Zhou
L
,
Song
X
,
Zhang
X
,
Wu
D
, et al
Multifactorial analysis of prognostic factors and survival rates among 706 mucosal melanoma patients
.
Ann Surg Oncol
2018
;
25
:
2184
92
.
13.
Curtin
JA
,
Fridlyand
J
,
Kageshita
T
,
Patel
HN
,
Busam
KJ
,
Kutzner
H
, et al
Distinct sets of genetic alterations in melanoma
.
N Engl J Med
2005
;
353
:
2135
47
.
14.
Bai
X
,
Kong
Y
,
Chi
Z
,
Sheng
X
,
Cui
C
,
Wang
X
, et al
MAPK pathway and TERT promoter gene mutation pattern and its prognostic value in melanoma patients: a retrospective study of 2,793 cases
.
Clin Cancer Res
2017
;
23
:
6120
7
.
15.
Hayward
NK
,
Wilmott
JS
,
Waddell
N
,
Johansson
PA
,
Field
MA
,
Nones
K
, et al
Whole-genome landscapes of major melanoma subtypes
.
Nature
2017
;
545
:
175
80
.
16.
Kong
Y
,
Si
L
,
Zhu
Y
,
Xu
X
,
Corless
CL
,
Flaherty
KT
, et al
Large-scale analysis of KIT aberrations in Chinese patients with melanoma
.
Clin Cancer Res
2011
;
17
:
1684
91
.
17.
Yu
S
,
Xu
T
,
Dai
J
,
Ma
M
,
Tang
H
,
Chi
Z
, et al
TERT copy gain predicts the outcome of high-dose interferon alpha-2b therapy in acral melanoma
.
Onco Targets Ther
2018
;
11
:
4097
104
.
18.
Yan
J
,
Wu
X
,
Yu
J
,
Yu
H
,
Xu
T
,
Brown
KM
, et al
Analysis of NRAS gain in 657 patients with melanoma and evaluation of its sensitivity to a MEK inhibitor
.
Eur J Cancer
2018
;
89
:
90
101
.
19.
Yu
H
,
Ma
M
,
Yan
J
,
Xu
L
,
Yu
J
,
Dai
J
, et al
Identification of coexistence of BRAF V600E mutation and EZH2 gain specifically in melanoma as a promising target for combination therapy
.
J Transl Med
2017
;
15
:
243
.
20.
Kong
Y
,
Sheng
X
,
Wu
X
,
Yan
J
,
Ma
M
,
Yu
J
, et al
Frequent genetic aberrations in the CDK4 pathway in acral melanoma indicate the potential for CDK4/6 inhibitors in targeted therapy
.
Clin Cancer Res
2017
;
23
:
6946
57
.
21.
Sheppard
KE
,
McArthur
GA
. 
The cell-cycle regulator CDK4: an emerging therapeutic target in melanoma
.
Clin Cancer Res
2013
;
19
:
5320
8
.
22.
Kwong
LN
,
Costello
JC
,
Liu
H
,
Jiang
S
,
Helms
TL
,
Langsdorf
AE
, et al
Oncogenic NRAS signaling differentially regulates survival and proliferation in melanoma
.
Nat Med
2012
;
18
:
1503
10
.
23.
Zhang
J
,
Bu
X
,
Wang
H
,
Zhu
Y
,
Geng
Y
,
Nihira
NT
, et al
Cyclin D-CDK4 kinase destabilizes PD-L1 via cullin 3-SPOP to control cancer immune surveillance
.
Nature
2018
;
553
:
91
5
.
24.
Deng
J
,
Wang
ES
,
Jenkins
RW
,
Li
S
,
Dries
R
,
Yates
K
, et al
CDK4/6 inhibition augments antitumor immunity by enhancing T-cell activation
.
Cancer Discov
2018
;
8
:
216
33
.
25.
Yu
J
,
Wu
X
,
Yan
J
,
Yu
H
,
Xu
L
,
Chi
Z
, et al
Anti-GD2/4-1BB chimeric antigen receptor T cell therapy for the treatment of Chinese melanoma patients
.
J Hematol Oncol
2018
;
11
:
1
.
26.
Sai
J
,
Owens
P
,
Novitskiy
SV
,
Hawkins
OE
,
Vilgelm
AE
,
Yang
J
, et al
PI3K inhibition reduces mammary tumor growth and facilitates antitumor immunity and anti-PD1 responses
.
Clin Cancer Res
2017
;
23
:
3371
84
.
27.
Lin
S
,
Huang
G
,
Cheng
L
,
Li
Z
,
Xiao
Y
,
Deng
Q
, et al
Establishment of peripheral blood mononuclear cell-derived humanized lung cancer mouse models for studying efficacy of PD-L1/PD-1 targeted immunotherapy
.
mAbs
2018
;
10
:
1301
11
.
28.
Roh
W
,
Chen
PL
,
Reuben
A
,
Spencer
CN
,
Prieto
PA
,
Miller
JP
, et al
Integrated molecular analysis of tumor biopsies on sequential CTLA-4 and PD-1 blockade reveals markers of response and resistance
.
Sci Transl Med
2017
;
9
:
pii: eaah3560
.
29.
Miao
D
,
Margolis
CA
,
Gao
W
,
Voss
MH
,
Li
W
,
Martini
DJ
, et al
Genomic correlates of response to immune checkpoint therapies in clear cell renal cell carcinoma
.
Science
2018
;
359
:
801
6
.
30.
Tang
B
,
Yan
X
,
Sheng
X
,
Si
L
,
Cui
C
,
Kong
Y
, et al
Safety and clinical activity with an anti–PD-1 antibody JS001 in advanced melanoma or urologic cancer patients
.
J Hematol Oncol
2019
;
12
:
7
.
31.
Cho
J
,
Ahn
S
,
Yoo
KH
,
Kim
JH
,
Choi
SH
,
Jang
KT
, et al
Treatment outcome of PD-1 immune checkpoint inhibitor in Asian metastatic melanoma patients: correlative analysis with PD-L1 immunohistochemistry
.
Invest New Drugs
2016
;
34
:
677
84
.
32.
Le
DT
,
Durham
JN
,
Smith
KN
,
Wang
H
,
Bartlett
BR
,
Aulakh
LK
, et al
Mismatch repair deficiency predicts response of solid tumors to PD-1 blockade
.
Science
2017
;
357
:
409
13
.
33.
Rizvi
NA
,
Hellmann
MD
,
Snyder
A
,
Kvistborg
P
,
Makarov
V
,
Havel
JJ
, et al
Cancer immunology. Mutational landscape determines sensitivity to PD-1 blockade in non-small cell lung cancer
.
Science
2015
;
348
:
124
8
.
34.
Snyder
A
,
Makarov
V
,
Merghoub
T
,
Yuan
J
,
Zaretsky
JM
,
Desrichard
A
, et al
Genetic basis for clinical response to CTLA-4 blockade in melanoma
.
N Engl J Med
2014
;
371
:
2189
99
.
35.
Hugo
W
,
Zaretsky
JM
,
Sun
L
,
Song
C
,
Moreno
BH
,
Hu-Lieskovan
S
, et al
Genomic and transcriptomic features of response to anti–PD-1 therapy in metastatic melanoma
.
Cell
2017
;
168
:
542
.
36.
Jamal-Hanjani
M
,
Wilson
GA
,
McGranahan
N
,
Birkbak
NJ
,
Watkins
TBK
,
Veeriah
S
, et al
Tracking the evolution of non-small-cell lung cancer
.
N Engl J Med
2017
;
376
:
2109
21
.
37.
Bambury
RM
,
Bhatt
AS
,
Riester
M
,
Pedamallu
CS
,
Duke
F
,
Bellmunt
J
, et al
DNA copy number analysis of metastatic urothelial carcinoma with comparison to primary tumors
.
BMC Cancer
2015
;
15
:
242
.
38.
Peng
W
,
Chen
JQ
,
Liu
C
,
Malu
S
,
Creasy
C
,
Tetzlaff
MT
, et al
Loss of PTEN promotes resistance to T cell-mediated immunotherapy
.
Cancer Discov
2016
;
6
:
202
16
.
39.
George
S
,
Miao
D
,
Demetri
GD
,
Adeegbe
D
,
Rodig
SJ
,
Shukla
S
, et al
Loss of PTEN is associated with resistance to anti–PD-1 checkpoint blockade therapy in metastatic uterine leiomyosarcoma
.
Immunity
2017
;
46
:
197
204
.
40.
Turner
NC
,
Slamon
DJ
,
Ro
J
,
Bondarenko
I
,
Im
SA
,
Masuda
N
, et al
Overall survival with palbociclib and fulvestrant in advanced breast cancer
.
N Engl J Med
2018
;
379
:
1926
36
.
41.
Teh
JL
,
Purwin
TJ
,
Greenawalt
EJ
,
Chervoneva
I
,
Goldberg
A
,
Davies
MA
, et al
An in vivo reporter to quantitatively and temporally analyze the effects of CDK4/6 inhibitor-based therapies in melanoma
.
Cancer Res
2016
;
76
:
5455
66
.
42.
Tang
B
,
Sheng
X
,
Kong
Y
,
Chi
Z
,
Si
L
,
Cui
C
, et al
Palbociclib for treatment of metastatic melanoma with copy number variations of CDK4 pathway: case report
.
Chinese Clin Oncol
2018
;
7
:
62
.
43.
Luoto
S
,
Hermelo
I
,
Vuorinen
EM
,
Hannus
P
,
Kesseli
J
,
Nykter
M
, et al
Computational characterization of suppressive immune microenvironments in glioblastoma
.
Cancer Res
2018
;
78
:
5574
85
.
44.
Sun
SC
. 
The non-canonical NF-kappaB pathway in immunity and inflammation
.
Nat Rev Immunol
2017
;
17
:
545
58
.
45.
Sun
SC
. 
The noncanonical NF-kappaB pathway
.
Immunol Rev
2012
;
246
:
125
40
.
46.
Dejardin
E
. 
The alternative NF-kappaB pathway from biochemistry to biology: pitfalls and promises for future drug development
.
Biochem Pharmacol
2006
;
72
:
1161
79
.
47.
Abramson
J
,
Anderson
G
. 
Thymic epithelial cells
.
Annu Rev Immunol
2017
;
35
:
85
118
.
48.
Abiko
K
,
Matsumura
N
,
Hamanishi
J
,
Horikawa
N
,
Murakami
R
,
Yamaguchi
K
, et al
IFN-gamma from lymphocytes induces PD-L1 expression and promotes progression of ovarian cancer
.
Br J Cancer
2015
;
112
:
1501
9
.
49.
Ayers
M
,
Lunceford
J
,
Nebozhyn
M
,
Murphy
E
,
Loboda
A
,
Kaufman
DR
, et al
IFN-gamma-related mRNA profile predicts clinical response to PD-1 blockade
.
J Clin Invest
2017
;
127
:
2930
40
.
50.
Yu
J
,
Zhou
X
,
Nakaya
M
,
Jin
W
,
Cheng
X
,
Sun
SC
. 
T cell-intrinsic function of the noncanonical NF-kappaB pathway in the regulation of GM-CSF expression and experimental autoimmune encephalomyelitis pathogenesis
.
J Immunol
2014
;
193
:
422
30
.
51.
Elewaut
D
,
Shaikh
RB
,
Hammond
KJ
,
De Winter
H
,
Leishman
AJ
,
Sidobre
S
, et al
NIK-dependent RelB activation defines a unique signaling pathway for the development of V alpha 14i NKT cells
.
J Exp Med
2003
;
197
:
1623
33
.
52.
Brenner
S
,
Prosch
S
,
Schenke-Layland
K
,
Riese
U
,
Gausmann
U
,
Platzer
C
. 
cAMP-induced Interleukin-10 promoter activation depends on CCAAT/enhancer-binding protein expression and monocytic differentiation
.
J Biol Chem
2003
;
278
:
5597
604
.
53.
Wen
AY
,
Sakamoto
KM
,
Miller
LS
. 
The role of the transcription factor CREB in immune function
.
J Immunol
2010
;
185
:
6413
9
.
54.
Kim
JH
,
Choi
YJ
,
Lee
BH
,
Song
MY
,
Ban
CY
,
Kim
J
, et al
Programmed cell death ligand 1 alleviates psoriatic inflammation by suppressing IL-17A production from programmed cell death 1-high T cells
.
J Allergy Clin Immunol
2016
;
137
:
1466
76
.
55.
Tsukamoto
H
,
Fujieda
K
,
Miyashita
A
,
Fukushima
S
,
Ikeda
T
,
Kubo
Y
, et al
Combined blockade of IL6 and PD-1/PD-L1 signaling abrogates mutual regulation of their immunosuppressive effects in the tumor microenvironment
.
Cancer Res
2018
;
78
:
5011
22
.
56.
Zhang
N
,
Zeng
Y
,
Du
W
,
Zhu
J
,
Shen
D
,
Liu
Z
, et al
The EGFR pathway is involved in the regulation of PD-L1 expression via the IL-6/JAK/STAT3 signaling pathway in EGFR-mutated non-small cell lung cancer
.
Int J Oncol
2016
;
49
:
1360
8
.
57.
Teh
JLF
,
Aplin
AE
. 
Arrested developments: CDK4/6 inhibitor resistance and alterations in the tumor immune microenvironment
.
Clin Cancer Res
2019
;
25
:
921
7
.
58.
Schaer
DA
,
Beckmann
RP
,
Dempsey
JA
,
Huber
L
,
Forest
A
,
Amaladas
N
, et al
The CDK4/6 inhibitor abemaciclib induces a T cell inflamed tumor microenvironment and enhances the efficacy of PD-L1 checkpoint blockade
.
Cell Rep
2018
;
22
:
2978
94
.
59.
Vilgelm
AE
,
Johnson
CA
,
Prasad
N
,
Yang
J
,
Chen
SC
,
Ayers
GD
, et al
Connecting the Dots: therapy-induced senescence and a tumor-suppressive immune microenvironment
.
J Natl Cancer Inst
2016
;
108
:
djv406
.
60.
Goel
S
,
DeCristo
MJ
,
Watt
AC
,
BrinJones
H
,
Sceneay
J
,
Li
BB
, et al
CDK4/6 inhibition triggers anti-tumour immunity
.
Nature
2017
;
548
:
471
5
.
61.
Syedbasha
M
,
Egli
A
. 
Interferon lambda: modulating immunity in infectious diseases
.
Front Immunol
2017
;
8
:
119
.