No date

Session Item

Sunday

August 29

16:45 - 17:45

Room 1

Proffered papers 25: Upper GI

Chair: , Belgium;

Chair: , Austria

Session Code: 1770

Session Type: Proffered papers

Track: Clinical

17:05 - 17:15

Definition Of LOcal REcurrence Site in resected pancreatic cancer: a multicentric study (DOLORES-1)

Alessandra Arcelli, Italy

Presentation Number: OC-0413

Abstract

Abstract Title:

Definition Of LOcal REcurrence Site in resected pancreatic cancer: a multicentric study (DOLORES-1)

Authors:

Alessandra Arcelli^1,2, Federica Bertini^1,2, Silvia Strolin³, Gabriella Macchia⁴, Francesco Deodato^4,5, Savino Cilla⁶, Salvatore Parisi⁷, Aldo Sainato⁸, Michele Fiore⁹, Pietro Gabriele¹⁰, Domenico Genovesi¹¹, Francesco Cellini^5,12, Alessandra Guido¹, Silvia Cammelli^1,2, Milly Buwenge^1,2, Emiliano Loi³, Matteo Renzulli¹³, Rita Golfieri^2,13, Alessio Giuseppe Morganti^1,2, Lidia Strigari³

¹Radiation Oncology, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy; ²Department of Experimental, Diagnostic and Specialty Medicine (DIMES), Alma Mater Studiorum, Bologna University, Bologna, Italy; ³Medical Physics, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy; ⁴Radiation Oncology Unit, Gemelli Molise Hospital – Università Cattolica del Sacro Cuore, Campobasso, Italy; ⁵Istituto di Radiologia, Università Cattolica del Sacro Cuore, Roma, Italy; ⁶Medical Physics Unit, Gemelli Molise Hospital – Università Cattolica del Sacro Cuore, Campobasso, Italy; ⁷Unit of Radiation Therapy, IRCCS “Casa Sollievo della Sofferenza”, San Giovanni Rotondo, Italy; ⁸Radiation Oncology, Pisa University Hospital, Pisa, Italy; ⁹Radiation Oncology, Campus Bio-Medico University, Rome, Italy; ¹⁰Radiation Therapy, Candiolo Cancer Institute - FPO, IRCCS Candiolo, Candiolo, Italy; ¹¹Department of Radiation Oncology, SS. Annunziata Hospital, G. D'Annunzio University of Chieti, Chieti, Italy; ¹²UOC di Radioterapia, Dipartimento di Scienze Radiologiche, Radioterapiche ed Ematologiche, Fondazione Policlinico Universitario A. Gemelli, IRCCS, Roma, Italy; ¹³Radiology Unit, IRCCS Azienda Ospedaliero-Universitaria di Bologna, Bologna, Italy

Show Affiliations

Purpose or Objective

To generate a local failures (LF) risk map in resected pancreatic cancer (PC) patients, and validate the results of the previous studies on this topic. We also aimed to propose new guidelines for postoperative target delineation in PC patients.

Material and Methods

A retrospective, multicentric, observational study, on behalf of AIRO (Italian Association of Radiation and Clinical Oncology) was conducted collecting data and imaging (contrast enhanced Computed Tomography, (CT)) of resected PC patients with LF from six Italian centers. A radiologist specialized in gastrointestinal tumors delineated the LF on the follow-up contrast enhanced CT and reported the recurrence areas on the CT images of a representative patient (Figure 1A-B). The 70% of LF points were randomly extracted from the clinical target volume (CTV) based on RTOG guidelines and combined to the 30% of points randomly obtained from the LF database. Based on a Kernel density estimation the 3D distribution map of LF points was generated and compared with the results of two previously published studies using the Dice index.

Results

Sixty-four patients were included in this analysis. Most patients (59.4%) underwent adjuvant treatment after surgery. Twenty-one (32.8%) patients experienced LF closer to the root of the celiac axis (CA) and forty-three patients (67.2%) experienced LF closer to the root of the superior mesenteric artery (SMA). The mean (± standard deviation) distance of LF points to CA and SMA was 21.5 ±17.9 mm and 21.6 ±12.1 mm, respectively. The Dice values comparing the isolevel of risk map corresponding to the 80% and 90% probabilistic density and the CTV80s and CTV90s proposed in the previous publications were 0.45-0.53 and 0.58-0.60, respectively.

Figure 1A-1B: Local recurrence sites (blue symbols) plotted with respect to the Celiac Axis (yellow) and the Superior Mesenteric Artery (cyan). CTV90 (light blue) and CTV80 (green line) proposed in previuos pubblications

Conclusion

The adoption of Kernel density approach is feasible, automatic, and might represent a robust tool for the identification/adoption of a probabilistic definition of the CTV.