Automated digital TIL analysis (ADTA) adds prognostic value to standard assessment of depth and ulceration in primary melanoma.

Accurate prognostic biomarkers in early-stage melanoma are urgently needed to stratify patients for clinical trials of adjuvant therapy. We applied a previously developed open source deep learning algorithm to detect tumor-infiltrating lymphocytes (TILs) in hematoxylin and eosin (H&E) images of early-stage melanomas. We tested whether automated digital (TIL) analysis (ADTA) improved accuracy of prediction of disease specific survival (DSS) based on current pathology standards. ADTA was applied to a training cohort (n = 80) and a cutoff value was defined based on a Receiver Operating Curve. ADTA was then applied to a validation cohort (n = 145) and the previously determined cutoff value was used to stratify high and low risk patients, as demonstrated by Kaplan-Meier analysis (p ≤ 0.001). Multivariable Cox proportional hazards analysis was performed using ADTA, depth, and ulceration as co-variables and showed that ADTA contributed to DSS prediction (HR: 4.18, CI 1.51-11.58, p = 0.006). ADTA provides an effective and attainable assessment of TILs and should be further evaluated in larger studies for inclusion in staging algorithms.

Distinguishing melanophages from tumor in melanoma patients treated with talimogene laherparepvec.

Response to talimogene laherparepvec (T-Vec) is difficult to assess as pigmented macrophages that have ingested melanoma cells ('melanophages') persist after injection, mimicking melanoma. We used quantitative immunofluorescence (qIF) to (1) distinguish melanophages from melanoma in biopsies from two patients treated with T-Vec and (2) evaluate the tumor microenvironment pretreatment and posttreatment. Tissues were stained with 4',6-diamidino-2-phenylindole, cluster of differentiation (CD) 3, CD8, CD68, human leukocyte antigen-DR isotype (HLA-DR), and SRY-Box Transcription Factor 10 (SOX10), and multispectral images were analyzed. Post-T-Vec samples showed melanophages with cytoplasmic costaining of CD68, SOX10, and HLA-DR, without nuclear SOX10 expression. qIF revealed a dense immune infiltrate of CD3, CD8, and CD68 cells in post-T-Vec samples. Melanophages from tumors post-T-Vec stain the nuclear melanoma marker SOX10 in their cytoplasms as compared to melanoma cells that stain nuclear SOX10. This novel finding highlights the phagocytosis of melanoma cell components by macrophages after treatment with T-Vec. qIF may assist pathologists in determining whether lesions treated with immunotherapy contain residual viable melanoma.

Linking Transcriptomic and Imaging Data Defines Features of a Favorable Tumor Immune Microenvironment and Identifies a Combination Biomarker for Primary Melanoma.

Patients with resected stage II-III melanoma have approximately a 35% chance of death from their disease. A deeper understanding of the tumor immune microenvironment (TIME) is required to stratify patients and identify factors leading to therapy resistance. We previously identified that the melanoma immune profile (MIP), an IFN-based gene signature, and the ratio of CD8+ cytotoxic T lymphocytes (CTL) to CD68+ macrophages both predict disease-specific survival (DSS). Here, we compared primary with metastatic tumors and found that the nuclei of tumor cells were significantly larger in metastases. The CTL/macrophage ratio was significantly different between primary tumors without distant metastatic recurrence (DMR) and metastases. Patients without DMR had higher degrees of clustering between tumor cells and CTLs, and between tumor cells and HLA-DR+ macrophages, but not HLA-DR- macrophages. The HLA-DR- subset coexpressed CD163+CSF1R+ at higher levels than CD68+HLA-DR+ macrophages, consistent with an M2 phenotype. Finally, combined transcriptomic and multiplex data revealed that densities of CD8 and M1 macrophages correlated with their respective cell phenotype signatures. Combination of the MIP signature with the CTL/macrophage ratio stratified patients into three risk groups that were predictive of DSS, highlighting the potential use of combination biomarkers for adjuvant therapy. SIGNIFICANCE: These findings provide a deeper understanding of the tumor immune microenvironment by combining multiple modalities to stratify patients into risk groups, a critical step to improving the management of patients with melanoma.

Deep Learning Based on Standard H&E Images of Primary Melanoma Tumors Identifies Patients at Risk for Visceral Recurrence and Death.

PURPOSE: Biomarkers for disease-specific survival (DSS) in early-stage melanoma are needed to select patients for adjuvant immunotherapy and accelerate clinical trial design. We present a pathology-based computational method using a deep neural network architecture for DSS prediction. EXPERIMENTAL DESIGN: The model was trained on 108 patients from four institutions and tested on 104 patients from Yale School of Medicine (YSM, New Haven, CT). A receiver operating characteristic (ROC) curve was generated on the basis of vote aggregation of individual image sequences, an optimized cutoff was selected, and the computational model was tested on a third independent population of 51 patients from Geisinger Health Systems (GHS). RESULTS: Area under the curve (AUC) in the YSM patients was 0.905 (P < 0.0001). AUC in the GHS patients was 0.880 (P < 0.0001). Using the cutoff selected in the YSM cohort, the computational model predicted DSS in the GHS cohort based on Kaplan-Meier (KM) analysis (P < 0.0001). CONCLUSIONS: The novel method presented is applicable to digital images, obviating the need for sample shipment and manipulation and representing a practical advance over current genetic and IHC-based methods.

Biomarkers Predictive of Survival and Response to Immune Checkpoint Inhibitors in Melanoma.

Immunotherapy has revolutionized the treatment of melanoma. Targeting of the immune checkpoints cytotoxic T-lymphocyte-associated protein 4 and programmed cell death protein 1 has led to improved survival in a subset of patients. Unfortunately, the use of immune checkpoint inhibitors is associated with significant side effects and many patients do not respond to treatment. Thus, there is an urgent need both for prognostic biomarkers to estimate risk and for predictive biomarkers to determine which patients are likely to respond to therapy. In this review, prognostic and predictive biomarkers that are an active area of research are outlined. Of note, certain transcriptomic signatures are already used in the clinic, albeit not routinely, to prognosticate patients. In the predictive setting, programmed cell death protein ligand 1 expression has been shown to correlate with benefit but is not precise enough to be used as an exclusionary biomarker. Future investigation will need to focus on biomarkers that are easily reproducible, cost effective, and accurate. The use of readily available clinical material, such as serum or hematoxylin and eosin-stained images, may offer one such path forward.

Tissue-Resident Memory T Cells Mediate Immune Homeostasis in the Human Pancreas through the PD-1/PD-L1 Pathway.

Non-recirculating tissue-resident memory T cells (TRMs) are the predominant T cell subset in diverse tissue sites, where they mediate protective immune responses in situ. Here, we reveal a role for TRM in maintaining immune homeostasis in the human pancreas through interactions with resident macrophages and the PD-1/PD-L1 inhibitory pathway. Using tissues obtained from organ donors, we identify that pancreas T cells comprise CD8+PD-1hi TRMs, which are phenotypically, functionally, and transcriptionally distinct compared to TRMs in neighboring jejunum and lymph node sites. Pancreas TRMs cluster with resident macrophages throughout the exocrine areas; TRM effector functions are enhanced by macrophage-derived co-stimulation and attenuated by the PD-1/PD-L1 pathways. Conversely, in samples from chronic pancreatitis, TRMs exhibit reduced PD-1 expression and reduced interactions with macrophages. These findings suggest important roles for PD-1 and TRM-macrophage interactions in controlling tissue homeostasis and immune dysfunctions underlying inflammatory disease, with important implications for PD-1-based immunotherapies.

Immune microenvironment modulation unmasks therapeutic benefit of radiotherapy and checkpoint inhibition.

BACKGROUND: Immune checkpoint inhibitors (ICIs) for solid tumors, including those targeting programmed cell death 1 (PD-1) and cytotoxic T lymphocyte-associated antigen 4 (CTLA-4), have shown impressive clinical efficacy, however, most patients do not achieve durable responses. One major therapeutic obstacle is the immunosuppressive tumor immune microenvironment (TIME). Thus, we hypothesized that a strategy combining tumor-directed radiation with TIME immunomodulation could improve ICI response rates in established solid tumors. METHODS: Using a syngeneic mouse model of human papillomavirus (HPV)-associated head and neck cancer, mEER, we developed a maximally effective regimen combining PD-1 and CTLA-4 inhibition, tumor-directed radiation, and two existing immunomodulatory drugs: cyclophosphamide (CTX) and a small-molecule inducible nitric oxide synthase (iNOS) inhibitor, L-n6-(1-iminoethyl)-lysine (L-NIL). We compared the effects of the various combinations of this regimen on tumor growth, overall survival, establishment of immunologic memory, and immunologic changes with flow cytometry and quantitative multiplex immunofluorescence. RESULTS: We found PD-1 and CTLA-4 blockade, and radiotherapy alone or in combination, incapable of clearing established tumors or reversing the unfavorable balance of effector to suppressor cells in the TIME. However, modulation of the TIME with cyclophosphamide (CTX) and L-NIL in combination with dual checkpoint inhibition and radiation led to rejection of over 70% of established mEER tumors and doubled median survival in the B16 melanoma model. Anti-tumor activity was CD8+ T cell-dependent and led to development of immunologic memory against tumor-associated HPV antigens. Immune profiling revealed that CTX/L-NIL induced remodeling of myeloid cell populations in the TIME and tumor-draining lymph node and drove subsequent activation and intratumoral infiltration of CD8+ effector T cells. CONCLUSIONS: Overall, this study demonstrates that modulation of the immunosuppressive TIME is required to unlock the benefits of ICIs and radiotherapy to induce immunologic rejection of treatment-refractory established solid tumors.