Citations

Abstract Globe-sparing surgery for orbit-involving sinonasal neoplasms (OISNs) is increasingly an option since the introduction of effective neoadjuvant treatments. Even if vision is preserved, loss of orbital support structures can cause major ocular sequelae. This study analyzed long-term ophthalmic outcomes following globe-sparing maxillectomy (GSM). We retrospectively reviewed all cases of GSM for OISNs with at least 1 year of ophthalmology follow-up at a tertiary care multisite institution from 2005 to 2024. Patient demographics, surgical details, ophthalmic findings, and final oncologic status were recorded. Orbital involvement was graded based on preoperative computed tomography imaging, and outcomes were compared across orbital involvement grades. Thirty-three patients (39% female) were included, with a mean age of 56 ± 18 years and a median follow-up of 2 years (range 1–12 years). Eleven patients (33%) had Grade I, 19 (58%) had Grade II, and 3 (9%) had Grade III involvement. With higher grades of orbital involvement, patients had a higher likelihood of receiving a preoperative ophthalmic exam (p < 0.001) and experiencing preoperative globe malposition (p = 0.015). Patients with higher-grade orbital involvement had a greater frequency of most ophthalmic symptoms postoperatively. Final best visual acuity was 20/40 or better in 82% of Grade I patients and 71% of Grade II/III patients (p = 0.681). GSM for OISNs offers a favorable visual prognosis. Ocular structural and functional impairment were observed at high frequency after surgery, even with lower-grade orbital involvement. Establishing ophthalmic care for all OISN patients is critical to adequately manage ocular sequelae.

Abstract To quantify lateral extension over the optic canal (OC) through endoscopic endonasal approaches (EEAs) and present techniques to maximize lateral extension. On 20 sides of 10 specimens, transplanum EEAs were performed. The OC roof was drilled maximally laterally. The lateral anterior cranial fossa (ACF) dura accessed was marked; its distance from lateral OC was measured at the proximal and distal OC. To enhance lateral drilling, we performed a posteromedial orbital decompression (POD)—removal of the posteromedial lamina papyracea and orbital apex bone—and continued lateral drilling. New lateral limits were marked and measured. At proximal OC, the transplanum approach accessed the full width of ACF dura overlying OC in 55% of sides; POD addition provided access in 100% of sides (p < 0.01). At proximal OC, the transplanum approach provided access to ACF dura an average of 0.7 mm lateral to OC; POD addition provided an average access of 6.0 mm lateral to OC (p < 0.01). At distal OC, the transplanum approach accessed the full width of ACF dura overlying OC in 30% of sides dissected; POD addition afforded access in 100% of sides (p < 0.01). At distal OC, the transplanum approach provided access to ACF dura an average of 1.1 mm medial to OC; adjunct POD provided an average access of 6.4 mm lateral to OC (p < 0.01). A transplanum approach affords inconsistent access to the full width of ACF dura overlying OC; adjunct POD consistently affords access to ACF dura lateral to OC and significantly increases lateral reach.

Background : Diabetic-foot osteomyelitis (DFO) is a severe complication of diabetic-foot ulcers associated with delayed healing, amputation, and increased mortality. Machine learning (ML) has increasingly been explored for diagnosis and clinical decision-making in diabetic-foot care; however, the current evidence specific to DFO has not been systematically synthesized. Methods : We performed systematic searches of MEDLINE, Embase, Scopus, and Web of Science for studies published between January 2000 and May 2026. Studies that developed or evaluated ML models for DFO diagnosis, prediction, or outcomes were included. Data extraction was performed using the CHARMS framework, reporting followed PRISMA 2020 and TRIPOD-SRMA recommendations, and methodological quality was evaluated using PROBAST + AI. Due to considerable heterogeneity, narrative synthesis was performed. Results : A total of 10 studies published from 2022 to 2026 met the inclusion criteria. Five studies developed or evaluated ML models with a direct DFO diagnostic endpoint, one additional study evaluated model explainability, and four evaluated complementary DFO-relevant diabetic-foot applications. Reported discrimination ranged from 0.879 to 0.998, but comparisons were limited by differences in clinical endpoints, reference standards, input modalities and validation strategies. Only one direct DFO model was externally validated, and no study developed a multimodal model combining clinical and imaging data. Conclusion : Current evidence is limited by heterogeneous methodologies, limited external validation, and inconsistent reporting despite promising discrimination reported by several models. Future studies should evaluate whether multimodal models integrating clinical and imaging data provide incremental value over unimodal approaches, with external validation, calibration, and clinical utility assessment before clinical implementation.

Abstract [18F]FFNP is a promising investigational radiopharmaceutical for PET imaging of progesterone receptor (PR) expression in breast cancer. [18F]FFNP PET imaging enables noninvasive assessment of PR expression and serves as a valuable diagnostic tool for determining the functional endocrine status of ER+ breast tumors to guide endocrine therapy. In this study, we report a robust and fully automated radiosynthesis of [18F]FFNP using a commercially available radiosynthesizer. The ratio of the mesylate precursor to the K2.2.2/K2CO3 complex was found to be critical for optimizing radiochemical yield. The optimized and automated radiosynthesis process was completed within 50 min and successfully passed three consecutive validation runs, producing [18F]FFNP activity in 19.52–23.91% decay-corrected radiochemical yields at the end of synthesis. Approximately 215–265 mCi (7.96–9.81 GBq) of [18F]FFNP activity was obtained with high radiochemical purity (99.65–99.72%) and molar activity (16.83–37.40 Ci/μmol [623–1384 GBq/μmol]) in our cGMP-compliant PET Radiochemistry facility. All drug product specifications passed quality control (QC) testing and met the requirements for clinical investigational use.

ABSTRACT Background and Objectives Among patient with early‐stage breast cancer, breast conservation surgery (BCS) is the most performed surgery, utilized in approximately two‐thirds of patients. While 5‐year overall‐survival rates are strong—typically above 90%—additional re‐excisions can occur in up to 25% of patients who undergo index BCS and some patients will require subsequent surgery for new breast cancer or recurrences. As such, these additional procedures can complicate patients’ pathway to survivorship, negatively impact quality of life, and lead to excess healthcare costs. Methods This retrospective cohort study used linked claims and an electronic healthcare record database to (1) characterize patients at higher risk for additional breast oncologic surgery (aBOS) following BCS, and (2) quantify the long‐term financial impact of aBOS after BCS. In total, 87,190 patients with breast cancer, who had an index BCS between January 1, 2012, to March 31, 2023, were followed for up to 5‐years post‐operatively. Patients with and without aBOS were analyzed, comparing demographics, tumor characteristics, healthcare utilization, and per‐person‐per‐month (PPPM) costs. Results Twenty‐three percent ( n = 20,212) of patients underwent aBOS. These patients were younger, had higher tumor grades and larger tumor sizes, and incurred greater healthcare costs compared to patients who did not undergo a secondary surgery. Conclusion This study highlights significant factors influencing BCS outcomes and healthcare utilization, underscoring the need for tailored therapeutic strategies for patients with breast cancer.

Purpose This study aimed to develop and validate a tool to predict the primary tumor site in patients with metastatic cancer of unknown primary using demographic and clinical variables, including imaging-derived sites of metastasis, without pathology or laboratory data. Methods In this population-based study, data were extracted from the SEER database. Patients diagnosed between 2010 and 2020 with a known primary cancer site, metastatic staging, and metastasis in the bone, liver, brain, and lungs were included. Patients were randomly split into training (70%) and/or testing (30%) sets. A total of 27 primary tumor sites were included. A gradient boosting model was trained to predict the primary tumor site, with random forest and classification tree models evaluated as comparators. Random guess and frequency-weighted guessing models were used for baseline comparison. Results A total of 689,715 patients were included from 9,208,295 queried. The most common primary cancer sites were non–small cell lung cancer (NSCLC) (n = 211,338), intestines (n = 64,187), and pancreas (n = 60,310). The baseline models achieved Top-5 accuracies of 0.179 for a random guess and 0.476 for a frequency-weighted guess. The gradient boosting model achieved Top-1, Top-3, and Top-5 accuracies of 0.40 ± 0.001, 0.680 ± 0.001, and 0.790 ± 0.001, respectively, across 100 Monte-Carlo replicates. Top-5 accuracy was highest for primary NSCLC (0.990 ± 0.001) and intestinal cancer (0.900 ± 0.003). Our model was most influenced by sex (gain = 0.27) and metastasis to bone (gain = 0.20). A web app was constructed: primarycancerpredictor.netlify.app. Conclusion This study developed and evaluated a gradient boosting machine learning model and a user-friendly web tool to aid in the diagnosis of cancer of unknown primary.

Introduction The purpose of this study is to evaluate the agreement, repeatability, and testing efficiency of Melbourne Rapid Fields-Web (MRF-Web), a browser-based, device-independent perimeter, compared with the Octopus 600 perimeter in glaucoma patients. Methods This retrospective, cross-sectional, single-center observational study included 113 patients undergoing routine visual field assessment for glaucoma or glaucoma suspicion at Marengo Asia Hospitals, Gurgaon, India. MRF-Web testing was performed on a calibrated 14-inch laptop with Hindi voice guidance. Agreement between MRF-Web and Octopus 600 for Mean Deviation (MD) and pattern-based global indices was assessed using intraclass correlation coefficients (ICCs) and Bland–Altman analysis. Test-retest repeatability was evaluated in 51 patients who underwent repeat MRF-Web testing. Results MRF-Web showed strong group-level concordance with Octopus 600 for MD, with an ICC of 0.88 (95% CI, 0.81–0.92), Bland–Altman bias of -0.78 dB, and 95% limits of agreement from -4.72 to 3.17 dB. Concordance for the pattern-based indices was moderate, with an ICC of 0.72 (95% CI, 0.58–0.81), bias of -0.59 dB, and 95% limits of agreement from -4.41 to 3.24 dB. MRF-Web demonstrated good test-retest concordance, with an ICC of 0.86 (95% CI, 0.69–0.94). Mean per-eye test duration was significantly shorter with MRF-Web than with Octopus 600 (232.9 ± 49.8 versus 342.4 ± 94.4 seconds). Conclusions MRF-Web demonstrated strong group-level concordance with Octopus 600 MD measurements and shorter per-eye testing times in this predominantly mild-disease cohort. The limits of agreement indicate that measurements from the two platforms should not be regarded as directly interchangeable, and further study is required in moderate and advanced glaucoma.

PURPOSE Patients with breast cancer requiring postmastectomy radiation therapy (PMRT) often desire breast reconstruction. Reconstructions can be adversely affected by PMRT, and fear of reconstruction complications has discouraged adoption of short-course radiation therapy used after lumpectomy. To fill the evidence gap, we conducted a noninferiority trial of hypofractionated versus conventional PMRT in patients undergoing breast reconstruction. The primary outcome was reconstruction complication rate, a composite of wound healing, readmission, capsular contracture, unplanned reoperation, and reconstruction failure. METHODS Women with unilateral invasive breast cancer with planned immediate or delayed reconstruction and PMRT were randomly assigned (1:1) to 50 Gy (25 fractions/5 weeks) or 42.56 Gy (16 fractions/3 weeks). This study was designed with a noninferiority margin of 10%, assuming a 2-year reconstruction complication rate of 25% with conventional PMRT. Randomization was stratified by immediate versus delayed and autologous versus implant-only reconstruction. RESULTS Between 2018 and 2021, 898 patients enrolled with a median follow-up of 3.9 years; 825 patients were evaluable for the primary end point. The 24-month reconstruction complication rates were 12.3% (52/422) with hypofractionated PMRT and 10.2% (41/403) with conventional PMRT, corresponding to an estimated difference of -2.1% (95% CI, -6.5 to 2.2). The complication rate was higher (regardless of arm) with implant-only versus autologous reconstruction (odds ratio = 1.65). There was no statistically significant difference in 36-month locoregional recurrence rates: 1.5% (95% CI, 0.7 to 3.2) with hypofractionated and 2.2% (95% CI, 1.1 to 4.3) with conventional PMRT. Acute and late toxicity rates were not statistically different. CONCLUSION A 16-fraction course of hypofractionated PMRT is safe and noninferior to conventional PMRT in women undergoing breast reconstruction after mastectomy.

In recent years, lipid nanoparticles (LNPs) have shown great promise as highly flexible delivery systems for cancer vaccines, which enable effective cargo delivery and allow for controlling their distribution, cell-targeting ability and immunogenicity. Innovations in ionizable lipid chemistry, biodegradable formulations, RNA engineering and LNP design resulted in improvements in RNA encapsulation, endosomal escape, delivery outside the liver, and targeting of lymphoid organs and antigen-presenting cells. All these achievements led to an expansion of strategies for developing cancer vaccines from using only mRNA to personal neoantigen vaccines, self-amplifying RNA (saRNA), circular RNA (circRNA), peptide/protein vaccines, and multi-purpose platforms that combine antigen delivery and modulation of the immune system. LNPs are also studied as tools for altering the tumor microenvironment (TME) and boosting anti-tumor immunity through local delivery of cytokines and innate immune modulators. Preliminary clinical results with personalized mRNA neoantigen vaccines show strong potential for this approach. However, inflammatory toxicity, suitability for repeated dosing, off-target delivery, difficulties with large-scale production, formulation stability, personalized production schedule and regulatory issues are still significant challenges. New biodegradable and PEG-free formulations, innovations in manufacturing and lyophilization technology, and AI-based approaches in lipid screening could help overcome some of these problems.

Background Bioprosthetic aortic valve replacement, specifically surgical (SAVR) and transcatheter (TAVR), is among the most frequently performed cardiac interventions globally, yet anticoagulation strategies after implantation remain incompletely defined. Atrial fibrillation, notably prevalent pre- and post-operatively in these patients, substantially amplifies thromboembolic risk and complicates decision-making in an already elderly, frail, and comorbid population. Objectives To critically evaluate current anticoagulation strategies after bioprosthetic aortic valve replacement across sinus rhythm and atrial fibrillation in both SAVR and TAVR, and to identify evidence gaps and guideline inconsistencies. Methods We conducted a narrative review of major international guidelines (ESC/EACTS 2025, ACC/AHA 2020, and NICE 2021) and key clinical trials on antithrombotic therapy in valve replacement populations. Focus was placed on bioprosthetic aortic valve replacement, with mechanical and non-aortic valves excluded. Results Aspirin monotherapy is the most consistently supported strategy after bioprosthetic TAVR in sinus rhythm, though guidelines diverge on duration and recommendation strength. In atrial fibrillation, all three guidelines apply uniform CHA 2 DS 2 -VASc-based frameworks irrespective of AF phenotype, procedural context, or detection modality, reflecting lacking evidence rather than deliberate clinical judgement. No trial has enrolled patients specifically with new-onset post-procedural AF or stratified by AF subtype, and subclinical AF after bioprosthetic AVR remains entirely uncharacterised. Conclusions Guideline divergence reflects a structurally inadequate evidence base rather than genuine clinical equipoise. Individualised decisions informed by procedure type, AF phenotype, frailty, and bleeding risk are required to guide clinical practice, alongside an adequately powered stratified randomised trial incorporating mandatory continuous rhythm monitoring.