Beyond Ammann's pain classification: multidimensional pain phenotyping and cluster analysis in chronic pancreatitis

Abstract Assessment of pain in chronic pancreatitis (CP) has largely focused on intensity and pattern, unable to address its complexity. To evaluate pain in a multidimensional fashion, we aimed to identify phenotypes based on the Comprehensive Pain Assessment Tool Short Form (COMPAT-SF) questionnaire and examine their associations with clinical factors. A cross-sectional study including 248 patients with painful CP from Asia, Europe, and the United States was performed. A cluster analysis including the 5 pain dimensions from the COMPAT-SF questionnaire (severity, fluctuation, provocative factors, spreading pain, and qualitative descriptors) identified pain phenotypes. The phenotypes were compared to demographic and clinical data, including patient-reported outcomes and quantitative sensory testing. Three phenotypes were identified in the cluster analysis: a low-burden phenotype, cluster 1 (n = 151); a high-intensity, constant pain phenotype, cluster 2 (n = 75); and a widespread pain, multidimensional phenotype, cluster 3 (n = 22). Quality of life and sleep scores were worse in cluster 3 than in the other phenotypes (all P < 0.001). The degree of anxiety, depression, and catastrophizing was also worse in cluster 3 (all P < 0.001). Cluster 3 showed increased hyperalgesia on sensory testing with a lower sum of pressure pain detection thresholds than cluster 1 ( P = 0.008) and higher temporal summation than cluster 2 ( P = 0.023). The COMPAT-SF questionnaire thereby identified 3 clinically relevant phenotypes in CP. Widespread, multidimensional pain correlated with increased hyperalgesia, higher psychological distress, and worse overall well-being. Phenotyping based on the COMPAT-SF questionnaire may prove helpful in guiding treatment plans and more accurately allocating patients in clinical trials.