Impact of pancreatic enzyme replacement therapy (PERT) on clinical outcomes in nonresected pancreatic cancer (PC): Initial results.

Abstract

400 Background: Current NCCN guidelines recommend PERT use in PC patients (pts) with symptoms of EPI. However, little evidence exists regarding clinical outcomes following PERT use in PC pts, especially those with nonresected disease. We present initial results from PERT use in this pt group regarding disease control and symptom improvement. Methods: Pts with initially nonresected PC were obtained for this study from the Virginia Mason PC Database. Eligibility criteria included:1) pathologically diagnosed nonresected adeno PC from January 2014-December 2019 ; 2) no prior PC anticancer therapy (PCRx) of any kind; 3) ≥2 PGSGA (Patient Generated Global Subjective Assessment ) forms completed with initial PGSGA reporting before 30 days following first PCRx. Pts were stratified by PERT vs non-PERT prescription as validated by the EHR. Clinicians tended to prescribe PERT based on either abnormal fecal elastase and/or clinical symptoms (e.g. diarrhea) as in the PGSGA; no formal criteria for PERT usage existed. Pts were considered to have received PERT if prescribed for the majority of time during the assessment period (i.e. > 30 days) and were analyzed based on an "intent to treat" basis without compliance validation. Results: 344 pts were identified via this method ; 207 (60%)/137 (40%) did/did not receive PERT. > 95% received Creon as PERT. Median time from 1st day PCRx to 1st reassessment was 60 days. 79% pts completed PGSGA prior to and 97% within 2 weeks of 1st day PCRx. Pt characteristics were balanced between PERT/ non PERT groups including race (90% white), age (median 68 yrs), sex (M/F 53%/47%), non-metastatic/metastatic disease 24%/76% . BMI distribution (< 18.5 3%, 18.5-25 40%, 25-30 34%, and >30 23%),and albumin distribution (<3 g/dL 6%, 3-3.4 g/dL 12% and ≥3.5 g/dL 82%). However, mean baseline PGSGA score was higher in the PERT vs non PERT group (9.9 overall, 10.9 (95% CI 10.1-11.7) vs 8.2 (95% CI 7.3 - 9.1), p< 0.01). At 1st reassessment, disease control (PR+SD) (83% overall) was greater in the PERT (87%) vs. non PERT (79%) group (p=0.04).Change from baseline PGSGA was favorable overall (Δ-5.0, 95% CI-5.7--4.3), and in all pt subsets, but with greater improvement in the PERT (Δ-6.0) vs. non PERT (Δ -3.4) group (p< 0.001) and in disease controlled (Δ-5.3) vs. non-disease controlled (Δ- 3.2) pts (p=0.033). Disease controlled vs. non-disease controlled PERT pts did not differ in their PGSGA response (Δ-6.2/-4.8, p=0.98) Conclusions: 1) PERT vs non-PERT pts were similar with respect to basic clinical or nutritional parameters (e.g. BMI, albumin) in this pt cohort. 2) Despite having more adverse baseline PGSGA scores, PERT pts were statistically superior to non PERT pts with respect to both frequency of disease control and magnitude of PGSGA response during initial Rx 3) Further investigation of the detail involving PERT usage and clinical outcomes in nonresected PC is warranted from these data.