Introduction

The WATERFALL randomized controlled trial¹ demonstrated that aggressive fluid resuscitation is associated with fluid overload without improving clinical outcomes compared to moderate fluid resuscitation. However, these fluid resuscitation regimens were administered up to 8 hours after disease diagnosis, and the effect of fluid resuscitation administered before recruitment in the emergency department (FRED) was not analyzed. We aimed to assess the effect of FRED on clinical and safety parameters among patients with acute pancreatitis who participated in the WATERFALL trial.

Aims & Methods

We conducted a post-hoc analysis of the WATERFALL clinical trial¹. FRED was the volume received until recruitment (up till 8 hours from diagnosis) and it was not determined by protocol in the trial but it was registered. Patients were categorized into two groups according to median FRED administered in the whole cohort: a) low volume group: patients who received ≤median FRED before recruitment b) high volume group: patients who received > median FRED before recruitment. The primary efficacy outcome was the development of moderately severe or severe AP during admission. The main safety outcome was the development of fluid overload. As secondary endpoints, we evaluated the influence of FRED in the development of severe AP, local complications, pancreatic or peripancreatic necrosis and its infection, systemic inflammatory syndrome (SIRS) at 24 hours, persistent SIRS (> 48 h), fluid overload symptoms, signs or imaging, intensive care unit (ICU) admission, death during admission and symptom intensity. Prediction of severity was calculated with the Bedside Index Severity in Acute Pancreatitis (BISAP) score, comorbidity with Charlson comorbidity score, and symptom intensity by Patient-Reported Outcome Scale in Acute Pancreatitis (PAN-PROMISE score). We performed a univariate and multivariate analysis adjusted by possible confounding factors. Age, center, sex, BISAP ≥3, gallstone etiology, urea and arm of treatment of the WATERFALL trial were included in the model (those which could potentially influence the administration of different volume rates of FRED and the ones with a p-value <0,1 between both groups in the baseline characteristics bivariate analysis).

Results

186 out of 249 patients of the WATERFALL trial were included for whom data on FRED was available. 93 of the patients were in the aggressive fluid resuscitation arm of treatment and 93 in the moderate fluid resuscitation arm of treatment. There were statistically significant differences amongst both groups in gallstone etiology and urea (p<0.05). Median FRED was 700 mL. 94 (50,5%) of the cohort received low volume FRED (FRED ≤700 ml) and 92 patients (49,5%) received high volume FRED (FRED >700 ml).There were no statistically significant differences in the main efficacy outcome with 20 patients (21,3%) with moderately severe to severe AP in low volume group and 18 (19,6%) in high volume group in the univariate or multivariate analysis with an adjusted Odds Ratio (IC 95%) of 0,85 (0,39-1,86) and p=0.68. No statistically significant differences were found regarding fluid overload, with 15 patients (16%) in low volume group and 9 patients (9.8%) in high volume group in the univariate or multivariate analysis with and adjusted Odds Ratio of 0.75 (0.27-2.10) and p=0.58. Regarding secondary outcomes, we did not find statistically significant differences except for symptom intensity at recruitment. Higher intensity symptoms were associated to high volume FRED at recruitment with an adjusted Odds Ratio of 2.12 (1.12-4.01) and p=0.02.

Conclusion

To conclude, high volume FRED is not linked to improved clinical outcomes in AP nor does it pose a higher risk of fluid overload.

References

1.- de-Madaria E, Buxbaum JL, Maisonneuve P, et al. Aggressive or Moderate Fluid Resuscitation in Acute Pancreatitis. N Engl J Med 2022;387:989-1000.

Introduction

Pancreatic cancer (PC) is characterized by an aggressive biology and high tumor heterogeneity causing considerable variations in therapy response. Patient-derived organoids (PDOs) reflect parental tumor features and may represent a powerful preclinical tool to predict drug response and advance personalized treatment.

Aims & Methods

We have generated a living organoid biobank with >100 PDO lines derived from treatment-naïve and pretreated PC patient primary tumor and metastases with a reliable efficacy, and previously developed a pharmacotyping-guided prediction model to prognosticate patient therapy response. Following up our initial feasibility trial (Beutel, 2021), we now validated the model accuracy in real-life in a higher number of PC patients (41 cases). Additionally, a limited set of 6 PDO pairs were derived from the same patients at two distinct timepoints, followed by pharmacotyping and whole exome sequencing (WES).

Results

Real-life chemotherapy responses in 41 patients matched to pharmacotyping-guided predictions with overall 80.0% accuracy, 89.3% sensitivity, and 64.7% specificity. Particularly, successful response prediction for first and second-line regimens culminated with an accuracy of 85.7% in naïve patients. Prediction power was nevertheless lower in pretreated patients with a precision of 58.3% for subsequent chemotherapy lines. Retrospective analysis of patient clinical data finally showed that the administration of a therapy prognosticated to be efficient ultimately translated into a longer progression-free survival. Moreover, the implementation of an automated pipetting system and the subsequent miniaturization of screenings enhanced our process capacity, conferring the possibility to extend our drug panel to a second list of approved targeted agents and significantly reducing the time before pharmacotyping.
The access to longitudinal biopsies allowed to conduct WES on 12 PDOs to capture a comprehensive genetic profiling over the time of treatment, notably revealing a CHEK2-mutated patient responding over time upon PARP inhibitor maintenance therapy, in line with our PDO-based prediction, highlighting the robustness of our method and algorithm. Tracking clonal evolution in longitudinal biopsies by WES revealed lower mutational burden and therapy-driven genetic alterations upon progression. Finally, to decipher molecular mechanisms driving PC chemoresistance with an unprecedented accuracy level, we took advantage of our phenotypically and therapeutically-profiled PDO living biobank and employed single-nucleus multiomic approaches, capturing simultaneously the transcriptome as well as the chromatin landscape, of chemoresponsive and non-responsive PC PDOs and of a unique longitudinally-collected case study. This allowed us to identify transcriptomic and epigenetic variabilities in a longitudinal setting across several sequential therapeutic lines.

Conclusion

We report a robust and clinically-relevant companion tool for therapy-response prediction, one more step towards a PDO therapeutic profiling-guided personalized medicine in clinical routine. Additionally, uncovering expression signatures driving chemoresistance may strikingly enhance patient stratification and therefore advance precision oncology in PC.