The Diagnosis of Chronic Pancreatitis

WP 1: Exocrine function testing

Aims:

Evaluating an endoscopic short test in patients with suspected pancreatic disease. Main focus on Chronic pancreatitis. (Collaborations to other patient groups in other protocols and to a group of healthy controls. )

Methods:

Secretintsimulated Endoscopic short test With duodenal sampling 30-45 minutes after secretin stimulation. Measurement of duodenal enzymes and duodenal bicarbonate concentration.

Inclusion:

Consecutive patients referred to a specialist outpatient clinic under suspicion of CP.

End Points:

Duodenal bicarbonate and enzyme concentrations. CP diagnosis by Layer/ Mayo score. Statistics: Simple comparishments of grpup means. Accuracy calculations by ROC curves. Sensitivity/ specificity calculations.

Sample size: Power and number of patients at baseline are calculated based on the following assumptions: The smallest difference between the pancreatic sufficient and the insufficient groups rejecting the null hypothesis are estimated to 25% regarding peak bicarbonate in EST. The worst case standard deviation is chosen 25%. Sample sizes of 14 patients in each group are expected to give the desired power of at least 0.80. Sample size: At least 15 patients in each group.

WP 2: Testing of Advanced ultrasound Methods

Aims:

Evaluate the diagnostic accuracy of transabdominal ultrasound (US) compared to CP diagnosis by Layer/ Mayo score, Endoscopic ultrasound (EUS), previous CT/ MRI examinations and exocrine pancreatic function.

Sample size:

Power and number of patients at baseline are calculated based on the following assumptions: The smallest difference between the pancreatic sufficient and the insufficient groups rejecting the null hypothesis are estimated to 35% regarding peak bicarbonate in traditional criteria counting. (average 4 criteria vs average 2 criteria) The worst case standard deviation is chosen 25%. Sample sizes of 14 patients in each group are expected to give the desired power of at least 0.80.

Sample size for interobserver variation analysis: Calculated by the principles of A Cantor(1). Assuming a relative error of 30% (Due to earlier observed large interobserver variations in these studies) , a clinical relevant kappa of 0.50 the estimated sample size must be above 44. We intended to include at least 60 patients with both US and EUS. In this setting kappa values below 0.4 must be interpreted with caution.

Hypothesis:

- US can demonstrate changes in CP

- US can diagnose CP with good accuracy.

- US can demonstrate minimal changes in CP

- US findings have good agreement compared to CP

- Changes by US correlate to severity by exocrine pancreatic function.

Inclusion criteria: : Consecutive adult patients (<16 years) referred to a specialized outpatient clinic with suspected chronic pancreatitis (CP) due to presenting symptoms or classical CP characteristics based on previous diagnostic imaging.

Exclusion criteria:

- Patients unable to give informed consent.

- Patients who had undergone extensive pancreatic surgery

In protocol exclusions pre-analysis:

- Subjects who did not fulfil the protocol for an adequate Mayo score

- Subjects with insufficient US visualization of the pancreas due to obesity, repeatedly overlying bowel air or other factors.

Methods:

- Collection of clinical data and previous CT/ MRI examinations from electrical patient journal. Nutritional screening with Height/ Weight/ BMI. Laboratory tests. Blood for DNA in Biobank,

- Exocrine function tests. Short endscopic secretintest/ Faecal Elastase

- Transabdominal Ultrasound: On inclusion and repeated 1-2 times within 3 months. Operators blinded to clinical data.

Registration of Rosemont criteria. Contrast enhanced ultrasound (CEUS) by Sonovue.

Endoscopic ultrasound: blinded to the results of the US. (Subjects offered conscious sedation with intravenous midazolam and pethidine or fentanyl hydrochloride during the test procedure) Pancreas Scan from head to tail. Parenchymal and ductal criteria registered.

Trial chronology:

1. In advance: Invited to participate. Written information in advance.

2. 1. Visit: Signed written consent. Anamnesis and nutritional score.

o Ultrasound and EST

3. 2. Visit: Information of result.

o Repeated ultrasound with registration of Rosemont criteria and CEUS

4. 3. Visit: EUS with Rosemont score.

5. 4. Visit: Follow up, Information of results. Plan further follow up and treatment.

6. Post analysis: Layer/ Mayo score. Rosemont/ traditional score of US and EUS

Endpoints:

CP diagnosis by Layer/ Mayo score Imaging scores: Traditional criteria count and Rosemont weighted score. Exocrine function results Endocrine function results Nutritional status results.

Statistical analysis:

Comparisons between groups are made using student t-tests or Mann-Whitney U-test as appropriate.

Accuracy data calculated from receiver operator curves (ROC). 5% level of statistical significance will be used. Interobserver variation for the separate criteria will be calculated as Cohens kappa.

Reliability for the ordinal data will be calculated as intraclass correlation coefficients (ICCs) in a random, two-way analysis. Data analyzed according to absolute agreement.

Agreement will be defined according to Landis and Koch: 0 = no agreement, 0 - 0.20 = slight agreement, 0.21 - 0.40 = fair agreement, 0.41 - 0.60 = moderate agreement, 0.61 - 0.80 = substantial agreement, and 0.81--1 = almost perfect agreement. A clinically relevant agreement usually requires a value >0.5.

Inclusion period: Starting from des 2009.

Reference List

1. Cantor A. Sample-Size calculations for Cohens Kappa. Psychological Methods 1996;1:150-153.