Prime time for immune-checkpoint targeted therapy at ASCO 2015

ABSTRACT

Cancer immunotherapy has been one of the dominant topics in oral presentations and abstracts during the 2015 annual meeting of the American Society of Clinical Oncology (ASCO). The renewed interest in immunotherapy is explained by the wide spectrum of activity, the durability of tumor responses and the rapid clinical development of immune-checkpoint targeted monoclonal antibodies. These new drugs are currently revolutionizing the field of oncology. Here we highlight what were to us the most important results announced during the annual meeting of ASCO held in Chicago, IL from May, 29th to June, 2nd 2015. In addition, we searched all the posters/published abstracts pertinent to the field of immunooncology from this year conference. Among more than 400 published abstracts on this topic, we have grouped and briefly summarized the most relevant ones.

Introduction

Cancer immunotherapy has been one of the dominant topics in oral presentations and abstracts during the 2015 annual meeting of the American Society of Clinical Oncology (ASCO). The renewed interest in immunotherapy is explained by the very rapid clinical development of immune-checkpoint targeted monoclonal antibodies (imAbs) such as the anti-CTLA-4 imAb ipilimumab (BMS, FDA approved in 2011) and the anti-PD-1 imAbs nivolumab and pembrolizumab (BMS and Merck MSD respectively, FDA approved in 2014). These new drugs are currently revolutionizing the field of oncology for three reasons. First, imAbs bring a disruptive innovation in cancer therapy where drugs are designed to target immune cells as opposed to cancer cells, with the aim of helping patients to fight their own cancer with their immune system. Second, the same imAbs show a broad spectrum of activity with objective tumor responses (ORR) across many cancer types, despite their many histology and site differences. Third, they provide benefits in overall survival (OS) in a significant proportion of patients whereas most of the progress obtained over the last decade with tumor-targeted therapies provided benefits in progression-free survival (PFS) in very genomically selected subsets of patients. Here we will highlight what were to us the most important imAbs results announced during the annual meeting of ASCO held in Chicago, IL from May, 29th to June, 2nd 2015.

In addition, we searched all the posters/published abstracts pertinent to the field of immuno-oncology and imABs from this year conference. Among more than 400 published abstracts on these topics, we have grouped and briefly summarized the most relevant in three different categories: (1) Results from ongoing clinical trials using imABs. (2) Preclinical analysis of novel immune therapies. (3) Determination of new biomarkers for patients treated with imABs. (Tables 1–3)

Table 1.

Ongoing clinical trials involving imAbs presented at ASCO 2015.

Clinical trial	Number of patients	Conclusions	Abstract # and authors
Melanoma
Vemurafenib+Ipi	10 MM	*2 Gr 3 elevation transaminase *5 PR, 2 SD, 3 PD	e20075 (Hassel JC)
Phase II Ipi+ILP* vs. Ipi alone	In transit stage IIIB-IIIC mel n = 100	Wieberdink Gr1–2 limb toxicity 4 patients enrolled	e20107 Cavalcanti A
Single arm Phase II Ipi 10mg/kg	Stage III-IV MM with preexisiting anti-NY-ESO-1 Ab	22 evaluable pts 23% irPR, 27%irSD, 32% irPD, 18% stopped (irAEs)	e20061 Haag GM
Multicenter Turkish study Ipi 3 mg/kg x 4 doses	MIPI-TURK Survival rate of 97 turkish MM	19.2% survival at 30 months (no CR) median OS : 6 months 3 Gr 3 irAE, 4 Gr4 irAE 2 deaths (hepatic toxicity)	e20072 Sevinc A
Ipilimumab+immunoembolisation (infused cytokines in liver mets)	18 metastatic uveal melanoma	13/15 SD of liver mets 8/13 secondary failure	e20015 Orloff MM
Ipi+radiotherapy vs. Ipi alone	Retrospective analysis of MM N = 22 Ipi+RT N = 26 Ipi	29% CR in Ipi+RT versus 8% Ipi alone	e20048
Ipi 3mg/kg ×4 doses	N = 248 MM, South Africa	In 119 pts, median OS = 8.8 months, 3 y survival rate : 12%	e20112
Pembro10mg/kg/2 weeks (10/17 previously Ipi treated)	17 MM with brain mets	12 evaluated: 3 PR, 2 SD, 7 PD (2 mixed, 1 pseudoprogression) BrM response duration : 7+, 6+, 3+ months	e9009
BRAFi+/− ICB (sequential) retrospective	26 Pts MM BRAFV600-mutant : 23 pts : BRAFi prior to ICB and 9 pts : ICB prior to BRAFi	RR with the sequence BRAFi+ICB> RR ICB+BRAFi but survival at 24mo idential (50%)	e20030 Aya F
Phase II BRAF-inhibitor followed by Ipi 10mg/kg Q3W*4 to prevent hepatotoxicity in MM	46	Unlike the combination of both medications, sequentially VEM + Ipi had no evidence of hepatotoxicity and satisfactory response (ORR: 30%)	9032. A. Amin
Phase II: P 10mg/kg Q2W in untreated brain metastasis from MM or lung cancer	64	- Interim analysis. PR 3/12, SD 2/12. Anticonvulsant and short course of prednisone were required for couple patients	9009. H.M. Kluger
Phase 1b/2: T-VEC + P vs. P in MM	110	- After encouraging data from T-VEC monotherapy in MM, this study is currently recruiting participants	TPS9081. A. Ribas
Retrospective study I in MM in a population excluded from clinical trials	12	Incidence of IrAE in patients with known autoimmune disorders on I. 50% experienced a flare of their underlying condition during treatment.	9019: D.B. Johnson
Talimogene laherparepvec (T-VEC) enhanced OS in MEL. Combination with I may lead to a synergetic effect in MM	19 pts MM: T-VEC Il <4 mL 106 PFU/mL week 1, then 108PFU/mL week 4 then q2weeks + starting at 6th week I 3mg/kg Q3W *4	-2/3 patients were alive at 18-month -PFS:10.6 -IrAE grade>3: 32%	9063.I. Puzanov:
Double T cells unleashing with Peg-IFN + PD-1 in MM	12pts: P 2mg/kg Q3W + Peg-IFN SC (1,2 &3mg/kg) at escalating dose	- No evidence of early toxicity in dose escalation of Peg-IFN + P	e20018. H. M. Zarour
MEL brain metastasis with Ipi+ brain radiosurgery (SRS) and/or whole brain treatment (WBRT) to assess radionecrosis	34 pts: I + SRS vs. I + WBRT vs. I + SRS + WBRT	- The incidence of radionecrosis was higher in SBS compare to WBRT (41% vs. 0%). One patient required surgery despite prednisone to treat the RN.	e20019: L. Khoja
lpi trial of SCIB1, a DNA vaccine targeting DC in melanoma pts	n = 16 resected stage III or IV melanoma i.m. electroporation	Epitope specific proliferation, γIFN Elispot response, PFS 78% and 72% for stage III and IV, OS 100%, little associated toxicity	9035 Poulam M
PROCLAIM national registry, prospective cohort: HD IL-2 followed by Ipi and/or anti-PD-1/PD-L1 therapy	n = 240 MM	1 y survival rate of 100%, 68% and 58%, mOS of 25.1 (n = 20), 18.4 (n = 75) and 14 mo (n = 112) for PD1α, ipi after HD IL-2 and IL-2 only group. High grade irAE for ipi (n = 2) and PD1α (n = 2).	e20071 Wong MKK
CALM: open-label, multicentre phase II of Coxsackievirus A21 oncolytic virus in pts with advanced melanoma	N = 57, i.t. , treated or untreated unresectable stage IIIc-IV M1c melanoma, primary endpoint irPFS at 6mos	irPFS at 6mos: 38,6% (21/57), median irPFS of 4.2 mos, irRECIST 28,1% (6/57), irRECIST ≥6 mo 19,3% (11/57), median time to response 2.8mos, 1-year survival rate 75.4% (43/57), median OS not reached (follow-up 16.5 mos), no grade 3 or 4 AE	9030 Andtbacka RHI
CheckMate 066: randomized, double-blind phase III of Nivo vs. DTIC in treatment-naïve melanoma pts, QoL analysis	n = 418 QoL questionnaire at baseline and Q6W: EORTC QLD-C30 EQ-5D	EORTC QLQ-C30 70.0% with NIVO, 64.9 with DTIC EQ-5D 69.5% with NIVO, 64.9% with DTIC Improvements from baseline for NIVO from week 7 (p = 0.011) through week 49 (p = 0.034), No change for DTIC from baseline,	9027 Long GV
KEYNOTE-029: phase I/II study of plus low-dose ipi or PEG-IFN in pts with MM or RCC	Pembro plus Ipi: n = 18	DLT 6/19 of grade 3 (1 colitis, 1 uveitis, 2 ALT/AST elevation, 1 lipase plus pneumonitis, 1 elevation pancreatic enzymes plus hyperthyroidism)	3009 Atkins MB
OPTiM, a randomized phase III trial of talimogene laherparepvec vs. GM-CSF in unresected stage IIIB/C and IV melanoma, post-hoc review of resectability criteria	T-VEC an oncolytic immunotherapy derived from herpes simplex virus type-1, n = 50	50 Pts: 39 unresectable, 7 deemed resectable, 1 disputed, 3 unevaluable	e20109 Faries MB
Ir-AE and association with survival outcomes for nivolumab in resected and unresectable melanoma	n = 148 (33 resected, 115 unresectable), nivolumab 1-, 3-, or 10mg/kg Q2W for 12 weeks, up to 2 additional years,	rash and vitiligo are associated with improved PFS (HR 0.237, 95%CI 0.098 to 0.573 p < 0.001; HR 0.121, 95%CI 0.024 to 0.620 p < 0.005) and OS (HR 0.098, 95%CI 0.006 to 1.711 p < 0.005; HR 0.146, 95%CI 0.009 to 2.488 p < 0.081). No difference were seen with other irAEs	9028 Freeman-Keller ML
SWOG S1404: phase III comparing high dose IFN to pembro in pts with high risk resected melanoma	Stratification, stage, PD-L1 status (+/−/unknown), n = 1,378, Randomization 1:1 IFNα-2b 20 MU/m2/days 1–5 weeks 1–4, followed by IFNα-2b 10 MU/m2/d SC days 1,3,5 weeks 5 -52, or and pembro 200 mg Q3W for 52 weeks.	Ongoing: OS, PFS, with interim analyses planned for both endpoints at 50% OS information. Safety, pharmacokinetics, and QoL will be studied	TPS9085 Grossmann KF
Lung cancer
Pembro in PDL1+ BrM NSCLC (untreated)	N = 10	44% OR in BrM (4/9 PR), systemic RR was 34%	8035 Goldberg SB
Randomized 2 :1 Phase II in stage IV non-squamous NSCLC : carboplatin :paclitaxel, bevacizumab +/−Imprime PGG (beta 1,3/1,6 glucan : a yeast PAMP)	in 92 pts at 2 :1	Combo with PGG mediated rapid, durable RR (60% vs. 43%, 12 weeks vs. 18 weeks, 10.3 mo vs. 5.6 mo), regardless of tumor location and burden, maintenance of regression after tt stop Safety fine	3070 Biothera Inc. Engel-Riedel W, Ann Oncol 2014, 25 (5)
Phase IIb/IV : Nivo in refractory NSCLC	N = 226, 74% NSQ, 50% >3 prior tt	179/226 remain on study at 4 mo, 1.8% Gr3–4 Out of 51 evaluable (week 9), 14% PR, O CR, 43%SD	3013 Bauer TM
Randomized Phase IIb : TG4010 Pox-MUC1-IL-2 in combination with chemotherapy	N = 222 stage IV NSCLC	PFS improved in low CD56+CD16+CD69+ (TrPAL), in non-squamous NSCLC, OS following these trends Risk reduction HR : 0.61 in low TrPAL/NSQ Vaccine still significant in PD-L1 negative tumors	3034 Quoix E Transgene
Phase I/II Nivo +/− Ipi in SCLC	75 pts : Nivo 3mg/kg Q2W vs. Nivo (dose escalation)+ Ipi Q3W *4	- ORR was 1.6 superior with combination with acceptable safety profile (Only 1 pneumonitis in each arm)	7503. S.C. Antonia
Renal cell Carcinoma
CheckMate 016 (Nivo+Ipi in mRCC)	Nivo = 3mg /kg + Ipi 1mg/kg Q3W *3 vs. Nivo1+Ipi3 vs. + Nivo3+13*4 then Nivo	- Nivo3+Ipi3 was too toxic - 41 pts in 2 other arms had similar ORR but 2 fold increase IrAE with Nivo1+Ipi3	4516. H.J. Hammers
Phase 1 Keynote -029: Combination lower Ipi dose + std Pembro in mRCC and MM	Pembro 2mg/kg + Ipi1mg/kg Q3W followed by Pembro	-19 pts enrolled in the safety run period: 32% of grade 3 IrAE were observed (54% in Postow et al. Nejm 2015;372:2006–17)	3009. M. B. Atkins
Pancreatic cancer
IMAGE1 Randomized trial in advanced pancreatic cancer Mycobacterium obuense (IMM-101)and gemcitabine	N = 92 N = 28 Gem alone N = 68 Gem + IMM-101	Increase in median OS from 4.4 (Gem) vs. 7 mo in Gem+IMM-101 P = 0 .01, HR : 0.64 Survival rate at 12 months : 11.5 % (Gem) vs. 22.4% (Gem+IM-101)	3051 Dalgleish A
CAR T mesothelin	10 chemorefractory pancreatic cancers	53/54 planned CAR T cell infusions 2 Gr3 toxicities 2/6 SD change SUV mav : -2% within 1 month PET/CT in one pt with liver mets negative after 1 month	3007 Beatty GL
Phase 1 b I + Gem in advanced pancreatic cancer	13 pts : Gem 750–1000mg/m2 Q1W *7 then q Q1W *3 + Ipi (3–10mgkg) on weeks 1,4,7 &10	-Diarrhea and hepatotoxicity DLT lead to MTD I 3 + Gemcitabine 1000 - Encouraging outcome	e15281. N.A. Mohindra
Prostate
PROSTVAC vaccine in advanced HRPC (hormonoresistant prostate cancers) : 2 Phase II and 1 Phase I with Ipilimumab	i) Randomized phase II with 125 men ii)Phase I with 32 HRPC iii)Phase I with 30 HRPC	i) OS Prostvac : 25.1 vs. 16.6 mo, HR : 0.56ii)26.6 vs. 17.4 (Halabi nomogram)iii)37.2mo (Prostvac+Ipi 10mg/kg) vs. 18.5	e14008 Gulley JL
Lymphoma
Phase I anti-4–1BB and rituximab in CD20+NHL	N = 35 pts, 22 FL, 5 mantle cell L, 3 diffuse LBCHL	Increased soluble 4–1 BB, Memory T cells, activated NK, PK PF-05082566 half life : 10 d ORR refractory pts : 29% with 2 CR (FL)	3004 Kohrt HE
Glioblastoma
Randomized trial Nivo vs. Nivo+ipi at first recurrence of GBM : CHECKMATE-143	Report on 20 pts with >6 mo follow up	80% Gr3–4 tox in combo leading to 50% discontinuation OS at 6 mo : 70% with Nico and 80% with combo	3010 Sampson JH
Phase I trial : HER2-CMVpp65 specific CART T cells in GBM	15 pts seropositive for CMV	5/15 with objectives responses : 1 PR, 62% reduction, >8 mo, 4 SD 4->24 mo	3008 Ahmed N
Gastro-esophageal cancer
Gene therapy Adp53 (endoscopic it route) and chemotherapy (CT) in esophageal cancer	60 patients randomized in 2 arms GT+CT vs. CT	Curative effect of the combo with 80%+/−16% vs. 68+/−17% shrinking at 12 weeks	e15097 Cui H
Cell immunotherapy (cytokine-induced autologous killer cells) and chemotherapy in gastric cancer	83 eligible but 58 analyzed : 30 in CIK+CT vs. 28 in CT	Median PFS significantly longer in CIK+CT (p = 0 .021) 2yyear PFS in stage III GC or N+ GC were 62%vs. 26% and 50% vs. 27% respectively.	e15106 Cui J
Cervical cancer
Phase I/II I in second-line therapy of women with cervical cancer treated with I 10mg/kg Q3W *4	42	Encouraging signal of immunotherapy activity with a PFS of 2.5 months.	3061. S. Lheureux
Advanced cancers
TLR5 agonist Entolimod Phase I trial	26 pts all comers advanced cancers	3 SD for >12 weeks neutralizing Ab the second week for all pts SNP TLR5 dictate transaminitis	3063 Bakhribah H,
Maintenance immunotherapy with IL-2 and retinoic acid in advanced cancer	500 pts, 4400 courses of chemo, 112 months follow up	No Gr3–4 toxicity, mild hypothyroidism in 5%, sustaine ddecrease of VEGF, increase of NK cells and unpredicted 5 year-survival rates	E14005 Recchia F
Phase I trial using CEA-targeted IL-2 variant with no CD25 binding property	11 advanced CEA+ cancers	89Zr-CEA-IL-2v at doses of 30 mg traffick to tumors NK and T cell accumulation in blood and tumors	3016 Schellens Roche Glycart
Phase I RG7155 anti-CSF1R Ab in advanced solid cancers and pigmented villonodular synovitis (PVNS)	N = 44 (29 PVNS)	PR and SD by PET-CT in 5/44 and 6/40 pts, 24/28 PVNS had OR, 18% Gr3–4 tox (fatigue, pyrexia), reduction of tumor TAM, CSFR1+ cells in tumors, CD14+CD16++ circulating monocytes	3005 Gomez-Roca CA, Roche Pharma

AE: adverse events, BrM: Brain metastasis, CAR: Chimeric antigen receptor, CMV: Cytomegalovirus, CR:Complete response, DTIC: Dacarbazine, FL: Follicular lymphoma, GBM: Glioblastoma, Gem: Gemcitabine, Gr: grade, GT: gene therapy, HR: Hazard ratio, ILP : isolated limb perfusion, i.m., intramuscular, Ir: Immune related, Ipi: Ipilimumab, i.t.:intratumoral, Mel: Melanoma, MM: Metastatic melanoma, mo:month, mRCC: metastactic renal cell carcinoma, Nivo:Nivolumab, NHL: non-Hodgkin lymphoma, NK: Natural killer, NSCLC: Non-small cell lung cancer, ORR: Objective response rate, OS: Overall survival, PD: Progressive Disease, PEG-IFN, pegylated interferon alfa-2b, Pembro: Pembrolizumab, PFS: Progression free survival, PK: Pharmacokinetics, PR: Partial Response QoL: Quality of life, RCC:renal cell carcinoma, RR: Response rate, RT: Radiotherapy, SCLC: Small cell lung cancer, SD: Stable Disease, TLR: Toll-like receptor,T-VEC: Talimogene laherparepvec

Table 2.

Preclinical studies including immunotherapy or combining imABs and conventional therapeutic modalities presented at ASCO 2015.

Combination regimen or IT	Preclinical model system	Conclusions	Abstract # and authors
Phosphatidylserine targeting Ab (bavituximab) combined with PD1 blockade	Melanoma and breast murine tumors and 3D human tumor structures	Accumulation of Th1, Lag3+CD8, TIls expressing Tc1 and GzB, PD1, less MDSC and Treg, synergistic antitumor effects PD1 as a biomarker of response to bavituximab	3059 and 3060 Huang X, Altiok S, Peregrine Pharma.
UCART19 allogeneic ATC	TCRa/CD52 gene deficient RQR8+ and CD19-CAR+T cells : no alloreactivity in a xeno-GvHD mouse model	Donor allogeneic T cells resistant to the lymphodepelting agent alemtuzumab and as a safety measure sensitive to rituximab. Ready for phase I in refractory CD19+ B cell leukemia	3069 Smith J, Cellectis
Myd88/CD40/iCasp9 CAR T cells	CD19 or Her2 targeted CARs incorporating costimulatory domains (CD28, 4–1BB, Myd88/CD40)	Ip injection of rimiducid induce CAR T apoptosis Higher proliferation, IL-2 release, survival and increased activity against Raji /Daudi and SK-BR3	3064 Bellicum Pharma. Foster A
Anti-PDL-1 avelumab (MSB0010718C)	ADCC mediating anti-PDl-1 Ab	Induction of lysis of many human tumor cell lines that correlated with MFI PDL-1 (increased with IFNγ or IL-12)	3038, 3044, TPS3101 Tsang KY Kelly K Heery CR
Anti-TNFR2 antagonist Ab	anti-Treg monoclonal antibody	Turns off NF-kB in Treg, proliferation and promotes cell death of Treg in vitro	E14020 Faustman DL
Bispecific antibody targeting CD47/CD19	Enhancing phagocytosis of CLL and ALL B malignancies	In vitro activity against rituximab reisstant leukemic marrow Safety profile in non-human primates	e14016 NovImmune SA
Blocking the Wnt5a-b catenin signaling pathway synergize with aCTLA4 Ab in BRAFV600E xPTEN KO mice	Reduction of Treg Enhancing DC functions Reinstating effector T cell responses	Both in mouse and human systems	3054 Hanks BA
IDO inhibitor INCB024360	DC immunogenicity (maturation, antigen presentation) and antitumor immunity (activation of MUC1-C specific CTL, Tregs), generated from human PBMC, maturation IFNΥ (50ng/ml) and LPS (1µg/ml) (48h)	No change in CD80, CD83, PD-1, PD-L1, or B7-DC (PD-L2) expression levels on IFNγ/LPS matured human DCs, higher MUC1-C peptide specific CTL, higher TNFα, GM-CSF and IL-8 production, higher levels of tumor cell lysis, decreased Treg proliferation	e14012 Jochems C
Scientific rational for HD-IL-2 + Anti-CTLA4 on peripheral T cells of pts with MM based on 6 pts on HD-IL2	- Upregulation of CTLA4 on Treg	- Flow cytometry analysis of PBMCs pointed out to an elevation of Treg at day 4 and more decisive increase of CTLA4 on Treg from baseline	9043. M.L Diller

Abbreviations: Same as Table 1 with the addition of: DC: Dendritic cell, IDO: Indoleamine 2,3-dioxygenase, GM-CSF: granulocyte-macrophage colony stimulating factor, HD: High dose, IFNγ : Interferon-gamma, TIL: Tumor infiltrating lymphocytes, TNFα : Tumor necrosis factor alpha, Treg: Regulatory T cell

Table 3.

Prognostic parameters or predictors of response to imABs presented at ASCO 2015.

Clinical trial design	Predictor(s)	Conclusions	Abstract # and authors
Immunohistochemistry
PD1 and PDL1 prognostic values in ccRCC	IHC study in 92 metastatic ccRCC after sunitinib	PDL-1 expression in 71% cases :PD-L1 high associated with shorter PFS and double positive worse clinical outcome	E14002 Brunot A
NSCLC	PDL1 IHC	PDL1 expression higher in females, non-squamous, smokers , high PDL1 expression associated with lower risk of recurrence and better OS	e7551 Owonikoko TK
2 series of NSCLC TMA Yale : n = 204 TMA Greece : n = 350	CD3, CD8, PDL1, IDO, B7-H4 multiplexed IHC	PDL1, IDO, B7-H4 differentially expressed Limited prognostic value PDL-1 and IDO1 correlated with increased TILs	e3067
PDL-1 expression and TILs density in chinese gastric cancer and OS	398 stage I-IV operable GC, PDL-1 and TIL in IHC PDL-1+ if >5% tumor cells	PD-L1 + in 14% PD-L1 positivity associated with TIL density (p = 0.002), high TIL densitiy correlated with OS (HR : 0.10)	4042 Geng R
CXCR4 expression on tumors and CTC	Randomized Phase II : CXCR4 antagonist (LY2510924) and chemotherapy in SCLC	High levels of CTCs and/or CXCR4 expression on CTC and/or tumor tissues at baselien predict benefit of adding LY to standard chemotherapy	7567 Salgia R
Quantity and quality of TILs predicting pCR/DFS/OS	N = 100 TNBC pre- and post- neoadjuvant chemotherapy (NAC)	TIL in pre-NAC (but not post-NAC) predict pCR and survival CD4/CD3 ratio lower in pre-NAC with pCR and lower in post-NAC with pCR CD8/CD4 ratio higher in pre-NAC with pCR and lower in post- NAC with pCR	3035 Altamirano C
PD-L1 expression in immune infiltrates (IC) vs. tumor cells (TC) in NSCLC	498 and 706 NSCLC from MPDL3280A and non-trial cohort using SP142 IHC assay	High PD-L1 expression on TC or IC associated with ORR, PFS and OS to anti-PDL1 Ab in NSCLC but different immunological context that may impact cliniacl outcome further	3015 and others Gettinger S, Spigel Spira, Horn
Stromal (S) versus intratumoral infiltrating CD8+ lymphocytes in resected NSCLC	CD8, CD4 and FoxP3 IHC of TMA 507 NSCLC primary tumor	54% AC, 35% SCC, 11% other. Increased DSS for CD8+ S infiltrate (HR 0.45, 95% CI 0.30–0.97 ; p = 0.04) in N+ pts. Improved OS for PD-L1 (HR 0.71, 95% CI 0.53–1.01 ; p = 0.06) in N0.	11051 Ameratunga M
TIL classification and survival in primary melanoma	TIL on H&E in 1241 primary cutaneous melanoma (absent, non-brisk and brisk); CD3, CD45 IHC in 200 tumors ; 365 SNV in 607 pts,	absent (n = 388 , 31%), non-brisk (n = 330 , 27%), brisk (n = 523 , 42%) TILs. RFS and OS (HR 0.56, p = 0.007 ; HR 0.54, p = 0.0017) for brisk TIL. CD3, CD45 not prognostic. Two germline variants of BCL10 associated with TIL grade (rs962409, OR = 2 .98, p = 0.0007)	e20042 Ann Weiss S
CXCR4 expression on tumor and CTCs as a predictive response marker for CXCR4 antagonist LY2510924 in combination with carboplatin-etoposide in SCLC in open-label, randomized phase II	IHC H-score ROC curves and their AUCs to determine optimal cutoff n = 90	high levels of CTCs (HR 0.403, 95% CI 0.151, 1.076) and/or CXCR4 expression on CTCs (HR 0.476, 95% CI 0.147, 1.548) and/or tumor tissue (HR 0.787, 95% CI 0.211, 2.933) at baseline are predictive for the response to the addition of LY	7567 Salgia R
Zonal difference in PD-1 expression in center of tumor vs. periphery in MSS and MSI CRC	CD8, PD-1 expression in tumor center and periphery n = 42 (28 MSI, 13 MSS)	MSI: CD8 count in tumor center and periphery correlate with earlier tumor size and stage, PD-1 positivity in the center 61% MSS: no PD-1 positivity in the center	3574 O’Kane G
IHC comparison between 75 breast cancer brain metastasis to 20 primary brain tumor	- Absence of TILs - PDL-1	- TILs were absent/low in 97% of breast met and significantly reduced compared to primary tumor -PDL-1 gene expression was negative in both tumors	2059. N.O. Williams
58 LNs surgical resection from Mel patients to characterize microenvironmental environment	-IHC PD-1 in regional LN	- Unlike VEGFA, PD-1 was mostly expressed in pN0 and associated to 5 y survival	9026. V.L. Vazquez
Multicenter trial of 78 patients with MM on I	- PD-L1 gene expression ineffective to predictive I effect	- Extraction from mRNA from formalin-fixed followed by RT-PCR once again showed no statistical difference in PD-L1 expression and response to I	9044. C. Brueggemann
Preclinical analysis looking at the PD-L1 expression in EBV+ associated nasopharyngeal carcinoma	-IHE evaluation of PD-L1	- Using a IHE PDL-1 cut-off of > 1%, 70% of EBV positive NPC expressed in a cohort of 119 patients	6070. K. Mahaprom
Stromal (S) vs. tumoral PD-L1 expression with CD8+ cell count in gastro-esophageal cancer	n = 34 PD-L1 (5H1 clone) status (5% cutoff) at the tumor cell and TIL/TAMs and stromal CD8+ density (50% cutoff) in gastric and gastrooesophageal junction AC	12% tumors PD-L1+, 45% TIL PD-L1+, high stromal CD8+ density (>500 cells/m2) is associated with PD-L1 positivity (89%, p = 0.004) and worse prognosis (PFS, HR 3.91 95%IC 1.32–11.59 p = 0.01; OS, HR 3.46 95%IC 1.09–10.96 p = 0.03)	4031 Kelly RJ
ELISA
Detectable soluble PDL-1 in LG or HG glioma patients or brain mets of solid cancers	EDTA plasma from 16 LG and 68 HG glioma (sandwich ELISA using Millipore ABF133 and clone 5H1)	Detectable in 37% LGG and 52% HGG but ns for concentrations Detectable in 10–50% of brain mets in breast, lung cancers, not in RCC, not in tumors with high PD-L1 membrane expression	e13039 and e13026 Berghoff AS
Plasma soluble PD-L1 in glioma pts and in brain metastatic (BM) pts from solid tumors (melanoma, lung, breast, RCC)	EDTA plasma, sandwich ELISA, rabbit polyclonal anti-PDL1/CD274 ab on a PD-L1 monoclonal ab (clone 5H1), limit of detection PD-L1 50pg/ml.	37.5% (6/16) low-grade and 52.9% (36/68) high-grade glioma, median (1232 vs. 411 pg/mL; p > 0.05), young age (50 vs. 61 ys; p = 0.006). Melanoma (n = 43), (2/4, 50%), breast (1/4, 25%), lung (3/29, 10.3%), RCC (0/6, 0%).	e13039 and e13026 Berghoff AS
Flow cytometry
Predicting response to Pembro in MM	Bim+PD1+ CD11ahighCD8+ circulating T cells	Higher frequencies at Baseline associated with clinical benefit at 4 cycles, dynamic decrease in serial PB samples in OR	9013 Dronca S
IFNα-2b dose reduction in melanoma: a pilot study	Jak-STAT pathway activation: pSTAT1 via flow cytometry, activation of ISGs via PCR analyses	Induction of pSTAT1 and ISGs are similar for IFNα-2b dose reduction	e20035 Levine KM
MHC-II expression to predict response to anti-PD1 therapy	MHC-I/II mRNA (Gene Set Analysis), flow cytometry, IHC (MHC-I, MHC-II, SOX10, CD4, CD8) basal and stimulated (IFN) 60 melanoma cell lines 26 pts’ tumor samples	MCH-I (n = 16) is not associated with response. MCH-II (n = 10) is associated with response and CD4/CD8 infiltrate	9041 Balko JM
29 MM patients on Pembro 2mg/kg on EAP	- BIM (BCL-2 interacting mediator of cell death)	- Flow cytometry evaluation of BIM on CD11a+,PD1+, CD8+. Responders after 4 P cycles could be discriminated based on a higher baseline BIM	9013. R.S. Dronca
Sequencing
Immune gene signatures and efficacy of MK-3475 (anti-PD1 Ab) in PDL1+ gastric cancers	33 RNA expression profiling of advanced gastric cancers (Phase I KEYNOTE 012)	IFNg, TCR signaling, expanded immune and denovo signatures all predicted ORR and PFS to anti-PD1 Ab	3026 Shankaran V
Association of EMT markers and elevated checkpoint inhibitors in lung adenocarcinoma	Integrated analysis of gene expression profiling in 2 large data sets from MD Anderson (PROSPECT, BATTLE-1) and TCGA	EMT phenotype (miR-200/ZEB1) coexist with PD-L1, PD-L2, PD1, Tim3, BTLA, CTLA4 gene expression, and Foxp3 Treg infiltration	3030 Lou Y
Molecular drivers and TILs in urothelial bladder cancer	267 samples: RNA sequencing and exome somatic mutation data	In 36% cases, high TIL infiltrates: Expression of CD8A positively correlated with PDL1, IDO1, FOXP3, TIM3, LAG3 In 33% : T cell exclusion associated with activation of b catenin, PPARG, FGFR pathways Mutational burden does not vary between these subsets	4511 Sweis RF
Intrinsic β-catenin signaling in melanoma	Exome sequencing and gene expression profiling (TCGA), genetically engineered mouse melanoma models (BrafV600E/PTEN-/-+/- active β-catenin)	T-cell exclusion by tumor-intrinsic β-catenin signaling, via failed production of CCL4, with failed recruitment of Batf3-lineage DC; resistance to anti-CTLA4/anti-PD-L1 mAb in mice	9014 Spranger S
Profiling exosomal mRNAs in Ipi treated pts for malignant melanoma	Exosome RNA profile, RNA extraction using exoRNeasy columns, from plasma baseline, 2 and 4 weeks, and 4 mos 586 mRNAs associated with inflammation and 21 mRNA controls by RT-qPCR using the OpenArray technology	n = 450 genes detected, differential expression in responding and non-responding pts, current candidate genes of interest include BMP2, PLCG1 and PDGFB, RNA expression profiling from circulating exosomes offers the ability to monitor changes in immune pathway genes	22159 Hurley J
Expression of immune related genes in metastatic mesothelioma to address immune escape mechanism	- Immune profile for potential patient that would benefit from ICB	- Out of 75 mesotheliomas, 2 clusters were described (High/Low) based on their immune cell related genetic signature.	7566. A. Khattri
41 PBMCs pre/post were obtained from different MM trials	-TCR diversity post Pembro	- Unlike Anti-CTL4, TCR sequencing revealed a greater diversity post-P, which correlated with patients’ outcome.	3027. J.M. Zaretsky
Functional
Ipi treatment in 10 pts with MM to evaluate effect on NK cells	Off target effect on NK from baseline and after 2nd and 4th dose	-2 folds upregulation of IL-2R α chain and improvement of in vitro cytotoxic NK activity in responder patients	9065. I.E.D.P Da Silva
T cell responses to anti-GD2 CH14.18 antibody in NB children	T cell and Ab responses to GD2 mimotopes and TAA antigens	Robust T cell responses to distinct TAA and GD2 mimotopes (not found in healthy volunteers) in 17 NB long term responders. No Ab responses.	3029 Kayser S
Blood biomarker
Early changes in LDH a marker of response to PD-1, PD-L1 inhibitors	Retrospective analysis of MM pts’ fluctuation in LDH from baseline to cycle 2 n = 65	Higher LDH at cycle 2 associated with non-reponse (64 vs. 36%) Similarly for patients with high baseline LDH level. (93 vs. 7%)	e20045 Shreders A
Retrospective observational study of 29 patients with MM treated with anti-PD-1	-Eosinophil count >100/mm3 at 3 weeks	Eosinophil counts increment of >100/mm3 and >400/mm3 at 3 weeks and 12 weeks respectively after PD-1 injection correlated with a superior clinical outcome	9069. L. Gaba
720 MM treated with Ipi 3mg/kg in an EAP	- neutro >7500 - neutro/lympho	- Non-responders could be predicted by an elevated baseline neutrophil count and neutro/lympho	9034. P.F. Ferrucci
MM on PD-1/PDL-1 blockage therapy NCT01295827	-Normal baseline LDH	According to these 2 trials of 110 and 54 pts respectively, normal baseline LDH is the most important clinical predictive marker of good response	9031 K.K. Tsai + e20029 D.W.Kim
Pharmacokinetic
KEYNOTE-001and-002 pharmacokinetic study to determine the incidence of anti-drug antibody	- Absence of anti-drug antibody	- Prembro resistance is not linked to formation of autoimmune antibody based in the analysis of 268 patients where less <1% developed antidrug antibody	3058. T.C. Gangadhar

Abbreviations: Refer to Table 1 with the addition of: ccRCC: clear cell Renal cell carcinoma, CRC: colorectal cancer CTC: Circulating tumor cells, EAP: extended access program, EBV: Epstein-Barr virus, EMT: Epithelial-mesenchymal transition, IHC: immunohistochemistry, LDH = Lactate dehydrogenase, Lympho: Lymphocyte, MSI: microsatellite unstable, MSS: microsatellite stable, Neutro: Neutrophils, PBMC: Peripheral blood mononuclear cell, RFS: relapse free survival, TCR: T cell receptor

Confirmation of efficacy in anti-PD-1 sensitive cancers

During the meeting, results from large clinical trials (# of patients > 100) have been reported in cancer types with previously demonstrated activity of anti-PD-1/PD-L1 imAbs.

Metastatic melanoma

The activity and benefits in OS provided by anti-PD-1 imAbs have been confirmed in patients with Metastatic Melanoma (MM). Dr Adil Daud (University of California, San Francisco, CA) reported in a pooled analysis of 655 patients, the long-term efficacy of pembrolizumab in patients with advanced melanoma enrolled in the KEYNOTE-001 study (NCT01295827). The objective response rate (ORR) was 34% (29% in patients previously treated with ipilimumab, 38% in ipilimumab naïve patients), with a 6% CR rate. The 2 y overall survival was 50% (46% in ipilimumab pre-treated patients, 53% in ipilimumab naïve patients).¹

Non-small cell lung cancer

Two randomized phase III trial results were reported during the meeting with the anti-PD-1 nivolumab in second line Non-Small Cell Lung Cancer (NSCLC), meaning after platinum-based doublet chemotherapy failures. Dr David R Spigel (Sarah Cannon Research Institute, Nashville, TN) reported the results of the CheckMate 017 trial (NCT01642004) which randomized nivolumab vs. docetaxel in previously treated advanced or metastatic squamous cell non-small cell lung cancer (SQ-NSCLC). The ORR was 20% (27/135 patients) for nivolumab vs. 9% (12/137 patients) for docetaxel. The 1 y OS was 42% in the nivolumab arm vs. 24% in the docetaxel arm (Hazard Ratio = 0.59).^2,3

Dr Luis Paz-Ares (Hospital Universitario Virgen Del Rocio, Sevilla, Spain) reported the results of the CheckMate 057 Phase III trial (NCT01673867) which randomized nivolumab (292 patients) vs. docetaxel (290 patients) in advanced non-squamous cell non-small cell lung cancer (non-SQ-NSCLC). The ORR was 19.2% for nivolumab vs. 12.4% for docetaxel. The 1 y OS was 50.5% in the nivolumab arm vs. 39% in the docetaxel arm (HR = 0.73).⁴

Interestingly, in the randomized nivolumab vs. docetaxel CheckMate 017 Phase III trial in SQ-NSCLC, the OS hazard ratios (HR) favored nivolumab regardless of PD-L1 expression and tumor PD-L1 status was neither prognostic nor predictive for efficacy endpoints compared to docetaxel.² However, in non-SQ-NSCLC, the OS HR favored nivolumab only in patients with PD-L1 expression ≥ 1% (Brahmer et al., 2015).

Dr Leora Horn also presented an update of the results of Roche/Genentech anti-PD-L1 atezolizumab (MPDL3280A) in NSCLC (NCT01375842). Out of the 88 patients evaluable for efficacy, the ORR of atezolizumab was 23% and the median duration of response was 17 months. Of interest, patients were concomitantly screened for PD-L1 expression with their internal IHC assay using the SP142 clone. As opposed to Merck (MSD) and BMS, Roche/Genentech scores not only tumor cells for PD-L1 expression but also the immune infiltrative cells. They reported that a subpopulation representing 13.6% (12/88) of patients with NSCLC have high expression of PD-L1 on immune infiltrative cells but no expression of PD-L1 on tumor cells. These patients had an ORR of 50%. The 1 y OS of the whole population was 63%.⁵

Another report of atezolizumab has been presented by Dr Spigel in PD-L1+ selected NSCLC patients (NCT01846416, phase II FIR trial). The ORR (with mRECIST) was 29% in 1st line (31 patients), 17% in 2nd line NSCLC without brain metastases (92 patients) and and 23% in 2nd line NSCLC with brain metastases (13 patients). The one year OS was 73%, 48% and 35% in 1st line, 2^nd line without brain mets and 2nd line with brain mets respectively.⁶

Head and neck squamous cell carcinoma

A confirmation of anti-PD-1 activity in recurrent or metastatic Head and Neck Squamous Cell Carcinoma (HNSCC) has been reported by Dr T.Seiwert (The University of Chicago, Chicago, IL) who shared an update on the KEYNOTE 012 (NCT01848834) trial. The ORR was 24.8% (29/117) in the whole population, 20.6% (7/34) in HPV+ and 26.3% (21/80) in HPV− patients.⁷

Ovarian cancer

Some tumor responses in Ovarian Cancer (OC) patients were seen in early anti-PD-1 phase I trials with a 6% ORR (1/17 patients) reported (Brahmer et al., 2012). Additional data have been presented this year. Dr Disis (University of Washington School of Medicine, Seattle, WA) reported the results of the anti-PD-L1 imAb avelumab (MSB0010718C/ PF-06834635) currently co-developed by Merck KGaA (EMD Serono) and Pfizer in patients with OC.⁸ Of note this anti-PD-L1 is a fully human, non-modified, IgG1 imAb with abilities to perform antibody derived cell cytotoxicity/phagocytosis (ADCC/ADCP) via FcγR positive cells (NK cells, myeloid cells,…). Patients with a median of four prior lines of therapy were treated with avelumab 10mg/kg Q2W (NCT01772004). Out of 75 patients without PD-L1 screening, avelumab showed a 10.7% ORR (8/75 patients) and a stable disease (SD) in 44% patients (33/75).

Dr Andrea Varga (Gustave Roussy, Villejuif, France) reported the results of the anti-PD-1 pembrolizumab (MK3475, Merck, MSD) in patients with OC (NCT02054806). Patients were selected for PD-L1 expression using an immuno-histo-chemistry (IHC) assay based on 1% threshold positivity with the anti-PD-L1 22C3 clone. In this cohort and with this assay/threshold, 51% of patients (49/96) were PD-L1+. In these PD-L1+ patients, an ORR of 11.5% (3/26) has been reported and a stable disease in 23.1% (6/26) of patients.⁹

Identification of new anti-PD-1 sensitive cancers

During this ASCO, four additional histologies have been added to the growing list of cancers with demonstrated sensitivity to anti-PD-1/PD-L1 therapy: Small Cell Lung Cancer (SCLC), Hepatocarcinoma (HCC), Colo-Rectal Cancers (CRC) with Micro-Satellite Instability (MSI), and Esophageal Cancers (_EC) (Fig. 1).

Figure 1.

imABs provide a different approach to treat cancer. Instead of being specific for the tumor cells, they target and unleash the immune system. Therefore, PD1/PDL-1 inhibition is currently being tested in a wide range of cancer subtypes.

Small cell lung cancer

Dr Patrick Ott (Dana-Farber Cancer Institute, Boston, MA) presented the results of pembrolizumab therapy in second line SCLC patients (NCT02054806). Patients were screened and selected for PD-L1 expression. Out of the 147 patients with Small Cell Lung Cancer (SCLC) screened, 42 (28.6%) had PD-L1+ tumors. Only 20 patients were subsequently eligible for the trial. The ORR was 35% (7/20) and stable disease of 5% (1/20).¹⁰

Hepatocarcinoma

Intra-tumoral PD-L1 expression was known to have a negative impact on survival of patients with HCC (Gao et al., 2009; Shi et al., 2011). This data supported the development of an anti-PD-1 based therapy in Hepatocarcinoma (HCC). Dr Anthony B. El-Khoueiry (University of Southern California Norris Comprehensive Cancer Center, Los Angeles, CA) reported the results of the CA209–040 trial testing nivolumab in patients with histologically confirmed advanced HCC with Child-Pugh (CP) score ≤ B7 and progressive disease (PD) on, intolerant of, or refusing sorafenib in patients with either no active hepatitis virus infection or virus-infected HCC patients. The ORR was 19% (8/42 patients) and the DCR was 67%. The 1 y OS (out of 47 patients) was 62%. Overall, in this small cohort of patients, there was no efficacy or safety difference between uninfected and HCV/HBV infected patients.¹¹

Esophageal cancer

Dr Toshihiko Doi (National Cancer Center Hospital East, Chiba, Japan) reported the results of the _EC cohort of the KEYNOTE-028 (NCT02054806) trial. This cohort included squamous cell carcinoma (SCC) or adenocarcinoma of the esophagus or gastroesophageal junction. Patients were selected for PD-L1 expression. With a 1% cancer cells PD-L1 positivity threshold (22C3 clone), 44.6% (37/83) of Esophageal Cancer (_EC) were PD-L1 positive. The ORR was 30.4% (7/23) and the DCR was 43.4% (10/23). The median duration of response was ~10 months.¹²

Mismatch repair deficiency

Recent data has suggested that the level of somatic mutations could be correlated to the immunogenicity of tumors and chances of response to single imAbs.^13-15 This hypothesis is supported by the excellent results of anti-PD-1 therapy in cancers with Mismatch repair deficiency (MMR) defects. Dr Dung Le (The Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins, Baltimore, MD) reported the results of a phase II trial of pembrolizumab in colo-rectal cancer patients (CRC) with either MMR-proficient (n = 25) or MMR-deficient (n = 25) cancers, but also in non-CRC with MMR-defects (n = 21) (NCT01876511). The MMR-deficient non-CRC cohort was made of four cholangiocarcinoma, two endometrial cancers, two small bowel cancers, one prostate and one gastric cancer. The ORR in MMR-deficient cancers was 62% and 60% in CRC and non-CRC respectively, whereas it was 0% in MMR-proficient CRC.¹⁶ The whole-exome sequencing of MMR-deficient tumors revealed a mean of 1782 somatic mutations per tumor as compared with 73 MMR–proficient tumors. The high somatic mutation loads were associated with prolonged progression-free survival.¹⁷

How to overcome immune checkpoint blockade resistance

Combination with chemotherapy

Response rates obtained with atezolizumab (MPDL3280A) in 2nd line NSCLC are around 20%, whereas first line platinum-based doublet chemotherapy usually gives ORR of around 30%. Dr Stephen Liu (Georgetown Univ Hosp, Washington, DC) reported the results of the Phase Ib study evaluating atezolizumab combined with 4–6 cycles carboplatin + either paclitaxel (Arm C), pemetrexed (Arm D) or weekly nab-paclitaxel (Arm E) in patients with first line locally advanced or metastatic NSCLC (NCT01633970). In this cohort of patients, 81% had non-squamous and 19% had squamous NSCLC. Thirty (30) patients were evaluable for efficacy (Arm C, 5; Arm D, 12; Arm E, 13).The ORR was 67% across all arms: 60% in Arm C (3 PRs), 75% in Arm D (9 PRs) and 62% in Arm E (6 PRs and 2 CRs). Tumor responses were seen in each arm independently of PD-L1 expression.¹⁸

The safety and efficacy of combining an anti-PD-1/PD-L1 imAb with chemotherapy in NSCLC has also been reported by Dr Vassiliki Papadimitrakopoulou (The University of Texas MD Anderson Cancer Center, Houston, TX). The KEYNOTE-021 trial evaluated the safety, tolerability and efficacy of pembrolizumab in combination with four cycles of platinum-doublet chemotherapy in NSCLC with Wild-type EGFR and negative ALK translocation status and any PD-L1 status (NCT02039674). Cohort A (n = 25) was treated with pembrolizumab, carboplatin and paclitaxel. Cohort C (n = 24) was treated with pembrolizumab, carboplatin and pemetrexed. Cohort A was composed of 52% (13/25) adenocarcinoma and 36% (9/25) squamous cell carcinoma, and 92% (23/25) current/former smokers. Cohort C was composed of 79% adenocarcinoma (19/24) but no squamous cell carcinoma, and 83% (20/24) current/former smokers. The ORR was 28% and DCR 84% in cohort A. The ORR was 58% and the DCR was 100% in cohort C.¹⁹

Although these two reports contained a limited number of patients they both suggest that combination with pemetrexed might be more synergistic with an anti-PD-1/PD-L1 imAb than with (nab) paclitaxel. However, we need to wait for data with more patients and longer survival follow-up to draw any conclusion.

Anti-PD-1 + anti-CTLA-4 combination therapy

Dr Amita Patnaik (START, San Antonio, TX) presented the results of the pembrolizumab + ipilimumab combination in patients with NSCLC. Patients with stage IIIB/IV NSCLC that recurred after ≤ two prior regimens received pembrolizumab + ipilimumab every 3 weeks for 4 cycles followed by maintenance pembrolizumab (NCT02039674). The population was composed of 56% adenocarcinoma and 22% squamous cell carcinoma (22% NOS). The ORR was 39% (7/18) and the DCR 83% (15/18 patients). No data were presented with further follow up (to see if additional SD transform into PR), notably no survival data was presented.²⁰

Dr Scott Antonia (Moffitt Cancer Center, Tampa, FL) reported the results of the nivolumab + ipilimumab combination trial in SCLC patients (Checkmate 032, NCT1928394). In the nivolumab alone arm (40 patients) the ORR was 18% and the DCR was 38%). In the nivolumab+ipilimumab arm (46 patients) the ORR was 17% and the DCR of 54%. Of note, out of the 17 patients with SD, 7 had subsequently a confirmed PR resulting in an updated ORR of 32.6% in the nivolumab + ipilimumab arm (no additional responses occurred in the nivolumab alone arm).²¹

In-situ immunization

In situ immunization consists in injecting immunostimulatory products into a tumor site in order to prime the antitumor immune response. Oncolytic viruses are versatile biotherapies with multiple antitumor effects into a single product: tumor cell cytotoxicity, immune adjuvant effect thanks to the viral pathogen associated molecular patterns and pro-inflammatory cytokines encoded by transgenes. T-VEC is an HSV derived oncolytic virus genetically modified to express GM-CSF. Dr Igor Puzanov (Vanderbilt University Medical Center, Nashville, TN) reported an update on the IV ipilimumab + IT T-VEC phase 1b/2 trial in stage IIIB-IV melanoma (NCT01740297). Out of 18 patients treated with more than 17 months of follow-up, the ORR was 56% (33% of CR) with 44% having responses during more than 6 months. On a per lesion level, 68.5% (24/35) injected index lesions and 50% (8/16) uninjected index lesions regressed more than 50%.²²

Another in situ strategy has been followed by the team of Dr Randy Christopher Bowen (University of Utah, Salt Lake City, UT) with intra-tumoral (IT) interleukin-2 and IT ipilimumab in patients with III/IV unresectable melanoma with at least one injectable lesion (NCT01672450). Twelve patients were treated, 10 were evaluable for response by irRC. The ORR was 40% although one PD was later found to be a CR by resection. The two non-evaluable patients had regression of multiple skin lesions. A response in at least one non-injected lesion (abscopal effect) was seen in 9/12 patients (75%).²³

Conclusion

Besides these major announcements, smaller results with significant impact on clinical practice have been presented. For instance, experiencing a severe irAE from anti-CTLA-4 might not preclude a patient with melanoma from subsequently receiving an anti-PD-1.²⁴ On the other hand, treating a melanoma patients with anti-CTLA4 after anti-PD-1 failure still have antitumor activity and do not provide additional protection.²⁵ Against common dogma, anti-PD-1 therapy might be also active in metastatic uveal melanoma. Out of small cohort of seven patients treated with anti-PD-1, 1 presented a CR, 1 a PR and 1 a SD.²⁶ Desmoplastic melanoma, another rare subtype of melanoma comprising approximately 1% of cases and with higher mutational load than other sub-types of melanoma presented an ORR of 70% with PD-1/PD-L1 imAbs in a retrospective analysis of over 1,000 melanoma patients.²⁷ Altogether, this additional set of data reinforces the validity of using imAb to treat multiple types of advanced cancers. Although some patients can respond to these new immunotherapies, a significant proportion of them do not respond and new combination therapies have to be found to address such primary refractoriness. Also, responding patients can subsequently relapse and it is not clear what is at stake biologically to explain these secondary immune resistances. These first immune targeted therapies have demonstrated that the most immunogenic cancers can be treated with a single immune checkpoint blockade and that combination of imAbs can overcome monotherapy resistance. Now the question is not to know if this immunotherapy phenomenon is limited to a subset of highly mutated cancers but rather to identify which additional immune checkpoints are critical for the tolerance of primary resistant or less mutated cancers. And these yet to be discovered immune checkpoints might not only be on T-cells but also on other subsets of the adaptive and innate immune cells.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.