Sep 29

SS 41 - Optimizing Radiation Therapy in Gynecologic Cancers: Fractionation, Targeting, and Toxicity Reduction

327 - Multi-Institutional Analysis of the Ideal Target Volume in MRI-Based Brachytherapy for Medically Inoperable Endometrial Cancer

04:25pm - 04:35pm ET

Room 162

Presenter(s)

Isaac Lasko, MD - Loyola University Medical Center, Maywood, IL

I. Lasko¹, A. A. Vera², B. J. Eckelmann³, B. Buehl-Soppe³, D. J. Carpenter⁴, E. Evani⁴, E. Merfeld³, H. B. Musunuru², L. Bendesky⁵, J. Corteville⁶, M. K. Memon⁷, O. M. Feliciano⁶, K. Stang¹, R. Mulherkar², B. P. Venkatesulu², C. R. Weil⁵, P. N. Barry², Z. D. Horne⁷, J. P. Chino⁴, K. A. Bradley Jr³, A. H. Klopp⁵, B. A. Erickson⁶, S. Beriwal⁷, W. Small Jr¹, and M. M. Harkenrider¹; ¹Department of Radiation Oncology, Stritch School of Medicine, Cardinal Bernardin Cancer Center, Loyola University Chicago, Maywood, IL, ²Department of Radiation Oncology, UPMC Hillman Cancer Center, Pittsburgh, PA, ³Department of Human Oncology, University of Wisconsin Hospitals and Clinics, Madison, WI, ⁴Department of Radiation Oncology, Duke University Medical Center, Durham, NC, ⁵Division of Radiation Oncology, The University of Texas MD Anderson Cancer Center, Houston, TX, ⁶Department of Radiation Oncology, Medical College of Wisconsin, Milwaukee, WI, ⁷Allegheny Health Network Cancer Institute, Pittsburgh, PA

Purpose/Objective(s):

Patients with medically inoperable endometrial cancer (MIEC) are curable with brachytherapy (BT) +/- external beam radiation (EBRT). The ideal target volume in MRI-based brachytherapy is not well established. Two common volumes used are 1) the uterus + cervix ± proximal vagina (CTV) or 2) the gross tumor + endometrium ± cervix (HRCTV). We aim to compare dosimetry, disease outcomes, and toxicity between patients treated with these two approaches.

Materials/Methods:

We conducted a multi-institutional retrospective analysis of patients undergoing definitive MRI-based BT, with or without EBRT, for MIEC of T-stage =2, using a planning target volume of CTV or HRCTV, from 7 institutions. Patients planned with HRCTV had CTV retrospectively contoured. We collected patient, tumor, treatment, and EQD2 dosimetric data. Two-sample T tests, Fisher’s exact tests, and multivariable cox proportional hazards model for local recurrence-free survival (LRFS) and freedom from local recurrence (FFLR) were performed.

Results:

220 patients were included. Median follow up was 36 months (IQR 16-57). Median age was 69 years. ECOG PS was 0-1 in 57%. Disease was confined to the uterus in 90%. Most patients (89%) were node negative. EBRT was used in 67% of patients. With CTV planning (40%, n=87), median D90 (IQR) were 49.5 (39.2-55.8) with BT alone and 69.3 (59.4-71.7) with EBRT+BT. With HRCTV planning (60%, n=133), median D90 (IQR) were 50.0 (47.9-51.6) with BT alone and 73.1 (69.7-86.3) with EBRT+BT. CTV D90 (contoured retrospectively in cases of HRCTV planning) was significantly higher when planning with CTV compared to HRCTV, whether patients underwent EBRT+BT (66.6 vs 59.8 Gy, p<.01) or BT alone (46.8 vs 27.6 Gy, p<.01). 3-year local control (LC) with EBRT+BT was 92% with CTV vs 91% with HRCTV; and with BT alone was 86% with CTV vs 81% with HRCTV. On MVA, use of HRCTV vs CTV did not significantly correlate with LRFS (HR 0.87, p=0.8), while the size of the CTV (HR 1.27 per +100 cc, p=0.02) and CTV D90 (HR 0.73 per +10 Gy, p=0.05) independently predicted LRFS. Among patients planned with HRCTV target: with EBRT+BT, D90 =70 Gy led to 3-year LC of 95% vs 85%, and with BT alone, D90 =50 Gy led to 3-year LC of 84% vs 74%. HRCTV D90 did not significantly correlate with LRFS or FFLR on MVA. The overall incidences of grade 3+ toxicity were 6.7% sigmoid, 2.9% GU, 2.6% rectal, 2.4% vaginal, and 0.9% small bowel. Sigmoid D2cc was significantly higher with CTV vs HRCTV (EBRT+BT: 66.1 vs 62.7, p<.01; BT alone: 42.2 vs 36.5, p=0.03), as was bladder D2cc (EBRT+BT: 75.5 vs 65.3, p<.01; BT alone: 55.9 vs 33.5, p<.01). Grade 3+ sigmoid (p=0.003) and genitourinary (p=0.002) and vaginal (p=0.02) toxicity were more common with CTV planning.

Conclusion:

In MRI-based BT for MIEC, using “HRCTV” as planning target volume was associated with lower dose to the CTV, sigmoid, and bladder. High-grade toxicity were less with this approach, with no observed significant difference in local control.