Image-Guided Cholecystostomy Tube Placement PDF

Va s c u l a r a n d I n t e r ve n t i o n a l R a d i o l o g y O r i g i n a l R e s e a r c h Beland et al....

Va s c u l a r a n d I n t e r ve n t i o n a l R a d i o l o g y O r i g i n a l R e s e a r c h Beland et al. Transhepatic Versus Transperitoneal Cholecystostomy Tube Vascular and Interventional Radiology Original Research Image-Guided Cholecystostomy Downloaded from www.ajronline.org by UAB Libraries on 03/05/24 from IP address 164.111.77.114. Copyright ARRS. For personal use only; all rights reserved Tube Placement: Short- and Long- Term Outcomes of Transhepatic Versus Transperitoneal Placement Michael D. Beland1 OBJECTIVE. Image-guided percutaneous cholecystostomy may be performed by a trans­ Lakir Patel2 hepatic or transperitoneal approach. We compared the short- and long-term outcomes of per- Sun H. Ahn1 cutaneous cholecystostomy related to route of catheter placement. David J. Grand1 MATERIALS AND METHODS. A retrospective observational study of image-guided percutaneous cholecystostomy was performed from 2004 to 2016. A search of the hospital’s Beland MD, Patel L, Ahn SH, Grand DJ radiology information service was performed using the keywords “percutaneous cholecys- tostomy,” “gallbladder drain,” and “cholecystostomy tube” and the relevant Current Proce- dural Terminology codes. All search results were reviewed to identify the cohort of 373 pa- tients who underwent initial percutaneous cholecystostomy catheter placement. Imaging was reviewed to determine the method and route of percutaneous cholecystostomy and compli- cations. A chart review was performed to determine clinical outcomes. Differences were ex- amined using a generalized linear model assuming a binary distribution and logit function. RESULTS. Percutaneous cholecystostomy catheter placement was performed using ultra- sound guidance alone in 229 patients, ultrasound access with fluoroscopic guidance in 129 pa- tients, CT guidance in 14 patients, and fluoroscopic guidance in one patient. The trocar tech- nique was used for 183 patients, and the Seldinger technique was used for 190 patients. Two hundred eighteen percutaneous cholecystostomy catheters were placed via the transhepatic route, and 153 were placed via the transperitoneal route. The most common catheter sizes used were 8.5 French (n = 234) and 10 French (n = 124). No significant differences were observed between transperitoneal and transhepatic placement with regard to the frequency of pain, clogging, skin infection, bleeding, biloma, cholangitis, leakage, abscess, unplanned catheter removal, or need for replacement (p > 0.05). CONCLUSION. No evidence of a difference in outcomes was observed for transhepatic cholecystostomy tube placement over transperitoneal placement. The route that appears safer Keywords: cholecystitis, cholecystostomy, gallbladder, and less technically challenging should therefore be chosen. gallbladder drainage lthough laparoscopic cholecys- der procedures performed, a sixfold increase A doi.org/10.2214/AJR.18.19669 tectomy is considered the stan- from 1994. Percutaneous cholecystos- Received February 3, 2018; accepted after revision dard treatment for acute chole- tomy is a minimally invasive image-guided June 2, 2018. cystitis, surgical intervention can procedure that has been shown to be safe in carry significant morbidity and mortality in patients with multiple comorbidities [5, 6]. Based on a presentation at the Radiological Society of patients who are poor surgical candidates. Placement is generally done with imaging North America 2017 annual meeting, Chicago, IL. For these patients, percutaneous cholecystos- guidance, using fluoroscopy, ultrasound, CT, 1 Department of Diagnostic Imaging, The Alpert Medical tomy has become a widely used treatment or a combination of these. The percutaneous School of Brown University, Rhode Island Hospital, 593 option for management of acute cholecystitis. cholecystostomy catheter may be placed di- Eddy St, Providence, RI 02903. Address correspondence Percutaneous cholecystostomy was first rectly into the gallbladder (trocar technique) to M. D. Beland ([email protected]). described in 1979 as a treatment for obstruc- or by initially using a smaller catheter with 2 Department of Diagnostic Imaging, University of tive jaundice and later in 1980 as a treat- exchange over a wire for placement of the Maryland, Mercy Medical Center, Baltimore, MD. ment for gallbladder infection. Since larger diameter percutaneous cholecystosto- these reports, percutaneous cholecystosto- my catheter (Seldinger technique). AJR 2019; 212:201–204 my has gained acceptance as a temporary The percutaneous cholecystostomy course 0361–803X/19/2121–201 treatment of acute cholecystitis in poor sur- may be transhepatic, where the catheter tra- gical candidates. In 2009, percutaneous cho- verses the liver before entering the gallblad- © American Roentgen Ray Society lecystostomy represented 3% of all gallblad- der, or gallbladder access may be directly AJR:212, January 2019 201  Beland et al. transperitoneal without crossing the inter- cutaneous cholecystostomies were performed by sistent clinical symptoms suggesting incomplete vening liver parenchyma. The choice of ap- a total of 34 different attending physicians during treatment. Length of hospital stay and frequency proach may be dictated by the location of the the study period, with or without assistance from of death were recorded. gallbladder, degree of gallbladder distention, a resident or fellow. Eighty percent (n = 299) of Differences in outcomes between transhepatic presence of intervening bowel, and patient percutaneous cholecystostomies were performed and transperitoneal were examined using a gener- body habitus. However, in many cases, both by 13 attending physicians (at least 12 percutane- alized linear model assuming a binary distribu- Downloaded from www.ajronline.org by UAB Libraries on 03/05/24 from IP address 164.111.77.114. Copyright ARRS. For personal use only; all rights reserved the transhepatic and transperitoneal routes ous cholecystostomies each). Percutaneous chole- tion and logit function with SAS software (ver- may be available to the operator. Despite cystostomy was generally performed using local sion 9.4, SAS) with the GLIMMIX procedure. small series in the literature showing simi- anesthesia and conscious sedation with monitor- Alpha was established at the 0.05 level and 95% lar outcomes between the techniques, most ing by dedicated nursing personnel. The choice of CIs were calculated. authors and textbooks continue to state that imaging guidance, use of trocar versus Seldinger the transhepatic approach is preferred [8– technique, and route of placement were chosen at Results 10]. Rationales for preferring the transhepat- the discretion of the attending radiologist at the Of the 373 patients, percutaneous cho- ic approach have included a decreased likeli- time of the procedure. lecystostomy catheter placement was per- hood of bile peritonitis at the time of catheter Postprocedural hospital course and follow-up formed using ultrasound guidance only for placement or during removal, a decreased data focused on tube maintenance and duration of 229 patients, initial ultrasound access fol- likelihood of catheter dislodgment, and drainage, imaging, return to care, and complica- lowed by fluoroscopic guidance for 129 pa- quicker maturation of a catheter track [7, 10]. tions. Specific complications identified during re- tients, CT guidance for 14 patients, and fluo- Because of the continued insistence on the cord review included site pain, low drain output or roscopic guidance only for one patient. The superiority of transhepatic placement, oper- clogged catheter, skin infection, biloma, cholan- trocar technique was used in 183 cases, and ators may feel compelled to attempt it even gitis, pericatheter leakage, bleeding, abscess, bile the Seldinger technique was used in 190 cases. when a simpler approach exists via a trans- peritonitis, and unplanned catheter removal. Site Two hundred eighteen percutaneous chole- peritoneal course. The purpose of this study pain and skin infection were observations gen- cystostomies were performed via the trans­ is to review the outcomes and complications erally made after discharge. Presentations to the hepatic route, and 153 were performed by the of image-guided percutaneous cholecystos- emergency department or clinic for pain at the lo- transperitoneal route. We were unable to de- tomy catheter placement by either approach cation of the percutaneous cholecystostomy were termine route of placement in two patients over a period of 13 years. considered site pain. Skin infection was record- who were excluded from analysis. The per- ed as presentations to the emergency department cutaneous cholecystostomy catheter sizes were Materials and Methods or clinic after discharge that resulted in antibiot- 6.5 French (n = 1), 7 French (n = 1), 8.5 French This retrospective HIPAA-compliant study ic prescription for skin infection. Biloma was re- (n = 234), 10 French (n = 124), and 12 French was approved by the institutional review board of corded when an imaging study performed after (n = 13). Patient characteristics for transperi- Rhode Island Hospital. A search of the hospital’s percutaneous cholecystostomy revealed a new toneal versus transhepatic groups are report- radiology information service was performed us- collection that was described as a likely biloma. ed in Table 1. In summary, no significant dif- ing the keywords “percutaneous cholecystostomy,” Low drain output or clogged catheter referred ferences were seen for age, sex, body mass “gallbladder drain,” and “cholecystostomy tube” to a postplacement observation of no significant index (weight in kilograms divided by the from January 2004 through December 2016. To en- (< 10 mL) output in 24 hours for which subsequent square of height in meters), presence of gall- sure inclusion of all patients, a search of relevant drain exchange was performed. Pericatheter leak- stones, anticoagulation, laboratory values, Current Procedural Terminology codes used at our age was recorded if patients presented after per- duration that the percutaneous cholecystos- hospital for percutaneous cholecystostomy catheter cutaneous cholecystostomy complaining of non- tomy catheter remained in place, or initial placement was also performed over the same pe- bloody fluid draining to the skin around the site volume drained. There was a significant dif- riod. Search results were then manually reviewed of their cholecystostomy tube. In our study, bleed- ference in trocar versus Seldinger technique to identify all instances of percutaneous cholecys- ing was an observation made after discharge. We and percutaneous cholecystostomy catheter tostomy catheter placement. This yielded the study did not find any immediate or early postproce- size between the two groups. Transperitoneal cohort of 373 patients. Radiology reports and in- dure cases of recorded hemorrhage requiring fur- drains were more commonly placed via tro- traprocedural and postprocedural images were re- ther intervention. Patients presenting to the emer- car technique and were smaller, with a me- viewed to determine the technique and route of gency department with any visible bleeding at the dian size of 8.5 French, versus transhepatic percutaneous cholecystostomy. Chart review was site of their cholecystostomy tube were recorded drains, which were more commonly placed performed to determine the frequency of compli- as having bleeding. Abscess was a localized infec- via the Seldinger technique with median cations and clinical outcomes before and after per- tion that required a bedside incision and drainage. catheter size of 10 French (p < 0.0001). cutaneous cholecystostomy catheter removal. All Unplanned catheter removal was further record- Complication rates are presented in Table study patients had records available for review in ed as “dislodged” where the catheter remained in 2. In summary, no significant differences were the electronic medical record, including laboratory the patient but the pigtail had been dislodged from observed for transperitoneal versus transhe- results, imaging studies, and provider notes (inpa- the gallbladder lumen, as “fell out” where catheter patic route for pain, clogging, skin infection, tient service, emergency department, intervention- was entirely removed but the cause was unclear, bleeding, biloma, cholangitis, leakage, or ab- al radiology, and the medical and surgical clinics). or as “pulled out” where the patient forcefully re- scess (p > 0.05). There were no recorded cas- In our department, percutaneous cholecystosto- moved the entire catheter. After catheter removal, es of bile peritonitis after percutaneous cho- my is performed by either the body interventional record review was performed to determine wheth- lecystostomy catheter placement. There was service or vascular interventional radiology. Per- er there was a need for drain replacement for per- no significant difference observed in the fre- 202 AJR:212, January 2019  Transhepatic Versus Transperitoneal Cholecystostomy Tube TABLE 1: Patient Characteristics, Laboratory Values, and Percutaneous ence in the frequency of short- or long-term ­Cholecystostomy Data for Transperitoneal Versus Transhepatic complications between catheters that were Percutaneous Cholecystostomy Catheter Placement placed via the transhepatic versus transperi- Variable Transperitoneal Transhepatic p toneal routes. Our results do not support the traditional teaching that the transhepatic ap- Age (y) 77 (65, 86) 77 (66, 85) 0.656 proach should be preferred to decrease po- Downloaded from www.ajronline.org by UAB Libraries on 03/05/24 from IP address 164.111.77.114. Copyright ARRS. For personal use only; all rights reserved Sex (%) 0.831 tential complications. Female 42.2 43.3 Despite the traditional dogma that trans­ Male 57.8 56.7 hepatic placement minimizes complications and improves outcomes, there is a paucity Body mass indexa 26.6 (23.6, 32.5) 27.1 (24.1, 32.1) 0.828 of data specifically comparing the two ap- Calculous cholecystitis (%) 70.8 68.7 0.896 proaches. A 1994 study by Garber et al. Taking anticoagulants (%) 27.6 27.1 0.906 showed safe placement of cholecystostomy catheters via the transperitoneal approach in Prothrombin time (s) 14.5 (12.9, 16.9) 14.7 (13.3, 16.9) 0.562 34 consecutive patients using the trocar tech- International normalized ratio 1.3 (1.2, 1.6) 1.3 (1.2, 1.6) 0.792 nique with no evidence of bile peritonitis or Partial thromboplastin time (s) 31.7 (28, 36) 30.8 (27.5, 34.9) 0.415 biloma. In 2010, Loberant et al. found no Platelet count (× 10 9 /L) 205 (127, 291) 194 (136, 287) 0.931 significant difference between approaches in 173 patients. We believe that our study is the WBC count (× 10 9 /L) 16.6 (11.9, 21.8) 16.6 (13, 21.7) 0.811 largest to date and is concordant with previ- Bilirubin level (mg/dL) 1.2 (0.8, 2.6) 1.4 (0.9, 2.7) 0.244 ous smaller studies that found that the trans- Cholecystostomy catheter size (French), median 8.5 10 < 0.0001 peritoneal approach to percutaneous chole- Duration that percutaneous cholecystostomy catheter 42 (23, 58) 42 (23, 64) 0.668 cystostomy is equivalent in all ways to the remained in place (d) transhepatic approach. Trocar technique (%) 69.5 34.9 < 0.0001 Bile peritonitis is a painful complication of percutaneous cholecystostomy that occurs Initial volume drained (mL) 50 (20, 100) 50 (20, 100) 0.725 when bile leaks from the gallbladder into the Note—Except where noted otherwise, data are median (upper quartile, lower quartile). aWeight in kilograms divided by the square of height in meters. peritoneal cavity at the time of either cath- eter placement or removal. The transhepat- TABLE 2: Complication Rates for Transperitoneal Versus Transhepatic ic approach is classically thought to decrease ­Cholecystostomy Catheter Placement the risk of bile leak compared with the trans- peritoneal approach because the gallbladder Complication Transperitoneal Transhepatic p bed would theoretically contain the leaking Site pain 4.55 (2.18–9.25) 4.13 (2.16–7.76) 0.845 bile. However, we saw no significant dif- ference in the frequency of site pain, leak- Low output 2.6 (0.98–6.74) 1.84 (0.69–4.80) 0.620 age around the catheter, or biloma forma- Skin infection 3.25 (1.35–7.58) 0.92 (0.23–3.61) 0.128 tion between approaches, and there were Biloma 0.65 (0.09–4.49) 1.38 (0.44–4.19) 0.514 no documented cases of post–percutaneous Cholangitis 1.95 (0.63–5.89) 5.05 (2.81–8.90) 0.137 cholecystostomy bile peritonitis. We there- fore conclude that a transperitoneal approach Catheter leakage 1.95 (0.63–5.89) 2.75 (1.24–6.00) 0.621 does not increase the likelihood of bile leak Bleeding 1.30 (0.32–5.06) 3.21 (1.54–6.59) 0.254 at the time of percutaneous cholecystostomy Abscess 1.30 (0.32–5.06) 0.46 (0.06–3.20) 0.394 placement or removal. Catheter dislodged 5.84 (3.06–10.87) 5.51 (3.15–9.46) 0.889 Another postulated advantage to the trans­ hepatic approach is a decreased likelihood of Catheter fell out 3.90 (1.76–8.42) 4.59 (2.48–8.33) 0.747 catheter dislodgment. We considered three Catheter pulled out 9.74 (5.95–15.55) 5.96 (3.49–10.01) 0.179 general categories of premature catheter re- Needed drain replaced 3.25 (1.35–7.58) 3.67 (1.84–7.18) 0.827 moval: either the patient pulled the catheter Note—Except for p values, data are percentage (95% CI). out, the catheter fell out on its own, or the catheter became dislodged from the gallblad- quency of unplanned catheter removal or significantly different between the groups der but remained in the patient. We found no the need for replacement. Subsequent chole- (transhepatic, 1.8 years; transperitoneal, 2.8 significant difference in the frequency of cystectomy was performed in 31.5% of pa- years; p = 0.12). premature catheter removal by any of these tients in the transperitoneal group and 42.0% methods, nor did we find any significant dif- of patients in the transhepatic group (p = Discussion ference in the frequency of premature cathe- 0.083). Median length of hospital stay was 7 In reviewing 373 patients who underwent ter removal when the three types of removal days for both the transhepatic and transperi- percutaneous cholecystostomy over a period were pooled together. We also reviewed the toneal groups (p = 0.96). Survival was not of 13 years, we found no significant differ- medical record to determine whether a sec- AJR:212, January 2019 203  Beland et al. ond percutaneous cholecystostomy catheter the rationale was not regularly recorded. References was placed after premature catheter remov- Therefore, we are unable to retrospectively 1. Akyürek N, Salman B, Yüksel O, et al. Manage- al. We found no significant difference in the compare outcomes in patients on the basis ment of acute calculous cholecystitis in high-risk need for a second catheter placement on the of who may have had both trans­hepatic and patients: percutaneous cholecystotomy followed by basis of the approach of the initial percuta- transperitoneal approaches available ver- early laparoscopic cholecystectomy. Surg ­Laparosc neous cholecystostomy. Interestingly, there sus only one approach. Not all patients were Endosc Percutan Tech 2005; 15:315–320 Downloaded from www.ajronline.org by UAB Libraries on 03/05/24 from IP address 164.111.77.114. Copyright ARRS. For personal use only; all rights reserved was a significant difference in catheter diam- equally suitable for both the transhepatic 2. Elyaderani M, Gabriele OF. Percutaneous cholecys- eter between the two groups, with the median and transperitoneal approach, so there may tostomy and cholangiography in patients with ob- catheter size of 8.5 French for the transperi- have been operator selection bias only for structive jaundice. Radiology 1979; 130:601–602 toneal approach and 10 French for the trans­ patients with both transhepatic and trans- 3. Radder RW. Ultrasonically guided percutaneous hepatic approach. Although it may be sug- peritoneal windows available. Given the catheter drainage for gallbladder empyema. ­Diagn gested that the larger diameter catheter used relatively similar numbers of patients in Imaging 1980; 49:330–333 in the transhepatic technique may help pre- each group, this bias should be minimized 4. Duszak R Jr, Behrman SW. National trends in per- vent premature catheter removal, transperi- and similarly distributed between the two cutaneous cholecystostomy between 1994 and toneal catheters in our series were not more groups. In addition, there was no significant 2009: perspectives from Medicare provider frequently dislodged. difference in terms of body mass index, claims. J Am Coll Radiol 2012; 9:474–479 The literature has suggested that the trans­ age, or sex between the two groups. We did 5. Anderson JE, Chang DC, Talamini MA. A nation- hepatic approach may allow quicker matura- find a significant difference in percutaneous wide examination of outcomes of percutaneous tion of the catheter track [11, 12]. We found cholecystostomy catheter size between the cholecystostomy compared with cholecystectomy no difference in the length of time the cathe- two groups, with the median catheter size for acute cholecystitis, 1998-2010. Surg Endosc ter was left in place until purposeful removal. being larger for the transhepatic group than 2013; 27:3406–3411 Although the selection of when to remove the the transperitoneal group (10 vs 8.5 French). 6. Wang CH, Wu CY, Yang JC, et al. Long-term out- catheter was based on multiple factors and Not surprisingly, the trocar technique was comes of patients with acute cholecystitis after the length of time to track maturation cannot favored for the smaller catheter diame- successful percutaneous cholecystostomy treat- be assessed in this retrospective study, there ter because the largest ultrasound needle ment and the risk factors for recurrence: a decade was no significant difference in length of time guide available at our institution was 8.5 experience at a single center. PLoS One 2016; until catheter removal or complications, such French, and placing a larger diameter cath- 11:e0148017 as biloma or bile peritonitis, which could be eter would require using the Seldinger tech- 7. Venara A, Carretier V, Lebigot J, Lermite E. Tech- seen with inadequate track maturation. Al- nique. The difference in catheter diameter is nique and indications of percutaneous cholecys- though the smaller percutaneous cholecys- important to note because it may affect the tostomy in the management of cholecystitis in tostomy catheter size in the transperitoneal likelihood of bleeding, tube dislodgement, 2014. J Visc Surg 2014; 151:435–439 group may theoretically decrease the risk of a or bile leak, as addressed already. Although 8. van Overhagen H, Meyers H, Tilanus HW, Jeekel bile leak, there was also no significant differ- all study patients had records available for J, Laméris JS. Percutaneous cholecystostomy for ence in leakage around the catheter, suggest- review in the electronic medical record, it patients with acute cholecystitis and an increased ing that the catheter tracks were adequately is possible that a patient may have received surgical risk. Cardiovasc Intervent Radiol 1996; opposed to the catheter wall. care related to the percutaneous cholecys- 19:72–76 Finally, a theoretic advantage to the tostomy at an outside institution without 9. Garber SJ, Mathieson JR, Cooperberg PL, transperitoneal route is a decreased risk of documentation in our system. Because we ­MacFarlane JK. Percutaneous cholecystostomy: bleeding because the liver is not traversed were able to determine the final disposition safety of the transperitoneal approach. J Vasc In- by a large-bore catheter, particularly in pa- of all patients who underwent percutane- terv ­Radiol 1994; 5:295–298 tients with liver disease or coagulopathy ous cholecystostomy in this study, we think 10. Loberant N, Notes Y, Eitan A, Yakir O, Bickel A. [12, 13]. Importantly, we saw no significant that this limitation had a minimal effect on Comparison of early outcome from transperito- difference in laboratory bleeding parame- our findings. Of note, percutaneous chole- neal versus transhepatic percutaneous cholecys- ters or presence of anticoagulation between cystostomy was performed by 34 different tostomy. Hepatogastroenterology 2010; 57:12–17 the two groups. Despite transhepatic cath- attending radiologists with or without the 11. Hatjidakis AA, Karampekios S, ­ Prassopoulos eters being significantly larger in diameter assistance of a resident or fellow over the P, et al. Maturation of the tract after percutaneous than transperitoneal catheters, there was no course of our study. Having numerous phy- cholecystostomy with regard to the access route. significant difference in the frequency of sicians and a long study period limits bias Cardiovasc Intervent Radiol 1998; 21:36–40 catheter dislodgment between the groups, for a specific technique or outcome to be at- 12. Little MW, Briggs JH, Tapping CR, et al. Percuta- even though the smaller transperitoneal tributed to a particular physician. neous cholecystostomy: the radiologist’s role in catheter size might suggest a predisposition In conclusion, percutaneous cholecystos- treating acute cholecystitis. Clin Radiol 2013; to dislodgment. tomy can be safely performed by either the 68:654–660 Limitations of this study include its retro- transhepatic or transperitoneal route with 13. vanSonnenberg E, D’Agostino HB, ­ Goodacre spective design. Patients were not random- perhaps no significant difference in com- BW, Sanchez RB, Casola G. Percutaneous gall- ized to the transhepatic or transperitone- plication rates or outcomes. The choice of bladder puncture and cholecystostomy: results, al approach. The decision was made at the route should be at the operator’s discretion complications, and caveats for safety. ­Radiology time of procedure by the proceduralist, and and based on the patient’s anatomy. 1992; 183:167–170 204 AJR:212, January 2019