Osteosarcoma (OSA) is a rare, but disproportionately lethal cancer that predominantly affects children. Sadly, discovery of new therapies for OSA has largely been unsuccessful in the past 30 years; there is an urgent need to identify new treatments for OSA. Pet dogs with naturally-occurring OSA represent a unique comparative “model” to discover new treatments for OSA. Unlike humans, in which fewer than 1,000 cases of OSA occur each year, there are nearly 50,000 new cases each year of OSA in dogs. In addition, dogs have an intact immune system, a shared environment with humans, and more rapid progression of disease. Together these factors make dogs an important comparative model for new therapies for OSA. The purpose of this study was: 1) to validate this mouse-dog-human pipeline for drug discovery and 2) to validate CRM1 as a novel target for ostesoarcoma treatment. We developed patient-derived cell lines and xenografts of OSA from both dogs and humans and applied these models to identify new therapies for OSA using high-throughput drug screens in vitro followed by in vivo validation. Whole exome sequencing was performed on the patient-derived models and original tumors to identify potential driver mutations. A high-throughput screen in both dog and human OSA identified CRM1 inhibitors as effective at killing dog and human OSA patient-derived cell lines in vitro. In vivo, CRM1 inhibition led to significant tumor growth inhibition in patient-derived xenografts from dogs and humans. Western blotting demonstrated increased levels of CRM1 protein expression across nine different dog and human OSA cell lines compared to non-transformed human osteoblasts. CRM1 upregulation in OSA cells was further verified by immunofluorescence staining. Increased CRM1 expression was prognostic for poorer metastasis-free survival and poorer overall survival. Our cross-species personalized medicine pipeline identified CRM1 as a potential therapeutic target to treat OSA in both dogs and humans. Future studies are focused on testing CRM1 inhibitors in canine clinical trials.
For soft tissue sarcoma patients receiving preoperative radiation therapy, wound complications are common and potentially devastating; they may result in multiple subsequent surgeries and significant patient morbidity. The purpose of this study was to assess the feasibility of intraoperative indocyanine green fluorescent angiography (ICGA) as a predictor of wound complications in resections of irradiated soft tissue sarcoma of the extremities. A consecutive series of patients of patients with soft tissue sarcoma of the extremities or pelvis who received neoadjuvant radiation and a subsequent radical resection received intraoperative ICGA with the SPY PHI device (Stryker Inc, Kalamazoo MI) at the time of closure. Three fellowship trained Orthopaedic Oncologic Surgeons were asked to prospectively predict likelihood of wound complications based on fluorescence. Retrospective analysis of fluorescence signal along multiple points of the wound length was performed and quantified. The primary endpoint was wound complication, defined as delayed wound healing or wound dehiscence, within 3 months of surgery. An a priori power analysis demonstrated that 5 patients were necessary to achieve statistical significance. Univariate and multivariate statistical analyses were performed to identify predictors of wound complications. 14 patients were consecutively imaged. The diagnosis was undifferentiated pleomorphic sarcoma in 9 (64.3%) of patients; 11 (78.6%) tumors were high grade. There were 6 patients with wound complications classified as “aseptic” in 5 cases and secondary to hematoma in 1 case. Using the ICGA, blinded surgeons correctly predicted wound complications in 75% of cases. In the area of wound complication, the mean % of maximal signal for wound complications was 49% during the inflow phase and 48% during the peak phase. The mean % maximal signal for peri-incisional tissue without wound complications was 77% during the inflow phase and 83% during the peak phase (p=0.003 and p<0.001). During the inflow phase, a mean ratio of normal of 0.62 maximized the area under the curve (AUC=0.90) for predicting wound complications with a sensitivity of 100% and specificity of 77.4%. During the peak phase, a mean ratio of normal of 0.55 maximized the area under the curve (AUC=0.95) for predicting wound complications with a sensitivity of 88.9% and a specificity 100%. Intraoperative use of indocyanine green fluorescent angiography may help to predict wound complications in patients undergoing resection of preoperatively irradiated soft tissue sarcomas of the extremities and pelvis. Future studies are necessary to validate this technology in a prospective manner and to determine if interventions can be instituted to prevent predicted wound complications.
