Pancreatic Secretory Trypsin Inhibitor (SPINK1) Gene Mutation in Patients with Acute Alcohol Pancreatitis (AAP) Compared to Healthy Controls and Heavy Alcohol Users without Pancreatitis

Round 1

Reviewer 1 Report

Acute pancreatitis represents one of the most frequently encountered gastrointestinal disease around the world. Despite its wide degree of prevalence, etiologies for this condition are multiple and differ in terms incidence on a regional basis; historically, the most frequent form were alcohol-related, with gallstones, cancer and hypertriglyceridemia following. Nowadays, a steep spike in biliary and hypetriglyceridemicforms is being observed, but despite that alcohol consumption still presents some diagnostic dilemmas that need to be addressed. The hypothesis behind this study is based on the surprisingly low prevalence of acute or recurrent acute alcoholic pancreatitis (AAP) among populations of heavy drinkers, standing somewhere in between 3-5%, suggesting how additional triggers and/or factors of susceptibility might exist. One of them is a mutation – previously investigated in the literature - the SPINK1-related gene, involved into the inactivation of prematurely activated trypsin from the pancreatic ducts.

The present study firstly aimed at detecting the prevalence of the SPINK1 mutation, by a prospective model of analysisdivided into 3 main arms, in a population of 103 patients that presented with either acute (64) or recurrent acute (43) alcoholic pancreatits in a tertiary hospital in Finland. Beside the study population, 98 individuals with a recorded history of heavy alcohol consumption and negative history for episodes of pancreatitis were enrolled into the study protocol from a local rehab clinic, together with a control group of 1914 individuals coming from 5 University Hospital districts. Inclusion criteria clearly defining the diagnosis of AAP are implemented from the revised Atlanta classification of 2012, while chronic alcohol consumption was defined either by positive history or/and by paraclinical investigations, lowering bias of interpretation. For the same purpose, Odds ratios were calculated using 95% confidence interval. 

Effect of SPINK1 mutation as an independent factor for AAP was evaluated using binomial logistic regression model adjusted for potential confounders. SPINK1 mutation was found in 8.7% of the patients with AAP. The prevalence was significantly lower in healthy controls (3.4 %, OR 2.72;1.32–5.64) and very low in alcoholics without pancreatitis (1.0 %, OR 9.29;1.15–74.74). In a comparison adjusted for potential cofounders between AAP patients and alcoholics, SPINK1 was found to be an independent marker for AAP. Results are firstly presented in both text and table formats for prompt retrieval of data, and then extensively discussed, also pointing out the potentially innovative character of the study, being the first to actually find statistical significance with both a healthy control population (OR 3.16; 1.31–7.61) and a heavy drinker with no history of pancreatitis control group (OR 10.78; 1.26–91.89). The results attained are further strengthened through an additional targeted review of the literature aimed at focusing the attention of the reader to the background of the SPINK1 mutation-related current level of knowledge. As pointed out in the conclusion, SPINK-1 N34DS testing might become routine and standard in the appropriate healthcare and resource setting for patients for recurrent episodes of AAP, as it is currently already recommended for cases with recurrent idiopathic panreatitis, patients under 25 years with RAP, and patients with a postive family history. Nevertheless, further research is warranted in order to set such a standard of diagnostic workup.

Author Response

The authors would like to thank the reviewer for an extensive overview, commentary and insights regarding the manuscript and its broader research field.

Reviewer 2 Report

This paper deals with SPINK1 mutations in AAP. This is a new theme in AAP.

1. There are some knowns SPINK1 mutations, such as p.N34S, c.194+1G>A, and c.194+2T>C. The authors studied only p.N34S, but it is necessary to determine other mutations.

2. On page 5, lines 171-173, alcohol intake per day showed a negative correlation. What is the reason? It contradicts the general relationship. 

3. Page 6, line 185, a possible reason could be the low number of subjects. From the results, SPINK1 mutation seems to be significant when combining alcohol intake. In that case, the reason may not be the low number of subjects.

4. Explain in detail the ‘alcohol portions per sitting’.

5. In Drug use (Table 1), what drugs were taken?

Author Response

The authors would like to thank the reviewer for the revisions and valuable comments.

Replies to each specific comment:

Comment number 1: Thank you for your comment. In our study, we decided to investigate the most widely studied SPINK1 mutation, N34s which is also the most clinically used as well. As we proposed in the manuscript, it would be interesting to investigate a broader sample of genetic variants - preferably in a genome-wide association study. This was further discussed in page 6, paragraph 4 of the manuscript (discussion section) as well as mentioned as a limitation of our study.

Comment number 2: Thank you for the observation. In the alcoholics group, the alcohol consumption was much higher than in the AAP group (alcohol g/day 149.7 vs 60.7, p<0.001, table 1), which explains the negative correlation. The alcoholics group is a selected population with no history of AAP. The selection was done specifically to study the prevalence of SPINK1 in this population - do they have a low number of SPINK1 mutations which decreases their risk for AAP? This was our hypotheses which our study showed to be accurate.

Comment number 3: Indeed, SPINK1 did not reach statistical significance in the multivariate analysis when comparing AAP patients to the control population. The need for an accumulation of risk factors to initiate AAP was further discussed in page 6, paragraph 3 in the discussion section. A conclusion that we made in Page 6, line 185 "a possible reason could be the low number of subjects" is not sufficiently supported by evidence as the reviewer pointed out, so we decided to remove this sentence from the manuscript.

Comment number 4: Thank you for pointing this out. To elucidate this, we revised the manuscript accordingly, see page 2, paragraph 2 in the materials and methods sections: "- - reported portions per sitting; amount of alcoholic portions containing an equivalent of 14g of pure alcohol on an average day in the last 4 weeks - -"

Comment number 5: We collected self-reported data on illicit drug use from all groups, the specific kind of illicit drug that was used was not specified. To clarify this, we revised the manuscript in page 2, paragraph 2 in the materials and methods section: "- - illicit drug (any illicit drug use, self-reported) and tobacco use - -". 

We hope that we replied to the reviewers’ comments sufficiently and would be happy discuss them further if necessary.