Comparison of emergency cholecystectomy and delayed cholecystectomy after percutaneous transhepatic gallbladder drainage in patients with acute cholecystitis: a systematic review and meta-analysis

Laparoscopic cholecystectomy and percutaneous transhepatic gallbladder drainage (PTGBD) are common treatments for patients with acute cholecystitis. However, the safety and efficacy of emergency laparoscopic cholecystectomy (ELC) and delayed laparoscopic cholecystectomy (DLC) after PTGBD in patients with acute cholecystitis remain unclear. The PubMed, EMBASE, and Cochrane Library databases were searched through October 2019. The quality of the included nonrandomized studies was assessed using the Methodological Index for Nonrandomized Studies (MINORS). The meta-analysis was performed using STATA version 14.2. A random-effects model was used to calculate the outcomes. A total of fifteen studies involving 1780 patients with acute cholecystitis were included in the meta-analysis. DLC after PTGBD was associated with a shorter operative time (SMD − 0.51; 95% CI − 0.89 to − 0.13; P = 0.008), a lower conversion rate (RR 0.43; 95% CI 0.26 to 0.69; P = 0.001), less intraoperative blood loss (SMD − 0.59; 95% CI − 0.96 to − 0.22; P = 0.002) and longer time of total hospital stay compared to ELC (SMD 0.91; 95% CI 0.57–1.24; P < 0.001). There was no difference in the postoperative complications (RR 0.68; 95% CI 0.48–0.97; P = 0.035), biliary leakage (RR 0.65; 95% CI 0.34–1.22; P = 0.175) or mortality (RR 1.04; 95% CI 0.39–2.80; P = 0.933). Compared to ELC, DLC after PTGBD had the advantages of a shorter operative time, a lower conversion rate and less intraoperative blood loss.


Introduction
Acute cholecystitis (AC), an inflammatory condition of the gallbladder, is also a common disease that involves hospitalization and surgical treatment [1]. Patients with AC may present with a wide spectrum of inflammation, which may progress to empyema, perforation, and abscess formation, with an overall mortality rate of approximately 0.6% [2][3][4]. Since the first cholecystectomy was performed with the use of an operative laparoscope 30 years ago [5], laparoscopic cholecystectomy (LC) is currently recognized as a standard treatment for AC. In most AC patients, laparoscopic cholecystectomy can rapidly attenuate inflammatory symptoms and signs. Nevertheless, LC may precipitate certain complications, such as biliary leakage, bile duct injury and intraabdominal abscess [6,7], especially in elderly AC patients who undergo emergency laparoscopic cholecystectomy, Shao-Zhuo Huang, Hao-Qi Chen and Wei-Xin Liao contributed equally to this work. which may cause high morbidity and mortality rates [8]. Several studies have also documented that LC is associated with a high rate of conversion to open cholecystectomy and a long length of hospital stay [9,10].
Percutaneous transhepatic gallbladder drainage (PTGBD), which was first applied by Radder in 1980 [11], is a minimally invasive operation performed to relieve gallbladder tension through external drainage under ultrasound or CT guidance [3]. According to the 2018 Tokyo guidelines for drainage management of AC, PTGBD can alleviate inflammation caused by edema of the gallbladder wall and pericholecystic adhesions, and it has the advantage of a lower risk of adverse events compared with cholecystectomy, which is an alternative to surgical treatment in highrisk AC patients [6,12,13]. Previous studies have reported that LC performed after PTGBD has several advantages, such as early symptom remission, surgery facilitation and patient stabilization [14][15][16]. However, PTGBD may also cause complications related to the procedure, such as bile leakage and pain at the puncture site [16,17]. For severe AC patients, studies have shown that PTGBD is related to a high mortality rate and prolonged hospital stay [18]. In addition, biliary obstruction and cholestasis caused by gallstones are the major causes of acute cholecystitis, which is difficult to treat with PTGBD [19]. Besides, PTGBD may lead to complications associated with the drainage tube, such as obstruction of the drainage by stones and debris, pierced gallbladder and slipping off drainage tube [17].
The effects of emergency laparoscopic cholecystectomy (ELC) and delayed laparoscopic cholecystectomy (DLC) after PTGBD in AC patients remain unclear. Therefore, we conducted this meta-analysis to better understand and compare the safety and efficacy of ELC and PTGBD followed by DLC in AC patients in terms of the operative time, rate of conversion to open surgery, length of hospital stay, intraoperative blood loss, postoperative complications, biliary leakage and mortality.

Data source
This study was carried out according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 2009 guidelines [20]. Conducted by two authors (H.S. and C.H.), the search was performed using PubMed (1980 to October 2019), EMBASE (1988 to October 2019), and the Cochrane Central Register of Controlled Trials. The following medical subject terms were used in the search: ("cholecystectomy" OR "cholecystectomies") AND ("drainage") AND ("cholecystitis" OR "gallbladder Inflammation" OR "inflammation, gallbladder" OR "empyema, gallbladder" OR "gallbladder empyema" OR "empyema, gall bladder" OR "gall bladder empyema"). Neither IRB approval nor written consent was required for this study.

Study selection
Duplicates were removed. Titles and abstracts were independently screened by the reviewers (H.S. and C.H.) to assess the relevance of the publications. Subsequently, full-text articles were retrieved and checked. The remaining articles were surveyed by cross-referenced searches to detect studies that might have been overlooked. All studies concerned patients treated with emergency cholecystectomy and delayed cholecystectomy after percutaneous transhepatic gallbladder drainage for acute cholecystitis. The criteria for the diagnosis of acute cholecystitis had to be defined in the article, and acute cholecystitis had to be proven either by ultrasound or histologically. No language or publication types were restricted. Studies only aimed at severe cases or special population were excluded. Studies lacking necessary data or control groups were excluded. In studies with multiple publications from the same population, only the most recent one was included. Letters, posters, conference abstracts, expert opinions, review articles, case reports, animal experiments, and in vitro studies were also excluded. The selection process of the studies is shown in Fig. 1.

Data extraction
Data were extracted independently from each study by two investigators (H.S. and C.H.). The following items recorded for each study were extracted: first author, year of publication, age, country, number of controls and cases, study period, and time between PTGBD and LC. However, some studies expressed data with medians and ranges. We changed these data means with standard deviations [21]. Quantitative statistics were reported as the mean ± SD.

Quality assessment
All of the studies were retrospective. The data quality of nonrandomized studies was assessed using the Methodological Index for Nonrandomized Studies (MINORS) ( Table 1) [22]. MINORS, an index designed to assess the methodological quality of nonrandomized surgical studies, was developed by a group of surgeons because of the problems faced by clinicians as to the lack of randomized surgical trials and the large number of observational studies in surgery. By considering 12 items (8 for noncomparative and 4 for comparative studies), the total score was calculated by summing the values attributed as follows: 0 (not reported), 1 (reported but inadequate), and 2 (reported and adequate). The global ideal score for noncomparative studies was 16 and for comparative studies was 24.

Statistical analysis
Data from the eligible studies were extracted independently. Statistics Analysis (STATA version 14.2) (Stata Corporation; College Station, TX, USA) was used to perform the data analysis. Differences between groups were expressed as RRs with 95% CIs. The random-effect models were used to calculate the outcomes [23]. All statistical analyses used in this study were two-sided, and P < 0.05 was regarded as statistically significant.
Cochran's Q test and I 2 statistic were used to evaluate the heterogeneity across the studies. Cochran's Q < 0.10 or I 2 > 50% was regarded as significant heterogeneity across studies [24,25]. To evaluate the extent of publication bias, Egger's test and Begg's test were used [26,27]. Subgroup analysis was used to decrease the heterogeneity among studies.

Study characteristics and quality
The characteristics of the 15 included studies are shown in

Publication bias analysis
In this study, potential publication bias was investigated using Begg's and Egger's tests. The two plots for biliary leakage did not show obvious visual asymmetry (Fig. 9), and the P values of Egger's test were also greater than 0.05 (P = 0.632). Moreover, the other P values of the index of the test were greater than 0.05 (data not shown). Therefore, there was no significant publication bias in this meta-analysis.

Discussion
AC patients may present with a wide spectrum of disease severities linked to the operative difficulty, leading to the possibility of bile duct injury when AC patients undergo LC [2][3][4]. Several studies have also endorsed the usefulness of PTGBD for AC with a high success rate of efficacious drainage and a low complication rate, making it is a suitable treatment for AC patients [29,30,36]. However, no randomized controlled trials have been performed to clarify the advantages of both approaches. In addition, in clinical practice, many physicians decide their treatment strategies mostly based on local experience and personal preferences owing to the ambiguous indications for PTGBD [3,34]. This leads to a controversial comparison between ELC and DLC after PTGBD. Given this background, we performed a metaanalysis to further clarify the safety and efficacy of ELC and PTGBD followed by DLC in AC patients. The pooled analysis of the 15 included retrospective studies provided moderate quality evidence in favor of DLC after PTGBD for the treatment of acute cholecystitis. Our results showed that DLC after PTGBD had the advantages of a short operative time, a low conversion rate and limited intraoperative blood loss. However, there were no significant differences in several aspects including the postoperative length of hospital stay, postoperative complications, biliary leakage and mortality.
According to our analysis, DLC after PTGBD demonstrated a shorter operation duration than ELC (SMD − 0.51; 95% CI − 0.89 to − 0.13; P = 0.008). This finding could be related to the alleviation of local inflammation after PTGBD. As a minimally invasive procedure, PTGBD decompresses gallbladder distention and alleviates edema of the gallbladder wall and pericholecystic inflammation [6]. The clinical symptoms of AC in patients with good preoperative conditions can be immediately relieved by PTGBD [37]. According to a study from Hu et al. [30], 35 AC patients who underwent successful PTGBD had decompressed gallbladders and normal body temperatures within 72 h. Our result is also consistent with that of Chikamori et al. [16], who found that the duration of LC was shortened when cholecystectomy was performed as soon as possible after PTGBD. In addition, PTGBD can be employed for cholangiography to reveal the biliary tract anatomy and provide clear information about the surgical site, which might facilitate operative procedures [38].
Our analysis of the conversion rate suggested that DLC had a significantly lower potential to cause conversion to open surgery than ELC (RR 0.43; 95% CI 0.26-0.69; P = 0.001). Previous studies concluded that the root cause of conversion was repetitively progressive inflammation with distended and edematous walled gallbladders [39,40]. The low conversion rate in the DLC group may be attributed to PTGBD, as PTGBD has the ability to relieve inflamed gallbladder adhesions [15]. On the other hand, emergency laparoscopic cholecystectomy is regarded as a challenging procedure in AC patients due to the frequent presence of adhesions around the acutely inflamed gallbladder and the high incidence of bile duct injuries [41]. In addition, in clinical practice, due to restrictions involving ethics, there might have been bias in terms of the selection of patients who had undergone PTGBD, as the patients in the DLC group had more severe preoperative inflammation, a higher risk of ASA classification and a poorer general condition than the patients in the ELC group [6,15,42]; DLC patients may have had severe inflammation and dense adhesions with an increased risk of conversion to laparotomy [43]. However, our conversion rate results still showed that DLC was superior to ELC, regardless of the effect of PTGBD. Our subgroup analysis also showed that there was no significant difference in the rate of conversion to open surgery when LC was performed within 7 days of or 7 days after PTGBD, similar to the results of previous studies [44,45].
Our analysis of intraoperative hemorrhage revealed that DLC demonstrated less blood loss than ELC (SMD − 0.59; 95% CI − 0.96 to − 0.22; P = 0.002). PTGBD can immediately relieve the decompression of swollen gallbladders and pericholecystic inflammation to prevent the development of fibrosis in Calot's triangle [37,46]. With the help of PTGBD, the operation field of Calot's triangle is much clearer, which facilitates laparoscopic cholecystectomy and reduces blood loss during the operation. In addition, surgeons have less information about patients who undergo ELC than what is routinely required. In contrast, during the preoperative period, the surgeon can identify the DLC patient's underlying disease and the status of the biliary system, together with sufficient laboratory and radiological test results and surgical planning, which can enhance the patient's safety [15,34].
Our analysis of the hospital stay suggested that DLC after PTGBD had a significantly longer time of the total hospital stay than ELC (SMD 0.91; 95% CI 0.57 to 1.24; P < 0.001). Because of hospitalization for PTGBD, the time of total hospital stay was much longer in the PTGBD + DLC group. In clinical practice, patients in the PTGBD + DLC group may have poor general conditions and severe preoperative inflammation, which may need more time to retain the drainage [6,15,42]. Lo et al. [47] reported that PTGBD had the adverse outcome of a longer hospital stay for the management of gallbladder perforation, while urgent LC without PTGBD had similar surgical outcomes as that of elective LC with PTGBD. But previous studies suggested that AC patients treated with ELC may have a shorter postoperative hospital stay and lower hospital costs than those who underwent DLC without PTGBD [48][49][50].
Our analysis showed that there was no significant difference in terms of complications, biliary leakage or mortality. Apart from the complications caused by LC, PTGBD can also lead to complications related to the procedure. In the 49 patients with PTGBD from Ke's study [17], there were 23 patients had postoperative complications after PTGBD and 12 cases among were associated with the drainage tube, such as bile leakage, hepatapostema and common bile duct stone. The complications caused by PTGBD may be associated with the long-time drainage and the underlying disease. However, patients in the EC group have higher incidence of respiratory failure and admission to the ICU, which contrarily indicates the effect of PTGBD on reducing severe complications after cholecystectomy in AC patients [17]. Besides, Giger et al. [43] analyzed more than 20,000 patients treated with LC and suggested that emergency surgery may be an independent risk factor for possible perioperative complications. In addition, Jackson et al. [51] and Jia et al. [29] suggested that postoperative complications may be associated with operative time and intraoperative blood loss, while our analysis showed a shorter operative time and less intraoperative blood loss in the DLC group than in the ELC group, which may imply a better outcome in terms of postoperative complications.
In our subgroup analysis, we divided studies into two groups based on the time interval between PTGBD and LC. Although it was not significant, probably because of the rough treatment of PTGBD time, the PTGBD time may affect the outcome of AC patients. Choi et al. [15] and Han et al. [52] reported that patients who underwent LC within 72 h after PTGBD had a worse outcome than those who underwent LC more than 72 h after PTGBD. These results can be explained by the fact that the short duration of PTGBD may cause the incompleteness of fistula formation around the PTGBD tube, and inflammation is not alleviated well enough for cholecystectomy [16].
Undoubtedly, there are several limitations in this metaanalysis. The quality of the included studies was deemed moderate, and all of the included studies were retrospective in nature, with the dearth of randomized controlled trials performed for ELC and PTGBD + DLC. Furthermore, limited by ethics, there was a bias in the selection of patients treated with PTGBD + DLC, as the patients in the DLC group may have worse conditions for cholecystectomy [6,15,42]. And patients who received PTGBD failed to undergo DLC are not reported in the literatures we included, which may lead to a publication bias in our meta-analysis. Finally, most of the included studies were performed in Asia, and the data regarding ELC and DLC in Europe and

3
America were unclear. Randomized controlled trials and multicenter studies with a large sample size are needed to verify the outcomes of this meta-analysis.
In conclusion, our meta-analysis suggested that delayed cholecystectomy after PTGBD may be preferred over emergency cholecystectomy, as it had a shorter operative time, a lower conversion rate and less intraoperative blood loss. Therefore, delayed laparoscopic cholecystectomy after percutaneous transhepatic gallbladder drainage might be a good approach for patients with acute cholecystitis.
Author contributions S-ZH, H-QC and W-CL contributed to study design; S-ZH and H-QC contributed to manuscript writing and study selection; W-XL and HZ contributed to data analysis; W-YZ and J-HC contributed to proofreading; BL is a senior author and conceptualized the research project; K-PH is the corresponding author and supervised the research project.
Funding Funding supported by Grants from the National Natural Science Foundation of China (81702375) and the Training program for young teachers of Sun Yat-sen University (19ykpy32).

Compliance with ethical standards
Conflict of interest All authors have no conflicts of interest or financial ties to disclose.

Research involving human participants and/or animals This article
does not contain any studies with human participants or animals performed by any of the authors.
Informed consent For this type of study, formal consent is not required.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat iveco mmons .org/licen ses/by/4.0/.