Purpose While multi‐parametric magnetic resonance imaging (MRI) shows great promise in assisting with prostate cancer diagnosis and localization, subtle differences in appearance between cancer and normal tissue lead to many false positive and false negative interpretations by radiologists. We sought to automatically detect aggressive cancer (Gleason pattern ≥ 4) and indolent cancer (Gleason pattern 3) on a per‐pixel basis on MRI to facilitate the targeting of aggressive cancer during biopsy. Methods We created the Stanford Prostate Cancer Network (SPCNet), a convolutional neural network model, trained to distinguish between aggressive cancer, indolent cancer, and normal tissue on MRI. Ground truth cancer labels were obtained by registering MRI with whole‐mount digital histopathology images from patients who underwent radical prostatectomy. Before registration, these histopathology images were automatically annotated to show Gleason patterns on a per‐pixel basis. The model was trained on data from 78 patients who underwent radical prostatectomy and 24 patients without prostate cancer. The model was evaluated on a pixel and lesion level in 322 patients, including six patients with normal MRI and no cancer, 23 patients who underwent radical prostatectomy, and 293 patients who underwent biopsy. Moreover, we assessed the ability of our model to detect clinically significant cancer (lesions with an aggressive component) and compared it to the performance of radiologists. Results Our model detected clinically significant lesions with an area under the receiver operator characteristics curve of 0.75 for radical prostatectomy patients and 0.80 for biopsy patients. Moreover, the model detected up to 18% of lesions missed by radiologists, and overall had a sensitivity and specificity that approached that of radiologists in detecting clinically significant cancer. Conclusions Our SPCNet model accurately detected aggressive prostate cancer. Its performance approached that of radiologists, and it helped identify lesions otherwise missed by radiologists. Our model has the potential to assist physicians in specifically targeting the aggressive component of prostate cancers during biopsy or focal treatment.

Purpose: Magnetic resonance imaging (MRI) has great potential to improve prostate cancer diagnosis; however, subtle differences between cancer and confounding conditions render prostate MRI interpretation challenging. The tissue collected from patients who undergo radical prostatectomy provides a unique opportunity to correlate histopathology images of the prostate with preoperative MRI to accurately map the extent of cancer from histopathology images onto MRI. We seek to develop an open-source, easy-to-use platform to align presurgical MRI and histopathology images of resected prostates in patients who underwent radical prostatectomy to create accurate cancer labels on MRI. Methods: Here, we introduce RAdiology Pathology Spatial Open-Source multi-Dimensional Integration (RAPSODI), the first open-source framework for the registration of radiology and pathology images. RAPSODI relies on three steps. First, it creates a three-dimensional (3D) reconstruction of the histopathology specimen as a digital representation of the tissue before gross sectioning. Second, RAPSODI registers corresponding histopathology and MRI slices. Third, the optimized transforms are applied to the cancer regions outlined on the histopathology images to project those labels onto the preoperative MRI. Results: We tested RAPSODI in a phantom study where we simulated various conditions, for example, tissue shrinkage during fixation. Our experiments showed that RAPSODI can reliably correct multiple artifacts. We also evaluated RAPSODI in 157 patients from three institutions that underwent radical prostatectomy and have very different pathology processing and scanning. RAPSODI was evaluated in 907 corresponding histpathology-MRI slices and achieved a Dice coefficient of 0.97 AE 0.01 for the prostate, a Hausdorff distance of 1.99 AE 0.70 mm for the prostate boundary, a urethra deviation of 3.09 AE 1.45 mm, and a landmark deviation of 2.80 AE 0.59 mm between registered histopathology images and MRI. Conclusion: Our robust framework successfully mapped the extent of cancer from histopathology slices onto MRI providing labels from training machine learning methods to detect cancer on MRI.

Cancer genotyping has identified a large number of putative tumor suppressor genes. Carcinogenesis is a multistep process, but the importance and specific roles of many of these genes during tumor initiation, growth, and progression remain unknown. Here we use a multiplexed mouse model of oncogenic KRAS–driven lung cancer to quantify the impact of 48 known and putative tumor suppressor genes on diverse aspects of carcinogenesis at an unprecedented scale and resolution. We uncover many previously understudied functional tumor suppressors that constrain cancer in vivo. Inactivation of some genes substantially increased growth, whereas the inactivation of others increases tumor initiation and/or the emergence of exceptionally large tumors. These functional in vivo analyses revealed an unexpectedly complex landscape of tumor suppression that has implications for understanding cancer evolution, interpreting clinical cancer genome sequencing data, and directing approaches to limit tumor initiation and progression. Significance: Our high-throughput and high-resolution analysis of tumor suppression uncovered novel genetic determinants of oncogenic KRAS–driven lung cancer initiation, overall growth, and exceptional growth. This taxonomy is consistent with changing constraints during the life history of cancer and highlights the value of quantitative in vivo genetic analyses in autochthonous cancer models. This article is highlighted in the In This Issue feature, p. 1601

Compelling evidence from animal studies has demonstrated that allospecific FoxP3+CD4+ regulatory T (Treg) cells expanded ex vivo can be used as effective therapeutic tools in the treatment of allograft rejection and graft-vs-host disease. Despite the promising results from animal studies, there remain major barriers to developing Treg cell-based immunotherapy in humans. Currently, no effective approach has been established for selective expansion of human allospecific Treg cells ex vivo. Additionally, the very low frequency of Treg cells present in human peripheral blood could pose a formidable challenge to obtaining a sufficient number of Treg cells from a single donor for ex vivo expansion for therapeutic utilization. Extending our recent finding that mouse B cells preferentially induce expansion of alloreactive Treg cells, we report herein that human Treg cells can be expanded ex vivo with allogeneic B cells. The expanded Treg cells express very high levels of FoxP3, maintain anergic phenotype, and are potent suppressors capable of inhibiting the alloproliferation of third-party responder T cells at very low Treg-to-T effector cell ratio in an alloantigen-specific manner. The alloantigen specificity demonstrated by B cell-expanded Treg cells is not determined by the HLA haplotypes of the Treg cells, but it is induced and determined by the haplotype of the B cells used to expand them. Our findings represent a significant advance in the development of Treg cell-based immunotherapy in humans and raise the possibility of using third-party Treg cells for therapeutic applications.

Cancer survivors often relapse due to evolving drug-resistant clones and repopulating tumor stem cells. Our preclinical study demonstrated that terminal cancer patient’s lymphocytes can be converted from tolerant bystanders in vivo into effective cytotoxic T-lymphocytes in vitro killing patient’s own tumor cells containing drug-resistant clones and tumor stem cells. We designed a clinical trial combining peginterferon α-2b with imatinib for treatment of stage III/IV gastrointestinal stromal tumor (GIST) with the rational that peginterferon α-2b serves as danger signals to promote antitumor immunity while imatinib’s effective tumor killing undermines tumor-induced tolerance and supply tumor-specific antigens in vivo without leukopenia, thus allowing for proper dendritic cell and cytotoxic T-lymphocyte differentiation toward Th1 response. Interim analysis of eight patients demonstrated significant induction of IFN-γ-producing-CD8+, -CD4+, -NK cell, and IFN-γ-producing-tumor-infiltrating-lymphocytes, signifying significant Th1 response and NK cell activation. After a median follow-up of 3.6 years, complete response (CR) + partial response (PR) = 100%, overall survival = 100%, one patient died of unrelated illness while in remission, six of seven evaluable patients are either in continuing PR/CR (5 patients) or have progression-free survival (PFS, 1 patient) exceeding the upper limit of the 95% confidence level of the genotype-specific-PFS of the phase III imatinib-monotherapy (CALGB150105/SWOGS0033), demonstrating highly promising clinical outcomes. The current trial is closed in preparation for a larger future trial. We conclude that combination of targeted therapy and immunotherapy is safe and induced significant Th1 response and NK cell activation and demonstrated highly promising clinical efficacy in GIST, thus warranting development in other tumor types.