Long-Term Outcomes of Non-Vestibular Cranial Nerve Schwannomas Treated with Fractionated Stereotactic Radiotherapy and Stereotactic Radiosurgery

PURPOSE/OBJECTIVE(S)

Non-vestibular cranial nerve schwannomas (NVCNS) are rare tumors that account for approximately 10% of cranial nerve schwannomas. They are commonly treated with radiation therapy (XRT) due to their location often precluding safe resection. We examined the long-term outcomes of NVCNS treated with XRT as primary management and for post-operative salvage.

MATERIALS/METHODS

We conducted a retrospective review of patients with NVCNS treated with fractionated stereotactic radiation therapy (FSRT) or Gamma Knife Stereotactic Radiosurgery (GK-SRS) from 1996 to 2018 at our institution. We examined patient demographics, cranial nerve (CN) involvement, CN function pre-/post-XRT, treatment volume (TV), toxicity, surgery pre-XRT, and local control. Kaplan-Meier analysis was performed for evaluation of local control.

RESULTS

We identified 66 patients (38 female, 28 male) with NVCNS, a portion of whom had tumors involving more than one cranial nerve. Forty-six (69.7%) were treated with FSRT (median dose 50.4 Gy in 1.8 Gy/fraction; range 45-54 Gy), and 20 (30.3%) with GK-SRS (median dose 12 Gy; range 12-15 Gy). Median follow-up time was 92.5 months (5-306). Median Karnofsky Performance Status was 90 (70-100). Median age at start of XRT was 45 years old (15-92). Prior to XRT, 34.8% (23) of patients had surgical resection, with median time from surgery to XRT of 4.25 months (0.5-130 months). Median treatment volume was 4.72 cc (0.26-29). The cranial nerve most commonly involved was CN V (48.4%), followed by CN X (15.2%), CN VII (13.6%), CN VI (6.1%), CN XII (6.1%), CN III (6.1%), and CN IX (3%). Twenty-nine (43.9%) patients experienced grade 1 acute toxicity during treatment. Six (9%) patients experienced grade 1 chronic toxicity. No grade 2 or higher acute or chronic toxicity was observed. No significant difference in rates of acute or chronic toxicity was observed between patients treated with GK-SRS vs. FSRT. Post-XRT, 37 patients (56.1%) had improvement in CN function/symptom, 24 patients (36.3%) had stable function/symptoms, and 5 patients (7.6%) had worsening function/symptoms. Local control at one and five years was 100%. In-field recurrence was observed in one patient (1.5%), at 9 years post-XRT. For salvage this patient was treated with a second course of FSRT to the recurrent tumor.

CONCLUSION

Our large institutional series with long term follow up showed excellent local control of NVCNS treated with FSRT or GK-SRS both for primary management and post-operative salvage. Treatment is well tolerated, with high rates of preservation or improvement of CN function, and minimal acute and chronic toxicity.

Novel Deep Learning Segmentation Models for Accurate GTV and OAR Segmentation in MR-Guided Adaptive Radiotherapy for Pancreatic Cancer Patients

PURPOSE/OBJECTIVE(S)

MR-guided adaptive radiotherapy (MRgART) improves target coverage and organ-at-risk (OAR) sparing in pancreatic cancer radiation therapy (RT). Inter-fractional changes in patients undergoing RT require time intensive re-delineation of gross tumor volume (GTV) and OARs prior to adaptive optimization. Accurate automatic segmentation has the potential to significantly improve efficiency of the adaptive workflow. We hypothesized that state-of-the-art deep learning (DL) segmentation models could adequately segment GTV and OARs in both planning and daily fractional MR scans.

MATERIALS/METHODS

The study included 21 patients with pancreatic cancer treated with MRgART (10 Gy x 5 fractions). The planning MR as well as all daily MR images and registrations were collected (6 image sets per patient and a total of 126 image sets). The planning MR and fraction 1-4 image sets were used as the training set (N = 105), while the test set (N = 21) comprised images for fraction 5, to simulate the last step of incremental learning from planning to final fraction. Evaluated contours included the GTV, Small Bowel, Large Bowel, Duodenum, Left and Right Kidney, Liver, Spinal Cord, and Stomach. To mimic clinical conditions, contour accuracy was evaluated within the ring structure surrounding the PTV, inside of which daily adaptive re-contouring is applied (2 cm expansion in the cradio-caudal direction, 3 cm expansion otherwise). We evaluated three DL model architectures: SegResNet, SegResNet 2D, and SwinUNETR to autosegment GTV and OARs. The segmentation models were trained on the training set using 5-fold cross-validation (CV) and quantitatively analyzed by comparing against clinically used contours with DICE scores. Qualitative analysis was performed by a radiation oncologist using a scoring scale: 1 = perfect, 2 = minor discrepancy, 3 = moderate discrepancy, and 4 = rejected.

RESULTS

Overall, the DL segmentations were in acceptable agreement with clinical contours. The best performing model was the SwinUNETR model with overall training DICE = 0.88±0.06, test DICE = 0.78±0.11, and qualitative score of 1.6±0.8. The agreement between the DL model and clinical segmentation for the GTV was 0.79±0.08, with a qualitative score of 2.2±0.9. The highest and lowest OAR DICE scores were for the Left Kidney (DICE = 0.93) and Small Bowel (DICE = 0.68), respectively. The highest qualitative OAR scores were for the Kidney, Liver, and Spinal Cord (score = 1.0) and the lowest qualitative score was for the Duodenum (score = 2.3) CONCLUSION: We report here the most comprehensive work on DL segmentation for pancreatic cancer MRgART, including quantitative and clinically-pertinent qualitative evaluations of 126 image sets and 3 DL architectures. Our data show good quantitative agreement between DL and clinical contours, and acceptable clinician evaluations for the majority of GTVs and OARs. The current work has great potential to significantly reduce a major bottleneck in the MRgART workflow for pancreatic cancer patients.

Early Outcomes of MR-Guided SBRT for Patients with Recurrent Pancreatic Adenocarcinoma

PURPOSE/OBJECTIVE(S)

Local treatment options for patients with locally recurrent pancreatic adenocarcinoma (L-PAC) are limited, with expected median survival time (MST) of 8-11 months (mo) following recurrence. MRI-guided radiation therapy (MRgRT) provides the ability to dose escalate while sparing normal tissue. The literature for MR-guided Stereotactic Body Radiotherapy (MRgSBRT) for L-PAC is sparse. Here we report on the early outcomes of MRgSBRT in patients with L-PAC.

MATERIALS/METHODS

Patients with prior resection of pancreatic adenocarcinoma with post-operative chemotherapy as indicated followed by local recurrence of disease at prior surgical site and treated with MRgSBRT at a single tertiary referral center from 5-2021 to 8-2022 for L-PAC were identified from our prospective database. MRgSBRT was delivered to 40-50 Gy in 4-5 fractions with target and OAR delineation per institutional standards. Descriptive analysis of the patient, disease, and treatment characteristics were performed. Endpoints included local control, defined as absence of tumor progression per RECIST criteria, distant failure, overall survival (OS), and acute and chronic toxicities per Common Terminology Criteria for Adverse Events (CTCAE), version 5.

RESULTS

Eleven patients with L-PAC were identified with median follow-up of 10.7 mo (3.2 - 22.3). Ten of those underwent surgical resection at the treating radiation facility and one patient underwent preoperative radiation for 50.4 Gy in 28 fractions followed by surgical resection at an outside hospital. MRgRT was delivered a median of 18.8 mo (3.5 - 48.0) following resection. There were 5 females and 6 males, with a median age of 72 years (52-83) and median KPS of 80 (60-100). OS rates following initial diagnosis at 12, 18 and 24 mo were 100%, 82%, and 61%, respectively, with an MST of 25.3 mo (12.4-53.1). OS rates following recurrence at 6 and 12 mo were 82% and 52%, respectively, with an MST of 10.7 mo (3.2 - 21.9). One patient experienced local failure at 7.8 mo, and 9 patients experienced distant failure at a median of 3.4 mo (0.3 - 21.9) following MRgSBRT. Five patients experienced distant failure less than 3 mo following radiation. Grade 1 or 2 acute GI toxicity was noted in 45% of patients and chronic GI toxicity, in 18% of patients. No Grade≥3 AEs were noted.

CONCLUSION

MRgSBRT for recurrent pancreatic adenocarcinoma demonstrates good local control with acceptable acute and chronic toxicity as well as reasonable overall survival. Distant failure remains a substantial problem with a significant number of patients demonstrating metastases immediately following radiation, suggesting the presence of micro-metastatic disease prior to local therapy. Adequate patient selection for MRgSBRT, and proper integration of systemic therapy in this patient population remains a topic of discussion that requires further exploration.