AI-based pathology predicts origins for cancers of unknown primary.

Cancer of unknown primary (CUP) origin is an enigmatic group of diagnoses in which the primary anatomical site of tumour origin cannot be determined1,2. This poses a considerable challenge, as modern therapeutics are predominantly specific to the primary tumour3. Recent research has focused on using genomics and transcriptomics to identify the origin of a tumour4-9. However, genomic testing is not always performed and lacks clinical penetration in low-resource settings. Here, to overcome these challenges, we present a deep-learning-based algorithm-Tumour Origin Assessment via Deep Learning (TOAD)-that can provide a differential diagnosis for the origin of the primary tumour using routinely acquired histology slides. We used whole-slide images of tumours with known primary origins to train a model that simultaneously identifies the tumour as primary or metastatic and predicts its site of origin. On our held-out test set of tumours with known primary origins, the model achieved a top-1 accuracy of 0.83 and a top-3 accuracy of 0.96, whereas on our external test set it achieved top-1 and top-3 accuracies of 0.80 and 0.93, respectively. We further curated a dataset of 317 cases of CUP for which a differential diagnosis was assigned. Our model predictions resulted in concordance for 61% of cases and a top-3 agreement of 82%. TOAD can be used as an assistive tool to assign a differential diagnosis to complicated cases of metastatic tumours and CUPs and could be used in conjunction with or in lieu of ancillary tests and extensive diagnostic work-ups to reduce the occurrence of CUP.

Pan-cancer integrative histology-genomic analysis via multimodal deep learning.

The rapidly emerging field of computational pathology has demonstrated promise in developing objective prognostic models from histology images. However, most prognostic models are either based on histology or genomics alone and do not address how these data sources can be integrated to develop joint image-omic prognostic models. Additionally, identifying explainable morphological and molecular descriptors from these models that govern such prognosis is of interest. We use multimodal deep learning to jointly examine pathology whole-slide images and molecular profile data from 14 cancer types. Our weakly supervised, multimodal deep-learning algorithm is able to fuse these heterogeneous modalities to predict outcomes and discover prognostic features that correlate with poor and favorable outcomes. We present all analyses for morphological and molecular correlates of patient prognosis across the 14 cancer types at both a disease and a patient level in an interactive open-access database to allow for further exploration, biomarker discovery, and feature assessment.

Deep learning-enabled assessment of cardiac allograft rejection from endomyocardial biopsies.

Endomyocardial biopsy (EMB) screening represents the standard of care for detecting allograft rejections after heart transplant. Manual interpretation of EMBs is affected by substantial interobserver and intraobserver variability, which often leads to inappropriate treatment with immunosuppressive drugs, unnecessary follow-up biopsies and poor transplant outcomes. Here we present a deep learning-based artificial intelligence (AI) system for automated assessment of gigapixel whole-slide images obtained from EMBs, which simultaneously addresses detection, subtyping and grading of allograft rejection. To assess model performance, we curated a large dataset from the United States, as well as independent test cohorts from Turkey and Switzerland, which includes large-scale variability across populations, sample preparations and slide scanning instrumentation. The model detects allograft rejection with an area under the receiver operating characteristic curve (AUC) of 0.962; assesses the cellular and antibody-mediated rejection type with AUCs of 0.958 and 0.874, respectively; detects Quilty B lesions, benign mimics of rejection, with an AUC of 0.939; and differentiates between low-grade and high-grade rejections with an AUC of 0.833. In a human reader study, the AI system showed non-inferior performance to conventional assessment and reduced interobserver variability and assessment time. This robust evaluation of cardiac allograft rejection paves the way for clinical trials to establish the efficacy of AI-assisted EMB assessment and its potential for improving heart transplant outcomes.

Tumor fraction-guided cell-free DNA profiling in metastatic solid tumor patients.

BACKGROUND: Cell-free DNA (cfDNA) profiling is increasingly used to guide cancer care, yet mutations are not always identified. The ability to detect somatic mutations in plasma depends on both assay sensitivity and the fraction of circulating DNA in plasma that is tumor-derived (i.e., cfDNA tumor fraction). We hypothesized that cfDNA tumor fraction could inform the interpretation of negative cfDNA results and guide the choice of subsequent assays of greater genomic breadth or depth. METHODS: Plasma samples collected from 118 metastatic cancer patients were analyzed with cf-IMPACT, a modified version of the FDA-authorized MSK-IMPACT tumor test that can detect genomic alterations in 410 cancer-associated genes. Shallow whole genome sequencing (sWGS) was also performed in the same samples to estimate cfDNA tumor fraction based on genome-wide copy number alterations using z-score statistics. Plasma samples with no somatic alterations detected by cf-IMPACT were triaged based on sWGS-estimated tumor fraction for analysis with either a less comprehensive but more sensitive assay (MSK-ACCESS) or broader whole exome sequencing (WES). RESULTS: cfDNA profiling using cf-IMPACT identified somatic mutations in 55/76 (72%) patients for whom MSK-IMPACT tumor profiling data were available. A significantly higher concordance of mutational profiles and tumor mutational burden (TMB) was observed between plasma and tumor profiling for plasma samples with a high tumor fraction (z-score≥5). In the 42 patients from whom tumor data was not available, cf-IMPACT identified mutations in 16/42 (38%). In total, cf-IMPACT analysis of plasma revealed mutations in 71/118 (60%) patients, with clinically actionable alterations identified in 30 (25%), including therapeutic targets of FDA-approved drugs. Of the 47 samples without alterations detected and low tumor fraction (z-score<5), 29 had sufficient material to be re-analyzed using a less comprehensive but more sensitive assay, MSK-ACCESS, which revealed somatic mutations in 14/29 (48%). Conversely, 5 patients without alterations detected by cf-IMPACT and with high tumor fraction (z-score≥5) were analyzed by WES, which identified mutational signatures and alterations in potential oncogenic drivers not covered by the cf-IMPACT panel. Overall, we identified mutations in 90/118 (76%) patients in the entire cohort using the three complementary plasma profiling approaches. CONCLUSIONS: cfDNA tumor fraction can inform the interpretation of negative cfDNA results and guide the selection of subsequent sequencing platforms that are most likely to identify clinically-relevant genomic alterations.

Development of Genome-Derived Tumor Type Prediction to Inform Clinical Cancer Care.

IMPORTANCE: Diagnosing the site of origin for cancer is a pillar of disease classification that has directed clinical care for more than a century. Even in an era of precision oncologic practice, in which treatment is increasingly informed by the presence or absence of mutant genes responsible for cancer growth and progression, tumor origin remains a critical factor in tumor biologic characteristics and therapeutic sensitivity. OBJECTIVE: To evaluate whether data derived from routine clinical DNA sequencing of tumors could complement conventional approaches to enable improved diagnostic accuracy. DESIGN, SETTING, AND PARTICIPANTS: A machine learning approach was developed to predict tumor type from targeted panel DNA sequence data obtained at the point of care, incorporating both discrete molecular alterations and inferred features such as mutational signatures. This algorithm was trained on 7791 tumors representing 22 cancer types selected from a prospectively sequenced cohort of patients with advanced cancer. RESULTS: The correct tumor type was predicted for 5748 of the 7791 patients (73.8%) in the training set as well as 8623 of 11 644 patients (74.1%) in an independent cohort. Predictions were assigned probabilities that reflected empirical accuracy, with 3388 cases (43.5%) representing high-confidence predictions (>95% probability). Informative molecular features and feature categories varied widely by tumor type. Genomic analysis of plasma cell-free DNA yielded accurate predictions in 45 of 60 cases (75.0%), suggesting that this approach may be applied in diverse clinical settings including as an adjunct to cancer screening. Likely tissues of origin were predicted from targeted tumor sequencing in 95 of 141 patients (67.4%) with cancers of unknown primary site. Applying this method prospectively to patients under active care enabled genome-directed reassessment of diagnosis in 2 patients initially presumed to have metastatic breast cancer, leading to the selection of more appropriate treatments, which elicited clinical responses. CONCLUSIONS AND RELEVANCE: These results suggest that the application of artificial intelligence to predict tissue of origin in oncologic practice can act as a useful complement to conventional histologic review to provide integrated pathologic diagnoses, often with important therapeutic implications.

A rectal cancer organoid platform to study individual responses to chemoradiation.

Rectal cancer (RC) is a challenging disease to treat that requires chemotherapy, radiation and surgery to optimize outcomes for individual patients. No accurate model of RC exists to answer fundamental research questions relevant to patients. We established a biorepository of 65 patient-derived RC organoid cultures (tumoroids) from patients with primary, metastatic or recurrent disease. RC tumoroids retained molecular features of the tumors from which they were derived, and their ex vivo responses to clinically relevant chemotherapy and radiation treatment correlated with the clinical responses noted in individual patients' tumors. Upon engraftment into murine rectal mucosa, human RC tumoroids gave rise to invasive RC followed by metastasis to lung and liver. Importantly, engrafted tumors displayed the heterogenous sensitivity to chemotherapy observed clinically. Thus, the biology and drug sensitivity of RC clinical isolates can be efficiently interrogated using an organoid-based, ex vivo platform coupled with in vivo endoluminal propagation in animals.

Multicenter Prospective Phase II Trial of Neoadjuvant Dose-Dense Gemcitabine Plus Cisplatin in Patients With Muscle-Invasive Bladder Cancer.

Purpose Neoadjuvant chemotherapy followed by radical cystectomy (RC) is a standard of care for the management of muscle-invasive bladder cancer (MIBC). Dose-dense cisplatin-based regimens have yielded favorable outcomes compared with standard-dose chemotherapy, yet the optimal neoadjuvant regimen remains undefined. We assessed the efficacy and tolerability of six cycles of neoadjuvant dose-dense gemcitabine and cisplatin (ddGC) in patients with MIBC. Patients and Methods In this prospective, multicenter phase II study, patients received ddGC (gemcitabine 2,500 mg/m2 on day 1 and cisplatin 35 mg/m2 on days 1 and 2) every 2 weeks for 6 cycles followed by RC. The primary end point was pathologic downstaging to non-muscle-invasive disease (< pT2N0). Patients who did not undergo RC were deemed nonresponders. Pretreatment tumors underwent next-generation sequencing to identify predictors of chemosensitivity. Results Forty-nine patients were enrolled from three institutions. The primary end point was met, with 57% of 46 evaluable patients downstaged to < pT2N0. Pathologic response correlated with improved recurrence-free survival and overall survival. Nineteen patients (39%) required toxicity-related dose modifications. Sixty-seven percent of patients completed all six planned cycles. No patient failed to undergo RC as a result of chemotherapy-associated toxicities. The most frequent treatment-related toxicity was anemia (12%; grade 3). The presence of a presumed deleterious DNA damage response (DDR) gene alteration was associated with chemosensitivity (positive predictive value for < pT2N0 [89%]). No patient with a deleterious DDR gene alteration has experienced recurrence at a median follow-up of 2 years. Conclusion Six cycles of ddGC is an active, well-tolerated neoadjuvant regimen for the treatment of patients with MIBC. The presence of a putative deleterious DDR gene alteration in pretreatment tumor tissue strongly predicted for chemosensitivity, durable response, and superior long-term survival.

Cytokine profile in Egyptian hepatitis C virus genotype-4 in relation to liver disease progression.

AIM: To observe the imbalance between T helper cell Th1 and Th2 cytokines in several chronic hepatitis disease at different stages of disease progression. METHODS: We measured the cytokine levels of Th1 (IL-2 and IL-2R), Th2 (IL-10) and the pro-inflammatory cytokines (IL-6 and IL-6R and TNF and TNF-RI and II) by the ELISA technique in the sera of 33 hepatocellular carcinoma (HCC) patients and 20 chronic liver disease (CLD) patients. In addition, 20 asymptomatic hepatitis C virus carriers and 20 healthy subjects negative for hepatitis C virus(HCV) markers served as controls. RESULTS: Anti-HCV antibodies were found to be positive in 94% of HCC cases and 75% of CLD cases. On the other hand, HCV viremia was detected using RT-PCR in 67% of HCC cases and 65% of CLD cases. HBsAg was positive in 9% of HCC cases and 30% of CLD cases. Also bilharzial-Ab was positive in 55% of HCC cases, 65% of CLD cases and in 70% of asymptomatic carriers (ASC). HCC patients had significantly higher values of IL-2R, TNF-RII (P<0.001), and TNF-RI (P>0.05), but lower TNFalpha (P<0.001) and IL-6 (P = 0.032) in comparison to ASC. But, in comparison to non-cancer controls, HCC patients had higher values of IL-2R, IL-6R, TNF-RI and TNF-RII, but lower TNF-alpha (P<0.001). CLD patients had higher IL-2R, TNF-RI, and TNF-RII (P<0.001) than ASC. But, in comparison to non-cancer controls, CLD patients had higher values of IL-2R, TNF-RI and TNF-RII, but lower TNF-alpha (P<0.001). IL-10 was higher (though not significantly) in HCC and CLD patients than in symptomatic carriers and non-cancer controls. CONCLUSION: Liver disease progression from CLD to HCC due to HCV genotype-4 infection is associated with an imbalance between Th1 and Th2 cytokines. IL-2R, TNF-RI, and TNF-RII could be used as potential markers.