Memorial Sloan Kettering Cancer Center

To develop a deep learning-based auto-navigation technique for free-breathing golden-angle radial MRI named RANGR (Respiratory Auto-Navigator for Golden-angle Radial free-breathing abdominal MRI). RANGR computes a one-dimensional (1D) respiratory motion signal from 1D projections along the superior-inferior dimension (z) extracted directly from the acquired golden-angle stack-of-stars k-space data. The motion signal is used to retrospectively sort k-space data into different undersampled motion states, which are reconstructed using Movienet, a neural network developed for dynamic reconstruction. RANGR was trained using PCA (Principal Component Analysis) as a reference in cases where PCA was successful. The performance of RANGR is evaluated against PCA using a dynamic phantom with programmable motion waveforms on a 1.5 T MR-Linac system and free-breathing imaging of patients with abdominal tumors. In vivo image quality was qualitatively scored by body radiologists. For phantom studies, RANGR can track motion with less than 1 mm error with respect to the ground-truth. For in vivo studies, RANGR generalizes to cases where PCA failed due to limited liver coverage and/or the presence of high intensity regions outside the liver dome. RANGR outscores PCA on all qualitative image criteria as evaluated on a 5-point Likert scale by two expert body radiologists (p < 0.05). RANGR estimates motion faster than PCA on GPU (1.7 ± 0.3 vs. 168.7 ± 172.4 ms, p < 0.005). The total time from motion estimation to Movienet reconstruction is 2.91 ± 1.07 s. RANGR presents a robust auto-navigation solution based on deep learning for free-breathing MRI using golden-angle radial MRI acquisition.

Breast cancer patients treated with axillary lymph node dissection (ALND) have an approximate 20-30% lifetime risk of developing lymphedema, with multiple factors implicated in the pathogenesis. In patients undergoing mastectomy, ALND can be performed through the mastectomy incision or through a separate axillary incision. It is unknown whether the latter causes more lymphatic disruption resulting in higher lymphedema rates. We aimed to assess whether ALND through the mastectomy incision versus a separate incision impacts rates of breast cancer-related lymphedema (BCRL). A retrospective chart review was performed of patients who underwent mastectomy and ALND from 2017 to 2020. Patients were grouped by ALND via mastectomy or a separate incision. The primary outcome of interest was development of BCRL as defined by ICD-10 codes. A total of 1,036 patients were included; 483 via a separate axillary incision and 553 underwent ALND via the mastectomy incision. Median time to lymphedema development was 15months. The BCRL rates between patients who had ALND via a separate incision were not significantly different from those who had ALND via the mastectomy incision (29 vs. 30%, p = 0.77). Multivariable logistical regression showed patients with ALND performed through a separate incision did not have significantly greater odds of developing lymphedema compared with those with ALND performed through the mastectomy incision (odds ratio 0.89; 95% confidence interval 0.65-1.21; p = 0.45). Patients who received ALND via a separate axillary incision as compared to the mastectomy incision do not have significantly greater rates of BCRL. Oncologic safety should be prioritized when considering lymph node retrieval technique.

Computed tomography (CT) is extensively used for accurate visualization and segmentation of organs and lesions. While deep learning models such as convolutional neural networks (CNNs) and vision transformers (ViTs) have significantly improved CT image analysis, their performance often declines when applied to diverse, real-world clinical data. Although foundation models offer a broader and more adaptable solution, their potential is limited due to the challenge of obtaining large-scale, voxel-level annotations for medical images. In response to these challenges, prompting-based models using visual or text prompts have emerged. Visual-prompting methods, such as the Segment Anything Model (SAM), still require significant manual input and can introduce ambiguity when applied to clinical scenarios. Instead, foundation models that use text prompts offer a more versatile and clinically relevant approach. Notably, current text-prompt models, such as the CLIP-Driven Universal Model, are limited to text prompts already encountered during training and struggle to process the complex and diverse scenarios of real-world clinical applications. Instead of fine-tuning models trained from natural imaging, we propose OpenVocabCT, a vision-language model pretrained on large-scale 3D CT images for universal text-driven segmentation. Using the large-scale CT-RATE dataset, we decompose the diagnostic reports into fine-grained, organ-level descriptions using large language models for multi-granular contrastive learning. We evaluate our OpenVocabCT on downstream segmentation tasks across 14 public datasets and 1 institutional dataset for organ and tumor segmentation, demonstrating the superior performance of our model compared to existing methods. All code, datasets, and models will be publicly released at https://github.com/ricklisz/OpenVocabCT.

Although validated prognostic models exist for men with metastatic castration-resistant prostate cancer (mCRPC), current tools do not incorporate genomic biomarkers such as circulating tumor DNA (ctDNA) aneuploidy or pathogenic genetic alterations (PGAs). This study aimed to estimate the prevalence of PGAs in ctDNA, assess their correlation with ctDNA aneuploidy fraction, and evaluate their association with overall survival (OS). Additionally, we developed and validated a clinical-genetic (CG) model to predict OS. We analyzed ctDNA from 776 patients enrolled in the Alliance phase 3 trial (A031201). PGAs were derived using the AR-ctDETECT assay. The net reclassification improvement (NRI) evaluated the added value of the CG model, and the time-dependent area under the receiver operating characteristic curve (tAUC) assessed the accuracy of the OS model. Feature selection using random survival forest identified gains in androgen receptor (AR), AR enhancer, MYC, RSPO2, CCND1, BRAF, and MET; losses or pathogenic variants in ZBTB16, PTEN, MSH6, PPP2R2A, NKX3-1, ZFHX3, TP53, ZNFR3, RB1, FANCA, CHECK1, APC, and CHD1; and a pathogenic AR variant. The CG model significantly outperformed the clinical model, with an average tAUC of 0.77 (95% confidence interval [CI]: 0.73-0.79) for the CG model compared with the tAUC of 0.72 (95% CI: 0.69-0.75, p=0.01) for the clinical model, and NRI of 0.29 between the models. Patients were categorized by the predicted risk scores into poor-, intermediate-, and low-risk groups with median OS of 19.6, 33.6, and 60.8mo, respectively. Incorporation of PGAs from ctDNA into a CG model improved OS prediction by nearly 30% over a clinical model. This model can classify patients into risk groups and is useful for selecting patients in future mCRPC trials.

Chimeric antigen receptor (CAR)-T cell therapy has shown limited success in the treatment of solid tumors, reinforcing the need to elucidate the invivo biodistribution of these engineered T cells. Here, we integrate the anti-DOTA huC825 reporter ("Thor") platform into newly developed human anti-interleukin-13 receptor α-2 (IL13Rα2)-single-cell fragment variable (scFv)-derived CAR-T cells and investigate its utility for mapping CAR-T cell distribution in a xenograft mouse model of melanoma. Methods: We engineered anti-IL13Rα2-scFv-derived CAR-T cells expressing huC825 (KLG3BBz-huC825), evaluated detection sensitivity, and monitored CAR-T cell biodistribution via weekly [86Y]Y-aminobenzyl-DOTA PET/CT and therapeutic efficacy. Results: KLG3BBz-huC825 T cells demonstrated potent antigen-specific cytotoxicity and cytokine release invitro. The Thor radiohapten capture platform offered exquisite detection sensitivity of only 3,000 engineered T cells and enabled prolonged spatiotemporal assessment of CAR-T cell kinetics up to 7 wk after infusion, corroborated by histopathology. Treatment with KLG3BBz-huC825 resulted in an overall survival benefit. Conclusion: The Thor platform offers a versatile and highly sensitive approach to study the real-time kinetics of CAR-T cells invivo.

TP53mutations occur in 80-90% of triple-negative breast cancers (TNBCs) and drive genomic instability and metastatic progression. Poly (ADP-ribose) polymerase (PARP) is critical for DNA repair and replication fork stability. How oncogenic signaling influences PARP function to sustain proliferation during replication stress remains unclear. Mutant p53 (mtp53) R273H associates tightly with chromatin, forms complexes with PARP, and enhances PARP recruitment to replication forks [1–3]. The C-terminal region of mtp53 mediates mtp53-PARP and mtp53-Poly (ADP-ribose) (PAR) interactions that facilitate S phase progression [4, 5]. The PARP inhibitor talazoparib (TAL) combined with the alkylating agent temozolomide (TMZ) produces synergistic cytotoxicity selectively in mtp53, but not wild-type p53 (wtp53), breast cancer cells and organoids. Herein we evaluated the mechanism of mtp53-associated cell death and tested if this could translate to a preclinical xenograft model. We found that TMZ+TAL treatment induced elevated cleaved PARP and γH2AX and reduced the metastasis-promoting oncoprotein MDMX. In orthotopic xenografts expressing mtp53 R273H, but not wtp53, combination therapy significantly decreased circulating tumor cells (CTCs) and lung metastases. Transcriptomic profiling of tumors from combination treated animals demonstrated downregulation ofMDMX,VEGF,andNF-κB, consistent with the observed suppression of CTCs and lung metastasis, and increased γH2AX, indicative of replication stress in mtp53 xenografts. Inhibition of metastasis was also observed in mtp53 R273H WHIM25 and p53-undetectable WHIM6 TNBC patient-derived xenografts (PDX). The mtp53 C-terminal domain (347-393) demonstrated a critical tumor promoting function, as CRISPR-mediated deletion impaired replication fork progression, tumor growth, and metastatic dissemination. DNA fiber combing showed that expression of full-length mtp53 R273H, but not C-terminal deleted Δ347-393, supported sustained single-stranded DNA gaps (ssGAPs) following Poly (ADP-ribose) glycohydrolase (PARG) inhibition. These findings support that mtp53 uses C-terminal amino acids to exploit PARP to enable replication stress adaptation and that mtp53 is a predictive biomarker for combined PARP inhibitor and DNA damaging therapies targeting TNBC.Significance statementTP53mutations are the most common genetic alterations in TNBC and a major driver of replication stress and metastasis. This study shows that missense mutant p53 uses C-terminal amino acids to reprogram PARP activity to maintain tumor cell survival under replication stress. We demonstrate that p53 status governs the response to combined PARP inhibitor (PARPi) and DNA-damaging chemotherapy, establishing an additional molecular basis beyond BRCA1 mutations for treating TNBC with PARPi therapy. These findings reveal a previously unrecognized mechanism by which the mutant p53-PARP axis enables replication stress tolerance and drives cancer metastasis. We show mutation of p53 in TNBC provides an additional biomarker-guided framework to improve PARPi therapeutic outcomes.