Introduction: Smoking has been shown to have a dose-dependent, adverse association with chronic pancreatitis with several, smaller studies showing worsening of disease activity, increased opioid analgesic use, and recurrent hospital admissions. We aimed to compare clinical outcomes in chronic pancreatitis patients with and without a history of smoking.
Methods: We queried clinical data from TriNetX, a large national database which aggregates healthcare data from 59 healthcare organizations in the USA comprising over 70 million patients. TriNetX uses ICD-10 codes as well as CPT codes to identify diagnoses and procedures. All patients were adults ≥ 18 years with chronic pancreatitis (due to any cause). Cases had a history of smoking whereas controls were never smokers. Greedy 1:1 nearest neighbor propensity score matching was utilized to match cases and controls for age, gender, race, BMI ≥ 25, type 2 diabetes, hypertension, chronic kidney disease, COPD, alcohol dependence, hyperlipidemia, cholelithiasis and family history of GI malignancies. Odds ratios (OR) and 95% confidence intervals (CI) were calculated for the outcomes of interest.
Results: In the unmatched group, there were 41,432 chronic pancreatitis patients who had a smoking history (cases), in comparison to 66,852 adults with chronic pancreatitis who had no prior smoking history (controls). After propensity matching, cases and controls were evenly matched in terms of age, BMI, alcohol use, and other comorbidities, though a higher number of cases were long term users of opiate analgesics (18.8% vs. 8.4%, p < 0.001). After propensity matching, there was a higher incidence of acute pancreatitis flare (12% vs. 8%, OR = 1.52, 95% CI = 1.43, 1.62). Similarly, the incidence of walled-off pancreatic necrosis, inpatient admissions, ICU admissions and adjusted 5-year all-cause mortality were all higher among cases when compared to controls (Table 1). We did not find a significant difference in the incident risk of diabetes, however the risk of pancreatic cancer was surprisingly lower among cases versus controls (Table 1).
Discussion: Patients with chronic pancreatitis who are smokers have worse outcomes when compared to chronic pancreatitis patients who are not smokers. The reason for the lower incidence of pancreatic cancer among smokers with chronic pancreatitis is unclear and is in contrast with existing literature. Larger long term prospective studies would be necessary to clarify these associations.
