AUTHOR=Pyo Dae Hee , Choi Joon Young , Lee Woo Yong , Yun Seong Hyeon , Kim Hee Cheol , Huh Jung Wook , Park Yoon Ah , Shin Jung Kyong , Cho Yong Beom TITLE=A Nomogram for Predicting Pathological Complete Response to Neoadjuvant Chemoradiotherapy Using Semiquantitative Parameters Derived From Sequential PET/CT in Locally Advanced Rectal Cancer JOURNAL=Frontiers in Oncology VOLUME=11 YEAR=2021 URL=https://www.frontiersin.org/journals/oncology/articles/10.3389/fonc.2021.742728 DOI=10.3389/fonc.2021.742728 ISSN=2234-943X ABSTRACT=

We evaluated the predictive value of semiquantitative volumetric parameters derived from sequential PET/CT and developed a nomogram to predict pathological complete response (pCR) in patients with rectal cancer treated by neoadjuvant chemoradiotherapy (nCRT). From April 2008 to December 2013, among the patients who underwent nCRT, those who were taken sequential PET/CT before and after nCRT were included. MRI-based staging and semiquantitative parameters of PET/CT including standardized uptake value (SUV), metabolic tumor volume (MTV), and total lesion glycolysis (TLG) were evaluated before and after nCRT. Multivariable analysis was performed to select significant predictors to construct a nomogram. Sensitivity, specificity, accuracy, and area under the receiver operating characteristics curve (AUC) of the model were evaluated to determine its performance. Among 137 eligible patients, 17 (12.4%) had pCR. All post-PET/CT parameters showed significant differences between pCR and non-pCR groups. Patients were randomly assigned to a training group (91 patients) and a validation group (46 patients). In multivariable analysis with the training group, post-CEA, post-MRI T staging, post-SUVmax, and post-MTV were significantly associated with pCR. There was no significant pre-nCRT variable for predicting pCR. Using significant predictors, a nomogram was developed. Sensitivity, specificity, accuracy, and AUC of the nomogram were 0.882, 0.808, 0.848, and 0.884 with the training group and 0.857, 0.781, 0.783, and 0.828 with the validation group, respectively. This model showed the better performance than other predictive models that did not contain PET/CT parameters. A nomogram containing semiquantitative post-PET/CT could effectively select candidates for organ-sparing strategies.