Double-balloon enteroscopy as a rescue technique for failed direct percutaneous endoscopic jejunostomy when using conventional push enteroscopy (with video)

CASE STUDIES

Double-balloon enteroscopy as a rescue technique for failed direct percutaneous endoscopic jejunostomy when using conventional push enteroscopy (with video) Louis M. Wong Kee Song, MD, Todd H. Baron, MD, FASGE, Atif Saleem, MD, David H. Bruining, MD, Jeffrey A. Alexander, MD, Elizabeth Rajan, MD Rochester, Minnesota, USA

Direct percutaneous endoscopic jejunostomy (DPEJ) tube placement is indicated in patients who require prolonged enteral feeding but are not suitable candidates for PEG or when attempted PEG placement fails. Although PEG with transgastric jejunal tube extension (PEG-J) is another option, the delivery rate of enteral feeds is limited by the narrowbore extension tube, which is also prone to frequent obstruction and migration back into the stomach.1,2 Furthermore, PEG-J may not decrease the risk of recurrent aspiration pneumonia as has been shown with DPEJ.3 However, DPEJ is technically challenging, with lower rates of successful placement than PEG. DPEJ is usually performed by using colonoscopes or push enteroscopes, Abbreviations: BAE, balloon-assisted enteroscopy; BMI, body mass index; CVA, cerebrovascular accident; DBE, double-balloon enteroscopy; DPEJ, direct percutaneous endoscopic jejunostomy; GETA, general endotracheal anesthesia; LLQ, left lower abdominal quadrant; LUQ, left upper abdominal quadrant; MAC, monitored anesthesia care; PEG-J, PEG with jejunal tube extension; PEJ, percutaneous endoscopic jejunostomy. DISCLOSURE: Fujinon, Inc., provided research funds and equipment for this study. L. M. Wong Kee Song and T. H. Baron receive research support from Fujinon, Inc. No other financial relationships relative to this publication were disclosed. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2012.04.448 Received January 29, 2012. Accepted April 12, 2012. Current affiliations: Division of Gastroenterology and Hepatology, Mayo Clinic, Rochester, Minnesota, USA. Presented in part at Digestive Disease Week, Chicago, Illinois, May 2011 (Gastrointest Endosc 2011;73:AB451). Reprint requests: Louis M. Wong Kee Song, MD, Division of Gastroenterology and Hepatology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905.

This video can be viewed directly from the GIE website or by using the QR code and your mobile device. Download a free QR code scanner by searching ‘‘QR Scanner’’ in your mobile device’s app store.

www.giejournal.org

with or without overtubes.4 In 1 large series, DPEJ failed in nearly a third of 307 attempts because of lack of transillumination and inability to pass into the jejunum.5 Double-balloon enteroscopy (DBE) may increase the likelihood of finding a suitable site for percutaneous endoscopic jejunostomy (PEJ) placement by allowing deeper access into the small intestine. We hypothesized that DBEassisted DPEJ placement would succeed when failure occurred with use of a colonoscope.

METHODS Patients In this prospective pilot study, approved by the Institutional Review board, patients in whom DPEJ with push enteroscopy failed were offered DBE-assisted DPEJ placement. Patients or their legal guardians provided written informed consent. Inclusion criteria were as follows: (1) age older than 18 years, (2) appropriate indication and approval for jejunostomy feeding tube placement by the nutrition service, (3) absence of a bleeding disorder, (4) platelet count over 50 ⫻ 10(9) per liter, and (5) International Normalized Ratio below 1.5. Exclusion criteria were as follow: (1) allergy to latex, (2) enteroenteric anastomosis less than 4 weeks old, (3) ascites, (4) smallbowel obstruction, and (5) pregnancy. Intravenous antibiotic prophylaxis was administered.

Procedures Conventional DPEJ was performed with variablestiffness pediatric colonoscopes (PCF-Q180AL, Olympus America, Center Valley, PA) with CO2 insufflation by 4 experienced endoscopists uninvolved with subsequent DBE procedures. Conventional procedures were performed by using moderate sedation (fentanyl and midazolam) or monitored anesthesia care (MAC). Procedure tolerance was graded as excellent, good, fair, or poor by the endoscopist and registered nurse responsible for sedation and monitoring, and this information was entered into the GI database at the end of the procedure. The DBE system (EN-450T5, Fujinon, Inc., Saitama, Japan) is located within a surgical suite at one Mayo Clinic hospital. Volume 76, No. 3 : 2012 GASTROINTESTINAL ENDOSCOPY 675

Double-balloon enteroscopy for failed direct percutaneous endoscopic jejunostomy

All DBE procedures are performed with anesthesia support. The type of sedation used (MAC or general endotracheal anesthesia [GETA]) is at the discretion of the anesthesiologist. CO2 is used for insufflation. DBE-DPEJ was performed by 1 of 2 experienced DBE endoscopists (L. M. W. K. S. and E. R.). Fluoroscopy was used only in patients with surgically altered gut anatomy to help identify the afferent and efferent limbs. One of 2 registered nurses who assist with all tube placements performed the skin portion of the procedure under the endoscopist’s direction. The DBE was performed by using the standard pushpull technique after the balloon enteroscope was advanced beyond the ligament of Treitz. We limited insertion of the enteroscope to the midjejunum (ⱕ150 cm from the ligament of Treitz). During advancement, a site in the jejunum was sought for PEJ tube placement by transillumination and finger indentation. If an adequate site was not identified, it was sought during instrument withdrawal. After a suitable site was identified, DPEJ placement was performed by using the Ponsky pull PEG technique and kit (MIC PEG Kit, Kimberly-Clark, Roswell, GA). The safe tract technique was used.6 To prevent jejunal loop displacement during trocar insertion, the anesthetic needle was grasped with a long-length polypectomy snare to anchor the jejunum to the abdominal wall.7 The trocar was inserted alongside the needle into the jejunum. The looped wire from the PEG kit was inserted through the trocar and grasped with the snare to exit the mouth. Over the wire, a 20F feeding tube was pulled into position (Video 1, available online at www.giejournal.org). The balloon enteroscope was not reinserted to evaluate the PEJ site. In patients with surgically altered gut anatomy, tattooing of either the afferent or the efferent limb was performed before DPEJ to facilitate subsequent endoscopic procedures (Video 1).

Data collection and follow-up Primary outcome measures were technical success and adverse event rates related to DBE-assisted DPEJ. Data collection also included patient demographics, indications for DPEJ, gut anatomy (native or altered), causes of failed standard DPEJ, type of sedation administered, and procedure time—from endoscope insertion to PEJ tube placement (for successful procedures) or to endoscope extubation (for failed procedures). Patients were contacted 24 hours later and 1 month later to assess tube function and adverse events. Adverse events were defined according to accepted criteria.8

Statistical analysis Descriptive statistics were reported as frequency (percentage) or mean (range), as appropriate. Comparison of mean procedure times for failed DPEJ by push enteroscopy and DBE-assisted DPEJ were performed by the Wilcoxon signed rank test because of the small sample size and non-Gaussian distribution. Statistical tests were 2-sided, and P values ⬍ .05 were considered statistically significant. 676 GASTROINTESTINAL ENDOSCOPY Volume 76, No. 3 : 2012

Song et al

RESULTS Of 33 patients referred for DPEJ tube placement, DPEJ with a pediatric colonoscope failed in 10 (30%), who then consented to DBE-assisted DPEJ. Tables 1 and 2 summarize patient characteristics and outcomes. The mean age was 62 years (range, 23-89 years), and the mean body mass index (BMI) was 25 kg/m2 (range, 14-40 kg/m2). Four patients had surgically altered gut anatomy: pylorus-preserving pancreaticoduodenectomy (n ⫽ 1), Roux-en-Y gastric bypass (n ⫽ 2), and Roux-en-Y esophagojejunostomy (n ⫽ 1). Conventional DPEJ was performed with the patients under moderate sedation in 9 of 10 patients; the mean doses of fentanyl and midazolam were 126 ␮g (range, 37.5-225 ␮g) and 6.7 mg (range, 1.5-14 mg), respectively. The procedure tolerance for conventional DPEJ was rated as good to excellent in 90% of patients and fair in 1 patient. DBE-assisted DPEJ was performed with the patients under GETA in 9 patients and with MAC in 1 patient. Causes of failed conventional DPEJ were inability to transilluminate and identify a site for tube placement in patients with normal gut anatomy (n ⫽ 5) or efferent limb in altered anatomy (n ⫽ 3), and limited endoscope advancement because of small-bowel angulation/fixation (n ⫽ 1) or excessive gastric looping (n ⫽ 1). DBE-assisted DPEJ was performed on the same day of failed standard DPEJ in 2 patients, the next day in 6, and 1 week later in 2. The estimated DPEJ tube location was proximal jejunum (n ⫽ 5) and midjejunum (n ⫽ 1) in patients with native gut anatomy. In patients with altered gut anatomy, the feeding tube was placed in the efferent jejunal limb (n ⫽ 3) and the distal Roux limb (n ⫽ 1) of a Roux-en-Y gastric bypass. The feeding tube location was left lower abdominal quadrant (n ⫽ 6) and left upper quadrant (n ⫽ 4). Glucagon was used in 4 of 10 patients (mean dose, 0.43 mg; range, 0.25-0.75 mg) during conventional push endoscopy and in 5 of 10 patients (mean dose, 0.6 mg; range, 0.25-1 mg) during DBE-assisted DPEJ, to decrease bowel motility. The mean procedure times for failed DPEJ by push enteroscopy and DBE-assisted DPEJ were similar: 31 (10-49) minutes and 29 (10-48) minutes, respectively (P ⫽ 1.00). Successful PEJ placement with use of DBE was achieved in all 10 patients. One moderately severe adverse event (peristomal cellulitis) developed less than 24 hours after the procedure and resolved with intravenous antibiotics during a 4-day hospitalization. No additional adverse events occurred either 24 hours after the procedure or at the 1-month followup. All patients tolerated DPEJ feedings. Although the study was designed for inclusion of 20 patients, recruitment was terminated early because of the success seen in 10 of 10 (100%) patients. This decision was made by the principal investigator (L. M. W. K. S.) and communicated to the Institutional Review Board, which approved. www.giejournal.org

Song et al

Double-balloon enteroscopy for failed direct percutaneous endoscopic jejunostomy

TABLE 1. Patient features, indications, and causes of failed standard DPEJ Patient

Age

Sex

BMI

1

89

F

28

2

76

M

3

73

4

Gut anatomy

Indications for DPEJ

Causes of standard DPEJ failure

Normal

CVA with failed swallow study, failed PEG due to intrathoracic stomach

Lack of transillumination in proximal jejunum

23

Pylorus-preserving pancreaticoduodenectomy

Gastroparesis

Lack of transillumination in efferent jejunal limb

M

22

Normal

Altered mentation, aspiration pneumonia, frequent PEG-J tube replacements

Lack of transillumination in proximal jejunum

67

F

14

Normal

Gastroparesis

Examination to ligament of Treitz, excessive instrument loop in stomach

5

52

F

25

RYGB

Inadequate oral intake, rapid weight loss

Lack of transillumination in Roux limb

6

54

F

22

Normal

Neurologic swallowing disorder, aspiration pneumonia

Lack of transillumination in proximal jejunum

7

55

F

25

RYGB takedown; 12 abdominal surgeries

Gastroparesis

Limited instrument advancement in jejunum caussec by small-bowel angulation and fixation

8

23

F

22

Normal

Neuropathic motility disorder, gastroparesis

Lack of transillumination in proximal jejunum

9

48

F

40

Normal

Aspiration pneumonia

Lack of transillumination in proximal jejunum

10

85

F

27

Roux-en-Y esophagojejunostomy

CVA and dysphagia

Lack of transillumination in jejunum

BMI, Body mass index; CVA, cerebrovascular accident; DPEJ, direct percutaneous endoscopic jejunostomy; PEG-J, PEG with jejunal tube extension; RYGB, Rouxen-Y gastric bypass.

DISCUSSION The DPEJ procedure was described more than 20 years ago, but placement techniques are not standardized.2,9 DPEJ provides superior results over PEG-J.1 In most cases, a standard adult or pediatric colonoscope is used for DPEJ placement, particularly in patients with nonsurgically altered gut anatomy. However, the success rate for DPEJ is limited by the ability to find an appropriate site based on transillumination and finger indentation. This, in turn, is limited by the length of small bowel traversed, because a suitable area for PEJ placement may be beyond the maximal point of insertion achieved by the colonoscope. Indeed, in the largest published study to date, we reported a success rate of only 68% out of 307 DPEJ attempts by using a colonoscope.5 This is similar to the 71% success rate found during a recent audit of 28 conventional DPEJ procedures performed at our institution in 2010 (unpublished data). Although this success rate is lower than in other series, it remains consistent in our practice over a long period of time, with a higher patient volume than in other centers. It is possible that we are seeing a more difficult patent population or that there is publication bias in www.giejournal.org

other smaller series. In addition, 4 of 10 patients in our study in whom conventional DPEJ failed had surgically altered anatomy, although in other series altered gut anatomy has been associated with improved success rates. This difference is likely due to the various types of surgically altered anatomy and a more difficult patient population (selection bias) seen at our center. With the advent of balloon-assisted and spiral enteroscopy, the depth of small-bowel insertion is substantially greater than what is achievable with a colonoscope. Thus, we hypothesized that DBE would allow successful feeding tube placement in patients in whom attempted conventional DPEJ placement with colonoscopies had failed. Although DBE-assisted DPEJ has been described,10,11 no patients in these reports experienced failure of attempted conventional DPEJ; thus, a benefit of DBE over colonoscopes was not confirmed. In our study, we targeted a difficult patient cohort—patients in whom DPEJ had failed with the use of variable-stiffness colonoscopes, which have been shown to provide greater insertion depth than dedicated push enteroscopes,12—and showed the benefit of DBE for DPEJ placement. Indeed, since March 2011, only balloon-assisted Volume 76, No. 3 : 2012 GASTROINTESTINAL ENDOSCOPY 677

Double-balloon enteroscopy for failed direct percutaneous endoscopic jejunostomy

Song et al

TABLE 2. Outcomes of DBE-assisted DPEJ as rescue technique for failed standard DPEJ DBE-assisted DPEJ adverse events

Standard DPEJ

DBE-assisted DPEJ

Standard DPEJ

DBE-assisted DPEJ

DBE-assisted DPEJ technical success

1

M

GETA

49

29

Yes

LLQ

No

No

2

M

GETA

38

44

Yes

LLQ

No

No

3

M

GETA

40

10

Yes

LLQ

No

No

4

M

GETA

31

30

Yes

LUQ

No

No

5

MAC

MAC

21

40

Yes

LLQ

No

No

6

M

GETA

39

20

Yes

LLQ

No

No

7

M

GETA

43

48

Yes

LLQ

No

No

8

M

GETA

15

20

Yes

LUQ

No

No

9

M

GETA

10

20

Yes

LUQ

Yes*

No

10

M

GETA

22

29

Yes

LUQ

No

No

Sedation Patient

Procedure time (minutes)

DBE-assisted DPEJ site

At 24 hrs follow-up

At 1 month follow-up

DBE, Double-balloon enteroscopy; DPEJ, direct percutaneous endoscopic jejunostomy; GETA, general endotracheal anesthesia; LLQ, left lower abdominal quadrant; LUQ, left upper abdominal quadrant; M, moderate sedation (fentanyl and midazolam); MAC, monitored anesthesia care. *Peristomal cellulitis.

enteroscopy (BAE) has been used for DPEJ placement in our practice. Although one consideration might be to attempt DPEJ with a colonoscope and change to DBE assistance if necessary, logistic issues (need to potentially switch between 2 different endoscope systems) and patient safety issues (eg, prolonged procedure) may offset any potential gain starting with the standard colonoscope. Since the completion of this study, 29 patients (4 with altered gut anatomy) have undergone DBE-assisted DPEJ, with successful placement in 93%. In 2 patients, DBE advancement beyond the distal duodenum was not feasible because of small-bowel fixation with significant luminal narrowing from dense adhesions and an inflammatory mesenteric mass, necessitating surgical jejunostomy. PEJ placement via single-balloon enteroscopy may also be more successful than conventional DPEJ.13 There are additional advantages for using BAE over standard push endoscopes for DPEJ placement. In contrast to push enteroscopy, BAE minimizes instrument looping within the stomach and small bowel, which can result in marked patient discomfort. Despite the fact that the BAE procedures in this study were performed with anesthesia support, we believe that BAE is associated with less discomfort than push enteroscopy and that BAE-assisted DPEJ placement can be performed with the patient under moderate sedation or MAC rather than GETA. Another advantage of BAE is the flexibility of the tip of the endoscope, which can more easily navigate through smallbowel angulation and narrowing (as seen in patients with prior abdominal surgery) than a colonoscope. 678 GASTROINTESTINAL ENDOSCOPY Volume 76, No. 3 : 2012

The success rates (60%-73%) for standard DPEJ placement in overweight and obese patients (BMI ⱖ25 kg/m2) are less than in patients who are underweight (96%) or who have normal BMI (81%), and with a trend toward more frequent major adverse events.14 In our study, 5 of 10 patients had a BMI of 25 kg/m2 or higher, and the only adverse event (peristomal cellulitis) occurred in the patient with the highest BMI (40 kg/m2). A study limitation is the difference in sedation used for DPEJ by push enteroscopy and DBE-assisted DPEJ. However, the causes of failed DPEJ placement by push enteroscopy were lack of transillumination and limited instrument advancement unrelated to sedation issues, because procedure tolerance was rated as good to excellent in 9 of 10 patients. Additionally, inherent study bias may have been present because of the small sample size. In conclusion, DBE-assisted DPEJ is highly effective as a salvage technique for patients in whom DPEJ with variablestiffness pediatric colonoscopies has failed. Our findings suggest that when DPEJ is undertaken, one should consider using DBE assistance, although further studies comparing the efficacy and safety of BAE with those of conventional push enteroscopy for DPEJ placement are needed before formal changes in practice can be recommended.

REFERENCES 1. Fan AC, Baron TH, Rumalla A, et al. Comparison of direct percutaneous endoscopic jejunostomy and PEG with jejunal extension. Gastrointest Endosc 2002;56:890-4.

www.giejournal.org

Penn et al

Stents for EUS-guided transmural drainage of pancreatic pseudocysts

2. Rumalla A, Baron TH. Results of direct percutaneous endoscopic jejunostomy, an alternative method for providing jejunal feeding. Mayo Clin Proc 2000;75:807-10. 3. Panagiotakis PH, DiSario JA, Hilden K, et al. DPEJ tube placement prevents aspiration pneumonia in high-risk patients. Nutr Clin Pract 2008;23:172-5. 4. Kwon RS, Banerjee S, Desilets D, et al. Enteral nutrition access devices. Gastrointest Endosc 2010;72:236-48. 5. Maple JT, Petersen BT, Baron TH, et al. Direct percutaneous endoscopic jejunostomy: outcomes in 307 consecutive attempts. Am J Gastroenterol 2005;100:2681-8. 6. Foutch PG, Talbert GA, Waring JP, et al. Percutaneous endoscopic gastrostomy in patients with prior abdominal surgery: virtues of the safe tract. Am J Gastroenterol 1988;83:147-50. 7. Varadarajulu S, Delegge MH. Use of a 19-gauge injection needle as a guide for direct percutaneous endoscopic jejunostomy tube placement. Gastrointest Endosc 2003;57:942-5. 8. Cotton PB, Eisen GM, Aabakken L, et al. A lexicon for endoscopic adverse events: report of an ASGE workshop. Gastrointest Endosc 2010;71:446-54.

9. Baron TH. Direct percutaneous endoscopic jejunostomy. Am J Gastroenterol 2006;101:1407-9. 10. Despott EJ, Gabe S, Tripoli E, et al. Enteral access by double-balloon enteroscopy: an alternative method of direct percutaneous endoscopic jejunostomy placement. Dig Dis Sci 2011;56:494-8. 11. Mönkemüller K, Vormbrock K, Kassalik M, et al. Direct percutaneous endoscopic jejunostomy tube placement using double-balloon enteroscopy. Gastrointest Endosc 2012;75:463-5. 12. Harewood GC, Gostout CJ, Farrell MA, et al. Prospective controlled assessment of variable stiffness enteroscopy. Gastrointest Endosc 2003; 58:267-71. 13. Aktas H, Mensink PB, Kuipers EJ, et al. Single-balloon enteroscopyassisted direct percutaneous endoscopic jejunostomy. Endoscopy 2012;44:210-2. 14. Mackenzie SH, Haslem D, Hilden K, et al. Success rate of direct percutaneous endoscopic jejunostomy in patients who are obese. Gastrointest Endosc 2008;67:265-9.

Prospective evaluation of the use of fully covered self-expanding metal stents for EUS-guided transmural drainage of pancreatic pseudocysts D. Eli Penn, MD, Peter V. Draganov, MD, Mihir S. Wagh, MD, Chris E. Forsmark, MD, Anand R. Gupte, MD, Shailendra S. Chauhan, MD Gainesville, Florida, USA

Endoscopic drainage has become the procedure of choice for the management of symptomatic pancreatic pseudocysts in high-volume centers.1 In recent years, the use of EUS guidance for transmural pseudocyst drainage has gained popularity because of its ability to locate a suitable puncture site in patients without obvious extrinsic gastric or duodenal compression and its avoidance of intramural vessels during the initial pseudocyst puncture.1-6 Although various techniques have been described, the basic transmural drainage is performed by first accessing the pseudocyst; cauterizing the tract, dilating it, or both; and finally inserting multiple plastic stents to

Abbreviation: CSEMS, fully covered self-expanding metal stent. DISCLOSURE: The following author disclosed financial relationships relevant to this publication: P. V. Draganov: consultant to Boston Scientific and Cook. All other authors disclosed no financial relationships relevant to this publication. Copyright © 2012 by the American Society for Gastrointestinal Endoscopy 0016-5107/$36.00 http://dx.doi.org/10.1016/j.gie.2012.04.457 Received January 19, 2012. Accepted April 17, 2012. Current affiliations: Division of Gastroenterology, Hepatology, and Nutrition, Department of Internal Medicine, University of Florida College of Medicine, Gainesville, Florida, USA. Reprint requests: Shailendra S. Chauhan, MD, Division of Gastroenterology, Hepatology, and Nutrition, Department of Medicine, University of Florida College of Medicine, 1600 SW Archer Road, Gainesville, FL 32610-0214.

www.giejournal.org

facilitate drainage and maintain tract patency.7,8 Once the patient has improved clinically and the pseudocyst is resolved radiographically, the stents are removed. However, the placement of multiple plastic stents can be technically difficult and tedious because of the need to access the cyst cavity multiple times or the need to use 2 wires simultaneously to maintain access. Furthermore, 10F plastic stents can be hard to deploy through the relatively small 3.7-mm channel of the therapeutic linear echoendoscope scope. Recently, groups have reported the use of fully covered self-expanding metal stents (CSEMSs) for pseudocyst drainage.9 CSEMSs offer some advantages in that only a single stent may be required rather than multiple plastic stents. Moreover, they provide a larger diameter than do plastic stents. Therefore, they can theoretically allow for faster drainage and a decreased risk of occlusion, which might reduce the need for repeated procedures. In addition, pseudocyst drainage with CSEMSs eliminates the need to access the cyst cavity multiple times and the simultaneous use of 2 guidewires to secure cyst access. To our knowledge, none of these potential advantages of CSEMSs have been rigorously evaluated by prospective studies. Furthermore, there is concern that because of the presence of silicone coating on CSEMSs, they may have a higher migration rate. In this study, we prospectively evaluated the technical feasibility of EUS-guided single-access pseudocyst drainage with a CSEMS anchored with a double pigtail plastic stent inserted through the metal stent lumen. Additionally, we assessed for pseudocyst resolution and adverse events. Volume 76, No. 3 : 2012 GASTROINTESTINAL ENDOSCOPY 679