Optimal timing of cholecystectomy after percutaneous gallbladder drainage for severe cholecystitis

Background The Tokyo guideline for acute cholecystitis recommended percutaneous transhepatic gallbladder drainage followed by cholecystectomy for severe acute cholecystitis, but the optimal timing for the subsequent cholecystectomy remains controversial. Methods Sixty-seven patients who underwent either laparoscopic or open cholecystectomy after percutaneous transhepatic gallbladder drainage for severe acute cholecystitis were enrolled and divided into difficult cholecystectomy (group A) and non-difficult cholecystectomy (group B). Patients who had one of these conditions were placed in group A: 1) conversion from laparoscopic to open cholecystectomy; 2) subtotal cholecystectomy and/or mucoclasis; 3) necrotizing cholecystitis or pericholecystic abscess formation; 4) tight adhesions around the gallbladder neck; and 5) unsuccessfully treated using PTGBD. Preoperative characteristics and postoperative outcomes were analyzed. Results The interval between percutaneous transhepatic gallbladder drainage and cholecystectomy in Group B was longer than that in Group A (631 h vs. 325 h; p = 0.031). Postoperative complications occurred more frequently when the interval was less than 216 h compared to when it was more than 216 h (35.7 vs. 7.6%; p = 0.006). Conclusions Cholecystectomy for severe acute cholecystitis was technically difficult when performed within 216 h after percutaneous transhepatic gallbladder drainage. Electronic supplementary material The online version of this article (doi:10.1186/s12876-017-0631-8) contains supplementary material, which is available to authorized users.


Background
Acute cholecystitis (AC) is a common disease for which laparoscopic cholecystectomy (LC) has become the standard treatment [1]. However, in patients with severe AC, the rate of complications such as bile leak, common bile duct injury, and bowel injury is high after LC [2], suggesting an association between severity of inflammation and difficulty of LC. Therefore, evaluation of the severity of AC is important in determining the appropriate treatment.
The Tokyo Guidelines in 2007 was issued as the first international guidelines for the diagnosis and treatment of AC and has been recently revised as the Tokyo Guidelines 2013 (TG13) [3,4]. The TG13 suggested that the criteria for AC should be based on clinical symptoms, physical examination, blood tests, and imaging findings and classified AC into three grades of inflammation: mild (grade I), moderate (grade II), and severe (grade III) ( Table 1). The TG13 also recommended appropriate therapy depending on the grade of AC. Grade I patients are candidates for immediate LC; grade II patients could undergo either LC or percutaneous transhepatic gallbladder drainage (PTGBD); and grade III patients are strongly recommended to undergo immediate PTGBD.
In the early 1980s, Radder introduced the procedure of PTGBD to immediately improve the symptoms of AC [5]. In some patients who undergo PTGBD and subsequent LC, technical difficulties and postoperative complications associated with severe fibrosis and gallbladder adhesion may be encountered [6]. Therefore, it is very important to determine the optimal timing of cholecystectomy after PTGBD; however, this is still controversial [4].
The aim of this study was to assess the effect of the interval between PTGBD and cholecystectomy on the technical difficulty of cholecystectomy and postoperative complications. We also aimed to determine the optimal timing of cholecystectomy after PTGBD.

Methods
The medical records of the 77 study patients who underwent either laparoscopic or open cholecystectomy after PTGBD for AC between 2002 and 2015 were retrospectively reviewed (Fig. 1). Ten patients who met one of the conditions below were excluded from this study: 1) hemorrhage associated with liver cirrhosis, 2) relapsed cholecystitis after PTGBD tube removal, and 3) need for procedures other than cholecystectomy. At least one board-certified surgeon of the Japan Surgical Society participated in all surgeries.
Since there was no quantitative evaluation for the difficulty of cholecystectomy, we employed objective and subjective criteria in this study. Difficult cholecystectomy was defined as the presence of either one of these conditions: 1) conversion from laparoscopic to open cholecystectomy; 2) atypical cholecystectomy, such as subtotal cholecystectomy and/or mucoclasis; 3) necrotizing cholecystitis or pericholecystic abscess formation; 4) tight adhesions around the gallbladder neck; and 5) a patient unsuccessfully treated using PTGBD. In grade I and II cholecystites, timing of surgery following PTGBD was decided by each surgeon. In grade III cholecystitis, surgery was performed after recovery of patient from organ dysfunction. Decision of conversion during surgery depended on each surgeon. There were no definite criteria for performing either open cholecystectomy or laparoscopic cholecystectomy.
Necrotizing cholecystitis was confirmed by intraoperative and pathologic findings. The remaining 67 patients were divided into two groups according to the criteria mentioned above: the difficult cholecystectomy group (group A, n = 35) and the non-difficult cholecystectomy group (group B, n = 32). Preoperative characteristics and postoperative outcomes were analyzed. Detailed data of patient characteristics used in this study was shown in Additional file 1.
Data are shown as mean ± standard error mean (SEM). Categorical variables were analyzed by Chisquare test, whereas continuous data were analyzed by either two-tailed Student's t test or Wilcoxon test, according to the result of the Shapiro-Wilk test. Statistical significance was defined as P < 0.05. The optimal cut-off value was calculated using a receiver operating characteristic (ROC) curve and was defined as the number that indicated the highest sum of the sensitivity and specificity on the ROC curve. Statistical analysis was performed with JMP Pro 11 software (SAS Institute, Cary, NC, U.S.A).  ROC curve analysis yielded a value of 216 h as the optimal cut-off interval between PTGBD and cholecystectomy, in relation to the difficulty of cholecystectomy (Fig. 2). Using this cut-off value, patients were divided into the short interval (SI) (n = 14) and the long interval (LI) (n = 53) groups. The perioperative outcomes are shown in Table 5. The number of patients who underwent difficult cholecystectomy was significantly greater in the SI group than in the LI group (85.7 vs. 43.4%; p = 0.005). Compared with the LI group, the SI group had significantly longer operating time (143 ± 10.9 min vs. 117 ± 5.6 min; p = 0.05); higher rates of necrosis and/or abscess formation (66.7 vs. 30.2%; p = 0.016) and adhesions around the gallbladder neck (50.0 vs. 22.6%; p = 0.043); and higher incidence of postoperative complications, except surgical site infection (35.7 vs. 7.6%; p = 0.006). In particular, intraabdominal abscess developed more frequently in the SI group than in the LI group (21.4 vs. 0%; p = 0.006). Duration of postoperative hospital stay did not differ between the two groups (SI group vs. LI group, 10.6 ± 1.6 days vs. 8.6 ± 0.81 days; p = 0.269).

Discussion
Recently, immediate LC has become standard treatment for AC [7,8]. However, we sometimes experience difficulty during LC in AC patients with severe local inflammation; this can increase the rate of postoperative complications, such as bile leak, common bile duct injury, and bowel injury [2]. Furthermore, some investigators reported that the mortality rate was as high as 18-50% in elderly patients or those with severe comorbidities [9][10][11].
On the other hand, PTGBD is not a difficult procedure with a very low mortality rate [12]. Therefore, PTGBD has been performed as a safer substitute for cholecystectomy in such high-risk patients [13,14]. In this study, the rate of unsuccessful treatment using PTGBD was 3.9%, and this result was acceptable compared with that in a previous report. The TG13 has indicated the severity criteria for AC and its appropriate management, including PTGBD [3,4], but the optimal timing for subsequent cholecystectomy after PTGBD was not mentioned. Chikamori et al. suggested that early LC following PTGBD was safe and effective [15]. On the other hand, Kim et al. suggested that delayed LC following PTGBD decreased the rates of open conversion and complications [16]. Yamada et al. suggested that the interval between PTGBD and cholecystectomy was not related to the amount of blood loss, operating time, and open conversion rate [17]. Therefore, the optimal timing for subsequent cholecystectomy after PTGBD is still controversial.
In this study, grade III AC patients who have organ dysfunction were enrolled. Although organ dysfunction seemed to negatively affect postoperative morbidity and mortality, all of grade III AC patients underwent operation after complete recover from organ dysfunction. Therefore, organ dysfunction did not affect postoperative outcomes.
At the beginning of this retrospective study, we assessed whether the interval between PTGBD and cholecystectomy was a risk factor for a difficult surgery. Based on our results, the interval between PTGBD and cholecystectomy was short in patients who underwent difficult cholecystectomy. Comparison of outcomes according to a cut-off interval of 216 h showed a significantly higher rate of postoperative complications, especially intraabdominal abscess, and prolonged operating time in the SI group than the LI group. These results indicated that cholecystectomy should be considered at least 216 h after PTGBD.
The proper definition of difficult cholecystectomy is crucial in this study. Although conversion to open cholecystectomy reflects a difficult LC, it could also depend on the surgeon's ability. In addition to conversion rate, objective and subjective criteria for difficult cholecystectomy were employed in this study based on results of previous reports. In particular, gallbladder inflammation was a risk factor for technical difficulty and conversion from LC to open cholecystectomy [18][19][20]. We believed that our criteria for difficult cholecystectomy were appropriate. Although BMI, CRP, and the interval between symptom onset and PTGBD were suggested as independent risk factors for difficult cholecystectomy [17][18][19][20][21], these parameters did not differ between difficult and non-difficult cholecystectomy cases in this study.
In this study, early (within 72 h) cholecystectomy following PTGBD was performed only in seven cases. Therefore, the usefulness of early cholecystectomy could not be adequately assessed. AC progresses from edematous cholecystitis through necrotizing cholecystitis to sub-acute cholecystitis in about 10 days, and edematous phase, which is a phase before proceeding tight adhesion, lasts for 72 h after symptom onset [22]. Necrotizing cholecystitis is also reported to be one of the risk factor for difficult LC [23,24]. Thus, the Tokyo Guidelines recommend early LC. Accordingly, early cholecystectomy following PTGBD might reduce the incidence of postoperative complications. Some investigators reported rapid symptomatic improvement in more than 90% patients after cholecystectomy within  [25][26][27]. Indeed, some patients can safely undergo cholecystectomy soon after PTGBD. However, early LC following PTGBD for grade III cholecystitis seems challenging because of the accompanying severe organ dysfunction. Therefore, in patients who do not recover from organ dysfunction within 72 h, we should not perform early LC. One limitation of our study was inclusion of patients who underwent three different procedures (LC, conversion from LC to open cholecystectomy, and open cholecystectomy). Because operating time and blood loss could be confounded by these procedures, these two parameters had to be excluded from the criteria for difficult cholecystectomy. Surgical site infection was not included as a postoperative complication because of the same reason. The combination of three different procedures might also explain the same length of postoperative hospital stay among the patients, regardless of the interval between PTGBD and cholecystectomy.
In this study, the cut-off value of 216-h interval between PTGBD and cholecystectomy was similar with that of a previous report (7 days) [16]. In general, when inflammation occurs, cytokines are produced and they cause fibrosis in 7 days [28]. It is also reported that further fibrosis does not occur after 7 days [29]. Therefore, there is a possibility that waiting for longer time leads to a safer cholecystectomy. However, there is no evidence regarding this, and further examination is required.

Conclusions
Cholecystectomy should be performed later than 216 h after PTGBD in patients with severe AC, although the possibility of a safe outcome of early LC within 72 h of PTGBD remains. Authors' contributions TU contributed to the conception and revised the work critically for important intellectual content and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy and integrity of any part of the work were appropriately investigated and resolved. ON contributed to the design of the work and drafted the work and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy. DD contributed to the design of the work and drafted the work and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy. KS contributed to the design of the work and drafted the work and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy. SG contributed to the design of the work and drafted the work and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy. MT contributed to the design of the work and drafted the work and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy. CS contributed to the conception and revised the work critically for important intellectual content and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy. HN contributed to the conception and revised the work critically for important intellectual content and approved the version to be published and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy and integrity of any part of the work were appropriately investigated and resolved.