Needle knife fistulotomy in flat and diverticular papillae: Is it time for redemption?

Jo?o Frnns , , Jorg Cnn , , , Mrt Morir , Gon?lo Alxnrino , Luís Figuiro , Trísio Arújo , Luís Lourn?o , Dvi Hort , , Luís Lops , ,

a Department of Gastroenterology, Santa Luzia Hospital - Unidade Local de Saúde Alto Minho, Viana do Castelo, Portugal

b Department of Gastroenterology, Professor Doutor Fernando Fonseca Hospital, Amadora, Portugal

c Department of Gastroenterology, Nova Medical School/Faculty of Medical Sciences, Lisbon, Portugal

d University Center of Gastrenterology - Hospital Cuf Tejo, Lisbon, Portugal

e Life and Health Sciences Research Institute (ICVS), School of Medicine, University of Minho, Braga, Portugal

f ICVS/3B’s - PT Government Associate Laboratory, Braga/Guimar?es, Portugal

Keywords: Endoscopic retrograde cholangiopancreatography Needle-knife fistulotomy Small papilla Diverticular papilla Complications

ABSTRACT Background: European Society of Gastrointestinal Endoscopy ( ESGE) recommends needle-knife fistulo- tomy (NKF) as the preferred precut technique in cases when standard cannulation techniques fail. Despite scarce scientific evidence, flat and diverticular papillae are thought not to be ideal for NKF, as they are associated with poor outcomes. The present study aimed to determine the outcomes of the use of NKF in relation to flat and intradiverticular papillae. Methods: This prospective multicenter study enrolled consecutive patients, evidencing na?ve flat (group A, n = 49) or diverticular papilla (group B, n = 28), who underwent NKF after failure of standard cannu- lation techniques. Diverticular morphology was subdivided into intradiverticular (group B1, n = 14) and diverticular border papillae (group B2, n = 14), using a previously validated endoscopic classification of the major papilla. The success of biliary cannulation at initial endoscopic retrograde cholangiopancreatog- raphy (ERCP), overall biliary cannulation, overall cannulation time, and the rate of adverse events were assessed in the study. Results: The initial cannulation rates were 93.9%, 64.3% and 71.4% for group A, B1, and B2, respectively ( P = 0.005); overall cannulation rates after a second ERCP were 98.0%, 92.9% and 85.7%, respectively ( P = 0.134). Adverse events occurred in 11.7% of patients, with post-ERCP pancreatitis (PEP) being the most common adverse event (10.4%). Although there was a trend towards a higher incidence of PEP in flat papillae, univariate and multivariate analyses did not show any significant relationship between pan- creatitis and trainee involvement, papillary morphology, nor overall cannulation time. Conclusions: Although flat papillae are associated with high success rates of biliary cannulation using NKF, the rate of PEP is not negligible. NKF is feasible in diverticular papillae, but it is associated with a modest success rate in the initial ERCP.

Introduction

Endoscopic retrograde cholangiopancreatography (ERCP) is an advanced medical procedure widely used in the diagnosis and treatment of a variety of benign and malignant pancreatobiliary disorders [ 1 , 2 ]. Selective cannulation of CBD is the most critical and challenging step in a biliary ERC [ 3 , 4 ]. However, even in ex- perienced hands, biliary cannulation may fail in up to 15%-35% of cases in the first ERCP when standard methods alone are used [5] . In this subset of patients, additional cannulation techniques are necessary to access the CBD in order to continue with the ERCP.

Precut is the most common strategy used by experienced endo- scopists when conventional methods have failed [4] . Needle-knife fistulotomy (NKF) and conventional precut are the two most com- mon variants. Technically NKF is the creation of an artificial fistula between the most protuberant portion of the papilla, representing the intraduodenal portion of CBD, and the biliary tract, by using a diathermic cutter in a freehand fashion [ 3 , 6-12 ]. The major ad- vantage of this technique over conventional precut is that it avoids contact between the cutting device and the papillary orifice, and hence precludes direct thermal injury to the pancreatic duct, re- ducing the risk of post-ERCP pancreatitis (PEP).

Recently published guidelines recommend opting for NKF, as evidence suggests a lower risk of adverse events, especially pan- creatitis, when NKF is used early in the biliary cannulation algo- rithm [ 4 , 6 ]. Even so, it remains a controversial topic since many endoscopists only employ NKF in papillae with enlarged oral pro- trusions [13–15] . The reason for this is that, intuitively, endo- scopists tend to correlate endoscopic findings, such as the oral pro- trusion, with the CBD diameter. Flat papillae, without oral protru- sion, are assumed to be associated with thin CBDs, and tend to be approached by cannulation methods other than NKF due to the high risk of post-procedural adverse events such as perforation and pancreatitis [13] .

The same happens with diverticular papillae, where the papil- lary orifice lies inside the diverticulum or in its border. This mor- phology is known to hinder biliary cannulation [16] . The weak- ness of the duodenal wall, intrinsic to the diverticulum, is prone to perforation, so it naturally dissuades many endoscopists from performing NKF. Another explanation, one devoid of subjectivity, is that the papillary orifice is often hidden within the dome-shaped wall of the diverticulum, precluding an ideal approach for perform- ing a freehand incision.

To the best of our knowledge, no studies have addressed the influence of these two morphologies in the feasibility, success, and complications of NKF. Therefore, the present study aimed to deter- mine the outcomes of NKF in flat and diverticular papillae.

Methods

Patients and setting

This was a prospective multicenter (two ERCP high volume cen- ters) clinical study of all consecutive patients with a na?ve small or diverticular papilla who had undergone NKF for biliary access be- tween March 2018 and June 2020. Patients with surgically altered anatomy or tumors of the papilla were excluded.

Papillary morphology was classified according to a previously validated endoscopic classification of the major papilla, called Viana Classification [17] . It comprises 7 categories ( Fig. 1 ). Type I: flat type, without oral protrusion; type IIA: prominent tubular non-pleated, with oral protrusion and ≤1 transverse fold over the oral protrusion; type IIB: prominent tubular pleated, with oral pro- trusion and ≥2 transverse folds over the oral protrusion; type IIC: prominent bulging, with an enlarged and bulging oral protru- sion; type IIIA: diverticular-intradiverticular, papillary orifice inside the diverticulum; type IIIB: diverticular-diverticular border, papil- lary orifice less than 2 cm from the diverticular border; type IV: unusual aspect, papilla with no morphology classified in the other categories. For the sole purpose of this study, we selected patients with type I (assigned as group A) or type IIIA/B, (assigned as group B1 and B2 respectively) papillary morphologies among all the NKF procedures.

Fig. 1. Viana classification of the major papilla.

The data collected included patient demographics, indication for ERCP, underlying biliary pathology, papillary morphology, inter- ventions throughout the procedure, rate of success of NKF in the first ERCP and overall after a second ERCP in initial failures, total ERCP time until biliary cannulation, intraprocedural adverse events, and post-procedural adverse events (30-day follow-up).

Outcomes and definitions

The primary outcomes were as follows: 1) NKF cannulation rate at the first ERCP for flat, intradiverticular and diverticular border papillae; and 2) adverse events rate. The secondary endpoints in- cluded the overall cannulation rate after a second ERCP attempt if the first ERCP failed, and total ERCP time until biliary cannulation. There was no protocol regarding the initial method of cannulation, including the specific timing throughout the ERCP. These decisions were left within the complete discretion of the senior endoscopist.

Early NKF was defined as the use of NKF immediately after one of the followings: more than 5 contacts with the papilla while attempting to cannulate; more than 5 min spent attempting to cannulate following visualization of the papilla or more than one unintended pancreatic duct cannulation or opacification, which is the recommendation of European Society of Gastrointestinal En- doscopy ( ESGE) for defining difficult biliary cannulation [4] . Late NKF was defined as the use of NKF for biliary cannulation after at least 10 min of unsuccessful standard biliary cannulation attempts as suggested elsewhere [ 18 , 19 ]. If NKF was undertaken between 5 and 10 min of the procedure (not early nor late), it was considered to be regular NKF.

There was no specific protocol for the second ERCP after an ini- tial failure of ERCP, but the endoscopists who performed NKF (J. Canena and L. Lopes) started the second ERCP through the previ- ous existing incision of NKF and if necessary, increased the depth of the incision.

Overall cannulation time was defined as the length of time be- tween the first contact with the papilla and fluoroscopic visualiza- tion of a guidewire into the biliary duct.

Post-ERCP complications were classified and graded according to consensus guidelines [ 4 , 20 ].

Intervention, endoscopists, and PEP prevention

All ERCP procedures were performed with the patients in the prone position. Sedation was achieved with administration of propofol by an anesthesiologist. The NKF procedures were exclu- sively performed by two expert endoscopists (J. Canena and L. Lopes), who achieved selective biliary cannulation in more than 80% of the patients using standard access techniques. Furthermore, the aforementioned endoscopists have an annual caseload above 400 ERCPs and have performed more than 800 NKFs in their ca- reer and more than 60 NKF/year in the past 5 years.

The details of the NKF technique have been extensively de- scribed elsewhere [3–12] . Prophylaxis of PEP was undertaken in all patients. Rectal administration of 100 mg of indomethacin or di- clofenac immediately before the ERCP was performed routinely. In all cases of inadvertent guidewire passage into the pancreatic duct or pancreatic opacification, a prophylactic 5-Fr pancreatic stenting was placed as recommended [ 4 , 6 ].

Statistical analysis

Qualitative variables were summarized using numbers and fre- quencies, and quantitative variables were summarized using the mean and standard deviation or the median and interquartile range (IQR), depending on their distribution profiles. The normal- ity of the quantitative variables was assessed using the histogram distribution. Relations between categorical variables were assessed using Chi-square test or Fisher’s exact test. Differences between two or more groups of quantitative variables were evaluated using a Kruskal Wallis test. To predict pancreatitis, both univariate and multivariate analyses were performed. Three predictor variables were used in the multivariate model: trainees, papillary morphol- ogy, and cannulation time. The null hypothesis was rejected when the test statisticsPvalues were＜0.05. Statistical analysis, sample size calculation, and graphics were performed using Stata software (StataCorp. 2015. Stata Statistical Software: Release 14. College Sta- tion, TX: StataCorp LP).

Results

Patients

During the study period, 1970 na?ve papillae were submitted to ERCP. Of these patients, 310 underwent NKF as a rescue method for biliary cannulation, among which 77 were carried out in small and diverticular papillae. Therefore, 77 patients, including 31 males and 46 females, were enrolled in the study, with a mean age of 72.6 years. The patients’ demographic data, trainee involvement in the procedure (prior to the NKF), and ERCP findings are summarized according to papillary morphology in Table 1 .

Primary outcomes

Primary outcomes are shown in Table 2 . Regarding the cannula- tion rate at the initial ERCP, access to the biliary tree was obtained in 65/77 (84.4%) of the patients. In the flat papillae group (group A), biliary access was obtained in 46/49 (93.9%) of the patients, compared to 9/14 (64.3%) in the intradiverticular group (group B1) and to 10/14 (71.4%) in the diverticular border group (group B2) (P= 0.005).

Table 1 Patient’s demographics, group distribution and group characteristics

Table 2 Primary and secondary outcomes of the study

Table 3 Univariate analysis to predict pancreatitis

Table 4 Multivariate analysis to predict pancreatitis

Early NKF was undertaken in 5 of the flat papillae group and in 4 of the diverticular group (2 of group B1 and 2 of group B2). For the remaining, NKF was undertaken as regular or intermediate, meaning that NKF was performed between the 5th and the 10th minute of the procedure. All of the failed NKF during initial ERCP was undertaken as regular. Timing of NKF was not statistically as- sociated with success (P= 0.121).

Adverse events occurred in 9/77 (11.7%) of the patients. Pan- creatitis was the most common complication, with an overall fre- quency of 10.4% regardless of papilla morphology. Most adverse events, 7/9 (77.8%), occurred in group A, with all being cases of pancreatitis, except for one case of bleeding conservatively treated without blood transfusion. The pancreatitis rate difference among the three groups was not statistically significant (12.2% vs. 0% vs. 14.3%,P= 0.555). The univariate analysis to predict pancreatitis, as shown in Table 3 , showed no correlation between PEP and trainee involvement (P= 0.703), papillary morphology (P= 0.555), nor overall cannulation time (P= 0.665). A multivariate analysis was performed showing similar results as univariate analysis ( Table 4 ). No patients developed perforation nor other complications during or following the procedure.

Secondary outcomes

In the second ERCP, following a failed first attempt, success- ful biliary cannulation increased to 73/77 (94.8%) of the patients: 4 8/4 9 (98.0%) in group A, 13/14 (92.9%) in group B1, and 12/14 (85.7%) in group B2 (P= 0.134). All the 8 patients with successful cannulation at the second ERCP were obtained with primary NKF. These second procedures were performed after a median of 7 days (IQR 5-10).

The median cannulation time ( Fig. 2 ) of the three groups was 10 min, with 13.7 min in group A, 8 min in group B1 and 9.5 min in group B2 (P= 0.184).

Both primary and secondary outcomes are summarized in Fig. 3 .

Fig. 2. Outcomes of needle-knife fistulotomy by papillary morphology.

Fig. 3. Overall cannulation time depending on the papillary morphology. ERCP: endoscopic retrograde cholangiopancreatography; NKF: needle-knife fistulotomy.

Discussion

In this study we observed that NKF was associated with a high success rate in flat papillae, although PEP was not negligible. NKF is feasible in diverticular papillae with modest success rates in the first ERCP, which can be increased to high success rates after a sec- ond ERCP attempt.

In the last two decades, NKF has undergone a remarkable evo- lution within the biliary cannulation algorithm. Once considered a technique of last resort, its use is increasingly advocated at an ear- lier stage throughout ERCP [ 4 , 6 ]. Even so, there are several uncer- tainties regarding this technique, namely its use in certain papil- lary morphologies. The degree of oral protrusion influences the de- cision of whether or not to perform NKF [13–15] since most endo- scopists tend to correlate the degree of oral protrusion to the com- mon bile duct diameter. In this way, flat papillae are not usually approached by NKF in light of the risk of cutting deviated from the duct axis. However, as our team has recently demonstrated, there is no relationship between the longitudinal and transverse dimen- sions of the papilla, including its oral protrusion, and the common bile duct diameter [1] . In other words, it is possible to have a flat papilla with a dilated bile duct, or a bulging papilla with a thin bile duct. On the other hand, diverticular papillae are found in 9%-32% of patients undergoing ERCP [21] , and are also often considered unsuitable for NKF. Factors like the fragility and slenderness of the duodenal wall often dissuade endoscopists from using a freehand cutting technique like NKF given the greater fear of complications such as perforation. Additionally, it is not uncommon to find the papillary orifice concealed in the diverticular dome.

Taking these facts into account, we set out to design a study to assess the effectiveness and safety of NKF in these papillary mor- phologies. Although there are studies that assess the influence of papillary morphology on biliary cannulation, this is one of the few that does so with a rescue technique such as NKF.

Previous attempts have undertaken to devise a classification of the major papilla. Nonetheless the great majority either has not been validated or lacks important morphological forms such as di- verticular and folded papillae [22] . One exception was reported in 2019 by Watanabe et al. [23] , which encompassed 8 types of papillae. Although internally validated by three experienced endo- scopists, the classification was not tested for inter- and intraob- server agreement. Furthermore, the classification is of such a de- gree of complexity that it seems unsuitable for routine clinical practice.

In the present study, the initial ERCP reached a cannulation rate of 84.4%, which is below the overall success reported in the lit- erature (87.4%-97.9%) [ 3 , 8 , 11 , 12 , 14 , 24 ]. However, the reported se- ries avoided flat and diverticular related papillae most of the time. There was a higher rate of initial cannulation in group A (93.9%) when compared to the other two groups (64.3% and 71.4%, respec- tively) (P= 0.005). As previously explained, the papillary orifice that is easily seen in the flat papilla is an important factor favor- ing the feasibility of NKF. In addition, having recently proved that there is no relationship between the dimensions of the papilla and the diameter of the CBD [1] may have acted as a confidence factor in this type of papillae, when compared to diverticular papillae and the inherent weakness of the duodenal wall. Although the number of papillae in group B1 and B2 was fairly small, a greater success in the execution of NKF in papillae at the diverticular border was to be expected, since in these, unlike intradiverticular papillae, the papillary orifice was always visible, and the biliary axis was easier to determine.

After an unsuccessful first ERCP in 12 patients, a second proce- dure proved to be useful in all groups, with success in 8 additional patients increasing the overall biliary cannulation rate from 84.4% to 94.8% (98.0%, 92.9% and 85.7%, respectively). These data are co- herent with the results obtained by Colan-Hernandez et al. [25] , who proved that a second ERCP attempt after precut failure in the first ERCP is safe and effective despite not taking papillary mor- phology into account.

Regarding complications, there were a total of 9 adverse events, corresponding to 8 cases of mild pancreatitis and one case of mild hemorrhage, conservatively treated. Our study, which encom- passed only flat and diverticular papillae, showed a PEP rate of 10.4%, compared to previous studies that showed a PEP rate of up to 11% irrespective of the papillary morphology [ 3 , 8 , 11 , 12 , 14 , 24 ]. With the aim of exploring variables associated with pancreatitis, we carried out a univariate analysis, as shown in Table 3 . Con- trary to Haraldsson et al. [26] , who stated that involvement of trainees increased the chances of difficult cannulation, our study did not show any relationship between the presence of trainees and the occurrence of pancreatitis. Of the 28 procedures in which trainees participated, only two cases of pancreatitis were regis- tered. On the other hand, although the majority of PEP was ob- served in the flat papillae, there was also no statistically signifi- cant relationship between papillary morphology and the develop- ment of pancreatitis. Finally, a higher median cannulation time was observed in the group of patients who had pancreatitis (24.6 min vs. 10.0 min), but the difference was not statistically significant. The same three predictors used in the univariate analysis were applied in the logistic regression ( Table 4 ) and showed no ability to predict PEP. Although the results of our study did not demon- strate it categorically, it seemed that a larger sample would most likely demonstrate a clear relationship between both the presence of trainees and a longer cannulation time with the development of pancreatitis.

Although the median cannulation time was longer in the flat papillae group compared to the intradiverticular group and to the diverticular border group, the difference was not significant. The longer cannulation time in the flat papillae group was due to the greater insistence on standard cannulation methods and conse- quently later use of NKF. The scarce participation of trainees in ERCP with diverticular papillae and the early use of NKF in this morphology may have contributed to a lower rate of PEP in group B1/2. It should be noted that there was no case of perforation even though this was intuitively the most feared complication of NFK in this type of papillae.

Taken together a speculative explanation for the higher rate of pancreatitis in flat papillae could be due to the longer cannula- tion time. Furthermore, NKF avoids thermal injury to the pancre- atic duct but perhaps in small and flat papillae, due to the smaller distance between the two orifices, some thermal injury may be conducted to the pancreatic orifice. Furthermore, NKF is beneficial specially as a primary tool or early in the cannulation algorithm which was not the case in this study.

In 2008 a group from Iran first reported primary fistulotomy as a cannulation method in a randomized clinical trial (RCT) compar- ing a population submitted to primary NKF versus a group of pa- tients submitted to traditional cannulation methods [27] . In 2016, a group from South Korea reported the utility of NKF as an ini- tial method of biliary access in a group of patients with one or more patient-related risk factors for PEP, which they considered to be a population at a high risk of PEP [7] . They enrolled 55 pa- tients and observed no pancreatitis without using indomethacin for PEP prevention. Four years later, the same group conducted an RCT comparing 87 patients submitted to the conventional cannula- tion method versus 96 patients allocated to primary NFK, and all of the patients were considered to be at high risk for pancreatitis [8] . The investigators observed a significant difference in technical suc- cess and complications between the two groups in favor of the NFK group. Again, they reported no pancreatitis in the NKF group, and none of the patients in either group was submitted to PEP prophy- laxis with non-steroidal anti-inflammatory drugs (AINEs). There- fore, primary NKF could be beneficial in patients with flat papilla and average or high PEP risk. However, in a flat papilla the distance between the two orifices is smaller and primary NKF may not be as successful for PEP prevention (due to some thermal injury of the pancreatic duct) as it is on a prominent and bulging papillae. Future studies of primary NKF including flat papillae could help to further clarify this issue.

The present study has several limitations. First was the small sample size. Even so, we have to take into account that the papillary morphologies included in this study are relatively un- common. If we combine this assumption with the fact that NKF is used mostly as a rescue technique, we realize the difficulty in reaching superior samples. Another potential limitation was that all the NKF procedures were performed by two experi- enced ERCP endoscopists. The reproduction of the results ob- tained in our study should therefore always take into account the degree of expertise of the endoscopist in relation to the NKF technique.

The strengths of our study include the fact that it uses a val- idated papilla classification that is the first to accommodate both diverticular and folded morphologies, and the fact that is the first study to evaluate the relationship between the success and feasi- bility of NKF with these types of papillae.

In conclusion, NKF is a feasible and safe rescue cannulation technique in both flat and diverticular papillae, despite a moderate success rate in diverticular papillae at first ERCP. The risk of post- ERCP complications tends to be higher in flat papillae, especially if NKF is delayed.

Acknowledgments

None.

CRediTauthorshipcontributionsstatement

Jo?oFernandes:Conceptualization, Data curation, Formal anal- ysis, Investigation, Methodology, Writing - original draft, Writing - review & editing.JorgeCanena:Conceptualization, Data cura- tion, Formal analysis, Investigation, Methodology, Project admin- istration, Supervision, Validation, Visualization, Writing - original draft, Writing - review & editing.MartaMoreira:Writing - review & editing.Gon?aloAlexandrino:Writing - review & editing.LuísaFigueiredo:Writing - review & editing.TarcísioAraújo:Writing - review & editing.LuísLouren?o:Writing - review & editing.DavidHorta:Writing - review & editing.LuísLopes:Conceptual- ization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Software, Supervision, Validation, Visualiza- tion, Writing - original draft, Writing - review & editing.

Funding

None.

Ethicalapproval

This study was approved by the Ethics Committee of the Santa Luzia Hospital - Unidade Local de Saúde Alto Minho. Written in- formed consent was obtained from all participants.

Competinginterest

No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the sub- ject of this article.