NEOADJUVANT IMMUNO-RADIOTHERAPY IN HEAD AND NECK CANCER: PHASE I/IB STUDY OF COMBINED NIVOLUMAB/SBRT PRIOR TO SURGICAL RESECTION (CTPL04_CT182)
Rom Leidner, MD (Providence Cancer Center, Portland, OR, USA) presented the phase I safety and efficacy data from the neoadjuvant immuno-radiotherapy (NIRT) trial of nivolumab in combination with stereotactic body radiation therapy (SBRT) in patients with p16+ oropharyngeal head and neck squamous cell carcinoma (HNSCC) (NCT03247712). This is the first study to evaluate the safety and efficacy of neoadjuvant nivolumab + SBRT for treatment of HNSCC patients. Standard treatment of locally advanced human papillomavirus (HPV)-associated oropharyngeal HNSCC includes definitive chemoradiotherapy or surgery followed by risk-adapted adjuvant radiotherapy +/- chemotherapy. In this study, patients received nivolumab 240mg IV Q2W three times prior to surgery, with SBRT to gross tumor volume (GTV) + 3 mm delivered between nivolumab doses 1 and 2, in two dose finding cohorts (n = 5 each): 8 Gy x 5 daily fractions and de-escalated 8 Gy x 3 fractions. Surgery (transoral robotic resection and/or neck dissection) was performed 5 weeks post-SBRT followed by adjuvant nivolumab 480mg IV Q4W three times, starting 4 weeks post-operation. In both cohorts, there were no unplanned surgical delays (primary safety endpoint). All patients had evidence of decreased tumor size prior to surgery, but none showed a CR by RECIST. However, the pathologic CR rate was 100% in the 8 Gy x 5 cohort and 80% in the 8 Gy x 3 cohort, with the remaining patient achieving MPR (major pathologic response; less than 10% residual viable tumor). Grade 3 delayed toxicity was observed after surgery in both cohorts but the rate was higher in the 8 Gy x5 cohort. Of note, lymphopenia, which is a common toxicity observed with conventional fractionated radiation, was not observed, nor was any grade 4/5 toxicity. Interestingly, grade 2 adrenal insufficiency caused by unknown mechanism(s) was observed in 50% (5/10). Together, neoadjuvant combined nivolumab/SBRT to GTV + 3 mm dosed at either 8Gy x5 or 8 Gy x3 did not delay HNSCC surgery in this phase I trial and pathologic anti-tumor response was observed in all cases (CR = 9, MPR = 10). A high response rate and lower toxicity profile favors the 8 Gy x3 cohort for future development. Analyses of blood and tumor tissues using RNAseq and tandem TCRseq are ongoing.
PHASE I/II STUDY OF MBG453 ± SPARTALIZUMAB IN PATIENTS WITH ADVANCED SOLID TUMORS (CTPL04_CT183)
Giuseppe Curigliano, MD, PhD (Istituto Europeo di Oncologia IRCCS, Milan, Italy) reported the dose escalation results from a phase I-Ib/II study of MBG453 ± spartalizumab in metastatic solid tumors (NCT02608268). MBG453 and spartalizumab are humanized IgG4 monoclonal antibodies to block binding of TIM-3 to Phosphatidylserine and PD-1 to PD-L1/2, respectively. Preclinical studies show synergistic anti-tumor activity of TIM-3 + PD-1 co-blockade. In this study, patients received MBG453 alone (80-1200 mg IV, Q2W or Q4W), or combination therapy (Q2W/Q4W) with MBG453 (20-800 mg/80-1200 mg) + spartalizumab (80-240 mg/80-400 mg). As of July 26, 2018, 87 patients received MBG453 alone and 86 patients received MBG453 + spartalizumab. Treatment-related AEs were reported in equal to or higher than 2% of patients treated with MBG453 and MBG453 + spartalizumab, and grade 3/4 AEs in 0% and 11% of patients, respectively. Maximum tolerated doses were not identified with the tested dose/schedule. MBG453 800 mg Q4W (n = 9) and MBG453 800 mg + spartalizumab 400 mg Q4W (n = 6) were declared as RP2Ds. Stable disease (SD) was seen in 25/87 (29%) patients treated with MBG453 alone. Of 86 MBG453 + spartalizumab treated patients, PR was seen in 4 patients (5%), SD was in 34/86 (40%). Additionally, RNAseq analysis of screening and on-treatment biopsies revealed a pharmacodynamic trend of increased IFN-gamma-associated gene signatures in patients with the combination treatment. Thus, MBG453 + spartalizumab was well tolerated and demonstrated preliminary signs of anti-tumor activity. MBG453 800 mg Q4W and MBG453 800 mg + spartalizumab 400 mg Q4W were selected as the recommended phase II doses. The dose expansion phase II study is ongoing in patients with melanoma or NSCLC resistant to anti-PD-1/PD-L1.
EFFECTOR B CELLS AND TERTIARY LYMPHOID STRUCTURES PREDICT RESPONSE TO IMMUNE CHECKPOINT BLOCKADE IN SOLID TUMORS (MS.IM02.02_4488)
Sangeetha Reddy, MD (University of Texas, MD Anderson Cancer Center, Houston, TX, USA. Current: UT Southwestern Medical Center, Dallas, T, USA) presented data from the phase II trial of neoadjuvant immune checkpoint blockade (ICB) in patients with high-risk resectable melanoma (NCT02519322). To date, studies on anti-tumor immunity and tissue biomarkers of response to ICB treatment have been focused on T cell markers (e.g. PD-L1) and cell-cell interactions (T cells, dendritic cells, tumor cells). However, other immune compartments also have a role in anti-tumor activity. For example, correlative analyses of the neoadjuvant melanoma ICB cohort identified infiltrated B cell markers as being predictive to response in addition to expected biomarkers of response (CD8, PD-L1, TMB). Also, whole transcriptome and immune cell expression analyzed by Microenvironment Cell Population (MCP) counter revealed that in responders, B cells had the most differentially expressed genes (MZB1, JCHAIN, IGLL5, FCRL5, p < 0.0001). In tumor and blood samples, non-overlapping B cell population was detected by CyTOF analysis, including class switched memory subsets and plasma-like cells (CD27+, IgD-, C38++, CD138-, CD20-). Responders tend to have a higher fraction of long-lived effector cells with less naïve cells in tumor than non-responders. Immunohistochemistry (IHC) and multiplex staining identified that B cells were present in organized tertiary lymphoid structures (TLS) in close proximity to T cells and follicular dendritic cells, and the area occupied by TLS was higher in responders (p = 0.037 baseline and p = 0.002 on-treatment). In summary, intratumoral B cells are distinct (activated, class switched effector cells) from peripheral B cell populations and these data suggest that B cells with activated effector phenotypes and TLS predict response to ICB treatment of solid tumors.
INTRATUMORAL IMMUNOLOGIC PROFILE ASSOCIATES WITH OVERALL SURVIVAL IN METASTATIC PANCREATIC CANCER PATIENTS TREATED WITH COMBINATION IMMUNOTHERAPY GVAX WITH OR WITHOUT NIVOLUMAB (MS.IM02.02_4491)
Annie A. Wu (Johns Hopkins Medical Institute, Baltimore, MD, USA) described the clinical testing of an irradiated, granulocyte-macrophage colony-stimulating factor (GM-CSF)-secreting, allogenic vaccine (GVAX) in patients with metastatic pancreatic ductal adenocarcinoma (PDAC). PDAC is considered a "nonimmunogenic" neoplasm and single-agent immunotherapies have failed to demonstrate significant clinical activity. Previously treated, metastatic PDAC patients (n = 88) were randomized 1:1 to receive the treatment of vaccine GVAX (5 x 10^8 cells, 6 intradermal injections) + cyclophosphamide (Cy; 200 mg/ml^2, IV) + CRS-207 (live attenuated listeria monocytogenes expressing mesothelin; 1 x 10^9 CFU, IV) or Cy/GVAX + nivolumab (3 mg/kg, IV). Biopsies (n = 22) were obtained at baseline and after 2 GVAX prime and 1 CRS-207 boost from 96 vaccinated patients. Multiplex IHC was used for longitudinal monitoring of intratumoral immune infiltrates. After prime-boost, favorable OS correlated with low CD68+ myeloid cell and high CD45+ lymphoid cell numbers. Expansion and higher functional status of CD8+ T cells and increased Th1/Th2 cell ratio was associated with longer OS. At baseline, fewer CSF1R+ tumor associated macrophages (TAMs), CD68+CD163+ and CD163- myeloid cells in tumors was associated with longer OS. Also, less exhausted and increased effector memory CD8+ T cells during treatment benefits OS. Tumors from nivolumab-treated patients demonstrated a decrease in CD68+ myeloid cells after prime-boost compared to baseline. Interestingly, CD163+ TAMs in tumors of nivolumab-treated patients expressed higher PD-L1 levels. These data suggest that induction of lymphoid-inflamed expression profiles with less exhausted and more effector memory CD8+ T cells during treatment is associated with longer OS, independent of nivolumab treatment. These findings also suggest that nivolumab induces PD-L1 expression on myeloid cells. Further identification of key tissue-based biomarkers correlating with OS may be useful for predicting therapeutic response.
Glossary
GM-CSF = granulocyte-macrophage colony-stimulating factor
GVAX = GM-CSF-secreting, allogenic vaccine
HNSCC = head and neck squamous cell carcinoma
HPV = human papillomavirus
ICB = immune checkpoint blockade
IHC = immunohistochemistry
IV = Intravenous
MCP = microenvironment cell population
NIRT = neoadjuvant immuno-radiotherapy
NSCLC = non-small cell lung cancer
OS = overall survival
PDAC = pancreatic ductal adenocarcinoma
PD-L1 = Programmed death-ligand 1
RP2D = recommended phase II dose
SBRT = stereotactic body radiation therapy
Q2W = every 2 weeks
Q4W = every 4 weeks
SD = stable disease
TLS = tertiary lymphoid structures
TMB = tumor mutational burden