Dataset Information

Machine learning highlights the deficiency of conventional dosimetric constraints for prevention of high-grade radiation esophagitis in non-small cell lung cancer treated with chemoradiation.

ABSTRACT:

Background and purpose

Radiation esophagitis is a clinically important toxicity seen with treatment for locally-advanced non-small cell lung cancer. There is considerable disagreement among prior studies in identifying predictors of radiation esophagitis. We apply machine learning algorithms to identify factors contributing to the development of radiation esophagitis to uncover previously unidentified criteria and more robust dosimetric factors.

Materials and methods

We used machine learning approaches to identify predictors of grade ? 3 radiation esophagitis in a cohort of 202 consecutive locally-advanced non-small cell lung cancer patients treated with definitive chemoradiation from 2008 to 2016. We evaluated 35 clinical features per patient grouped into risk factors, comorbidities, imaging, stage, histology, radiotherapy, chemotherapy and dosimetry. Univariate and multivariate analyses were performed using a panel of 11 machine learning algorithms combined with predictive power assessments.

Results

All patients were treated to a median dose of 66.6 Gy at 1.8 Gy per fraction using photon (89.6%) and proton (10.4%) beam therapy, most often with concurrent chemotherapy (86.6%). 11.4% of patients developed grade ? 3 radiation esophagitis. On univariate analysis, no individual feature was found to predict radiation esophagitis (AUC range 0.45-0.55, p ? 0.07). In multivariate analysis, all machine learning algorithms exhibited poor predictive performance (AUC range 0.46-0.56, p ? 0.07).

Conclusions

Contemporary machine learning algorithms applied to our modern, relatively large institutional cohort could not identify any reliable predictors of grade ? 3 radiation esophagitis. Additional patients are needed, and novel patient-specific and treatment characteristics should be investigated to develop clinically meaningful methods to mitigate this survival altering toxicity.

SUBMITTER: Luna JM

PROVIDER: S-EPMC7132156 | biostudies-literature | 2020 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Machine learning highlights the deficiency of conventional dosimetric constraints for prevention of high-grade radiation esophagitis in non-small cell lung cancer treated with chemoradiation.

Luna José Marcio JM Chao Hann-Hsiang HH Shinohara Russel T RT Ungar Lyle H LH Cengel Keith A KA Pryma Daniel A DA Chinniah Chidambaram C Berman Abigail T AT Katz Sharyn I SI Kontos Despina D Simone Charles B CB Diffenderfer Eric S ES

Clinical and translational radiation oncology 20200324

<h4>Background and purpose</h4>Radiation esophagitis is a clinically important toxicity seen with treatment for locally-advanced non-small cell lung cancer. There is considerable disagreement among prior studies in identifying predictors of radiation esophagitis. We apply machine learning algorithms to identify factors contributing to the development of radiation esophagitis to uncover previously unidentified criteria and more robust dosimetric factors.<h4>Materials and methods</h4>We used machi ...[more]

PMID: 32274426

Similar Datasets

Project description:PurposeTo determine the incidence and predictors of gastric bleeding after chemoradiation for esophageal or gastroesophageal junction cancer.Methods and materialsWe reviewed patients receiving chemoradiation to at least 41.4 Gy for localized esophageal cancer whose fields included the stomach and who did not undergo surgical resection. The primary endpoint was grade ≥3 gastric hemorrhage (GB3+). Comprehensive stomach dose-volume parameters were collected, and stomach dose-volume histograms were generated for analysis.ResultsA total of 145 patients met our inclusion criteria. Median prescribed dose was 50.4 Gy (range, 41.4-56 Gy). Median stomach Dmax was 53.0 Gy (1.0-62.7 Gy), and median stomach V40, V45, and V50 Gy were 112 cm3 (0-667 cm3), 84 cm3 (0-632 cm3), and 50 cm3 (0-565 cm3), respectively. Two patients (1.4%) developed radiation-induced GB3+. The only dosimetric factor that was significantly different for these patients was a higher stomach Dmax (58.1 and 58.3 Gy) than the cohort median (53 Gy). One of these patients also had cirrhosis, and the other had a history of nonsteroidal anti-inflammatory drug use. Five other patients had GB3+ events associated with documented tumor progression. A Cox proportional hazards model based on stomach Dmax with respect to the development of GB3+ was found to be statistically significant. Time-to-event curves and dose-volume atlases were generated, demonstrating an increased risk of GB3+ only when stomach Dmax was >58 Gy (P < .05).ConclusionsWe observed a low rate of GB3+ events in patients who received chemoradiation to a median dose of 50.4 Gy to volumes that included a significant portion of the stomach. These results suggest that when prescribing 50.4 Gy for esophageal cancer, there is no need to minimize the irradiated gastric volume or dose for the sake of preventing bleeding complications. Limiting stomach maximum doses to <58 Gy may also avoid bleeding, and particular caution should be taken in patients with other risk factors for bleeding, such as cirrhosis.

Project description:BACKGROUND: Hilar cholangiocarcinoma is a rare tumour which is best managed by an aggressive surgical approach. The role of adjuvant or neoadjuvant radiation therapy, chemotherapy or chemoradiation remains controversial, as no prospective randomized studies have been performed. METHODS: This review summarizes the recent literature regarding the role of radiation, chemotherapy and chemoradiation in hilar cholangiocarcinoma. The results of a biliary cancer questionnaire regarding current treatment strategies are also reported. RESULTS: A number of retrospective studies have shown that patients treated with adjuvant radiation therapy have prolonged survival compared with untreated patients. However, most of these reports did not control for tumour stage or performance status. A carefully controlled trial from the Johns Hopkins Hospital did not demonstrate any benefit for adjuvant radiation therapy. A number of phase II trials of chemotherapy have demonstrated modest response rates (20-40%). The best responses have been reported with 5-fluorouracil (5-FU) in combination with interferon-alpha or with leucovorin and mitomycin C. Recent non-randomized reports of chemoradiation with 5-FU with or without gemcitabine as the radiosensitizer suggest, but do not prove, improved survival. Adjuvant chemoradiation is currently being employed at specialized centres most often in the Americas (71%) and the Asia/Pacific region (55%) and to a lesser degree in Europe (29%). DISCUSSION: The only chance for long-term survival in patients with hilar cholangiocarcinoma is complete resection with negative margins. Neither radiation therapy nor chemotherapy alone has been proven to prolong survival in completely or partially resected patients or in unresected patients. Recent uncontrolled data suggest that chemoradiation may improve survival in resected and locally unresectable patients. However, prospective, randomized multicentre trials need to be performed to confirm efficacy.

Dataset Information

Machine learning highlights the deficiency of conventional dosimetric constraints for prevention of high-grade radiation esophagitis in non-small cell lung cancer treated with chemoradiation.

Background and purpose

Materials and methods

Results

Conclusions

Publications

Machine learning highlights the deficiency of conventional dosimetric constraints for prevention of high-grade radiation esophagitis in non-small cell lung cancer treated with chemoradiation.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets