Predictors of Response to EUS-Guided Celiac Plexus Blockade for Abdominal Pain in Patients with Chronic Pancreatitis

Abstract

Background: The pain from chronic pancreatitis can be debilitating for patients and a difficult symptom for physicians to treat. Celiac plexus blockade (CPB) is utilized for palliation of refractory abdominal pain from chronic pancreatitis. Existing studies have shown only a marginal efficacy of CPB for chronic pancreatitis, and only two studies have specifically investigated predictors of response to CPB. Aim: To evaluate specific patient characteristics associated with a clinical response to EUS-guided CPB in patients with CP and abdominal pain. Patients and Methods: Patients presenting for EUS-guided CPB for CP from 9/05-10/06 were identified by retrospective chart review. The study included patients with pain >3 months duration and objective evidence of CP, defined as at least one of the following: pancreatic endocrine or exocrine insufficiency, >4 criteria for CP on EUS, high suspicion of CP by ERCP, abnormal secretin stimulation study, or histologic evidence of CP. Clinical response to CPB was defined as a >3 point drop on a 10-point VAS scale or >30% improvement in baseline pain scores. Variables analyzed included pre and post procedure pain scores, morphine equivalents, EUS criteria for CP, CPB formulation, use of general anesthesia during CPB, prior CPB, alcohol history, illicit drug use, psychiatric history, chronic pain syndromes, prior pancreatic surgery, and histologic diagnosis of CP. Results: 44 EUS-guided CPBs were performed on 31 patients (11 male, mean age = 47). A significant improvement in pain score occurred in 20/31 (65%) patients. The post-procedure pain scores decreased on average by 48%. The mean duration of pain reduction was 11.3 days (range 0-80). Predictors of response included prior pancreatic surgery (odds ratio (OR) 8.36, p = 0.02), histologic diagnosis of chronic pancreatitis (OR 4.19, p = 0.04), and higher dosage steroid formulation (OR 5.0, p = 0.05). Responders tended to have higher pre-procedure narcotic requirements (p = 0.09). Non-responders tended to have longer duration of pain (p = 0.18), history of prior alcohol abuse (p = 0.19), and chronic pain syndromes (p = 0.07). Of standard EUS criteria, only lobularity showed a trend toward relevance as a predictor of response (p = 0.19). Conclusions: Our study demonstrated that previous pancreatic surgery and histologic diagnosis of chronic pancreatitis were strong predictors of clinical response to EUS-guided CPB. Additionally, increasing dosage of corticosteroid formulation led to significant increases in response. More work is needed in delineating which patients with CP are more likely to benefit from EUS-guided CPB. Further prospective studies are warranted.