obesity diminishes response to pd-1-based immunotherapies

1Boi SK, et al. J Immunother Cancer 2020;8:e000725. doi:10.1136/jitc-2020-000725

Open access

Obesity diminishes response to PD-1- based immunotherapies in renal cancer

Shannon K Boi,1 Rachael M Orlandella,1 Justin Tyler Gibson ,1 William James Turbitt,2 Gal Wald,3 Lewis Thomas,3 Claire Buchta Rosean,3 Katlyn E Norris,4 Megan Bing,3 Laura Bertrand,3 Brett P Gross,5 Amani Makkouk,5 Dmytro Starenki,6 Kristine I Farag,7 Robert E Sorge,8 James A Brown,3,9 Jennifer Gordetsky,10 Hesham Yasin,9,11 Rohan Garje ,9,12 Lakshminarayanan Nandagopal,13,14 George J Weiner,9,15 David M Lubaroff,3,9 Rebecca C Arend,14,16 Peng Li,17 Yousef Zakharia,9,18 Eddy Yang,14,19 Aliasger K Salem,9,20 Kenneth Nepple,3,9 Tatiana T Marquez- Lago,21,22 Lyse A Norian 2,14

To cite: Boi SK, Orlandella RM, Gibson JT, et al. Obesity diminishes response to PD-1- based immunotherapies in renal cancer. Journal for ImmunoTherapy of Cancer 2020;8:e000725. doi:10.1136/jitc-2020-000725

► Additional material is published online only. To view please visit the journal online (http:// dx. doi. org/ 10. 1136/ jitc- 2020- 000725).

SKB, RMO and JTG contributed equally.

Accepted 18 November 2020

For numbered affiliations see end of article.

Correspondence toDr Lyse A Norian; lnorian@ uab. edu

Original research

© Author(s) (or their employer(s)) 2020. Re- use permitted under CC BY- NC. No commercial re- use. See rights and permissions. Published by BMJ.

ABSTRACTBackground Obesity is a major risk factor for renal cancer, yet our understanding of its effects on antitumor immunity and immunotherapy outcomes remains incomplete. Deciphering these associations is critical, given the growing clinical use of immune checkpoint inhibitors for metastatic disease and mounting evidence for an obesity paradox in the context of cancer immunotherapies, wherein obese patients with cancer have improved outcomes.Methods We investigated associations between host obesity and anti- programmed cell death (PD-1)- based outcomes in both renal cell carcinoma (RCC) subjects and orthotopic murine renal tumors. Overall survival (OS) and progression- free survival (PFS) were determined for advanced RCC subjects receiving standard of care anti- PD-1 who had ≥6 months of follow- up from treatment initiation (n=73). Renal tumor tissues were collected from treatment- naive subjects categorized as obese (body mass index, ‘BMI’ ≥30 kg/m2) or non- obese (BMI <30 kg/m2) undergoing partial or full nephrectomy (n=19) then used to evaluate the frequency and phenotype of intratumoral CD8+ T cells, including PD-1 status, by flow cytometry. In mice, antitumor immunity and excised renal tumor weights were evaluated ±administration of a combinatorial anti- PD-1 therapy. For a subset of murine renal tumors, immunophenotyping was performed by flow cytometry and immunogenetic profiles were evaluated via nanoString.Results With obesity, RCC patients receiving anti- PD-1 administration exhibited shorter PFS (p=0.0448) and OS (p=0.0288). Treatment- naive renal cancer subjects had decreased frequencies of tumor- infiltrating PD-1highCD8+ T cells, a finding recapitulated in our murine model. Following anti- PD-1- based immunotherapy, both lean and obese mice possessed distinct populations of treatment responders versus non- responders; however, obesity reduced the frequency of treatment responders (73% lean vs 44% obese). Tumors from lean and obese treatment responders displayed similar immunogenetic profiles, robust infiltration by PD-1int interferon (IFN)γ+CD8+ T cells and reduced myeloid- derived suppressor cells (MDSC), yielding favorable CD44+CD8+ T cell to MDSC ratios.

Neutralizing interleukin (IL)-1β in obese mice improved treatment response rates to 58% and reduced MDSC accumulation in tumors.Conclusions We find that obesity is associated with diminished efficacy of anti- PD-1- based therapies in renal cancer, due in part to increased inflammatory IL-1β levels, highlighting the need for continued study of this critical issue.

BACKGROUNDAlthough immune checkpoint blockade has revolutionized immune- based cancer treat-ment, clinical responses remain suboptimal. In renal cell carcinoma (RCC) less than 50% of patients achieve durable responses to anti-programmed cell death (PD-1) monotherapy or dual blockade of PD-1 and cytotoxic T lymphocyte associated protein (CTLA)-4.1 As such, efforts are underway to identify the factors that limit treatment efficacy. Obesity may be one such factor.

Currently, over 39% of US adults have obesity (body mass index (BMI) ≥30 kg/m2).2 Obesity increases the risk of developing at least thirteen different types of cancer, including RCC,3 and prior reports suggest that over 60% of RCC subjects have overweight or obesity at diagnosis.4 Mounting evidence indicates that obesity negatively impacts immune function and immune- based interventions, such as vaccinations.5 Furthermore, multiple labo-ratories, including our own, have reported that obesity subverts antitumor immunity to facilitate tumor progression, particu-larly in preclinical models.6–11 Despite this, multiple retrospective studies have reported that obesity and/or an elevated BMI is asso-ciated with improved outcomes following chemotherapy, targeted therapies and/or

on Decem

ber 2, 2021 by guest. Protected by copyright.

http://jitc.bmj.com

/J Im

munother C

ancer: first published as 10.1136/jitc-2020-000725 on 22 Decem

ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

http://bmjopen.bmj.com/

http://orcid.org/0000-0003-2967-5338

http://orcid.org/0000-0002-7244-5602

http://orcid.org/0000-0003-1016-313X

http://crossmark.crossref.org/dialog/?doi=10.1136/jitc-2020-000725&domain=pdf&date_stamp=2020-11-22

http://jitc.bmj.com/




















2 Boi SK, et al. J Immunother Cancer 2020;8:e000725. doi:10.1136/jitc-2020-000725

Open access

immunotherapy administration in cancer subjects11–15—a paradox particularly evident in men with melanoma who received immune checkpoint inhibitors.12 Notably, several prior reports analyzed immunotherapy outcomes based partially or solely on results of clinical trials,12 15 16 raising the question of whether different outcomes may ensue with standard of care administration.

The effects of obesity on antitumor immunity and immune checkpoint blockade outcomes specifically in RCC patients remain unclear. Several prior studies exam-ining RCC outcomes used a BMI cut- point of >25 kg/m2 which examines the effects of overweight plus obesity, rather than obesity alone.14 17 One such study found that a BMI >25 kg/m2 was associated with a trending reduction in overall survival (OS) in RCC patients (n=42) treated with anti- PD-1 or programmed cell death ligand (PD- L)1, but improved OS in RCC patients receiving targeted thera-pies at the same institution.18 However, another 2019 study of 90 RCC patients reported no difference in BMI status between patients who derived clinical benefit from anti- PD-1/PD- L1 versus those who were resistant to therapy.19 A third study found that in RCC patients (n=203) treated with immunotherapy (any monotherapy or combination of anti- PD-1/PD- L1/CTLA-4 during clinical trials or stan-dard of care), obesity did not alter immune infiltration or improve survival after adjusting for risk score, but instead decreased PD- L1 expression and increased angiogenesis in renal tumors.16 Thus, further investigation of this crit-ical issue is needed.

Here, we used multiple complimentary methods to examine the effects of obesity on antitumor immunity and anti- PD-1 outcomes in both RCC subjects and mice bearing orthotopic renal tumors. In doing so, we provide one of the first reports on interactions between obesity and immune checkpoint blockade outcomes in RCC patients receiving anti- PD-1 monotherapy as standard of care.

METHODSHuman analysesRetrospective study of metastatic RCC patients treated with anti-PD-1An IRB- approved retrospective study of metastatic RCC patients treated with nivolumab or pembrolizumab at The University of Iowa (UI) and University of Alabama at Birmingham (UAB) Hospitals was conducted. Data were obtained via UI chart reviews or UAB mining of elec-tronic hospital records using the i2B2 (Informatics for Integrating Biology and the Bedside) database to iden-tify all patients (n=121) with kidney cancer who received nivolumab or pembrolizumab from December 2015 to June 2019. A chart review was conducted for all patients to obtain details on demographics, treatment specifics, date of progression and date of death (if expired) (online supplemental table 1). Only patients who had at least 6- months of follow- up from their first dose of

immunotherapy, a BMI >18.5 kg/m2, and an Eastern Cooperative Oncology Group (ECOG) performance status score of 0–2 at treatment initiation were included in data analyses (n=72). Response to treatment (complete response, partial response, stable disease, progressive disease, not determined) was obtained from review of documentation in the attending clinical encounter. Patients were classified according to their BMI on the day of first dose of therapy based on the World Health Orga-nization (WHO) standard definitions as either having obesity (BMI ≥30 kg/m2) or not (BMI 18.5–30 kg/m2).

Prospective study of treatment-naive RCC subjectsApproval for this study was granted by the IRB of The UI. Adult patients with suspicious renal masses who were scheduled to undergo either partial or full nephrectomy were offered enrollment. Informed consent was obtained for subjects with localized renal masses (stages T1–T3) as well as age- matched and BMI- matched tumor- free healthy donors. Subjects with all histological subtypes of RCC were consented; only peripheral blood specimens with sufficient numbers of cells to permit collection of ≥20 000 events during flow cytometry (n=82) were analyzed; fresh tumor samples were obtained and stained from 19 subjects. Exclusion criteria: active secondary malignancy, immune- modulating medications and metastatic disease. Demographic information (online supplemental tables 2 and 3) was obtained from electronic health records. BMI was calculated from subject’s height/weight at surgery. Patients were classified according to their BMI based on the WHO standard definitions as non- obese (NOB) BMI <30 kg/m2 or as having obesity (OB) BMI ≥30 kg/m2.

Animals/dietsFemale and male BALB/c mice (7–8 weeks of age) were purchased from the Charles River Frederick colony. Mice were maintained on a low- fat chow diet (LFD) for 1 week to acclimate, then randomized to either LFD or a high- fat diet (HFD; Research Diets; catalog #12492) for 20 weeks and were allowed ad libitum access to food and water throughout. After 20 weeks, lean, diet- induced obese (DIO) and OB- resistant (OB- Res) mice were identified as described,8 20 then randomized to treatment groups. Mice were housed five to a cage in standard caging under specific- pathogen- free conditions in 12:12 light:dark cycles at 22°C (72 °F average).

Cell linesRenca, a BALB/c syngeneic renal adenocarcinoma cell line, was purchased from American Type Culture Collection (ATCC), engineered by lentiviral transduc-tion to stably express firefly Luciferase, and cultured as described.21 Cells were used at the same passage number throughout the study to minimize experimental varia-tion. CMS5, a BALB/c syngeneic fibrosarcoma cell line, was a generous gift of Paul Allen (Washington University School of Medicine) and was maintained as described.22

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

https://dx.doi.org/10.1136/jitc-2020-000725





Open access

Cell lines were not validated after purchase/receipt. Renca and CMS5 cells were tested and identified as mycoplasma- negative.

Tumor challenge and therapeutic administrationsOrthotopic murine renal cancer model and AdTR/CpG/PD-1 therapeutic administrationIntrarenal tumor challenges were performed as reported.21 Renal tumor growth was confirmed at day 6 post- challenge via bioluminescent imaging (BLI). Seven days post- tumor challenge, mice were reinjected in the tumor- bearing kidney with either sterile saline or 100 µg CpG1826 (Integrated DNA Technologies) plus 109 pfu of replication- deficient adenovirus encoding a membrane- bound version of full- length murine tumor necrosis factor (TNF)- related apoptosis- inducing ligand (TRAIL) (AdTR) protein (UI Viral Vector Core). On days 10, 13 and 16 post- tumor challenge, mice received intraper-itoneal injections of saline, 250 µg of Rat IgG2a Isotype Control (Clone: 2A3, BioXCell), or 250 µg of anti- PD-1 (Rat IgG2a, clone: RMP1-14, BioXCell). In some exper-iments, tumor- challenged mice receiving AdTR/CpG/PD-1 were injected i.p. every 3 days with saline or neutral-izing antibody against CCL5 (50 µg/mouse, clone: 53405, R&D Systems) or interleukin (IL)-1β (20 mg/kg, clone: B122, BioXCell) beginning at day 5 or day −5 relative to tumor challenge, respectively.

Murine fibrosarcoma and adoptive transfer systemSubcutaneous tumor studies with CMS5 fibrosarcoma cells were completed as described.22 23 To track T cell proliferation, 1×106 naive, CFSE- labeled, tumor antigen- specific, T cell receptor (TCR) transgenic CD8+ DUC18 T cells were adoptively transferred on day 4 post- tumor challenge. Tumor outgrowth was monitored via calipers. Inguinal draining lymph nodes were harvested 4 days after transfer for T cell proliferation assays. The no- DUC18 T cell control endpoint was day 15; the DUC18 T cell recip-ient endpoint was day 21 post- tumor challenge.

Bioluminescent imaging (BLI)On day 6 post- tumor challenge, BLI was performed to confirm that tumor burdens were statistically equivalent across experimental groups prior to randomization to treatment. BLI was accomplished using an IVIS Lumina III (PerkinElmer) and analyzed as described.21 Tumor luminescence was calculated in radiance (photons/sec/cm2/sr) using the Living Image Software, V.4.5.2.

Sample preparation and flow cytometryMurine sample preparationMurine renal tumors were harvested and disrupted using a Miltenyi GentleMacs Dissociator (Miltenyi Biotech) in HBSS followed by enzymatic digestion in 5 ug/mL Libera-seTM (Roche;) and 37.5 ug/mL DNase I (Sigma- Aldrich) at 37°C with agitation for 30 min. Tumor- draining lymph nodes from CMS5 tumor- bearing mice were harvested and single cell suspensions were generated by disrupting with glass slides.

Human sample preparationHuman tumor samples obtained from surgical pathology on the day of nephrectomy were mechanically homoge-nized in Roswell Park Memorial Institute (RPMI) 1640 to a single cell suspension using a GentleMacs Dissociator, then stained without freezing. Human peripheral blood samples were taken in the preoperative area, processed over Ficoll to permit mononuclear cell harvest and frozen in autologous serum until use. Prior to staining, periph-eral blood mononuclear cells were thawed and suspended in complete media (RPMI basal medium plus 10% fetal calf serum).

Flow cytometric analysesTo block Fc receptors, human samples were incubated in 5% normal rat serum and murine samples were incu-bated in TruStain FcX (BioLegend) and normal rat serum then stained using antibodies listed in online supple-mental table 4. Intracellular staining of murine Foxp3 was performed according to manufacturer instructions using a Foxp3/Transcription factor staining set (Ther-moFisher). For murine IFNγ, TNF⍺ and perforin intra-cellular staining, bulk tumor cells were ex vivo stimulated for 4 hours with plate- bound, purified anti- CD3 and anti- CD28 (BioLegend). GolgiPlug protein transport inhibitor (BD Biosciences) was added for the final 2 hours. Cells were harvested and stained using BD Biosciences Fixa-tion/Permeabilization Solution Kit. Results were obtained using an Attune NxT (ThermoFisher) or BD LSR II (BD Biosciences) flow cytometer and analyzed with FlowJo Software. Dead cells were excluded via Zombie Aqua or Green Fixable Viability Dye (BioLegend). Fluorescence minus one controls were used to objectively determine gate boundaries for positive events.

RNA isolation/quantification/qualityMurine tumors were excised and stored in RNAlater stabilization solution (ThermoFisher) at −80°C until bulk RNA isolation using the Qiagen RNeasy Mini Kit. RNA integrity was determined using an Agilent 2100 Bioana-lyzer. Samples with RIN values >8 were used in iReper-toire analyses.

NanoString immunogenetic profilingAs a screening tool, immune- related gene expression patterns were interrogated on whole- tumor RNA using the nanoString murine PanCancer Immune Profiling Panel according to manufacturer’s instructions by the UAB nanoString facility.

Protein quantificationMurine renal tumors were harvested, weighed, and disrupted using a Miltenyi GentleMacs Dissociator or pestle (USA Scientific) in Dulbecco’s Phosphate Buffered Saline (DPBS) at a concentration of 0.2 g of tissue per 1 mL of DPBS. Tissue supernatant was collected and stored at −80°C until processing. Murine cytokines/chemokines in tumor homogenates and human plasma leptin were quantified using the Bio- Plex Multiplex System (Bio- Rad)

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from





Open access

and analyzed on a Bio- Plex 200 (Bio- Rad) or MAGPIX (ThermoFisher) instrument.

StatisticsStatistical analyses were performed utilizing Prism, V.7.00 (GraphPad), unless specifically noted else-where. Gaussian distribution was assessed using Shapiro- Wilk normality testing. Throughout, statis-tical differences between two groups were analyzed using parametric two- tailed unpaired Student’s t- tests or nonparametric two- tailed Mann- Whitney U tests, as appropriate. Statistical significance is denoted as asterisks (*) for parametric tests or as pound signs (#) for nonparametric tests. Statistical differences among multiple groups with one independent vari-able were analyzed using parametric or Kruskal- Wallis nonparametric one- way analysis of variance (ANOVA) with Dunn’s post hoc multiple comparison testing. In studies with multiple groups and two independent variables (ie, obesity status and therapy groups), two- way ANOVAs were performed with Bonferroni’s post hoc multiple comparison testing. Statistical differ-ences in studies examining survival outcomes in mice were calculated using Log- rank (Mantel- Cox) testing. To examine linear correlations, linear regression analyses were performed where indicated. To assess changes in tumor weights over time, mixed- effects modeling with a Bonferroni correction was used. Note that in all figures, statistical comparisons between all groups were performed; however, only comparisons that were significantly different or trending (p<0.15) were shown on figures to improve data clarity.

Statistical analysis for RCC patients receiving anti-PD-1Patients’ demographic and clinical characteristics were summarized as mean and SD or median and range for continuous variables, and as frequency and proportion for categorical variables. In the univar-iate analysis, Wilcoxon test and Fisher’s exact test were used for continuous variables and categorical variables, respectively. OS was defined as the interval from the first dose of immunotherapy to the date of death or to the last follow- up date if patients were still alive (censored). Progression- free survival (PFS) was defined as the interval from the first dose of immuno-therapy to the date of progression or the date of death and censored at the last follow- up date if the patients were still alive without progression. A multivariable Cox proportional hazard regression model was used to evaluate the association between OS/PFS and obesity status, controlling for age, sex, International Meta-static RCC Database Consortium (IMDC) risk score and the number of prior therapies. The associations between obesity status and other outcomes, such as deceased rate within 90 days, objective response rate and disease control rate, were evaluated using multi-variable logistic regression with Firth’s penalized like-lihood approach, controlling for age, sex, IMDC risk

score and the number of prior therapies. All analyses were performed using SAS V.9.4.

Identification of murine therapy non-responders versus respondersThe weights of excised tumor- bearing kidneys from individual lean treatment- naive mice at day 28 were recorded and the mean calculated. AdTR/CpG/PD-1 therapy reduced the mean day 28 tumor- bearing kidney weight by 75% relative to no therapy (NT) controls (see Figure 4 in the Results section). This value was, therefore, used as a threshold to objectively define treatment responders (ie, individual mice with a ≥75% reduction in excised tumor- bearing kidney weight relative to NT controls) vs treatment non- responders (NR) (ie, individual mice with a <75% reduction in excised tumor- bearing kidney weight relative to NT controls) in subsequent experiments. To control for slight, non- significant differences in day 28 tumor weights between lean, DIO and OB- Res mice, the 75% response threshold was defined for each group (ie, lean treated mice were compared with lean NT mice, DIO treated mice were compared with DIO NT mice, etc.).

Hierarchical clustering and visualization of expression (mRNA) resultsRaw nanoString- determined gene expression reads were subjected to unbiased hierarchical clustering analysis according to previously published methods24 25 across all 750 genes in 21 different samples of interest using Matlab (version R2016b). Differential clusters are presented in an annotated heatmap based on stan-dardized expression values, along with the resulting hierarchical clustering dendrogram. Differentially expressed genes were identified in nSolver Analysis Software (nanoString) as genes with an unadjusted p value of less than 0.05 vs the respective baseline control given the exploratory nature of this study. Cell type scores were generated through nSolver Advanced Analysis (nSolver V.4.0).

Data availabilityGene expression data from this study are available from the corresponding author on reasonable request.

Study approvalWritten informed consent was received from each subject prior to inclusion and participation in our study. All animal procedures were approved by the Institutional Animal Care and Use Committee at UAB, an AAALAC- accredited institution.

RESULTSObesity is associated with poorer PFS and OS in metastatic RCC patients receiving anti-PD- 1 as standard of careWe began by examining survival outcomes of RCC patients with and without obesity following anti- PD-1 treatment. Recent retrospective studies investigating

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from



Open access

the effects of obesity on immunotherapy outcomes in RCC have provided intriguing but conflicting results: that increased adiposity (defined as BMI >25 kg/m2) trended toward being detrimental,18 that obesity (BMI >30 kg/m2) had no impact on outcomes in multivar-iate analysis,16 or that obesity improved PFS but not OS in patients who showed primary clinical benefit.19

We examined this issue by evaluating outcome data for all RCC patients treated with standard of care anti- PD-1 who had ≥6 months of follow- up from treatment initiation at the UAB Hospital (n=54) and the UI Hospitals and Clinics (n=18). All patients had an ECOG performance status of 0–2 and none had received prior immunotherapies. Demographic infor-mation for these patients can be found in online supplemental table S1. The WHO- defined BMI cut point of 30 kg/m2 was used to identify patients with obesity (OB) or without (NOB). The mean BMI of the OB cohort was 38.74 kg/m2 vs 25.07 kg/m2 for the NOB. Obesity at treatment initiation was associ-ated with a significant reduction in PFS (p=0.0448), as well as OS (p=0.0288) in Kaplan- Meier analysis (figure 1A,B). Specifically, obesity was associated with a 6.5 month- reduction in median PFS and a 12- month reduction in OS. Notably, after controlling for patients’ age, sex, number of prior treatments received and IMDC risk score, the Cox proportional hazard regression on PFS indicated that the hazard of disease progression in patients with BMI <30 kg/m2 was 54% (95% CI (0.31% to 0.95%; p=0.0322)] of that

seen in patients with BMI ≥30 kg/m2; for OS the risk was 48% (95% CI 0.24% to 0.96%; p=0.0371). When BMI ≥25 kg/m2 was used as the cut- point instead of 30 kg/m2, to permit evaluation of the combined effects of overweight and obesity, the Kaplan- Meier curves for PFS and OS were no longer statistically different (p=0.5520 and p=0.8279, respectively) (online supple-mental figure S1). Therefore, host obesity was asso-ciated with unfavorable survival outcomes in RCC patients treated with anti- PD-1 immunotherapy.

Obesity is associated with reduced frequencies of PD-1high intratumoral CD8+ T cells in treatment-naive renal cancerWe next explored the possibility that underlying differ-ences in CD8+ tumor- infiltrating lymphocyte (TIL) phenotypes existed in the presence of obesity that might contribute to the divergent response to anti- PD-1. PD-1 expression on CD8+ T cells regulates T cell function and cancer immunotherapy outcomes.26 27 Although frequently regarded only as an indicator of T cell functional exhaustion, PD-1 is upregulated following antigen recognition during the earliest stages of T cell activation.28 Thus, in the tumor micro-environment, PD-1 expression is used as an indicator of T cell encounter(s) with antigens,27 29 as repeated TCR engagement increases PD-1 levels.26 30 31

We therefore analyzed PD-1 expression on CD8+ TILs from a subset of treatment- naive RCC subjects with (BMI ≥30 kg/m2; n=11) or without (BMI <30 kg/m2; n=8) obesity. Demographic information for these RCC subjects

Figure 1 Host obesity is associated with reduced PFS and OS in metastatic RCC patients receiving anti- PD-1 as standard of care. (A) PFS and (B) OS of metastatic RCC patients with at least 6 months of follow- up after initiation of anti- PD-1 therapy as standard of care, categorized by BMI status. Survival curves for PFS and OS across BMI categories were generated with the Kaplan- Meier method. HRs were calculated from a Cox model controlling for patients’ age, sex, IMDC risk score and number of prior therapies. BMI, body mass index; IMDC, International Metastatic RCC Database Consortium; OS, overall survival; PD-1, programmed cell death-1; PFS, progression- free survival; RCC, renal cell carcinoma.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from







Open access

can be found in online supplemental table S2. This anal-ysis revealed that obesity did not alter the activation status (CD45RA-HLA- DR+) of CD8+ TILs; however, we observed significantly reduced percentages of activated PD-1+CD8+ TILs and PD-1highCD8+ TILs with obesity (figure 2A, gating as per online supplemental figure S2A). No obesity- associated alterations in the frequency of acti-vated PD-1+CD8+ T cells in the peripheral blood of RCC subjects (n=82) or tumor- free donors (n=19) were found (figure 2B). Demographic information for tumor- free donors is in online supplemental table S3. RCC subjects

with obesity displayed the expected elevations in plasma leptin (figure 2C and online supplemental figure S2B), but we found negative correlations between the frequen-cies of circulating activated PD-1+CD8+ T cells and plasma leptin concentrations (figure 2D) or subject BMI (online supplemental figure S2C), contrasting with prior findings in melanoma.11

To address whether similar trends existed regarding PD-1 expression on CD8+ TILs in mice with renal tumors, we used our well- characterized model of DIO.8 20 In this model, BALB/c mice are randomized to either a HFD or

Figure 2 Host obesity is associated with decreased frequencies of activated PD-1hiCD8+ TILs in treatment- naive human and murine renal tumors. (A) Intratumoral and (B) peripheral blood CD8+ T cells from NOB and OB treatment- naive RCC subjects and tumor- free donors. Frequencies of (A) activated CD8+ TILs, activated PD-1- expressing CD8+ TILs, and (B) peripheral blood activated PD-1+CD8+ T cells from RCC subjects and tumor- free donors. (C) Plasma leptin levels from NOB and OB RCC subjects. (D) Linear regression of activated PD-1+CD8+ PBMCs and plasma leptin levels in NOB and OB RCC subjects. (E) Resulting body weights for lean low- fat diet (LFD) and diet- induced obese (DIO) high- fat diet (HFD)- fed mice. Dotted line indicates three SD above the mean weight of LFD fed animals and the threshold for classification as DIO. (F) CD8+ TILs from lean and DIO treatment- naive mice on day 28 post- tumor challenge. Frequencies of activated CD8+ TILs and activated PD-1- expressing CD8+ TILs are shown. (G) PD-1 expression on activated CD8+ TILs at days 15 and 28 post- tumor challenge. Murine data are pooled from at least two independent experiments. Data are presented as means ±SEM. Statistical differences were calculated using parametric t- tests (***p<0.001), nonparametric Mann- Whitney U tests (#p<0.05), linear regression, or two- way ANOVA followed by post hoc Bonferroni’s multiple comparisons tests (**p<0.01, ****p<0.0001). ANOVA, analysis of variance; BMI, body mass index; NOB, non- obese; OB, obese; PBMCs, peripheral blood mononuclear cells; PD-1, programmed cell death-1; RCC, renal cell carcinoma; TIL, tumor- infiltrating lymphocyte.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from









Open access

a LFD for 20 weeks prior to orthotopic renal tumor chal-lenge. Individual HFD- fed mice are identified as DIO only if their body weight is ≥3 SD above the mean body weight of the age- matched LFD- fed lean cohort (figure 2E). DIO mice identified in this manner display classic hallmarks of obesity (increased adiposity, serum leptin and insulin and adipose tissue crown- like structures).8 20 Following orthot-opic, syngeneic renal tumor challenge, an equivalent percentage of CD8+ TILs cells were activated (CD44+) in DIO and lean animals, with ~75% of these CD44+ T cells expressing PD-1, regardless of obesity status at a late day 28 time point (figure 2F, gating as per figure 3). Similar to what was observed in RCC subjects, DIO mice had decreased frequencies of activated PD-1highCD8+ TILs. Notably, PD-1 expression on activated CD8+ TILs from lean mice increased from day 15 to day 28 post- tumor chal-lenge (figure 2G), reflecting prior reports that sustained antigen presence and repeated TCR engagement drive heightened PD-1 expression.30–32 In contrast, CD8+ TILs from DIO mice did not display increased PD-1 expression over time, resulting in significantly reduced PD-1 expres-sion on CD8+ TILs from DIO mice versus lean mice at day 28 (figure 2G). Thus, we found that obesity was asso-ciated with reduced frequencies of PD-1highCD8+ TILs in both human subjects with RCC and an orthotopic mouse model of renal cancer, implying that CD8+ TIL responses are indeed altered by obesity and illustrating the clinical relevance of our murine tumor model for further study.

PD-1highCD8+ TILs retain effector function in the treatment-naive settingWe next assessed the functional capacity of PD-1highCD8+ TILs from treatment- naive mice as prior studies indicated that, depending on the tumor type,33 PD-1+CD8+ TILs can exist either as functional effectors or as exhausted cells that are resistant to re- invigoration with anti- PD-1. We found a higher frequency of PD-1highCD8+ TILs secreting perforin and tumor necrosis factor (TNF)α relative to PD-1int or PD-1neg subsets (figure 3A,B). This pattern was conserved within both lean and DIO cohorts. Similar patterns were observed for interferon (IFN)γ secretion in lean, but not DIO, mice (figure 3C). Thus, obesity is associated with reduced frequencies of functionally- competent, PD-1highCD8+ TILs in the treatment- naive setting, suggesting the presence of a diminished endog-enous T cell response.

Obesity compromises the response to a combinatorial PD-1-based immunotherapyGiven the obesity- associated decrease in functional PD-1highCD8+ TILs in our murine model, we hypothesized that host obesity would impair anti- PD-1 immunotherapy outcomes. However, single agent anti- PD-1 was unable to reduce day 28 excised renal tumor burdens in either lean or DIO mice (figure 4A,B, and online supplemental figure S3A, respectively). Notably, lean mice treated with anti- PD-1 did experience significant, but modest, improve-ments in survival relative to NT controls (figure 4C), an

outcome reflecting survival benefits detected in RCC patients who receive anti- PD-1 monotherapy. Conse-quently, we developed a combinatorial therapeutic

Figure 3 Tumor- infiltrating PD-1highCD8+ T cells retain potent effector function in both lean and DIO treatment- naive mice. On day 28, post- tumor challenge CD8+ TILs from lean and DIO treatment- naive mice were evaluated ex vivo for their expression of PD-1 and production of effector cytokines (A) perforin, (B) TNF⍺ and (C) IFNγ. Representative flow plots for each cytokine are shown. Graphs show data pooled from at least two independent experiments and presented as means ±SEM for indicated cell subpopulations based on PD-1 status. Statistical differences were calculated using two- way ANOVA with Bonferroni’s multiple comparisons tests (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001). ANOVA, analysis of variance; DIO, diet- induced obese; Hi, PD-1 high; IFNγ, interferon-γ; Int, PD-1 intermediate; Neg, PD-1 negative; PD-1, programmed cell death-1; TIL, tumor- infiltrating lymphocyte; TNF⍺, tumor necrosis factor ⍺.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from





Open access

approach to evaluate the effect of host obesity on immu-notherapy outcomes, wherein anti- PD-1 was administered downstream of our previously described in situ T cell priming therapy.8 34 This priming approach consists of adenovirus encoding murine TRAIL (AdTR) coadminis-tered with the toll- like receptor 9 agonist CpG (AdTR/CpG) at day 7 post- tumor challenge (figure 4A), after renal tumors are established and when tumor burdens

are statistically equivalent based on BLI (figure 4H).35 36 Both AdTR37 and CpG38 have been used in phase 1 clin-ical trials with low toxicity, illustrating potential transla-tional relevance. Here, combinatorial AdTR/CpG/PD-1 reduced renal tumors by 75% in young, lean mice and was more efficacious than either AdTR/CpG or anti- PD-1 alone (figure 4B), leading to prolonged survival through day 100 (figure 4C). A similar analysis in older, DIO mice

Figure 4 Obesity reduces the efficacy of a novel AdTR/CpG/PD-1 combinatorial immunotherapy in preclinical renal cancer. (A) Experimental design for panels (B, C). (B) Day 28 renal tumor weights for lean animals, treated as indicated. Percentage indicates change in tumor burden relative to untreated controls. (C) Percent survival. (D) Change in body weights of LFD and HFD- fed animals. (E) Experimental design for panels (F, G). (F) Responder threshold (dashed line) was calculated as a >75% reduction in therapy- treated tumor burden compared with untreated controls as determined in (B). Day 28 renal tumor weights for lean, DIO and obese resistant (OB- Res) animals. (G) Response rates for lean, DIO and OB- Res therapy- treated animals from (F). (H) Day 6 tumor burden prior to therapy administration as determined by bioluminescence. Data are pooled from (C) a single experiment (n=10/group) or (B, D, F–H) two or more independent experiments and presented as means±SEM, survival or individual animals shown. Statistical differences were calculated using (B) nonparametric Kruskal- Wallis test with uncorrected Dunn’s multiple comparisons test (#p<0.05, ##p<0.01, #### p<0.0001), (C) log- rank (Mantel- Cox) test (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001), (F) nonparametric Mann- Whitney U tests (###p<0.001) or (H) two- way ANOVA with Bonferroni’s multiple comparisons tests as appropriate. ANOVA, analysis of variance; DIO, diet- induced obese; HFD, high- fat diet; LFD, low- fat chow diet; ns, not significant; NT, no therapy; PD-1, programmed cell death-1; Tx, AdTR/CpG/PD-1 therapy.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from



Open access

revealed that neither single agent anti- PD-1, AdTR/CpG, nor the combination of AdTR/CpG/PD-1 led to a signif-icant reduction in excised renal tumor weights (online supplemental figure S3A). However, the combination of AdTR/CpG/PD-1 did reduce tumor weights by 42% in DIO mice, an improvement over the 3% and 4% reduc-tions detected in mice treated with either anti- PD-1 or AdTR/CpG, respectively.

We therefore used this AdTR/CpG/PD-1 approach to evaluate the impact of obesity on therapeutic efficacy in age- matched mice. Lean and DIO mice (figure 4D) were established, challenged with orthotopic renal tumors, then treated with AdTR/CpG/PD-1 (figure 4E). Because AdTR/CpG/PD-1 reduced excised tumor weights by an average of 75% in young, lean mice relative to untreated controls (figure 4B), we classified individual AdTR/CpG/PD-1- treated mice as responders if they exhibited a ≥75% reduction in excised tumor weight, relative to the mean tumor weight of their respective untreated controls (figure 4F). These objectively determined responder thresholds are indicated by dotted lines in figure 4F. Lean mice maintained on LFD for 20 weeks displayed signifi-cant reductions in tumor weights after receiving AdTR/CpG/PD-1 immunotherapy (figure 4F), translating to a response rate of 73% (figure 4G). In contrast, immuno-therapy did not significantly alter tumor weights in DIO mice (figure 4F) and a response rate of only 44% was achieved (figure 4G), demonstrating impaired immuno-therapeutic outcomes in DIO mice. Next, we evaluated changes in excised renal tumor weights over time, and found that at day 15 post- tumor challenge, tumor weights were not significantly different among lean and DIO mice, in the absence or presence of AdTR/CpG/PD-1 therapy (online supplemental figure S4). However, by day 28, lean mice receiving AdTR/CpG/PD-1 exhibited a significant reduction in renal tumor burdens relative to lean and DIO NT controls and DIO mice receiving AdTR/CpG/PD-1(online supplemental figure S4). In DIO mice, no significant reductions in tumor burdens were observed following therapy administration (online supplemental figure S4). These outcomes reflect our clin-ical observations in RCC patients treated with anti- PD-1 (figure 1), further highlighting the clinical relevance of our preclinical renal tumor model.

As lean and DIO mice are fed different diets, we investigated the possibility that HFD administration, rather than obesity, was impairing therapeutic efficacy in DIO mice. To do this, we used our previously char-acterized OB- Res BALB/c mice.20 OB- Res mice are DIO cage- mates maintained on the same HFD for 20 weeks, but they demonstrate no weight gain relative to LFD- fed controls (figure 4D). More importantly, we have reported that OB- Res mice are physiologically comparable (ie, adiposity, serum leptin and insulin) to lean LFD- fed mice.20 AdTR/CpG/PD-1 efficacy was maintained in OB- Res mice (figure 4F), resulting in a response rate of 73% (figure 4G), equivalent to that observed in lean mice. Therefore, the loss of therapeutic

efficacy in DIO mice was linked to obesity, rather than HFD use.

We then examined potential underlying factors that could contribute to decreased tumor clearance and response rates in DIO mice. These included: tumor burdens at treatment initiation (figure 4H), body weights of individual DIO mice at tumor challenge, the frequency of PD- L1+ CD45+ or CD45- cells within renal tumors, and day 28 excised renal tumor weights in untreated mice (online supplemental figure S3B–E). No significant changes in any of the above parameters were found. We also interrogated intratumoral T cell receptor (TCR) diversity via deep sequencing at day 28 in untreated lean and DIO mice, as it was previously reported that obesity compromises TCR repertoire diversity.39 However, whole tumor αβ T cell repertoire sequencing revealed no signif-icant differences in magnitude or diversity between DIO and lean mice (online supplemental figure S3F). Finally, as these studies were conducted using female mice, we asked if our observations extended to male mice. We determined that the reduced response rate to immuno-therapy also occurred in male DIO mice, as only 17% of DIO males but 78% of lean males responded to AdTR/CpG/PD-1 (online supplemental figure S3G), demon-strating that loss of immunotherapeutic efficacy in DIO mice was consistent in both males and females. There-fore, none of the above factors appeared to contribute to differential AdTR/CpG/PD-1 response rates in DIO versus lean mice.

Lean and OB AdTR/CpG/PD-1 responders display similar intratumoral gene expression profiles that are distinct from those of non-responders (NR)To identify potential mediators of obesity- associated immune dysfunction within tumors, we evaluated gene expression profiles (GEPs) from whole renal tumors in untreated lean and DIO mice. We performed nanoString Immune profiling of 750 immune- related genes in tumor- free or tumor- bearing kidneys from treatment- naive DIO versus lean mice. GEPs were remarkably conserved between lean and DIO treatment- naive tumors and tumor- free contralateral kidneys (online supplemental figure S5A,B), illustrating that underlying immune responses were minimally impacted by obesity.

Next, we asked if obesity altered immune- related GEPs following AdTR/CpG/PD-1 therapy. We therefore used nanoString immune profiling as a screening tool to iden-tify broad immunogenetic changes in whole renal tumors from individual lean and DIO untreated mice, as well as treatment responders and NR. Here, unbiased two- way hierarchical analysis revealed that all lean and DIO treat-ment responders clustered together (figure 5A) and were characterized by a conserved group of highly expressed genes (denoted by the yellow dendrogram). Strik-ingly, DIO NR clustered with lean and DIO no therapy (NT) mice, demonstrating their lack of robust immune remodeling in response to therapy administration. Lean

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from














Open access

NRs were characterized by a conserved group of highly expressed genes (denoted by the green dendrogram).

To more fully understand GEPs that characterized a productive response to therapy, we used volcano plots to identify genes that were differentially expressed (DE; p<0.05 significance threshold indicated by dotted line) in lean or DIO responders versus their respective NR group (figure 5B). We identified several gene expression changes conserved in both lean and DIO responders. Of the shared responder genes (online supplemental table S4), many of those upregulated were indicative of CD8+ T cell immunity (ie, CD3e, CD3g, Cd8a, Cd8b, Ccl5, Gzma, Prf1). Conserved downregulated genes included those related to myeloid- lineage population dynamics (ie, Csf2, Ccl2, S100a8, Cxcr2, Il1r1, Il1b and Arg1). These patterns were reinforced through our analysis of nanoString cell type scores, which showed that tumors from lean and DIO responders had increased CD8+ T cell (‘CD8 T’) and

cytotoxic cell scores (‘Cyto’, comprizing both T cells and NK cells) relative to NR. Responders also had simulta-neous reductions in dendritic cell (DC) and macrophage (“MΦ”) scores (figure 5C). As the majority of DC in Renca tumors from DIO mice are myeloid lineage cells with potent suppressor functions,8 our GEP analyses suggested that productive immune responses to AdTR/CpG/PD-1 were characterized by strong CD8+ T cell responses with concomitant reductions in myeloid cell responses.

Increased prevalence of effector PD-1intCD8+ TILs and concurrent ablation of myeloid-derived suppressor cells underlies successful therapeutic outcomes in both lean and DIO miceWe next sought to validate and expand on the identified immunogenetic changes (figure 5) via flow cytometric cellular profiling of day 28 tumors from lean and DIO therapy- naive and therapy- treated mice. The percentages

Figure 5 Lean and DIO AdTR/CpG/PD-1 responders share a conserved tumoral gene expression profile that is distinct from those of non- responders and treatment- naive mice. (A) Heatmap depicting unbiased hierarchical clustering and gene expression patterns for 750 immune- related genes in the tumors of therapy- naive and therapy- treated lean and DIO animals. (B) Volcano plots showing expression of 750 genes in (left) lean and (right) DIO responders versus respective non- responders. Dashed line indicates an exploratory unadjusted p value threshold (p=0.05) to screen differentially expressed (DE) target genes. Dot colors correspond to genes from green, blue and yellow hierarchical clusters in (A) that were also significantly (p<0.05) DE in both lean and DIO responders versus non- responders. (C) NanoString- generated cell type scores comparing relative gene expression- based population abundance in responding and non- responding lean and DIO therapy- treated animals. DC, dendritic cell; DE, differentially expressed; DIO, diet- induced obese; no therapy, NT; non- responder, NR; PD-1, programmed cell death-1; Res, responder.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from





Open access

of activated (CD44+) CD8+ TILs were robustly and significantly increased in both lean and DIO AdTR/CpG/PD-1 responders relative to respective untreated controls (figure 6A), confirming our nanoString data. In contrast, lean and DIO NR exhibited a weak intratu-moral CD44+CD8+ T cell presence. Furthermore, in lean and DIO responders to AdTR/CpG/PD-1 therapy, acti-vated CD8+ TILs had low PD-1 expression (figure 6B), likely due to effective tumor clearance limiting repeated TCR engagement.26 30 31 We also observed significant increases in the frequencies of IFNγ+CD8+ TILs in lean and DIO responders, relative to their respective

untreated controls and NR (figure 6C). We determined that reduced PD-1 expression in responders was driven by a relative shift toward PD-1intCD8+ TILs in both lean and DIO mice (figure 6D). In both lean and DIO responders, IFNγ, TNFα and perforin were produced primarily by PD-1intCD8+ TILs (figure 6E). Next, as CD4+ TILs are known to modulate the quality and magnitude of CD8+ TIL responses,40 we examined the CD4+ T cell response. Activated CD4+ TILs (online supplemental figure S6A) were elevated in both lean and DIO therapy responders (online supplemental figure S6B). However, CD4+Foxp3+ Treg percentages were unchanged by obesity or response

Figure 6 Successful response to immunotherapy in both lean and DIO mice is associated with a predominance of effector PD-1intCD8+ TILs and reduced MDSCs. Analysis of day 28 intratumoral leukocytes from indicated treatment groups. (A) Frequencies of activated CD8+ TILs, (B) PD-1 expression on activated CD8+ TILs, (C) frequencies of IFNγ+ CD8+ TILs, (D) activated CD8+ TILs based on PD-1 expression level, (E) cytokine producing CD8+ TILs based on PD-1 expression level and (F) MDSCs. (G) Ratio of activated CD8+ TILs to MDSCs. Data are pooled from at least two independent experiments and presented as means±SEM. Statistical differences were determined by two- way ANOVA followed by post hoc Bonferroni’s multiple comparisons tests (*p<0.05, **p<0.01, ***p<0.001, ****p<0.0001). ANOVA, analysis of variance; DIO, diet- induced obese; IFNγ, interferon-γ; MDSCs, myeloid- derived suppressor cells; NR, non- responder; NT, no therapy; PD-1, programmed cell death-1; Res, responder; TILs, tumor- infiltrating lymphocyte.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from





Open access

to therapy (online supplemental figure S6B). These results indicate that immunotherapy administration in lean and DIO responders initiated a robust expansion of intratumoral PD-1intCD8+ T cells that produced IFNγ, TNFα and perforin.

We previously reported that in this orthotopic Renca model, obesity is associated with heightened intratu-moral immune suppression via myeloid- lineage DC and myeloid- derived suppressor cell (MDSC) which actively suppress CD8+ T cell proliferation when sort- purified and functionally evaluated ex vivo.6 8 Here, DIO therapy responders exhibited significant reduc-tions in tumor- infiltrating CD11b+CD11chighMHC II+ DC (online supplemental figure S7A), relative to NR (online supplemental figure S7C). DC subtype analysis (online supplemental figure S7B) showed no alterations toward or away from conventional DC1 (XCR1+Sirpα) or DC2 (XCR1-Sirpα+) subsets with obesity or response to treat-ment (online supplemental figure S5D,E). Total intratu-moral MDSCs (CD45+/CD11b+/CD11c-/MHC II-/Ly6G+ or Ly6C+) (gating strategy as per online supplemental figure S7A) were also significantly reduced in DIO and lean responders to AdTR/CpG/PD-1 therapy, relative to respective NR and untreated controls (figure 6F). MDSC subtype analysis showed a predominance of polymor-phonuclear ‘PMN- MDSCs’ over monocytic ‘M- MDSCs’ (online supplemental figure S7F–G), although both subtypes are suppressive in this model.6 Notably, the ratio of CD44+CD8+ TILs to MDSCs was significantly and favorably increased in DIO and lean responders (figure 6G), confirming our nanoString data (figure 5). Thus, our cellular analyses illustrated that AdTR/CpG/PD-1 success in both lean and OB mice was defined by immune responses that culminate in similar, favorable ratios of activated CD8+ T cells to MDSCs within tumors. Therefore, obesity reduced the percentage of mice that respond favorably to this immunotherapy, without altering the quality of immune responses in DIO versus lean treatment responders.

Obesity-associated alterations in intratumoral chemokine signatures contribute to diminished therapeutic outcomesNext, we sought to identify factors responsible for the differential response rates to therapy in DIO versus lean mice. We examined intratumoral immune responses at an early day 15 time point in treatment- naive mice, with the goal of identifying obesity- associated drivers of the diver-gent therapy outcomes we had identified (figure 4F,G). At this time, DIO mice displayed a trending but non- significant reduction in the frequency of activated CD44+CD8+ TILs versus lean mice (figure 7A), suggesting that baseline T cell priming and trafficking were predom-inantly intact in DIO hosts.

To more directly evaluate the effects of obesity on CD8+ T cell priming in tumor- bearing mice, we used a comple-mentary CMS5 fibrosarcoma tumor model that permits tracking of tumor antigen- specific T cell responses to an endogenous tumor antigen.22 23 Equal numbers of naive,

tumor antigen- specific, TCR transgenic DUC18 CD8+ TILs from lean mice were CFSE- labeled and transferred into lean or DIO recipients with early established tumors. Comparable percentages of activated (CD25+) prolif-erative DUC18 T cells were present in tumor- draining inguinal lymph nodes from lean and DIO mice at day 4 post- T cell transfer (online supplemental figure S8A), illustrating that T cell priming was not impaired by host obesity in this model. However, analysis of endpoint CMS5 tumor areas showed tumor regression (ie, no palpable tumor detected) in 75% of lean recipients, as expected,22 23 but only 33% of DIO recipients (online supplemental figure S8B). Therefore, despite intact T cell priming, transferred DUC18 T cells were less able to control CMS5 tumors outgrowth in obese animals, reflecting our findings in mice with renal tumors.

Having found no obvious defects in T cell priming in DIO mice, we examined MDSCs within renal tumors. At day 15 post- tumor challenge, DIO mice had significantly increased frequencies of tumor- infiltrating MDSCs in the absence of therapy (figure 7B). We therefore examined known soluble mediators of MDSC responses. Treatment- naive DIO mice had increased intratumoral concen-trations of the MDSC- related proteins GM- CSF, IL-1β, CXCL1 and CCL26 41–43 at day 15, relative to lean mice (figure 7C). IL-1β is a known driver of MDSC genera-tion, accumulation and suppressive capacity,44 45 and was decreased at the gene expression level in both lean and DIO responder mice (figure 5B). Investigating the IL-1β pathway is clinically relevant, as high intratumoral IL-1β gene expression is an unfavorable prognostic in renal cancer patients (p<0.001, thehumanproteinatlas. org) and elevated IL-1β in human renal tumors is associated with increased MDSC accumulation.43

In our renal cancer model, intratumoral IL-1β concen-trations correlated positively with both tumor size and tumor- infiltrating MDSC abundance across all categories of lean and DIO treatment- naive and therapy- treated animals (figure 7D). We therefore asked if neutralization of IL-1β in AdTR/CpG/PD-1 treated DIO mice would improve treatment response rates. Indeed, administra-tion of anti- IL-1β in DIO mice receiving AdTR/CpG/PD-1 led to a significant reduction in tumor weights vs NT controls, and improved the mean response rate to 58% vs 33% in DIO AdTR/CpG/PD-1- treated mice with intact IL-1ß (figure 7E). Administration of anti- IL-1β in DIO mice receiving AdTR/CpG/PD-1 reduced intra-tumoral concentrations of IL-1ß (figure 7F) and also lowered concentrations of other myeloid- related proteins within tumors, including GM- CSF, CXCL1 and CCL2 (online supplemental figure S9A–C). Use of anti- IL-1β did not alter intratumoral concentrations of IL-2, IL-4, or CCL20 (online supplemental figure S9D–F), illustrating the specificity of this approach. Importantly, neutralizing IL-1β did reduce the frequencies of total MDSCs within renal tumors (figure 7G). Thus, reduced AdTR/CPG/PD-1 response rates in DIO mice were mediated in part by early, heightened elevations in intratumoral IL-1β

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from


















Open access

concentration and MDSCs, changes that negated the largely intact CD8+ T cell response present in OB hosts.

Finally, we asked if the mean response rate in lean mice could be impaired by neutralizing a chemoattrac-tant mediator of effector T cell trafficking into tumors. In concordance with our finding that host obesity did not

significantly alter day 15 CD44+CD8+ TIL accumulation, we observed no differences in the intratumoral concen-trations of T cell- related chemokines (CCL4, CCL5, CXCL10 and CXCL11) in DIO versus lean animals (online supplemental figure S10A). However, CCL5 was recently identified as a key regulator of CD8+ TIL trafficking in human renal tumors,41 and we found that elevated CCL5

Figure 7 Early intratumoral chemokine alterations in obese mice drive differential therapy response rates. On day 15, post- tumor challenge lean and DIO treatment- naive mice were evaluated for (A) activated CD8+ TILs, (B) MDSCs and (C) myeloid- associated cyto/chemokines within tumors. Linear regression of intratumoral (D) IL-1β concentrations versus (left) tumor weight and (right) MDSCs in lean and DIO therapy- naive and therapy- treated animals. DIO animals were treated with no therapy or AdTR/CpG/PD-1±anti- IL-1β neutralizing antibody. Day 28 endpoint (E) tumor weights, corresponding response rates, intratumoral (F) IL-1β concentrations and (G) MDSCs. Data are pooled from at least two independent experiments and presented as means±SEM or boxes defining 25th to 75th percentiles with line at median and whiskers extending to minimum and maximum points. Statistical differences were calculated using parametric t- tests (*p<0.05, **p<0.01, ****p<0.0001), nonparametric Mann- Whitney U tests (##p<0.01, ###p<0.001, ####p<0.0001), or linear regression analyses as appropriate. DIO, diet- induced obese; IL-1β, interleukin-1β; MDSCs, myeloid- derived suppressor cells; ns, not significant; NT, no therapy; PD-1, programmed cell death-1; Tx, AdTR/CpG/PD-1 therapy.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from





Open access

was associated with reduced tumor weights and increased CD44+CD8+ T cell infiltration into renal tumors of mice (online supplemental figure S10B). Neutralizing CCL5 in lean mice receiving AdTR/CpG/PD-1 therapy reduced the mean response rate from 86% (in mice with intact CCL5) to 46% (online supplemental figure S10C) and administration of isotype had no effect (online supple-mental figure S10D). Collectively, our results implicate IL-1β and CCL5 as important regulators of AdTR/CPG/PD-1 outcomes in lean and DIO mice with renal tumors.

DISCUSSIONHere, we report that that obesity, as defined by the WHO’s BMI cut point of ≥30 kg/m2, is associated with worse OS and PFS in metastatic RCC patients who received standard of care anti- PD-1, as evidenced by clinical outcome data from two independent institutions. These clinical trends are reflected in the reduced response rates we observed in DIO mice- bearing established renal tumors, wherein an anti- PD-1- based combinatorial immunotherapy had reduced efficacy. We demonstrate that this outcome arises in part from early, obesity- associated elevations in intratu-moral IL-1β concentration and MDSCs. Neutralization of IL-1β was sufficient to restore immunotherapy response rates in DIO mice to the level seen in lean animals. These immunotherapy outcomes are consistent with our flow cytometric data regarding the PD-1 status of CD8+ TILs from both human and murine renal tumors, which illus-trate that obesity is associated with reduced frequencies of activated PD-1highCD8+ TILs. Our data also concur with an earlier report by Thommen et al that linked elevated frequencies of PD-1highCD8+ TILs with improved responses to anti- PD-1 in a cohort of non- small cell lung carcinoma patients.26 In RCC, we find fewer PD-1highCD8+ TILs and worse responses to anti- PD-1 with obesity. Thus, our human and murine findings present a cohesive view regarding the detrimental effects of host obesity on anti- PD-1- based treatment outcomes, a provocative scenario given the current focus on obesity paradox paradigms in cancer immunotherapy.

Our findings are consistent with multiple preclinical studies that show obesity alters the tumor microenviron-ment and/or immune response to facilitate tumor progres-sion and dissemination. For example, we had reported previously that obesity was associated with increased frequencies of tumor- infiltrating DC that suppressed CD8+ T cells ex vivo, corresponding with reduced efficacy of AdTR/CpG.8 Additional obesity- associated mecha-nisms include: myeloid cell promotion of metastasis via IL-5 and GM- CSF,10 altered MDSC responses that impede CD8+ T cell immunity,6 46 decreased NK function,47 and increased production of proinflammatory cytokines from myeloid cells and adipocytes that promote tumor cell proliferation and immunosuppression.7 Additional studies showed that obesity triggers adverse changes in the T cell compartment, ranging from reduced T cell thymic progenitors39 to impaired metabolism and effector

function in CD8+ TILs.48 More recently, we demonstrated that DIO mice with mammary carcinoma display immu-notherapy resistance, due in part to CXCL1- driven accu-mulation of FasL+ MDSCs inducing apoptosis in effector CD8+ TILs.49 Our current study extends this body of literature by providing novel insight regarding the role for obesity- associated intratumoral IL-1β in the develop-ment of a tumor microenvironment that is more resistant to anti- PD-1- based immunotherapy (figure 7 and online supplemental figure S9).

Despite the numerous preclinical findings illustrating the negative effects of obesity on antitumor immunity, several retrospective clinical outcome studies report obesity paradox paradigms in patients with melanoma and other mixed tumor types. These studies indicated that obesity (BMI ≥30 kg/m2) is associated with improved outcomes following immune checkpoint blockade with anti- PD-1 or anti- PD- L1 in the presence or absence of anti- CTLA-4, administered either during clinical trials or as standard of care.11 12 15 However, published reports exam-ining the effects of obesity on immune checkpoint inhib-itor outcomes in RCC, specifically, are less supportive of this obesity paradox paradigm. For example, a recent meta- analysis reported beneficial effects of combined over-weight plus obesity (BMI ≥25 kg/m2), in terms of immune checkpoint blockade outcomes, across a variety of tumor types.17 Three independent RCC studies were included in this meta- analysis, but none examined immune check-point blockade outcomes using the WHO- defined cut- off of BMI ≥30 kg/m2 to determine the specific effects of obesity on clinical outcomes, as we did here. One of these studies, by Bergerot et al, found that elevated BMI was beneficial when RCC patients received targeted therapies such as VEGF or mTOR inhibitors, but found a detri-mental trend when immune checkpoint inhibitors were used at the same institution.18 A 2019 report by Sanchez et al examined retrospective outcomes from an observa-tional cohort of RCC patients (n=203) treated at Memo-rial Sloan Kettering between 2011 and 2018 with immune checkpoint inhibitors (monotherapy or any combination of anti- PD-1/PD- L1/CTLA-4).16 In this cohort, obesity was not associated with increased mortality after adjusting for IMDC risk score. This is in contrast to the authors reported results from the COMPARZ clinical trial of pazopanib vs sunitinib, wherein obesity was associated with improved OS.16 These results raise the possibility that obesity may differentially impact immune checkpoint inhibitor versus targeted therapy outcomes in RCC. Our analysis exam-ined outcomes only in RCC patients receiving anti- PD-1 monotherapy as standard of care and thus provides an insight that is distinct from that of prior studies, which performed analyses across multiple tumor types, immu-notherapeutic agents, and/or a combination of clinical trial and standard of care administration.11 12 14–16 There-fore, perhaps due to the these dissimilar nature of the analyses performed to date, no clear consensus regarding the effects of obesity on immune checkpoint inhibitor outcomes in RCC has yet emerged.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from









Open access

Our finding that treatment- naive RCC subjects with obesity and DIO mice had reduced percentages of PD-1highCD8+ TILs is intriguing, particularly because this TIL subset maintained potent effector function in untreated mice (figure 3). Given our findings that obesity promotes early MDSC accumulation in mice, we postulate a scenario in which obesity promotes increased IL-1β within tumors, thereby increasing the frequency of tumor- infiltrating MDSCs, resulting in an impaired ability of CD8+ TILs to engage with cognate antigen and respond by increasing PD-1 expression. In support of this model, we found that neutralizing IL-1β restored immu-notherapeutic efficacy in DIO animals (figure 7E). This reflects findings from other groups, wherein blocking IL-1β reduced tumor growth in the presence and absence of obesity,9 50 51 and improved response to anti- PD-1 immunotherapy in lean mice with mammary carci-noma.45 Furthermore, we observed that PD-1 expression on CD8+ TILs increased over time in untreated lean, but not DIO, mice (figure 2G). Other groups have found that high proportions of PD-1highCD8+ TILs prior to therapy predicted improved anti- PD-1 responses in patients with lung cancer,26 and elevated frequencies of PD-1+CD8+ T cells in the pretreatment setting correlated with response to anti- PD-1 in murine melanoma.11 Thus, the frequency of PD-1+CD8+ TILs and expression levels of PD-1 on those cells should be examined more closely for poten-tial associations with PD-1 outcomes in cancer patients, particularly in the context of obesity where heightened inflammation and MDSC accumulation occur.

Naturally, our study contains limitations. For example, our preclinical model of kidney cancer does not permit the examination of tumor antigen- specific T cell responses and is syngeneic to BALB/c mice, which develop obesity without insulin resistance.20 This is notable as type 2 diabetes may well influence immuno-therapy outcomes in clinical practice. Despite this limita-tion, our orthotopic Renca model resembles human RCC in terms of the immune responses to these renal tumors: the total frequency of activated CD8+ TILs is unchanged with obesity whereas the frequencies of PD-1highCD8+ TILs are reduced with obesity (figure 2), and in the significant but modest improvement in animal survival following anti- PD-1 monotherapy (figure 4). Our study also investigated immune- mediated mechanisms contrib-uting to impaired AdTR/CpG/PD-1 response with obesity, specifically focusing on obesity- associated IL-1β. However, obesity- associated inflammation is propagated through multiple adipokines and proinflammatory cyto-kines51 that we did not fully examine. Because obesity is a multifactorial disease, it is likely that multiple mecha-nisms will synergize to dictate immunotherapy success or failure in cancer patients. Additional in- depth transcrip-tomic studies such as those recently reported by Sanchez et al,16 as well as detailed cellular analyses, will permit a deeper understanding of when and why obesity is associ-ated with positive or negative outcomes in cancer patients receiving immunotherapy. Additionally, we do not know if

therapy- induced survival is also negatively impacted in our DIO mice. However, based on our prior results examining survival in DIO mice that received only AdTR/CpG,8 we fully expect that DIO mice would show significant reduc-tions in survival in response to AdTR/CpG/PD-1, relative to age- matched lean counterparts. Given the retrospec-tive nature of our survival outcomes data in RCC patients, we were unable to clinically validate our observations in mice regarding the roles of IL-1β and CCL5 in immu-notherapy outcomes; therefore, future studies should examine the abundance and function of these proteins in immunotherapy- treated RCC patients with and without obesity. Finally, although our survival outcomes analysis features RCC patients from two separate institutions, our cohort is small and it will be essential to perform similar studies with larger cohorts at other institutions, so that separate analyses for underweight, normal weight, over-weight and OB patients can be conducted.

CONCLUSIONSObesity is a health concern that is reaching epidemic proportions globally, and its increasing prevalence shows no sign of slowing. Our study provides important insights into the effects of obesity on anti- PD-1- based treatment outcomes: in mice receiving a combinatorial AdTR/CpG/PD-1 therapy, as well as in RCC patients receiving anti- PD-1 monotherapy as standard of care, obesity was associated with poorer treatment outcomes. These unified results add a critical element to our current understanding of the ways in which obesity impacts antitumor immunity and immune checkpoint blockade outcomes. However, given the uncertainty in the field, rigorous continued study of this critical issue is needed to better inform patient care.

Author affiliations1Graduate Biomedical Sciences, The University of Alabama at Birmingham, Birmingham, Alabama, USA2Department of Nutrition Sciences, The University of Alabama at Birmingham, Birmingham, Alabama, USA3Department of Urology, University of Iowa Roy J and Lucille A Carver College of Medicine, Iowa City, Iowa, USA4Honors Undergraduate Research Program, School of Health Professions, The University of Alabama at Birmingham, Birmingham, Alabama, USA5Interdisciplinary Program in Immunology, The University of Iowa, Iowa City, Iowa, USA6HudsonAlpha Institute for Biotechnology, Huntsville, Alabama, USA7Science and Technology Honors Program, College of Arts and Sciences, The University of Alabama at Birmingham, Birmingham, Alabama, USA8Department of Psychology, The University of Alabama at Birmingham, Birmingham, Alabama, USA9University of Iowa Holden Comprehensive Cancer Center, Iowa City, Iowa, USA10Departments of Pathology and Urology, Vanderbilt University Medical Center, Nashville, Tennessee, USA11Department of Internal Medicine, University of Iowa Roy J and Lucille A Carver College of Medicine, Iowa City, Iowa, USA12Genitourinary Oncology Program, Division of Hematology, Oncology and Blood and Marrow Transplantation, University of Iowa Roy J and Lucille A Carver College of Medicine, Iowa City, Iowa, USA13Division of Hematology and Oncology, The University of Alabama at Birmingham School of Medicine, Birmingham, Alabama, USA

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from



Open access

14The University of Alabama at Birmingham O'Neal Comprehensive Cancer Center, Birmingham, Alabama, USA15Department of Internal Medicine, Division of General Medicine, University of Iowa Roy J and Lucille A Carver College of Medicine, Iowa City, Iowa, USA16Department of Obstetrics and Gynecology, The University of Alabama at Birmingham School of Medicine, Birmingham, Alabama, USA17Department of Biostatistics, The University of Alabama at Birmingham School of Nursing, Birmingham, Alabama, USA18Genitourinary Oncology Program, Division of Hematology, Oncology and Blood and Marrow Transplantation, The University of Iowa, Iowa City, Iowa, USA19Department of Radiation Oncology, The University of Alabama at Birmingham School of Medicine, Birmingham, Alabama, USA20Division of Pharmaceutics and Translational Therapeutics, University of Iowa College of Pharmacy, Iowa City, Iowa, USA21Department of Genetics, The University of Alabama at Birmingham, Birmingham, Alabama, USA22Department of Cell, Developmental and Integrative Biology, The University of Alabama at Birmingham, Birmingham, Alabama, USA

Acknowledgements The authors would like to thank the following UAB cores and facilities: nanoString facility, and particularly Debbie Della Mana, the O’Neal Comprehensive Cancer Center’s Preclinical Imaging Shared Facility, the Heflin Center for Genomic Science Genomics Core, and the Animal Resources Program. We also thank the Genitourinary Molecular Epidemiology Research resource at UI’s Holden Comprehensive Cancer Center, and the Rock- N- Ride charity of Iowa. The authors would also like to thank the laboratory of Dr. Francis Lund for use of the MAGPIX Luminex plate reader.

Contributors LAN and SKB conceived the study. LAN, SKB, RMO, JTG and WJT designed experiments. LAN, SKB, RMO, JTG, WJT, GW, LT, CBR, DML and KIF performed experiments and analyzed data. KEN and TTM- L analyzed data. MB, LB and LT helped consent RCC patients into the prospective study and process biospecimens. JG, KN and JAB helped acquire renal tumor specimens. LAN and EY helped obtain i2B2 and outcome data for UAB RCC patients. HY, RG, and YZ performed chart reviews and acquired outcome data for UI RCC patients. PL performed statistical analyses of human subject data and provided statistical guidance on murine data analysis and interpretation. AM, BPG, RES, GJW, AKS, DML, RCA, JAB, EY, YZ, and KN provided scientific guidance. LAN, SKB, RMO, JTG, WJT, GW, AM, BPG, DML, TTM- L and AKS revised and edited the manuscript. All authors provided critical review and final approval of the manuscript. Authorship order among co- first authors was assigned accordingly: SKB designed, conducted and analyzed initial experiments; RMO and JTG designed, conducted and analyzed follow- up experiments. LAN and RMO wrote the manuscript and JTG created/compiled figures for publication.

Funding Research reported in this publication was supported by NIH Grant #5R01CA181088 to LAN, NIAID award #T32AI007051 and NCI award #R25CA047888 to SKB, NHLBI award #T32HL105349 and NCI award #T32CA047888 to RMO, NIGMS award #T32GM008111 to JTG, NIDDK award #T32DK062710 and NCI award #T32CA183926 to WJT, NIDDK award #P30DK056336 (to UAB Nutrition and Obesity Research Center), and NIH awards #P30CA013148 and #1S10OD021697 (to UAB O’Neal Comprehensive Cancer Center Preclinical Imaging Shared Facility). TTM- L was supported by her startup package through the UAB School of Medicine.

Disclaimer The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.

Competing interests CBR is currently employed by Immunomic Therapeutics. RCA serves as an advisory board member for Clovis, Tesaro and LEAP therapeutics. EY receives research funding from Eli Lilly and serves as an advisory board member for Bayer and AstraZeneca.

Patient consent for publication Not required.

Ethics approval Approval for our prospective human subjects study was granted by the IRB of The University of Iowa (UI). IRB approval for our retrospective study of metastatic RCC patients treated with anti- PD-1 was obtained from The University of Iowa and University of Alabama at Birmingham (UAB) Hospitals.

Provenance and peer review Not commissioned; externally peer reviewed.

Data availability statement Data are available on reasonable request. All data relevant to the study are included in the article or uploaded as online supplemental information. Gene expression data from this study are available from the corresponding author on reasonable request.

Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer- reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.

Open access This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY- NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non- commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non- commercial. See http:// creativecommons. org/ licenses/ by- nc/ 4. 0/.

ORCID iDsJustin Tyler Gibson http:// orcid. org/ 0000- 0003- 2967- 5338Rohan Garje http:// orcid. org/ 0000- 0002- 7244- 5602Lyse A Norian http:// orcid. org/ 0000- 0003- 1016- 313X

REFERENCES 1 Ribas A, Wolchok JD. Cancer immunotherapy using checkpoint

blockade. Science 2018;359:1350–5. 2 Hales CM, Fryar CD, Carroll MD, et al. Trends in obesity and severe

obesity prevalence in US youth and adults by sex and age, 2007-2008 to 2015-2016. JAMA 2018;319:1723.

3 Lauby- Secretan B, Scoccianti C, Loomis D, et al. Body Fatness and Cancer--Viewpoint of the IARC Working Group. N Engl J Med 2016;375:794–8.

4 Sanfilippo KM, McTigue KM, Fidler CJ, et al. Hypertension and obesity and the risk of kidney cancer in 2 large cohorts of US men and women. Hypertension 2014;63:934–41.

5 Paich HA, Sheridan PA, Handy J, et al. Overweight and obese adult humans have a defective cellular immune response to pandemic H1N1 influenza A virus. Obesity 2013;21:2377–86.

6 Hale M, Itani F, Buchta CM, et al. Obesity triggers enhanced MDSC accumulation in murine renal tumors via elevated local production of CCL2. PLoS One 2015;10:e0118784.

7 Incio J, Ligibel JA, McManus DT, et al. Obesity promotes resistance to anti- VEGF therapy in breast cancer by up- regulating IL-6 and potentially FGF-2. Sci Transl Med 2018;10:eaag0945. doi:10.1126/scitranslmed.aag0945

8 James BR, Tomanek- Chalkley A, Askeland EJ, et al. Diet- induced obesity alters dendritic cell function in the presence and absence of tumor growth. J Immunol 2012;189:1311–21.

9 Kolb R, Phan L, Borcherding N, et al. Obesity- associated NLRC4 inflammasome activation drives breast cancer progression. Nat Commun 2016;7:13007.

10 Quail DF, Olson OC, Bhardwaj P, et al. Obesity alters the lung myeloid cell landscape to enhance breast cancer metastasis through IL5 and GM- CSF. Nat Cell Biol 2017;19:974–87.

11 Wang Z, Aguilar EG, Luna JI, et al. Paradoxical effects of obesity on T cell function during tumor progression and PD-1 checkpoint blockade. Nat Med 2019;25:141–51. doi:10.1038/s41591-018-0221-5

12 McQuade JL, Daniel CR, Hess KR, et al. Association of body- mass index and outcomes in patients with metastatic melanoma treated with targeted therapy, immunotherapy, or chemotherapy: a retrospective, multicohort analysis. Lancet Oncol 2018;19:310–22.

13 Albiges L, Hakimi AA, Xie W, et al. Body mass index and metastatic renal cell carcinoma: clinical and biological correlations. J Clin Oncol 2016;34:3655–63.

14 Cortellini A, Bersanelli M, Buti S, et al. A multicenter study of body mass index in cancer patients treated with anti- PD-1/PD- L1 immune checkpoint inhibitors: when overweight becomes favorable. J Immunother Cancer 2019;7:57.

15 Kichenadasse G, Miners JO, Mangoni AA, et al. Association between body mass index and overall survival with immune checkpoint inhibitor therapy for advanced non- small cell lung cancer. JAMA Oncol 2020;6:512.

16 Sanchez A, Furberg H, Kuo F, et al. Transcriptomic signatures related to the obesity paradox in patients with clear cell renal cell carcinoma: a cohort study. Lancet Oncol 2020;21:283–93.

17 Xu H, Cao D, He A, et al. The prognostic role of obesity is independent of sex in cancer patients treated with immune

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

http://creativecommons.org/licenses/by-nc/4.0/

http://orcid.org/0000-0003-2967-5338

http://orcid.org/0000-0002-7244-5602

http://orcid.org/0000-0003-1016-313X

http://dx.doi.org/10.1126/science.aar4060

http://dx.doi.org/10.1001/jama.2018.3060

http://dx.doi.org/10.1056/NEJMsr1606602

http://dx.doi.org/10.1161/HYPERTENSIONAHA.113.02953

http://dx.doi.org/10.1002/oby.20383

http://dx.doi.org/10.1371/journal.pone.0118784

http://dx.doi.org/10.1126/scitranslmed.aag0945

http://dx.doi.org/10.4049/jimmunol.1100587

http://dx.doi.org/10.1038/ncomms13007

http://dx.doi.org/10.1038/ncomms13007

http://dx.doi.org/10.1038/ncb3578

http://dx.doi.org/10.1038/s41591-018-0221-5

http://dx.doi.org/10.1016/S1470-2045(18)30078-0

http://dx.doi.org/10.1200/JCO.2016.66.7311

http://dx.doi.org/10.1186/s40425-019-0527-y

http://dx.doi.org/10.1186/s40425-019-0527-y

http://dx.doi.org/10.1001/jamaoncol.2019.5241

http://dx.doi.org/10.1001/jamaoncol.2019.5241

http://dx.doi.org/10.1016/S1470-2045(19)30797-1



Open access

checkpoint inhibitors: a pooled analysis of 4090 cancer patients. Int Immunopharmacol 2019;74:105745.

18 Bergerot PG, Bergerot CD, Philip EJ, et al. Targeted therapy and immunotherapy: effect of body mass index on clinical outcomes in patients diagnosed with metastatic renal cell carcinoma. Kidney Cancer 2019;3:63–70.

19 Labadie BW, Liu P, Bao R, et al. BMI, irAE, and gene expression signatures associate with resistance to immune- checkpoint inhibition and outcomes in renal cell carcinoma. J Transl Med 2019;17:386.

20 Boi SK, Buchta CM, Pearson NA, et al. Obesity alters immune and metabolic profiles: new insight from obese- resistant mice on high- fat diet. Obesity 2016;24:2140–9.

21 Norian LA, Kresowik TP, Rosevear HM, et al. Eradication of metastatic renal cell carcinoma after adenovirus- encoded TNF- related apoptosis- inducing ligand (TRAIL)/CpG immunotherapy. PLoS One 2012;7:e31085.

22 Norian LA, Allen PM. Rapid maturation of effector T cells in tumors, but not lymphoid organs, during tumor regression. PLoS One 2007;2:e821.

23 Hanson HL, Donermeyer DL, Ikeda H, et al. Eradication of established tumors by CD8+ T cell adoptive immunotherapy. Immunity 2000;13:265–76.

24 Bar- Joseph Z, Gifford DK, Jaakkola TS. Fast optimal leaf ordering for hierarchical clustering. Bioinformatics 2001;17 Suppl 1:S22–9.

25 Eisen MB, Spellman PT, Brown PO, et al. Cluster analysis and display of genome- wide expression patterns. Proc Natl Acad Sci U S A 1998;95:14863–8.

26 Thommen DS, Koelzer VH, Herzig P, et al. A transcriptionally and functionally distinct PD-1+ CD8+ T cell pool with predictive potential in non- small- cell lung cancer treated with PD-1 blockade. Nat Med 2018;24:994–1004.

27 Siddiqui I, Schaeuble K, Chennupati V, et al. Intratumoral Tcf1+PD-1+CD8+ T Cells with Stem- like Properties Promote Tumor Control in Response to Vaccination and Checkpoint Blockade Immunotherapy. Immunity 2019;50:195–211. e10.

28 Ahn E, Araki K, Hashimoto M, et al. Role of PD-1 during effector CD8 T cell differentiation. Proc Natl Acad Sci U S A 2018;115:4749–54.

29 Gros A, Robbins PF, Yao X, et al. PD-1 identifies the patient- specific CD8⁺ tumor- reactive repertoire infiltrating human tumors. J Clin Invest 2014;124:2246–59.

30 Paley MA, Kroy DC, Odorizzi PM, et al. Progenitor and terminal subsets of CD8+ T cells cooperate to contain chronic viral infection. Science 2012;338:1220–5.

31 Zappasodi R, Budhu S, Hellmann MD, et al. Non- Conventional inhibitory CD4+Foxp3–PD- 1hi T cells as a biomarker of immune checkpoint blockade activity. Cancer Cell 2018;34:691.

32 Day CL, Kaufmann DE, Kiepiela P, et al. PD-1 expression on HIV- specific T cells is associated with T- cell exhaustion and disease progression. Nature 2006;443:350–4.

33 Egelston CA, Avalos C, Tu TY, et al. Human breast tumor- infiltrating CD8+ T cells retain polyfunctionality despite PD-1 expression. Nat Commun 2018;9:4297.

34 James BR, Anderson KG, Brincks EL, et al. CpG- mediated modulation of MDSC contributes to the efficacy of Ad5- TRAIL therapy against renal cell carcinoma. Cancer Immunol Immunother 2014;63:1213–27.

35 Walczak H, Miller RE, Ariail K, et al. Tumoricidal activity of tumor necrosis factor- related apoptosis- inducing ligand in vivo. Nat Med 1999;5:157–63.

36 Krieg AM. Development of TLR9 agonists for cancer therapy. J Clin Invest 2007;117:1184–94.

37 Holoch PA, Griffith TS. TNF- related apoptosis- inducing ligand (TRAIL): a new path to anti- cancer therapies. Eur J Pharmacol 2009;625:63–72.

38 Hofmann MA, Kors C, Audring H, et al. Phase 1 evaluation of intralesionally injected TLR9- agonist PF-3512676 in patients with basal cell carcinoma or metastatic melanoma. J Immunother 2008;31:520–7.

39 Yang H, Youm Y- H, Vandanmagsar B, et al. Obesity increases the production of proinflammatory mediators from adipose tissue T cells and compromises TCR repertoire diversity: implications for systemic inflammation and insulin resistance. J Immunol 2010;185:1836–45.

40 Pardoll DM, Topalian SL. The role of CD4+ T cell responses in antitumor immunity. Curr Opin Immunol 1998;10:588–94.

41 Dangaj D, Bruand M, Grimm AJ, et al. Cooperation between constitutive and inducible chemokines enables T cell engraftment and immune attack in solid tumors. Cancer Cell 2019;35:885–900. e10.

42 Li J, Byrne KT, Yan F, et al. Tumor cell- intrinsic factors underlie heterogeneity of immune cell infiltration and response to immunotherapy. Immunity 2018;49:178–93.

43 Najjar YG, Rayman P, Jia X, et al. Myeloid- Derived suppressor cell subset accumulation in renal cell carcinoma parenchyma is associated with intratumoral expression of IL1β, IL8, CXCL5, and Mip-1α. Clin Cancer Res 2017;23:2346–55.

44 Tannenbaum CS, Rayman PA, Pavicic PG, et al. Mediators of inflammation- driven expansion, trafficking, and function of tumor- infiltrating MDSCs. Cancer Immunol Res 2019;7:1687–99.

45 Kaplanov I, Carmi Y, Kornetsky R, et al. Blocking IL-1β reverses the immunosuppression in mouse breast cancer and synergizes with anti- PD-1 for tumor abrogation. Proc Natl Acad Sci U S A 2019;116:1361–9.

46 Clements VK, Long T, Long R, et al. Frontline science: high fat diet and leptin promote tumor progression by inducing myeloid- derived suppressor cells. J Leukoc Biol 2018;103:395–407.

47 Michelet X, Dyck L, Hogan A, et al. Metabolic reprogramming of natural killer cells in obesity limits antitumor responses. Nat Immunol 2018;19:1330–40.

48 Zhang C, Yue C, Herrmann A, et al. STAT3 Activation- Induced Fatty Acid Oxidation in CD8+ T Effector Cells Is Critical for Obesity- Promoted Breast Tumor Growth. Cell Metab 2020;31:148–61.

49 Gibson JT, Orlandella RM, Turbitt WJ, et al. Obesity- Associated myeloid- derived suppressor cells promote apoptosis of tumor- infiltrating CD8 T cells and immunotherapy resistance in breast cancer. Front Immunol 2020;11:590794.

50 Guo B, Fu S, Zhang J, et al. Targeting inflammasome/IL-1 pathways for cancer immunotherapy. Sci Rep 2016;6:36107.

51 Incio J, Liu H, Suboj P, et al. Obesity- Induced inflammation and desmoplasia promote pancreatic cancer progression and resistance to chemotherapy. Cancer Discov 2016;6:852–69.

on Decem


http://jitc.bmj.com

/J Im

munother C


ber 2020. Dow

nloaded from

http://dx.doi.org/10.1016/j.intimp.2019.105745

http://dx.doi.org/10.1016/j.intimp.2019.105745

http://dx.doi.org/10.3233/KCA-180047

http://dx.doi.org/10.3233/KCA-180047

http://dx.doi.org/10.1186/s12967-019-02144-7

http://dx.doi.org/10.1002/oby.21620



http://dx.doi.org/10.1016/S1074-7613(00)00026-1

http://dx.doi.org/10.1093/bioinformatics/17.suppl_1.S22

http://dx.doi.org/10.1073/pnas.95.25.14863

http://dx.doi.org/10.1038/s41591-018-0057-z

http://dx.doi.org/10.1016/j.immuni.2018.12.021

http://dx.doi.org/10.1073/pnas.1718217115

http://dx.doi.org/10.1172/JCI73639


http://dx.doi.org/10.1126/science.1229620

http://dx.doi.org/10.1016/j.ccell.2018.09.007

http://dx.doi.org/10.1038/nature05115

http://dx.doi.org/10.1038/s41467-018-06653-9

http://dx.doi.org/10.1038/s41467-018-06653-9

http://dx.doi.org/10.1007/s00262-014-1598-8

http://dx.doi.org/10.1038/5517



http://dx.doi.org/10.1016/j.ejphar.2009.06.066

http://dx.doi.org/10.1097/CJI.0b013e318174a4df

http://dx.doi.org/10.4049/jimmunol.1000021

http://dx.doi.org/10.1016/S0952-7915(98)80228-8

http://dx.doi.org/10.1016/j.ccell.2019.05.004

http://dx.doi.org/10.1016/j.immuni.2018.06.006

http://dx.doi.org/10.1158/1078-0432.CCR-15-1823

http://dx.doi.org/10.1158/2326-6066.CIR-18-0578

http://dx.doi.org/10.1073/pnas.1812266115

http://dx.doi.org/10.1002/JLB.4HI0517-210R

http://dx.doi.org/10.1038/s41590-018-0251-7

http://dx.doi.org/10.1016/j.cmet.2019.10.013

http://dx.doi.org/10.3389/fimmu.2020.590794

http://dx.doi.org/10.1038/srep36107

http://dx.doi.org/10.1158/2159-8290.CD-15-1177


1

Supplemental Materials and Methods

T Cell Repertoire Sequencing and Analysis

For TCR library preparation, we used the commercially available iRepertoire platform

(iRepertoire; Huntsville, AL) for nested amplicon arm-PCR (1) of CDR3 regions of the mouse

TCR β-chains, with addition of adapters for Illumina platform sequencing. Reverse transcription

of 1.5ug of RNA was conducted with the One-Step reverse transcription and amplification kit

(Qiagen). The PCR product was purified using Ampure XP magnetic beads (Beckman Coulter),

and secondary amplification of the resulting product was performed using GoTaq G2 PCR Kit

(Promega), allowing addition of Illumina adapter sequences. Libraries were purified with

Ampure XP magnetic beads and sequenced using Illumina MiSeq 250nt paired-end read-length.

TCR CDR3 sequences were extracted from the raw sequencing data via the iRmap pipeline (2).

TCR high-throughput sequencing (HTS) data were analyzed in R environment using the

tcR package v.2.2.2 (3) and common R routines. TCR repertoire diversity was assessed using

True Diversity indices and directly observed richness. Comparative analysis of TCR repertoire

richness was performed after normalization of the HTS depth. Total datasets were downsampled

to 250,000 randomly chosen sequencing reads using bootstrapping with 1000 iterations. For the

comparative analysis of TCR repertoire diversity indices, a downsampling of repertoire sizes to

the size of the smallest repertoire was performed using bootstrapping with 1000 iterations. A

median of each simulated richness and diversity distribution was used as estimated richness or

diversity, accordingly (4). The treatment groups were compared using Kruskal-Wallis ANOVA.

BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any relianceSupplemental material placed on this supplemental material which has been supplied by the author(s) J Immunother Cancer

doi: 10.1136/jitc-2020-000725:e000725. 8 2020;J Immunother Cancer, et al. Boi SK

2

Supplemental References:

1. Han J, et al. Simultaneous amplification and identification of 25 human papillomavirus

types with Templex technology. J Clin Microbiol. 2006;44(11):4157-62.

2. Wang C, et al. High throughput sequencing reveals a complex pattern of dynamic

interrelationships among human T cell subsets. Proc Natl Acad Sci U S A.

2010;107(4):1518-23.

3. Nazarov VI, et al. tcR: an R package for T cell receptor repertoire advanced data analysis.

BMC Bioinformatics. 2015;16:175.

4. Venturi V, Kedzierska K, Turner SJ, Doherty PC, and Davenport MP. Methods for

comparing the diversity of samples of the T cell receptor repertoire. J Immunol Methods.

2007;321(1-2):182-95.



3

Supplemental Data

Supplementary Figure 1. Analysis of metastatic RCC patients with combined overweight or

obesity (BMI ≥25kg/m2) reveals no alteration in PFS or OS, as compared to lean (BMI

<25kg/m2) metastatic RCC patients receiving anti-PD-1. The same patient cohort shown in

Figure 1, analyzed for (A) Progression-free (PFS) and (B) overall survival (OS) of metastatic

RCC patients treated withanti-PD-1, categorized by indicated BMI status. Survival curves for

PFS and OS across BMI categories were generated with the Kaplan-Meier method. Hazard ratios

were calculated from a Cox model controlling for patients’ age, sex, IMDC risk score, and number of prior therapies.

<25

≥25

<25

≥25

<25

≥25

<25

≥25

A B

BMI

Median

Survival

(Month)

6 Month

Survival

1 Year

Survival

2 Year

Survival

Hazard Ratio*

(<30 vs ≥30)

<25 11.4 66.7% 48.6% 32.4% 0.84 (0.39, 1.81)

≥25 8.3 61.7% 43.9% 17.7% P = 0.6622

BMI

Median

Survival

(Month)

6 Month

Survival

1 Year

Survival

2 Year

Survival

Hazard Ratio*

(<30 vs ≥30)

<25 24.3 100% 83.3% 60.8% 0.96 (0.37, 2.54)

≥25 23.0 100% 77.1% 43.8% P = 0.9406

<25

≥25

<25

≥25

<25

≥25



5

Supplementary Figure 3. Tumor burden at treatment initiation, initial body weight, day 28

PD-L1 expression on intratumoral CD45+ and CD45- cells, TCR repertoire and diversity,

and sex of mice do not predict or influence treatment outcomes. (A) Day 28 tumor weights

for DIO animals treated with NT, single agent anti-PD-1, AdTR/CpG, or AdTR/CpG/PD-1.

Regression analysis between day 28 tumor weights and (B) initial body weight at tumor

challenge or (C) initial tumor burden at the time of treatment initiation on day 7. Responders (R)

= green and non-responders (NR) = black. (D) Percentages of PD-L1+ cells at day 28 among

(left) CD45+ and (right) CD45- cells. (E) Day 28 tumor weights of NT Lean, DIO, and OB-Res

mice (repeated data from Fig. 3F). (F) Whole-tumor TCRβ CDR3 sequencing analysis from day 28 excised tumors. High throughput sequencing reads-normalized repertoire (left) richness and

(right) diversity. (G) Day 28 renal tumor weights for lean and DIO male mice, treated as

previously indicated, and resulting AdTR/CpG/PD-1 response rates. (B-E,G) Data are pooled

from at least two independent experiments or (A,F) a single experiment and presented as means

SEM. Statistical differences were determined using non-parametric Mann-Whitney U tests

(###p<0.001), linear regression analyses, or two-way ANOVA. NT = no therapy; NR = non-

responder; Res = responder; Tx = AdTR/CpG/PD-1.

0

2

4

6Tum

or

Weig

ht (g

)

DIO: Day 28 Tumor Weights

-AdTR/CpG

aPD-1

- +

- + -

- 3% - 4% - 42%

ns

+

+

NT NR Res NT NR Res0

5

10

15

% o

f C

D45

-PD-L1+

Lean DIO

ns

NT NR Res NT NR Res0

20

40

60

% o

f C

D45

+

PD-L1+

Lean DIO

ns

LeanNT

DIONT

OB-ResNT

0

2

4

6

Tum

or W

eig

ht (g

)

Day 28

ns

NT Tx NT Tx0

1

2

3

4

5

Tum

or

Weig

ht (g

)

Males: Day 28 Tumor Weights

# # #

Lean DIO

Lean

n = 9

78%

22%

DIO

n = 6

17%

83%

Non-Responder

Responder

105 106 107 108

0

2

4

6

Initial Tumor Burden (Radiance)

Tum

or W

eig

ht (g

)

r2 = 0.0227p = 0.6402

28 30 32 34 36 38 40

0

2

4

6

Body Weight (g)Tum

or

Weig

ht (g

)

r2 = 0.1344p = 0.2411

NRR

B C

D

F

A

G

LeanNT

DIONT

0.0

0.5

1.0

1.5

2.0

2.5

Num

ber of T

CR

b C

lonoty

pes

(x10

3)

TCR Repertoire Richness

ns

LeanNT

DIONT

0

5

10

15

20

Tru

e D

ivers

ity Index

TCR Repertoire Diversity

ns

E



7

Supplementary Figure 5. The immunogenetic landscape of intratumoral and tumor-free

kidneys is minimally altered by obesity in the absence of immunotherapy. Tumor samples

and contralateral tumor-free kidneys from lean and DIO treatment-naive mice were collected at

day 28 post-tumor challenge and analyzed using nanoString’s PanCancer Immune Profiling platform. (A) Analysis of differentially expressed (DE) genes within tumors indicates equivalent

gene expression in 97% of genes examined. (B) Analysis of DE genes in tumor-free kidneys

indicates equivalent gene expression in 94% of genes examined. Data are from one experiment.

Tx = AdTR/CpG/PD-1.

A B

28 Genes DE(6%)

n = 468

LeanNo

Tx

DIONo

Tx

440 Genes

(94%)

Tumor-free Kidney

17 Genes DE(3%)

n = 558

LeanNo

Tx

DIONo

Tx

541 Genes

(97%)

Tumor



12

Supplementary Figure 10. CCL5 neutralization impairs immunotherapy response in lean

mice. On day 15 post tumor challenge lean and DIO treatment-naive mice were evaluated for (A)

intratumoral T cell-associated cyto/chemokines. (B) Linear regression of intratumoral CCL5

concentrations versus (left) tumor weight and (right) associated leukocyte populations in lean

and DIO therapy-naive and therapy-treated animals. (C) Day 28 response rates for lean animals

receiving AdTR/CpG/PD-1 anti-CCL5 neutralizing antibody. (D) Day 28 comparison of lean

mice receiving no therapy or AdTR/CpG/PD-1 isotype control. Isotype administration had no

effect on therapeutic efficacy. Data are pooled from at least two independent experiments and

presented as means SEM. Statistical differences were calculated using parametric t-tests, non-

parametric Mann-Whitney U tests, linear regression analyses, or Kruskal-Wallis ANOVA as

appropriate. NT = no therapy and Tx = AdTR/CpG/PD-1 therapy.

Lean: Tx+αCCL5

n = 13

Responder

Non-Responder

46% 54%

Lean: Tx

n = 14

86%

14%C

CL

4

CC

L5

CX

CL

10

CX

CL

11

0

200

400

600

Conce

ntr

atio

n (pg/m

L)

Day 15 T Cell-AssociatedCyto/Chemokines

LeanDIO

p=0.13

A CB

0.0 0.5 1.0 1.5 2.0 2.5

0

200

400

600

800

CD8+CD44+ T Cells (% of Total Cells)

CC

L5 (

pg/m

L)

r2 = 0.5216p < 0.0001

0.0 0.5 1.0 1.5

0

200

400

600

800

Tumor Weight (g)

CC

L5 (

pg/m

L)

r2 = 0.1358p = 0.0193

Lean NT DIO NTLean Tx DIO Tx

NT Tx Tx + isotype

0

1

2

3

4

5

Tu

mo

r W

eig

ht (g

)

##

D



13

Supplemental Table 1. Demographics and characteristics of anti-PD-1 treated RCC patients. 1

BMI at 1st dose < 30 BMI at 1st dose 30

Variable N Mean SD Median Min Max Frequency (%)

N Mean SD Median Min Max Frequency (%)

p value*

Age 43 66.4 8.9 66.5 43 86

29 66.4 9.6 65.0 47.0 81.0

0.8277

Male 43

31 (72.1%) 29

22 (75.9%) 0.7219

Race 43

29

1.0000

White

38 (88.4%)

26 (89.7%)

AA

4 (9.3%)

2 (6.9%)

Other

1 (2.3%)

1 (3.5%)

Ethnic group 43

29

1.0000

Non-Hispanic/Latino

42 (97.7%)

28 (96.5%)

Unknown

1 (2.3%)

1 (3.5%)

Smoker 27

16

1.0000

Yes 14 (51.9%) 9 (56.3%)

No 13 (48.1%) 7 (43.7%)

BMI at time of 1st dose 43 26.5 2.7 27.1 20.7 29.4 29 36.2 5.1 34.3 30.3 48.8 <0.0001

Latest BMI 43 25.7 3.9 26.5 18.2 32.4

29 35.6 6.1 34.7 25.7 50.7

<0.0001

Diff of BMI (1st - latest) 43 0.9 2.7 0.2 -4.3 6.7

29 0.6 4.4 0.6 -11.9 7.4

0.7924

Follow up time (days) 43 638.8 316.5 585 187 1212

29 506.3 273.4 422.0 211.0 1269.0

0.0958

Number of doses of nivolumab 43 17.9 15.0 11.0 3.0 60

28 15.2 15.2 8.5 3.0 72.0

0.2635

Treatment duration (days) 41 302.5 287.5 196 28 1139

27 242.1 259.6 148.0 33.0 1269.0

0.4073

Histological subtype 43 29

Clear cell 37 (86.0%) 24 (82.8%) 0.7015

Papillary 4 (9.3%) 2 (6.9%)

Unclassified 2 (4.7%) 3 (10.3%)

IMDC risk score 42 29 0.7266

0 13 (31.0%) 8 (27.6%)

1 16 (38.1%) 11 (37.9%)

2 12 (28.6%) 7 (24.1%)

3 1 (2.4%) 2 (6.9%)

4 0 1 (3.5%)

Number of prior treatments 43 29 0.2785

0 3 (7.0%) 1 (3.5%)

1 24 (55.8%) 20 (69.0%)

2 12 (27.9%) 5 (17.2%)

3 4 (9.3%) 1 (3.5%)

4 0 2 (6.9%)

Autoimmune Hypothyroidism 41 28 0.6415

Yes

3 (7.3%)

1 (3.6%)

No

39 (92.9%)

27 (96.4%)

Deceased 43

18 (41.9%) 29

18 (62.1%) 0.0926

Best response 43

29

0.7613

SD

20 (46.5%)

15 (51.7%)

PD

13 (30.2%)

9 (31.0%)

PR

5 (11.6%)

1 (3.5%)

CR

1 (2.3%)

1 (3.5%)



14

2

OR present, not measured

1 (2.3%)

2 (6.9%)

Stopped for toxicity, death, or unknown

3 (7.0%)

1 (3.5%)

Objective response rate (ORR) 43

7 (16.3%) 29

4 (13.8%) 1.0000

Disease control rate (DCR) 43

27 (62.8%) 29

19 (65.5%) 0.8132

* Wilcoxon test for continuous variables and Fisher's exact test for categorical variables



15

Supplemental Table 2. Demographics and characteristics of treatment-naive renal 3

cancer subjects used for TIL analysis. 4

5 Renal Cancer Subjects

NOB OB

Subjects (n) 8 11

Sex M 75.00% F 25.00%

M 45.50% F 54.50%

Age (years, mean) 63.38 58.27

Age (years, range) 42 – 77 46 – 71

BMI (kg/m2, mean) 25.02 42.24

BMI (kg/m2, range) 19.80 – 28.49 30.66 – 62.78

Histologic Subtype

Clear Cell 100.00% 73.00%

Other 0.00% 27.00%

Stage (%)

T1a 12.50% 36.36%

T1b 25.00% 36.36%

T2a 12.50% 9.09%

T2b 12.50% 0.00%

T3a 37.50% 9.09%

T3b 0.00% 0.00%

N/A 0.00% 9.09%

Grade

1 0.00% 0.00%

2 50.00% 54.55%

3 25.00% 27.27%

4 25.00% 0.00%

NA 0.00% 18.18%



16

Supplemental Table 3. Demographics and characteristics of treatment-naive renal 6

cancer subjects and tumor-free donors used for peripheral blood analysis. 7

8

Renal Cancer Subjects Tumor-Free Subjects

NOB OB NOB OB

Subjects (n) 37 45 11 8

Sex M 70.27% F 29.73%

M 57.78% F 42.22%

M 27.27% F 72.73%

M 50.00% F 50.00%

Age (years, mean) 59.00 57.44 60.64 54.63

Age (years, range) 28 – 83 41 – 81 35 – 76 48 – 68

Weight (kg, mean) 76.34 118.42 72.67 100.05

BMI (kg/m2, mean) 25.07 38.74 25.16 32.58

BMI (kg/m2, range) 18.24 – 29.98 30.00 – 60.73 21.52 – 29.45 30.10 – 38.60

Diabetes (%) 13.51% 37.78% Unavailable Unavailable

Histologic Subtype

Clear Cell 84.00% 89.00%

Other 16.00% 11.00%

Stage (%)

T1a 54.05% 35.56%

T1b 13.51% 35.56%

T2a 2.70% 4.44%

T2b 2.70% 2.22%

T3a 24.32% 20.00%

T3b 2.70% 0.00%

N/A 0.00% 2.22%

Grade

1 0.00% 4.44%

2 32.43% 55.56%

3 45.95% 20%

4 16.22% 11.11%

NA 5.41% 8.89%



17

Supplemental Table 4. Differentially expressed (DE) genes shared between Lean 9

R vs NR and DIO R vs NR which overlap with yellow, blue, and green clusters 10

from Figure 5A. 11

12

Yellow Cluster Overlap (79 genes) 13

Abcb1a Cd200 Cd8a Cxcl12 Gpi1 Il17f Ltb Rps6 Adora2a Cd34 Cd8b1 Cxcl14 Gzma Il6st Map3k5 Sigirr Angpt1 Cd36 Cdh1 Cxcl16 H2-Aa Irf1 Mavs Syt17 App Cd3e Ceacam1 Cxcr3 H2-Ab1 Irf3 Nrp1 Tek C8a Cd3g Cfb Cxcr6 H2-DMb2 Irf8 Nt5e Thbd C8g Cd55 Chuk Dock9 Icam1 Itga6 Pml Thy1 Ccl19 Cd59b Cmpk2 Dpp4 Icam2 Jam3 Prf1 Tie1 Ccl21a Cd6 Ctsh Dusp6 Ifit3 Kit Psen1 Tnfsf10 Ccl27a Cd74 Cx3cl1 Ecsit Il15 Klrc1 Psen2 Xbp1 Ccl5 Cd79b Cxcl11 F2rl1 Il16 Klrd1 Rorc

14

Blue Cluster Overlap (96 genes) 15

Amica1 Ccl12 Cd84 Cxcl2 Il17rb Itga5 Nup107 Tgfb1 Atf1 Ccl2 Cd97 Cxcr2 Il18 Itgam Oas2 Thbs1 Bcl10 Ccl24 Cdk1 Dusp4 Il18r1 Itgb1 Osm Tlr4 Bcl2 Ccl3 Cfp F13a1 Il18rap Jak3 Ptgs2 Tlr5 Bcl6 Ccl7 Clec4a2 Fcer1g Il1b Lcn2 Pvr Tnfrsf10b Btk Ccr1 Clec4n Fcgr1 Il1r1 Litaf Rel Tnfrsf1b C3ar1 Cd163 Clec5a Fos Il1r2 Ly96 Runx1 Tnfsf18 C4b Cd200r1 Creb5 Gpr183 Il1rl1 Lyve1 Serpinb2 Trem1 C5ar1 Cd244 Csf2rb H60a Il1rn Lyz2 Siglec1 Trem2 C6 Cd33 Csf3r Hif1a Il4ra Msr1 Slc7a11 Ulbp1 Card9 Cd37 Ctsl Ifitm2 Il6 Mx2 Socs3 Usp18 Casp1 Cd68 Cxcl1 Il13ra2 Il6ra Nlrp3 Tank Vhl

16

Green Cluster Overlap (42 genes) 17

Anxa1 Ccl8 Chil3 Ifi35 Lgals3 Oasl1 Spn Arg1 Ccl9 Csf2 Ifngr1 Masp1 Plaur Tal1 Birc5 Cd40 Cxcl3 Il12rb1 Mst1r S100a8 Tnfrsf12a Bst1 Cd47 H2-M3 Irak4 Myc Serping1 Tnfrsf13c C1s1 Cdkn1a Hspb2 Isg15 Ncf4 Smn1 Tnfrsf9 Ccl6 Cebpb Ifi27 Lbp Oas3 Snai1 Vim

18



18

Supplemental Table 5. List of antibodies and dyes used in this study. 19

20

In Vivo Antibodies Antigen Clone Fluorochrome Catalog number Supplier

PD-1 (CD279) RMP1-14 none BE0146 BioXCell

Rat IgG2a Isotype Control 2A3 none BE0089 BioXCell

IL-1B B122 none BE0246 BioXCell

CCL5 53405 none MAB478 R&D Systems

Dyes for Flow Cytometry Dye Clone Fluorochrome Catalog number Supplier CFSE N/A N/A C34570 ThermoFisher

Zombie Aqua Viability Dye N/A N/A 423102 BioLegend

Zombie Green Viability Dye N/A N/A 423111 BioLegend

Murine Antibodies for Flow Cytometry Antigen Clone Fluorochrome Catalog number Supplier

CD3 145-2C11 APC, FITC 100312, 100305 BioLegend

CD4 GK1.5 APC/Fire750 100460 BioLegend

CD8 53-6.7 PE/Cy7, A488, APC 100722, 100723,

100712 BioLegend

CD11b M1/70 BV510, PE/Cy7 101263, 101216 BioLegend

CD11c N418 BV510, APC/Cy7 117338, 117324 BioLegend

CD19 6D5 FITC 115505 BioLegend

CD25 PC61 BV421 102033 BioLegend

CD26 H194-112 PE 137803 BioLegend

CD44 IM7 PerCP-Cy5.5 103032 BioLegend

CD45 30-F11 Pacific Blue, PerCP-Cy5.5 103126, 103132 BioLegend

CD45R (B220) RA3-6B2 FITC 103205 BioLegend

CD64 X54-5/7.1 FITC 139315 BioLegend

CD90.1 (Thy1.1) HIS51 APC 170900-82 eBioscience

CD161 PK136 FITC 108705 BioLegend

CD172a (Sirp) P84 PE-Dazzle 594 144016 BioLegend

CD274 (PD-L1) 10F.9G2 BV605 124321 BioLegend

CD279 (PD-1) 29F.1A12 BV421, APC/Fire750 135218, 135240 BioLegend

F4/80 BM8 FITC 123107 BioLegend

Foxp3 150D A647 320014 BioLegend

I-A/I-E (MHC II) M5/114.15.2 BV650 107641 BioLegend

I-Ad (MHC II) 39-10-8 Alexa Fluor 488 115008 BioLegend

IFN XMG1.2 BV421 505830 BioLegend

Ly6C HK1.4 BV605 128036 BioLegend

Ly6G 1A8 PerCP-Cy5.5 127615 BioLegend

Perforin S16009A PE 154306 BioLegend

TNFa MP6-XT22 PE/Cy7 506324 BioLegend

XCR1 ZET BV421 148216 BioLegend

Human Antibodies for Flow Cytometry Antigen Clone Fluorochrome Catalog number Supplier

CD3 HIT3a PE 300308 BioLegend

CD8 HIT8a PerCP-Cy5.5 300924 BioLegend CD45RA HI100 BV421 304130 BioLegend

CD45RO UCHL1 FITC 35-0457-T100 TONBO Biosci.

CD279 (PD-1) EH12.2H7 PE/Cy7 329918 BioLegend

HLA-DR L243 APC/Cy7 307618 BioLegend



obesity diminishes response to pd-1-based immunotherapies

Documents