Daphne Haas-Kogan
Professor and Chair
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Articles (10)
Year:
2026
Background Master clinical trial protocol structures offer administrative, procedural, and statistical advantages but have not been applied in assessing new radiotherapy devices. Herein, we report on a pooled analysis from a first-of-kind master trial evaluating stereotactic MRI-guided adaptive radiotherapy (SMART). Methods Subjects were enrolled in a prospective master protocol evaluating SMART for multiple oncologic indications. CTCAE v5 toxicities, patient-reported outcome measures (PROMs), and treatment efficiency metrics were assessed across the entire cohort and in anatomic subsets. Results One hundred and ninety three subjects were enrolled into 20 sub-protocols for different SMART indications. Data were analyzed from the first 161 subjects. The median time from subprotocol amendment submission to activation was 70.5 days (range: 63-93). All completed phase I sub-protocols (n = 9) met the primary endpoints of safety and feasibility. The risk of grade 3+ toxicity was 1.9% (95% CI 0.4% to 5.3%), 7.1% (95% CI 0.9% to 23.5%), 1.8% (95% CI 0.05% to 9.6%) and 0% (95% CI 0.0% to 4.7%) in the overall cohort and thoracic, abdominal and pelvic subsets, respectively. PROMs (PROMIS-10) during and after SMART were unchanged from baseline in all subsets. SMART delivery efficiency improved over the study period (first vs. final 3 months: 78.5 vs. 48.2 minutes, P < .001). Adaptive planning performed for 771 fractions resulted in clinically significant plan improvements in 93.0% of fractions (52.7% for organ-at-risk sparing, 20.5% for target coverage, and 19.8% for both). Conclusions SMART is feasible and safe for multiple thoracic, abdominal and pelvic radiotherapy indications. The master protocol platform offers an adaptable approach for assessment of new oncologic treatment technologies applicable to multiple indications. Clinical Trial NCT04115254.
Year:
2025
Background Pediatric low-grade gliomas (pLGGs) have heterogeneous clinical presentations, and given the morbidity of treatment, some patients receive observation with magnetic resonance (MR). The natural histories of untreated pLGGs remain understudied. We leveraged deep learning-based volumetrics to analyze longitudinal growth trajectories and progression risk factors for untreated pLGGs. Methods We conducted a pooled, retrospective study of radiographically diagnosed pLGG patients from two institutions diagnosed between 1992 and 2020 who were surveilled for at least 1 year post-diagnosis. Tumor segmentation was applied to longitudinal T2-weighted MR to calculate 3D tumor volumes. We assessed volume trajectories, disease progression, and associated risk factors using Cox-Hazards regression, survival analysis, and time-series forecasting with autoregressive integrated moving average (ARIMA). Patients were categorized based on volumetric changes into progression (≥25%), regression (≤−25%), or stability. Results Of 99 patients (970 scans; median follow-up: 7.0 years; median diagnosis age: 12.0 years), 55 (55.5%) had tumors that volumetrically progressed, 28 (28.3%) remained stable, and 16 (16.2%) regressed. 42 (42.4%) patients initiated treatment. Risk factors associated with progression included infancy/preschool age, cortical location, and female sex (p ≤ 0.05 for each). Most progressions occurred within five years of diagnosis (80.0%), most commonly in school-aged children (7-13 years old). Time-series forecasting predicted future tumor volume with a mean absolute error of 2.04 cm3. Conclusion Deep learning enables systematic, longitudinal, pLGG growth tracking and characterization of patients on surveillance, yielding insights into untreated tumor trajectories and progression risk. This pipeline is useful at population-level to study growth trends and at patient-level to guide personalized management.
Year:
2025
Supplementary Figure 3. Digital quantification from immunohistochemistry. Digital quantification from immunohistochemistry for T cell subsets and marker analysis from paired tumor samples for five patients enrolled to cohort A.
Year:
2025
Supplementary Figure 1. CODEX panel and experimental procedure. CODEX panel used to identify tumor cell subpopulations, immune cell populations and key cellular markers. (B) Experimental workflow for 5 pre- and post-treatment FFPE samples. ROIs were selected for multiplexed imaging and downstream analysis. (C) U Map clusters show major cell populations. Immune populations showed similar clustering in pre- and post- treatment samples.
Year:
2025
Supplementary Figure 6. Cell-cell interaction analyses. Cell-cell interaction comparison in pre- and post-treatment tumor samples using CODEX for patient #5. (A) All significant interacting cell types shown in the entire analyzed tumor region as annotated. (B) Example images of cell – cell interactions showing interaction of CD163+ macrophages with GZMB+ CD8 T cells, CD163+ macrophages with GZMB+ CD4 T cells, CD68+ macrophages with GZMB+ CD8 T cells. The plus sign indicates interactions that were in close proximity at a one-cell distance (20 pixels).
Year:
2025
Purpose: Radiotherapy may enhance antitumor immune responses by several mechanisms, including induction of immunogenic cell death. We performed a phase 2 study of pembrolizumab with re-irradiation in patients with recurrent glioblastoma. Patients and Methods: Sixty patients with recurrent glioblastoma received pembrolizumab with re-irradiation alone (cohort A, bevacizumab-naïve; n = 30) or with bevacizumab continuation (cohort B, n = 30). Dual primary endpoints, including the overall response rate and overall survival (OS) at either 12 (OS-12; cohort A) or 6 months (OS-6; cohort B), were assessed per cohort relative to historic benchmarks. Paired paraffin-embedded formalin-fixed tumor samples were assessed for immunologic biomarkers by IHC using digital quantification and co-detection-by-indexing (CODEX). Results: Study therapy was well tolerated, with most toxicities being grade ≤3. For cohort B, the primary endpoint of OS-6 was achieved (57%); however, survival was not improved for cohort A patients. The overall response rate was 3.3% and 6.7% for cohorts A and B, respectively. CODEX analysis of paired tumor samples from five patients revealed an increase of activated T cells in the tumor microenvironment after study therapy. Conclusions: Compared with historic controls, re-irradiation plus pembrolizumab seemed to improve survival among bevacizumab-refractory patients but not among bevacizumab-naïve patients. CODEX revealed evidence of intratumoral infiltration of activated immune effector cells. A randomized, properly controlled trial of PD-1 blockade plus re-irradiation is warranted to further evaluate this regimen for bevacizumab-refractory patients, but alternative approaches are needed for bevacizumab-naïve patients.
Year:
2025
Supplementary Figure 4. CODEX analysis of tumor PD-L1 expression patterns. (A) PD-L1 expression was detected in NCAM-1+ tumor cells surrounding by hypoxia region as detected by CA9 staining. (B) PD-L1 expression was also observed in GFAP+ tumor cells surrounded in the vicinity of apoptotic tumor cells (cleaved caspase-3 detection).
Year:
2025
Supplementary Figure 5. Cell Density Analyses. Comparison of defined cell populations in pre- and post-treatment tumor specimens via CODEX. (A) Each indicated cell type is shown as a percentage of total cells in each pre- and post-tumor specimen for all five patients. (B) Cell density is shown for T cell subpopulations in preand post-treatment tumor specimens for each patient. CD4 T cell populations are shown in different shades of blue; CD8 T cell populations in shades of red. Cell density was determined by cell count per microm2. T cell density increased substantially in patient #5. (C) Cell density changes are shown for paired pre- and posttreatment tumor samples. Patient #5 is highlighted with a blue dotted line as this patient showed an increase in most major immune populations in the post-treatment specimen. Of note, NK T cells were quite rare and not identified in patient 3. Paired t-test was used for statistical analysis: 0.05 *, 0.01 **, 0.001 ***. (D) Representative images show the increase in CD25+ activated CD4 T cells after treatment (indicated by yellow arrows).
Year:
2025
Background Postoperative recurrence risk for pediatric low-grade gliomas (pLGGs) is challenging to predict by conventional clinical, radiographic, and genomic factors. We investigated if deep learning (DL) of magnetic resonance imaging (MRI) tumor features could improve postoperative pLGG risk stratification. Methods We used a pretrained DL tool designed for pLGG segmentation to extract pLGG imaging features from preoperative T2-weighted MRI from patients who underwent surgery (DL-MRI features). Patients were pooled from 2 institutions: Dana Farber/Boston Children’s Hospital (DF/BCH) and the Children’s Brain Tumor Network (CBTN). We trained 3 DL logistic hazard models to predict postoperative event-free survival (EFS) probabilities with (1) clinical features, (2) DL-MRI features, and (3) multimodal (clinical and DL-MRI features). We evaluated the models with a time-dependent Concordance Index (Ctd) and risk group stratification with Kaplan–Meier plots and log-rank tests. We developed an automated pipeline integrating pLGG segmentation and EFS prediction with the best model. Results Of the 396 patients analyzed (median follow-up: 85 months, range: 1.5–329 months), 214 (54%) underwent gross total resection and 110 (28%) recurred. The multimodal model improved EFS prediction compared to the DL-MRI and clinical models (Ctd: 0.85 (95% CI: 0.81–0.93), 0.79 (95% CI: 0.70–0.88), and 0.72 (95% CI: 0.57–0.77), respectively). The multimodal model improved risk-group stratification (3-year EFS for predicted high-risk: 31% versus low-risk: 92%, P < .0001). Conclusions DL extracts imaging features that can inform postoperative recurrence prediction for pLGG. Multimodal DL improves postoperative risk stratification for pLGG and may guide postoperative decision-making. Larger, multicenter training data may be needed to improve model generalizability.
Year:
2024
Collaborators (15)
Ali Nabavizadeh
University of Pennsylvania
Hesham Elhalawani
Director of Pediatric/AYA in Radiation Oncology
Brigham and Women's Hospital
Maxime Meylan
Dana-Farber Cancer Institute
Ralph J. DeBerardinis
University of Texas Southwestern Medical Center
Yi Xiao
University of Texas Southwestern Medical Center
Samuel McBrayer
Assistant Professor
University of Texas Southwestern Department of Pediatrics
Kalil G Abdullah
University of Pittsburgh Medical Center
Tony J.C. Wang
Columbia University
Lin Xu
Assistant Professor (tenure track)
University of Texas Southwestern Medical School
Paul Catalano
Senior Lecturer on Biostatistics
Harvard University
Keith Ligon
Chief of Neuropathology/Principal Investigator
Dana-Farber Cancer Institute
Shyam Tanguturi
Vice Chair, Network Strategy & Business Development
Dana-Farber Cancer Institute
Kai Wucherpfennig
Professor and Department Chair
Dana-Farber Cancer Institute
Raymond Mak
Associate Professor
Harvard Medical School
Stephen J Bagley
University of Pennsylvania
Make your next move.
Prepare your application
Review drafts before you send them.