Society: AGA

Background Chronic pancreatitis (CP) may severely impact quality of life (QoL). Since CP is a chronic condition, multiple assessments of QoL are required to obtain a thorough understanding of its impact on patients. Such studies are currently lacking. Therefore, the aim of the present study is to gain insight into the course and predictors of both physical and mental QoL in patients with CP using prospective longitudinal data from a large cohort of patients.
Methods Post-hoc analysis of patients with CP registered in a prospective database between 2011 and 2019. Patient and disease characteristics, nutritional status, pain severity, medication usage, pancreatic function and pancreatic interventions were assessed from patients’ medical records and through standard follow-up questionnaires. The physical (PCS) and mental component summary (MCS) scales of the Short-Form 36 were used to assess QoL at baseline and during follow-up. The course of both physical and mental QoL and their associated factors were longitudinally assessed by using generalized linear mixed models (GLMM).
Results Overall, 1,165 patients with CP were included for this analysis. During 10-year follow-up, GLMM analyses revealed improvements in both PCS (41.6 to 45.1, p < .001) and MCS (45.8 to 48.5, P = 0.041). Older age, no alcohol consumption, unemployment, need for dietetic consultation, steatorrhea, higher Izbicki pain score and pain coping mechanism were negatively associated with physical QoL (P < .05). For mental QoL, a positive correlation was found between employment, non-alcoholic CP, no need for dietetic consultation, no steatorrhea, lower Izbicki pain score, pain coping mechanism and surgical treatment. No association was observed between disease duration and longitudinal QoL per patient.
Conclusions The present study provides insight into the dynamics of physical and mental QoL in patients with CP over time. Important and potentially influenceable factors to improve QoL in patients with CP are nutritional status, exocrine pancreatic function, employment status, and patients’ coping strategy.

Background: Central sensitization is a dimension of the pain experience in patients with painful chronic pancreatitis (CP) that has been identified by detection of specific patterns of hyperalgesia via pancreatic quantitative sensory testing (P-QST). Patients with hyperalgesia respond poorly to conventional therapy. Currently, there are no mechanism-based biomarker signatures for central sensitization. This pilot study hypothesized that specific biomarker profiles will associate with the presence of hyperalgesia in patients with painful CP.

Methods: Biomarkers (cytokines, chemokines, peptides, hormones) potentially associated with inflammation and metabolism were measured (all pg/mL) in serum samples from CP patients enrolled in the P-QST study. P-QST was used to classify patients based on the presence of hyperalgesia. Biomarker levels are reported as median and interquartile range. In cases where levels were below the level of detection (LOD), a binary variable (detectable v non-detectable) was created. P-values were calculated in the univariable analysis using the Wilcoxon rank-sum test, and Benjamini & Hochberg method for control of false discovery rate (FDR) was applied. Machine learning techniques including random forests and principal component analysis (PCA) were utilized to identify biomarker patterns (individual or groups) that correlate with presence of hyperalgesia.

Results: Of 39 patients, 22 (56%) had hyperalgesia on P-QST testing. Patients with hyperalgesia had significantly lower levels of Tumor Necrosis Factor-alpha (TNF-α) (1.6 [1.4: 2.4] v 3.2 [2.7: 4.9]; p = 0.0003), interleukin (IL) -17F (55.3 [0.0: 133.3] v 94.2 [54.9: 188.8]; p = 0.035), ghrelin (94.3 [34.4: 140.5] v 140.6 [121.4: 226.4]; p = 0.018), and IL - 27 (115.1 [80.5: 195.7] v 176.3 [114.4: 289.0]; p = 0.042). Using FDR-adjusted p-values, lower TNF-α was significantly associated with presence of hyperalgesia (p<0.01) and ghrelin was marginally significant (p=0.09). On random forests analysis, 56% of patients had TNF-α level <2.5, of those 86% had hyperalgesia; whereas among the 44% of patients whose TNF-α was ≥ 2.5, only 18% had hyperalgesia. On PCA analysis, 3 principal components were able to identify patients with hyperalgesia in increasingly higher proportions.

Conclusions: This pilot study suggests that distinct biochemical profiles may associate with presence of hyperalgesia in CP patients. Several cytokines and hormones with anti-nociceptive properties were higher in subjects that did not have hyperalgesia. Validation in larger populations and further work combining biomarker data with clinical characteristics will help in the development of precision medicine tools to optimize treatments for painful CP.

Introduction:
Pain in CP has multiple mechanisms including neural alterations. Understanding the pain type could guide administration of the appropriate treatment. It is often difficult to identify neuropathic pain in the clinic.
Pancreatic quantitative sensory testing (P-QST) in an evolving technique used to identify different types of neural sensitization. Since this is not routinely available, in this study we aimed to evaluate clinical surrogates that could identify sensitization patterns in CP.
Methods:
In this 3-center study, patients with CP of at least 3-yrs duration irrespective of the pain status were enrolled. Patients with acute exacerbation at presentation, malignancy, narcotic dependence, and antidepressant use were excluded. Included patients were subjected to through clinical evaluation including pain phenotyping (duration, intensity, frequency, patterns and triggers of pain, involvement of new areas, change in pain character), assessment of depression/anxiety (BDI II and HAD), quality of life (EORTC QLQc30), pancreatic morphology (duct size, calcification/calculi, atrophy).
P-QST was performed and the patients with pain were divided into those with segmental and widespread sensitization based on recently published criteria (PMID: 31787527)(Figure 1a). Principal component analysis (PCA) with analysis of similarity (ANOSIM) was performed to evaluate sensitization pattern. Logistic regression was employed to identify clinical surrogates for widespread sensitization. Data were expressed as odd’s ratio with 95% CI..
Results:
We included 424 patients (72 painless/347 painful) between October 2021 to October 2022. Overall, mean (SD) age was 35.01 (12.7)yrs and 276 (65.1%) were males. Patients with pain had significantly higher depression (HAD [p<0.0001]) and anxiety (HADS [p<0.0001]) scores, and, poorer global (p<0.0001), physical (p<0.0001), cognitive (p=0.04), emotional (p=0.04) and role (p=0.002) functions in the EORTC QLQc30 (Figure 1b-h).
P-QST was performed on 294 patients. While 57 (19.4%) patients did not have any sensitization, 74 (25.2%) and 163 (55.4%) had segmental and widespread sensitization respectively. There was significant clustering between the patients with and without widespread sensitization (Bonferroni corrected p =0.009 based on Bray Curtis Distance) (Figure 2a). Indicators of sensitization significantly positively correlated with anxiety and depression, while negatively with global, cognitive, and emotional functions (Figure 2b).
Logistic regression revealed the following surrogates for widespread sensitization: involvement of new areas of pain [OR(95%CI) 3.7 (1.79 to 8.70), p=0.03]; diabetes [OR(95%CI) 2.60(1.49-4.54), p=0.001]; and significant depression [OR(95%CI) 3.25(1.62-9.27); p=0.002].
Conclusion:
We identified clinical surrogates that can be used to suspect widespread sensitization, i.e., neuropathic pain in CP.

Subhaleena Sarkar

Appearances