Lena Specht

Background: Deep learning (DL) methods have shown strong performance in predicting dose distributions from patient anatomy and target structures; however, accurate dose prediction does not guarantee a deliverable plan after conversion into machine parameters and dose recalculation in the treatment planning system (TPS), motivating training objectives that explicitly couple prediction to deliverability. Lymphoma presents substantial clinical heterogeneity in target location, extent, and prescription/fractionation, so fast automated methods that produce deliverable plans are particularly valuable. Purpose: To investigate whether a deliverability‐aware training framework improves the end‐to‐end deliverability of DL‐generated intensity‐modulated radiotherapy (IMRT) plans for lymphoma. Methods: A retrospective single‐institution dataset of 619 plans (571 patients; 2009–2021) was used. Clinical volumetric modulated arc therapy (VMAT) plans were converted to standardized 15‐field coplanar sliding‐window IMRT reference plans in Eclipse using automated scripting. Two DL approaches were compared: a baseline dose‐only objective model (DL‐DO) and a multi‐objective model (DL‐MO) trained with dose supervision, fluence supervision, and a dose–fluence–dose cycle‐consistency constraint implemented via fixed inverse and forward networks. Predicted fluence maps were imported into the TPS, leaf‐sequenced, and recalculated with AcurosXB without further optimization; plan quality was evaluated on a held‐out test set ( n = 61) using planning target volume (PTV) D ₉₈ , D ₂ , and D _max with paired Wilcoxon testing and Holm correction. Results: DL‐MO produced significantly better machine‐deliverable plans than DL‐DO, improving coverage and reducing hotspots (median paired differences: +12.31 percentage points (pp) in D ₉₈ , −3.43 pp in D ₂ , and −3.18 pp in D _max ; Holm‐corrected p < 0.0001 for all). Mean ± SD PTV metrics were D ₉₈ = 90.10 ± 3.74% vs. 78.59 ± 5.92%, D ₂ = 106.28 ± 1.13% vs. 109.65 ± 2.40%, and D _max = 110.32 ± 2.20% vs. 113.27 ± 3.03% (DL‐MO vs. DL‐DO). DL‐MO approached the TPS reference IMRT plans ( D ₉₈ = 91.62 ± 3.45%) with residual hotspot elevation. Conclusions: Cycle‐consistency training improves the end‐to‐end deliverability of DL‐generated IMRT plans for heterogeneous lymphoma, yielding substantially better TPS‐recalculated, machine‐deliverable plans than conventional dose‐only supervision while maintaining sub‐second fluence‐map generation. However, statistically significant differences from the TPS reference remained for D ₉₈ , D ₂ , and D _max .

Background: Intensity‐modulated radiation therapy (IMRT) is widely used for lymphoma because a multileaf collimator (MLC) can sculpt dose around irregular targets. Planning, however, is time‐consuming and difficult to standardize. Fluence map prediction (FMP) using deep learning (DL) could potentially be used to automate planning, but it has not been evaluated in full‐body lymphoma, where prescriptions, organ‐at‐risk (OAR) sets, and anatomical presentations vary substantially. Here, we evaluate multiple DL configurations to determine whether they can be adapted to lymphoma and generate IMRT plans that are deliverable in a clinical treatment planning system (TPS). Purpose: To investigate the feasibility and limitations of DL–based dose and fluence prediction methods in the heterogeneous setting of lymphoma. Methods: A single‐institution lymphoma cohort (571 patients, 619 plans) was replanned from clinical volumetric modulated arc therapy (VMAT) to a consistent 15‐field IMRT reference using the Eclipse Scripting API (ESAPI), yielding 9285 beam‐level pairs. We implemented a two‐stage DL pipeline that first predicts beam‐wise dose from CT and planning target volume (PTV), then maps dose to fluence, and finally computes the corresponding dose in the TPS after leaf sequencing. We evaluated several architectural and training variants, from replication‐focused models to configurations with a clinical loss based on dose–volume histogram (DVH) aimed at moving beyond simple plan replication. For each configuration, we analysed how these choices influenced predicted dose realism and the quality of the resulting plans. Results: In Stage 1 (dose prediction), replication‐focused DL models were numerically inferior to the IMRT and VMAT references, while DVH‐driven variants yielded superior target DVHs. After fluence mapping and TPS sequencing, all configurations produced machine‐deliverable IMRT plans but with lower target coverage and higher hot spots than the references; fluence was largely preserved by MLC sequencing, indicating that current limitations mainly stem from the predicted dose distributions and the learned dose‐to‐fluence mapping. Conclusions: DL–based dose and fluence prediction can generate fast, machine‐deliverable IMRT plans for heterogeneous lymphoma that lie in the clinical ballpark, but they remain dosimetrically inferior to manual IMRT and VMAT, highlighting the current limitations of these methods for automated planning in this setting.

The diagnosis and treatment of hematologic malignancies has undergone significant advancements over the past few decades, resulting in enhanced outcomes. For hematologic malignancies diagnosed during pregnancy, this poses new questions. As possibilities continue to expand and the rising global incidence results in an increasing number of diagnoses within the pregnant population, there is a pressing need to gain a deeper understanding of the potential options for pregnant women and the impact on the newborn. As knowledge of the effects of cancer treatment increases, more women are receiving adequate cancer diagnosis and treatment during pregnancy. This paper presents an expert opinion on the use of diagnostic modalities and treatment options, including chemotherapy, radiation therapy, and immunotherapy, for acute leukemia, Hodgkin lymphoma, non-Hodgkin lymphoma, myeloproliferative neoplasms, aplastic anemia, and multiple myeloma during pregnancy.

We evaluated outcomes by management type for patients with stage IA nodular lymphocyte-predominant Hodgkin lymphoma (NLPHL) in the Global NLPHL One Working Group (GLOW) retrospective database of 2243 patients with stage I-IV disease diagnosed from 1992-2021 at 38 international institutions. 779 patients were stage IA with median age of 35 years (range=3-89) and median follow-up of 6.1 years. The 6-year PFS and OS were 86.3% and 97.7%, respectively. Outcomes were analyzed for two groups: complete resection and unresected disease. Patients with a complete resection and observation alone (n=99) had a 6-year PFS of 65.5% versus 90.5% for those who received RT (n=53). Patients with unresected disease (n=627; 80.5%) had a 6-year PFS of 62.0% for rituximab alone (n=31), 89.6% for RT alone (n=325), 76.8% for ABVD alone (n=40), and 94.3% for ABVD+RT (n=130). 127 patients relapsed (16.3%), of which 25 (19.7%) had transformation. Our analysis suggests 1) RT improves the PFS in patients with completely resected disease; 2) Rituximab or ABVD alone do not appear to achieve a durable response; 3) Chemotherapy was not observed to add additional PFS benefit when used in combination with RT. Thus, for stage IA NLPHL, RT alone is likely sufficient for definitive treatment.

Late cardiac toxicity after radiation therapy and/or systemic therapy with anthracyclines for lymphomas is a serious concern. To estimate the risks with different treatment combinations, reliable estimates of the dose‐response relationships are needed. We performed a literature‐based regression analysis of long‐term cardiac events after lymphoma treatment. The objective was to identify dose‐response relationships for anthracycline and radiotherapy regarding congestive heart failure (CHF), ischemic heart disease (IHD), and valvular heart disease (VHD). We included papers published January 2000‐December 2022 with data on long term cardiac outcomes in lymphoma patients, radiation doses to the heart, and anthracycline doses. Papers without dose/response information were excluded. We identified six eligible papers including 22,916 patients. The excess relative risk (ERR) of CHF for anthracyclines per 100 mg/m2 (corresponding to 2 cycles of ABVD or CHOP chemotherapy) was 92% (CI: 74%–101%), and for radiotherapy per Gray (Gy) of mean heart dose it was 6.1% (CI: 4.4%–7.6%). Corresponding numbers for the other endpoints were: IHD: No effect of anthracyclines, ERR = 4.4%/Gy (CI: 2.7%–6.1%) and VHD: ERR = 25%/100 mg/m2 (CI: 13%–37%) and ERR = 10%/Gy (CI: 6%–13%). Data agree with a linear no threshold dose‐response relationship for related endpoints, that is, there are no “safe” lower doses of either anthracyclines or radiotherapy. Late cardiac toxicity risks for all three endpoints can be assessed in each patient by a combined estimate from the cumulative doses of anthracyclines and radiation to the heart. This estimate can guide future treatment allocation in lymphoma patients.

Primary cutaneous lymphomas are highly radiosensitive. X-rays work well for localized cutaneous lymphomas. However, if disseminated in the skin and covering larger areas, as is commonly the case with the most common type, mycosis fungoides, x-ray therapy is not suited because the dose to underlying organs exceeds their tolerance. By contrast, electrons have a limited range of penetration, and are ideal for treating superficial lesions. Techniques have been developed to yield a fairly uniform dose to the entire skin surface and treating to a depth of about 1-1½ cm. Total skin electron beam therapy (TSEBT) is probably the most effective skin directed therapy for widespread primary cutaneous lymphomas. For many years the total dose used for mycosis fungoides was 30-36 Gy, given in small fractions. This treatment could only be repeated once. However, total doses of 10-12 Gy have now been shown to offer excellent response rates, and the treatment can be repeated up to 6 times, offering as much or probably even more palliation than the high-dose treatment. Today, most patients are treated with low-dose TSEBT, the higher doses reserved for patients with more resistant disease. Attempts have been made to use photon therapy for total skin irradiation, e.g., tomotherapy. However, even with the most meticulous of techniques there is too much dose in deeper structures, resulting in bone marrow toxicity even with low-dose treatment. This is never seen with electrons, even with high-dose therapy. Further research into optimizing TSEBT and exploring combinations with systemic treatments is ongoing.

PURPOSE: Nodular lymphocyte–predominant Hodgkin lymphoma (NLPHL) is a rare cancer, and large international cooperative efforts are needed to evaluate the significance of clinical risk factors and immunoarchitectural patterns (IAPs) for all stages of pediatric and adult patients with NLPHL. METHODS: Thirty-eight institutions participated in the Global nLPHL One Working Group retrospective study of NLPHL cases from 1992 to 2021. We measured progression-free survival (PFS), overall survival (OS), transformation rate, and lymphoma-specific death rate. We performed uni- and multivariable (MVA) Cox regression stratified by management to select factors for the lymphocyte-predominant international prognostic score (LP-IPS) validated by five-fold cross-validation. RESULTS: We identified 2,243 patients with a median age of 37 years (IQR, 23-51). The median follow-up was 6.3 years (IQR, 3.4-10.8). Most had stage I to II (72.9%) and few B symptoms (9.9%) or splenic involvement (5.4%). IAP was scored for 916 (40.8%). Frontline management included chemotherapy alone (32.4%), combined modality therapy (30.5%), radiotherapy alone (24.0%), observation after excision (4.6%), rituximab alone (4.0%), active surveillance (3.4%), and rituximab and radiotherapy (1.1%). The PFS, OS, transformation, and lymphoma-specific death rates at 10 years were 70.8%, 91.6%, 4.8%, and 3.3%, respectively. On MVA, IAPs were not associated with PFS or OS, but IAP E had higher risk of transformation (hazard ratio [HR], 1.81; P < .05). We developed the LP-IPS with 1 point each for age ≥45 years, stage III-IV, hemoglobin <10.5 g/dL, and splenic involvement. Increasing LP-IPS was significantly associated with worse PFS (HR, 1.52) and OS (HR, 2.31) and increased risk of lymphoma-specific death (HR, 2.63) and transformation (HR, 1.41). CONCLUSION: In this comprehensive study of all ages of patients with NLPHL, we develop the LP-IPS to identify high-risk patients and inform upcoming prospective clinical trials evaluating de-escalation of therapy for patients with low LP-IPS scores (<2).