JITC Editor Picks
Robert B Rebhun, Daniel York, Sylvia Margret Cruz, Sean J Judge, Aryana M Razmara, Lauren E Farley, Rachel V Brady, Eric G Johnson, Jenna H Burton, Jennifer Willcox, Luke A Wittenburg, Kevin Woolard, Cordelia Dunai, Susan L Stewart, Ellen E Sparger, Sita S Withers, Alicia A Gingrich, Katherine A Skorupski, Sami Al-Nadaf, Amandine T LeJeune, William TN Culp, William J Murphy, Michael S Kent, Robert J Canter
Journal for ImmunoTherapy of Cancer 2022;10:e004493 (9 June 2022)
RESEARCH
Summary:
Cytokine therapy with IL-15 has yielded disappointing results in clinical trials, with stable disease as the most common best response and high rates of systemic toxicity. Inhalation of cytokines limits systemic exposure and potential toxicity while allowing for direct delivery of therapy to the lungs, motivating Robert B Rebhun and colleagues to perform a first-in-dog phase I clinical trial evaluating inhaled recombinant human IL-15 in companion canines with melanoma or osteosarcoma. The primary endpoints were safety, dose-limiting toxicities, and maximum tolerated dose with secondary objectives including response rate, progression-free survival, overall survival, and correlative studies. A total of 21 dogs were enrolled: 11 with melanoma and 10 with osteosarcoma. At the initial dose level of 10 micrograms, no fevers, hypotension or other abnormalities were observed and the most frequently reported side effect was coughing. No dose-limiting toxicities were observed until the fifth dose level, at which point clinically evident necrosis or abscessation of confirmed metastatic melanoma was observed in the lymph nodes of two of six dogs. The maximum tolerated dose and recommended phase II dose was found to be 50 micrograms twice daily for 14 days. Among the 18 evaluable dogs, the overall clinical benefit rate was 39% including 1 complete response lasting more than 1 year, 1 partial response with resolution of multiple lesions, and 5 cases of stable disease. Median survival from the first treatment was 113.5 (range 34–407) and 82.5 (range 36–655) days for dogs with melanoma and osteosarcoma, respectively. Serum levels of IL-15 after inhalation were in the picomolar range, remaining relatively stable for 6 hours after treatment with a peak at the 4 hour time-point. No association was observed between response to treatment and plasma levels of endogenous cytokines. Lower baseline absolute lymphocyte counts and increased cytoxiticy of peripheral blood monocuclear cells post-treatment were statistically significantly associated with response. These data provide rationale for additional trials of inhaled cytokine therapy in canine and human patients.
Mirela Kremenovic, Alfred A Chan, Bing Feng, Lukas Bäriswyl, Steve Robatel, Thomas Gruber, Li Tang, Delphine J Lee, Mirjam Schenk
Journal for ImmunoTherapy of Cancer 2022;10:e004493 (22 June 2022)
RESEARCH
Summary:
Intralesional administration of the live bacterial strain Mycobacterium bovis Bacillus Calmette-Guérin (BCG) has been known to induce regression of injected melanoma since the 1970s, however, responses in distal lesions are rare. Reasoning that sustained delivery of pathogen-associated molecular patterns might induce proinflammatory tumor microenvironment capable of generating systemic responses, Mirela Kremenovic et al developed a novel BCG lysate for loading into a thermosensitive hydrogel. The BCG lysate led to more pronounced reduction in tumor growth than live BCG in murine B16F10 melanoma models. Tumors injected with lysate had high frequencies of infiltrating CD8+ and CD4+ T cells, natural killer cells, dendritic cells, and tumor-associated M1 macrophages, as well as elevated levels of proinflammatory cytokines, CXCR2-binding cytokines, and growth factors with an accompanying shift in gene expression profile toward an immune-inflamed tumor microenvironment. Notably, injection of BCG lysate-loaded hydrogel not only led to similar profiles of effector cells and macrophages in the treated tumors as lysate alone but also increased lymphocytic infiltration in untreated tumors on the contralateral flank. BCG hydrogel injection prolonged survival of treated mice by 24% in the B16F10 melanoma model and suppressed the development of pulmonary metastases in MC38 tumor models. Injection of BCG hydrogel led to significant upregulation of genes involved in antigen processing and presentation including several cathepsins, response to interferon gamma, and positive regulation of leukocyte chemotaxis by bulk tumor RNA sequencing as well as an increase in infiltrating T cell receptor diversity and concomitant decreased clonality. In the tumor-draining lymph nodes, increases in antigen-presenting cells, dendritic cells, CD4+ and CD8+ T cells with a decrease in naïve T cells were also seen. These findings suggest that hydrogel delivery of BCG lysate is capable of inducing systemic anti-tumor immune responses through a mechanism involving enhanced antigen processing and presentation.
Sakti Chakrabarti, Leslie Bucheit, Jason Scott Starr, Racquel Innis-Shelton, Ardaman Shergill, Hiba Dada, Regina Resta, Stephanie Wagner, Naomi Fei, Pashtoon Murtaza Kasi
Journal for ImmunoTherapy of Cancer 2022;10:e004493 (16 June 2022)
RESEARCH
Summary:
Tumors with high microsatellite instability (MSI-H) have excellent responses to checkpoint inhibition, however, tissue availability is often a barrier to genomic profiling. Building on reports of concordance between tissue- and circulating tumor DNA (ctDNA)-based measurements of MSI-H status, Sakti Chakrabarti et al retrospectively demonstrate a 77% overall response rate to checkpoint blockade for pancreatic cancer found to be MSI-H by liquid biopsy. Out of 6,000 patients with pancreatic ductal adenocarcinoma who underwent ctDNA profiling using the Guardant G360 74-gene panel, 52 were identified as MSI-H and clinical outcome data were available for 10, with 9 receiving ICI therapy. Among the six patients for whom both liquid and tumor biopsy results were available, only one case of non-concordance between ctDNA and tissue MSI-H testing was observed. Notably, the one tumor that was determined to be proficient in mismatch repair (the most common driver of microsatellite instability) by immunohistochemistry yet MSI-H by ctDNA analysis had a pathologic complete response to neoadjuvant anti-PD-1 plus anti-CTLA-4. Although small and retrospective, this analysis provides proof-of-concept for using validated ctDNA assays for identifying patients eligible for checkpoint blockade treatment through the tissue-agnostic indication for MSI-H tumors.
Michal Harel, Coren Lahav, Eyal Jacob, Nili Dahan, Itamar Sela, Yehonatan Elon, Shani Raveh Shoval, Galit Yahalom, Iris Kamer, Alona Zer, Ofer Sharon, David P Carbone, Adam P Dicker, Jair Bar, Yuval Shaked
Journal for ImmunoTherapy of Cancer 2022;10:e004493 (19 June 2022)
RESEARCH
Summary:
Tumor-intrinsic biomarkers such as mutational burden and PD-L1 expression have limited predictive power for benefit with checkpoint blockade and current requirements for biopsy tissue for evaluation of these biomarkers is an obstacle to capturing tumor heterogeneity as well as changes that occur during treatment. Using pre- and on-treatment plasma proteomic profiling of roughly 800 proteins combined with machine learning approaches, Michal Harel and colleagues identify a signature to predict response to anti-PD-(L)1 based therapy for patients with non-small cell lung cancer (NSCLC) based on age, sex, and on-treatment serum levels of CXCL18 and CXCL10. A retrospective cohort of 143 patients with NSCLC were included in the analysis, of whom 70 had no clinical benefit with checkpoint inhibitor-based treatment and 73 benefited from therapy, defined as stable disease, partial response, or complete response at 3 months on-treatment by RECIST criteria. The four-parameter signature had an area under the receiver operating characteristic curve of 0.79. Interestingly, although the vast majority of the patients received anti-PD-1 monotherapy, the predictive value of the signature decreased when the 12 patients who were treated with anti-PD-1 in combination with chemotherapy were excluded from the model. Clustering based on overall plasma proteomic profiles pre- and on-treatment revealed three distinct groups, each characterized by unique clinical features and enriched for response or non-response to therapy. Elevated levels of several neutrophil-associated proteins on-treatment were associated with lack of benefit with therapy, including CXCL8, as well as IL-6, PILRA, TNFRSF21, PLAUR, CDH3, GOLM1, GRO (CXCL1/CXCL2/CXCL3), SDC1, and CXCL10. This study supports further development of serum-based biomarkers to capture dynamic changes in immune cell responses to tumors during treatment and predict benefit with immunotherapy.