Joseph B. Shrager’s research while affiliated with Stanford University and other places

Stem cell therapy is a promising approach for tissue regeneration after traumatic injury, yet current applications are limited by inadequate control over the fate of stem cells after transplantation. Here we introduce a bioconstruct engineered for the staged release of growth factors, tailored to direct different phases of muscle regeneration. The bioconstruct is composed of a decellularized extracellular matrix containing polymeric nanocapsules sequentially releasing basic fibroblast growth factor and insulin-like growth factor 1, which promote the proliferation and differentiation of muscle stem cells, respectively. When applied to a volumetric muscle loss defect in an animal model, the bioconstruct enhances myofibre formation, angiogenesis, innervation and functional restoration. Further, it promotes functional muscle formation with human or aged murine muscle stem cells, highlighting the translational potential of this bioconstruct. Overall, these results highlight the potential of bioconstructs with orchestrated growth factor release for stem cell therapies in traumatic injury.

Background: We evaluate SAM 2 for surgical scene understanding by examining its semantic segmentation capabilities for organs/tissues both in zero-shot scenarios and after fine-tuning. Methods: We utilized five public datasets to evaluate and fine-tune SAM 2 for segmenting anatomical tissues in surgical videos/images. Fine-tuning was applied to the image encoder and mask decoder. We limited training subsets from 50 to 400 samples per class to better model real-world constraints with data acquisition. The impact of dataset size on fine-tuning performance was evaluated with weighted mean Dice coefficient (WMDC), and the results were also compared against previously reported state-of-the-art (SOTA) results. Results: SurgiSAM 2, a fine-tuned SAM 2 model, demonstrated significant improvements in segmentation performance, achieving a 17.9% relative WMDC gain compared to the baseline SAM 2. Increasing prompt points from 1 to 10 and training data scale from 50/class to 400/class enhanced performance; the best WMDC of 0.92 on the validation subset was achieved with 10 prompt points and 400 samples per class. On the test subset, this model outperformed prior SOTA methods in 24/30 (80%) of the classes with a WMDC of 0.91 using 10-point prompts. Notably, SurgiSAM 2 generalized effectively to unseen organ classes, achieving SOTA on 7/9 (77.8%) of them. Conclusion: SAM 2 achieves remarkable zero-shot and fine-tuned performance for surgical scene segmentation, surpassing prior SOTA models across several organ classes of diverse datasets. This suggests immense potential for enabling automated/semi-automated annotation pipelines, thereby decreasing the burden of annotations facilitating several surgical applications.

Introduction: Computer vision (CV) has had a transformative impact in biomedical fields such as radiology, dermatology, and pathology. Its real-world adoption in surgical applications, however, remains limited. We review the current state-of-the-art performance of deep learning (DL)-based CV models for segmentation and object detection of anatomical structures in videos obtained during surgical procedures. Methods: We conducted a scoping review of studies on semantic segmentation and object detection of anatomical structures published between 2014 and 2024 from 3 major databases - PubMed, Embase, and IEEE Xplore. The primary objective was to evaluate the state-of-the-art performance of semantic segmentation in surgical videos. Secondary objectives included examining DL models, progress toward clinical applications, and the specific challenges with segmentation of organs/tissues in surgical videos. Results: We identified 58 relevant published studies. These focused predominantly on procedures from general surgery [20(34.4%)], colorectal surgery [9(15.5%)], and neurosurgery [8(13.8%)]. Cholecystectomy [14(24.1%)] and low anterior rectal resection [5(8.6%)] were the most common procedures addressed. Semantic segmentation [47(81%)] was the primary CV task. U-Net [14(24.1%)] and DeepLab [13(22.4%)] were the most widely used models. Larger organs such as the liver (Dice score: 0.88) had higher accuracy compared to smaller structures such as nerves (Dice score: 0.49). Models demonstrated real-time inference potential ranging from 5-298 frames-per-second (fps). Conclusion: This review highlights the significant progress made in DL-based semantic segmentation for surgical videos with real-time applicability, particularly for larger organs. Addressing challenges with smaller structures, data availability, and generalizability remains crucial for future advancements.

Spatial omics is enabling unprecedented tissue characterization, but the ability to adequately compare spatial features across samples under different conditions is lacking. We propose a quantitative framework that catalogs significant, normalized, colocalizations between pairs of cell subpopulations, enabling comparisons among a variety of biological samples. We perform cell-pair colocalization analysis on multiplexed immunofluorescence images of assembloids constructed with lung adenocarcinoma (LUAD) organoids and cancer-associated fibroblasts derived from human tumors. Our data show that assembloids recapitulate human LUAD tumor-stroma spatial organization, justifying their use as a tool for investigating the spatial biology of human disease. Intriguingly, drug-perturbation studies identify drug-induced spatial rearrangements that also appear in treatment-naïve human tumor samples, suggesting potential directions for characterizing spatial (re)-organization related to drug resistance. Moreover, our work provides an opportunity to quantify spatial data across different samples, with the common goal of building catalogs of spatial features associated with disease processes and drug response.

Objective To test the hypothesis that patients undergoing minimally invasive pneumonectomy at high-volume minimally invasive lung surgery centers have improved survival compared with patients who undergo open pneumonectomy. Methods Patients from the National Cancer Database who underwent pneumonectomy for lung cancer between 2010 and 2020 were stratified into 3 groups according to surgical technique (open, minimally invasive, converted from minimally invasive to open). Institutions were categorized as low-, mid-, or high-volume minimally invasive lung surgery centers according to percentage of total anatomic lung resections performed by video-assisted or robotic-assisted thoracoscopic surgery. Outcomes were compared using Cox regression, Kaplan-Meier survival analysis, and propensity score matching. Results In total, 5750 patients from 850 facilities were included, with 4597 (79.9%) undergoing upfront open pneumonectomy. Among the 1153 attempted minimally invasive pneumonectomies, 364 (31.6%) required conversion to open pneumonectomy. Surgery at a non−high-volume center was associated with greater conversion risk (adjusted odds ratio, 4.16; P < .001), whereas neoadjuvant therapy was associated with lower risk (adjusted odds ratio, 0.60; P = .015). Similar 5-year overall survival was seen among the 3 groups (open 45.2%, minimally invasive 48.3%, converted 43.3%); however, conversion from minimally invasive to open pneumonectomy demonstrated a trend towards increased risk of death (hazard ratio, 1.16; P = .058). Conclusions Minimally invasive pneumonectomy for lung cancer had similar 5-year survival compared with open pneumonectomy. However, conversion from minimally invasive to open pneumonectomy showed a trend toward increased risk of death, and conversion rates were high irrespective of institutional minimally invasive lung surgery volume. Careful patient selection is necessary when considering minimally invasive pneumonectomy so that long-term outcomes are not compromised.

Objective Radiation after esophagectomy may cause conduit dysfunction with unclear oncologic benefits. We hypothesized that adjuvant chemoradiation does not improve survival over chemotherapy alone for patients with pathologic upstaging after primary surgery for cT1-2N0M0 esophageal adenocarcinoma. Methods The impact of adjuvant therapy after primary surgery for cT1-2N0M0 esophageal adenocarcinoma upstaged to pT3-4 or pN+ in the National Cancer Database (2004-2019) was evaluated with logistic regression, Kaplan–Meier analysis, and Cox modeling. Results A total of 574 patients met inclusion criteria, 300 (52.3%) who received adjuvant therapy (chemotherapy alone in 117 [39.0%], radiation alone in 15 [5.0%], chemoradiation in 168 [56.0%]) and 274 (47.7%) who did not. Adjuvant therapy was associated with improved 5-year survival (46.8% vs 32.7%, P < .001). In multivariate analysis controlling for age, year of diagnosis, Charlson Comorbidity Index, pT, pN, and positive margins, adjuvant chemotherapy was independently associated with improved survival (hazard ratio, 0.62, 95% CI, 0.46-0.84, P = .002); radiation use did not have a statistically significant association with survival (hazard ratio, 1.19, 95% CI, 0.86-1.63, P = .29). Among patients who received chemotherapy, independent predictors of also receiving radiotherapy included pathological T-upstaging (odds ratio, 2.01, 95% CI, 1.04-3.88, P = .037) and distance from facility less than 50 miles (odds ratio, 2.13, 95% CI, 1.05-4.33, P = .037). In univariate analysis of patients who received adjuvant therapy, chemotherapy alone was associated with significantly better 5-year survival compared with chemoradiation (54.2% vs 41.6%, P = .004). Landmark analyses at 3 and 6 months were consistent with the primary analysis. Conclusions Using radiation with chemotherapy as adjuvant therapy for patients upstaged after esophagectomy for cT1-2N0 esophageal adenocarcinoma is not associated with improved survival and should be considered only in select situations based on careful clinical evaluation.