Prognostic role of a novel clinical-nutritional index in pancreatic ductal adenocarcinoma: The Pancreatic Adenocarcinoma Nutritional-Clinical Index (PANCIN)

697

Background: Impaired nutritional status is often associated with Pancreatic Ductal Adenocarcinoma (PDAC) and poor prognosis. Little is known on the prognostic role of nutritional variables in PDAC patients (pts) receiving chemotherapy (CT). Methods: Locally advanced or metastatic PDAC pts enrolled at our Institute in a prospective observational study (PAC-MAIN) and treated with 1st-line CT between April 2019 and July 2021 were included in the analysis. Clinical and nutritional variables entailed biohumoral parameters, bioimpedance vector analysis (BIVA)- and Computed Tomography-derived body composition. Progression-free and Overall survival (PFS and OS) were calculated from CT start to progression or death. A Multivariate Cox proportional-hazards model for PFS prediction was generated by backward selection of features with a p-value (p) ≤ 0.06. The resultant index, named PANCIN, was calculated as linear combination of the covariates (Xi are the N) and the b Cox coefficients (bi), according to the formula: PANCIN = ∑Ni=1 bi Xi Kaplan–Meier test was performed to assess the ability of PANCIN to stratify pts according to its median value for PFS and OS prediction. Results: 74 pts were included in the study. The variables retained in the model were: serum Vitamin B12 (pg/ml) [Hazard Ratio (HR)= 1.001, 95% Confidence Interval (CI) 1.0004-1.0016, p=0.002]; BIVA-derived Body Cell Mass (%) [HR= 0.94, 95% CI 0.887-1.002, p=0.058]; ECOG Performance Status (0 vs 1-2) [HR= 3.25, 95% CI 1.048-10.077, p=0.041]; Albumin (g/L) [HR= 0.91, 95% CI 0.86-0.97, p=0.002]; FAACT Score [HR= 1.041, 95% CI 1.006-1.077, p=0.022]. Median PFS was 15.3 (95% CI 7.6-21.8) and 5.8 months (95% CI 2.7-9.0) for pts with PANCIN < or ≥ -1.7768 median value respectively [HR= 3.7, 95% CI 1.9-7.0, p=0.0001]. Median OS was 23.8 (95% CI 12.6-33.3) and 10.0 months (95% CI 5.7-12.7) for pts with PANCIN < or ≥ -1.7768 respectively [HR= 4.1, 95% CI 2.1-7.9, p<0.0001]. Conclusions: PANCIN is a novel nutritional-clinical index able to predict outcome of advanced PDAC pts receiving CT. If furtherly validated, it may represent a stratification tool both in clinical practice and in prospective trials. Our findings also support the relevance of a baseline comprehensive nutritional assessment, to define tailored nutritional interventions.[Table: see text]

Stomach and duodenum dose-volume constraints for locally advanced pancreatic cancer patients treated in 15 fractions in combination with chemotherapy

Purpose

To assess dosimetry predictors of gastric and duodenal toxicities for locally advanced pancreatic cancer (LAPC) patients treated with chemo-radiotherapy in 15 fractions.

Methods

Data from 204 LAPC patients treated with induction+concurrent chemotherapy and radiotherapy (44.25 Gy in 15 fractions) were available. Forty-three patients received a simultaneous integrated boost of 48-58 Gy. Gastric/duodenal Common Terminology Criteria for Adverse Events v. 5 (CTCAEv5) Grade ≥2 toxicities were analyzed. Absolute/% duodenal and stomach dose-volume histograms (DVHs) of patients with/without toxicities were compared: the most predictive DVH points were identified, and their association with toxicity was tested in univariate and multivariate logistic regressions together with near-maximum dose (D_0.03) and selected clinical variables.

Results

Toxicity occurred in 18 patients: 3 duodenal (ulcer and duodenitis) and 10 gastric (ulcer and stomatitis); 5/18 experienced both. At univariate analysis, V44cc (duodenum: p = 0.02, OR = 1.07; stomach: p = 0.01, OR = 1.12) and D_0.03 (p = 0.07, OR = 1.19; p = 0.008, OR = 1.12) were found to be the most predictive parameters. Stomach/duodenum V44Gy and stomach D_0.03 were confirmed at multivariate analysis and found to be sufficiently robust at internal, bootstrap-based validation; the results regarding duodenum D_0.03 were less robust. No clinical variables or %DVH was significantly associated with toxicity. The best duodenum cutoff values were V44Gy < 9.1 cc (and D_0.03 < 47.6 Gy); concerning the stomach, they were V44Gy < 2 cc and D_0.03 < 45 Gy. The identified predictors showed a high negative predictive value (>94%).

Conclusion

In a large cohort treated with hypofractionated radiotherapy for LAPC, the risk of duodenal/gastric toxicities was associated with duodenum/stomach DVH. Constraining duodenum V44Gy < 9.1 cc, stomach V44Gy < 2 cc, and stomach D_0.03 < 45 Gy should keep the toxicity rate at approximately or below 5%. The association with duodenum D_0.03 was not sufficiently robust due to the limited number of events, although results suggest that a limit of 45-46 Gy should be safe.

Prediction of the characteristics of aggressiveness of pancreatic neuroendocrine neoplasms (PanNENs) based on CT radiomic features

OBJECTIVES

To predict tumor grade (G1 vs. G2/3), presence of distant metastasis (M+), metastatic lymph nodes (N+), and microvascular invasion (VI) of pancreatic neuroendocrine neoplasms (PanNEN) based on preoperative CT radiomic features (RFs), by applying a machine learning approach aimed to limit overfit.

METHODS

This retrospective study included 101 patients who underwent surgery for PanNEN; the entire population was split into training (n = 70) and validation cohort (n = 31). Based on a previously validated methodology, after tumor segmentation on contrast-enhanced CT, RFs were extracted from unenhanced CT images. In addition, conventional radiological and clinical features were combined with RFs into multivariate logistic regression models using minimum redundancy and a bootstrap-based machine learning approach. For each endpoint, models were trained and validated including only RFs (RF_model), and both (radiomic and clinicoradiological) features (COMB_model).

RESULTS

Twenty-five patients had G2/G3 tumor, 37 N+, and 14 M+ and 38 were shown to have VI. From a total of 182 RFs initially extracted, few independent radiomic and clinicoradiological features were identified. For M+ and G, the resulting models showed moderate to high performances: areas under the curve (AUC) for training/validation cohorts were 0.85/0.77 (RF_model) and 0.81/0.81 (COMB_model) for M+ and 0.67/0.72 and 0.68/0.70 for G. Concerning N+ and VI, only the COMB_model could be built, with poorer performance for N+ (AUC = 0.72/0.61) compared to VI (0.82/0.75). For all endpoints, the negative predictive value was good (≥ 0.75).

CONCLUSIONS

Combining few radiomic and clinicoradiological features resulted in presurgical prediction of histological characteristics of PanNENs. Despite the limited risk of overfit, external validations are warranted.

KEY POINTS

• Histology is the only tool currently available allowing characterization of PanNEN biological characteristics important for prognostic assessment; significant limitations to this approach exist. • Based upon preoperative contrast-enhanced CT images, a machine learning approach optimized to favor models' generalizability was successfully applied to train predictive models for tumor grading (G1 vs. G2/3), microvascular invasion, metastatic lymph nodes, and distant metastatic spread. • Moderate to high discriminative models (AUC: 0.67-0.85) based on few parameters (≤ 3) showing high negative predictive value (0.75-0.98) were generated and then successfully validated.

High weekly integral dose and larger fraction size increase risk of fatigue and worsening of functional outcomes following radiotherapy for localized prostate cancer

Introduction

We hypothesized that increasing the pelvic integral dose (ID) and a higher dose per fraction correlate with worsening fatigue and functional outcomes in localized prostate cancer (PCa) patients treated with external beam radiotherapy (EBRT).

Methods

The study design was a retrospective analysis of two prospective observational cohorts, REQUITE (development, n=543) and DUE-01 (validation, n=228). Data were available for comorbidities, medication, androgen deprivation therapy, previous surgeries, smoking, age, and body mass index. The ID was calculated as the product of the mean body dose and body volume. The weekly ID accounted for differences in fractionation. The worsening (end of radiotherapy versus baseline) of European Organisation for Research and Treatment of Cancer EORTC) Quality of Life Questionnaire (QLQ)-C30 scores in physical/role/social functioning and fatigue symptom scales were evaluated, and two outcome measures were defined as worsening in ≥2 (WS2) or ≥3 (WS3) scales, respectively. The weekly ID and clinical risk factors were tested in multivariable logistic regression analysis.

Results

In REQUITE, WS2 was seen in 28% and WS3 in 16% of patients. The median weekly ID was 13.1 L·Gy/week [interquartile (IQ) range 10.2-19.3]. The weekly ID, diabetes, the use of intensity-modulated radiotherapy, and the dose per fraction were significantly associated with WS2 [AUC (area under the receiver operating characteristics curve) =0.59; 95% CI 0.55-0.63] and WS3 (AUC=0.60; 95% CI 0.55-0.64). The prevalence of WS2 (15.3%) and WS3 (6.1%) was lower in DUE-01, but the median weekly ID was higher (15.8 L·Gy/week; IQ range 13.2-19.3). The model for WS2 was validated with reduced discrimination (AUC=0.52 95% CI 0.47-0.61), The AUC for WS3 was 0.58.

Conclusion

Increasing the weekly ID and the dose per fraction lead to the worsening of fatigue and functional outcomes in patients with localized PCa treated with EBRT.

Radioluminescence imaging feasibility for robotic radiosurgery field size quality assurance

PURPOSE

To investigate the feasibility of radioluminescence imaging (RLI) as a novel 2D quality assurance (QA) dosimetry system for CyberKnife®.

METHODS

We developed a field size measurement system based on a commercial complementary metal oxide semiconductor (CMOS) camera facing a radioluminescence screen located at the isocenter normal to the beam axis. The radioluminescence light collected by a lens was used to measure 2D dose distributions. An image transformation procedure, based on two reference phantoms, was developed to correct for projective distortion due to the angle (15°) between the optical and beam axis. Dose profiles were measured for field sizes ranging from 5 mm to 60 mm using fixed circular and iris collimators and compared against gafchromic (GC) film. The corresponding full width at half maximum (FWHM) was measured using RLI and benchmarked against GC film. A small shift in the source-to-surface distance (SSD) of the measurement plane was intentionally introduced to test the sensitivity of the RLI system to field size variations. To assess reproducibility, the entire RLI procedure was tested by acquiring the 60 mm circle field three times on two consecutive days.

RESULTS

The implemented procedure for perspective image distortion correction showed improvements of up to 1 mm using the star phantom against the square phantom. The FWHM measurements using the RLI system indicated a strong agreement with GC film with maximum absolute difference equal to 0.131 mm for fixed collimators and 0.056 mm for the iris. A 2D analysis of RLI with respect to GC film showed that the differences in the central region are negligible, while small discrepancies are in the penumbra region. Changes in field sizes of 0.2 mm were detectable by RLI. Repeatability measurements of the beam FWHM have shown a standard deviation equal to 0.11 mm.

CONCLUSIONS

The first application of a RLI approach for CyberKnife® field size measurement was presented and tested. Results are in agreement with GC film measurements. Spatial resolution and immediate availability of the data indicate that RLI is a feasible technique for robotic radiosurgery QA.

Knowledge-based multi-institution plan prediction of whole breast irradiation with tangential fields

PURPOSE

To quantify inter-institute variability of Knowledge-Based (KB) models for right breast cancer patients treated with tangential fields whole breast irradiation (WBI).

MATERIALS AND METHODS

Ten institutions set KB models by using RapidPlan (Varian Inc.), following previously shared methodologies. Models were tested on 20 new patients from the same institutes, exporting DVH predictions of heart, ipsilateral lung, contralateral lung, and contralateral breast. Inter-institute variability was quantified by the inter-institute SD_int of predicted DVHs/Dmean. Association between lung sparing vs PTV coverage strategy was also investigated. The transferability of models was evaluated by the overlap of each model's geometric Principal Component (PC1) when applied to the test patients of the other 9 institutes.

RESULTS

The overall inter-institute variability of DVH/Dmean ipsilateral lung dose prediction, was less than 2% (20%-80% dose range) and 0.55 Gy respectively (1SD) for a 40 Gy in 15 fraction schedule; it was < 0.2 Gy for other OARs. Institute 6 showed the lowest mean dose prediction value and no overlap between PTV and ipsilateral lung. Once excluded, the predicted ipsilateral lung Dmean was correlated with median PTV D99% (R² = 0.78). PC1 values were always within the range of applicability (90th percentile) for 7 models: for 2 models they were outside in 1/18 cases. For the model of institute 6, it failed in 7/18 cases. The impact of inter-institute variability of dose calculation was tested and found to be almost negligible.

CONCLUSIONS

Results show limited inter-institute variability of plan prediction models translating in high inter-institute interchangeability, except for one of ten institutes. These results encourage future investigations in generating benchmarks for plan prediction incorporating inter-institute variability.

Knowledge-based automatic plan optimization for left-sided whole breast tomotherapy

Background/Purpose

Tomotherapy may deliver high-quality whole breast irradiation at static angles. The aim of this study was to implement Knowledge-Based (KB) automatic planning for left-sided whole breast using this modality.

Materials/Methods

Virtual volumetric plans were associated to the dose distributions of 69 Tomotherapy (TT) clinical plans of previously treated patients, aiming to train a KB-model using a commercial tool completely implemented in our treatment planning system. An individually optimized template based on the resulting KB-model was generated for automatic plan optimization. Thirty patients of the training set and ten new patients were considered for internal/external validation. Fully-automatic plans (KB-TT) were generated and compared using the same geometry/number of fields of the corresponding clinical plans.

Results

KB-TT plans were successfully generated in 26/30 and 10/10 patients of the internal/external validation sets; for 4 patients whose original plans used only two fields, the manual insertion of one/two fields before running the automatic template was sufficient to obtain acceptable plans. Concerning internal validation, planning target volume V_95%/D_1%/dose distribution standard deviation improved by 0.9%/0.4Gy/0.2Gy (p < 0.05) against clinical plans; Organs at risk mean doses were also slightly improved (p < 0.05) by 0.07/0.4/0.2/0.01 Gy for left lung/heart/right breast/right lung respectively. Similarly satisfactory results were replicated in the external validation set. The resulting treatment duration was 8 ± 1 min, consistent with our clinical experience. The active planner time per patient was 5–10 minutes.

Conclusion

Automatic TT left-sided breast KB-plans are comparable to or slightly better than clinical plans and can be obtained with limited planner time. The approach is currently under clinical implementation.

Atlas-Based Lung Segmentation Combined With Automatic Densitometry Characterization In COVID-19 Patients: Training, Validation And First Application In A Longitudinal Study

Purpose

To develop and validate an automated segmentation tool for COVID-19 lung CTs. To combine it with densitometry information in identifying Aerated, Intermediate and Consolidated Volumes in admission (CT1) and follow up CT (CT3).

Materials and Methods

An Atlas was trained on manually segmented CT1 of 250 patients and validated on 10 CT1 of the training group, 10 new CT1 and 10 CT3, by comparing DICE index between automatic (AUTO), automatic-corrected (AUTOMAN) and manual (MAN) contours. A previously developed automatic method was applied on HU lung density histograms to quantify Aerated, Intermediate and Consolidated Volumes. Volumes of subregions in validation CT1 and CT3 were quantified for each method.

Results

In validation CT1/CT3, manual correction of automatic contours was not necessary in 40% of cases. Mean DICE values for both lungs were 0.94 for AUTOVsMAN and 0.96 for AUTOMANVsMAN. Differences between Aerated and Intermediate Volumes quantified with AUTOVsMAN contours were always < 6%. Consolidated Volumes showed larger differences (mean: −95 ± 72 cc). If considering AUTOMANVsMAN volumes, differences got further smaller for Aerated and Intermediate, and were drastically reduced for consolidated Volumes (mean: −36 ± 25 cc). The average time for manual correction of automatic lungs contours on CT1 was 5 ± 2 min.

Conclusions

An Atlas for automatic segmentation of lungs in COVID-19 patients was developed and validated. Combined with a previously developed method for lung densitometry characterization, it provides a fast, operator-independent way to extract relevant quantitative parameters with minimal manual intervention.

Proposal for a New Pathologic Prognostic Index After Neoadjuvant Chemotherapy in Pancreatic Ductal Adenocarcinoma (PINC)

BACKGROUND

Limited information is available on the relevant prognostic variables after surgery for patients with pancreatic ductal adenocarcinoma (PDAC) subjected to neoadjuvant chemotherapy (NACT). NACT is known to induce a spectrum of histological changes in PDAC. Different grading regression systems are currently available; unfortunately, they lack precision and accuracy. We aimed to identify a new quantitative prognostic index based on tumor morphology.

PATIENTS AND METHODS

The study population was composed of 69 patients with resectable or borderline resectable PDAC treated with preoperative NACT (neoadjuvant group) and 36 patients submitted to upfront surgery (upfront-surgery group). A comprehensive histological assessment on hematoxylin and eosin (H&E) stained sections evaluated 20 morphological parameters. The association between patient survival and morphological variables was evaluated to generate a prognostic index.

RESULTS

The distribution of morphological parameters evaluated was significantly different between upfront-surgery and neoadjuvant groups, demonstrating the effect of NACT on tumor morphology. On multivariate analysis for patients that received NACT, the predictors of shorter overall survival (OS) and disease-free survival (DFS) were perineural invasion and lymph node ratio. Conversely, high stroma to neoplasia ratio predicted longer OS and DFS. These variables were combined to generate a semiquantitative prognostic index based on both OS and DFS, which significantly distinguished patients with poor outcomes from those with a good outcome. Bootstrap analysis confirmed the reproducibility of the model.

CONCLUSIONS

The pathologic prognostic index proposed is mostly quantitative in nature, easy to use, and may represent a reliable tumor regression grading system to predict patient outcomes after NACT followed by surgery for PDAC.