Surg Endosc DOI 10.1007/s00464-013-3368-7
and Other Interventional Techniques
Palliative venting percutaneous endoscopic gastrostomy tube is safe and effective in patients with malignant obstruction Rachel B. Issaka • David M. Shapiro • Neehar D. Parikh • Mary F. Mulcahy • Srinadh Komanduri • John A. Martin Rajesh N. Keswani
•
Received: 14 June 2013 / Accepted: 1 December 2013 Ó Springer Science+Business Media New York 2013
Abstract Background and study aims Obstructive symptoms are common in advanced malignancies. Venting percutaneous endoscopic gastrostomy (VPEG) tubes can be placed for palliation. The aim of this study was to determine the outcomes of VPEG placement in patients with advanced malignancy. Methods We retrospectively reviewed patients in whom a VPEG tube was attempted for a malignant indication from 1998 to 2010 at a tertiary care center. Clinical information, procedure details, and adverse events (AEs) were recorded. Results Ninety-six patients meeting the inclusion criteria were identified. Colorectal (27 %), pancreas (18.8 %), and gynecologic (17.7 %) malignancies were most common. Overall, 46.9 % of patients had ascites, with 35.6 % undergoing drainage prior to VPEG placement. VPEG placement was successful in 89 patients (92.7 %), and relief of obstructive symptoms was observed in 91.0 % of patients. Seven patients had refractory symptoms despite functioning VPEG tube. Ten post-procedural AEs were noted in nine patients, with one death. Infectious
R. B. Issaka (&) Division of Internal Medicine, Feinberg School of Medicine, Northwestern University, 251 E. Huron, Galter 3-150, Chicago, IL 60611, USA e-mail: [email protected]; [email protected] D. M. Shapiro N. D. Parikh S. Komanduri J. A. Martin R. N. Keswani Division of Gastroenterology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA M. F. Mulcahy Division of Hematology/Oncology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA
complications were more common in patients with ascites (12.2 %) versus those without (0 %; p = 0.02). There was a trend towards decreased infectious AEs when ascites was drained prior to VPEG (14.8 vs. 7.1 %; p = 0.64) in our patient cohort. We observed a decreased survival when AEs occurred (73 ± 47.8 days) compared with when they did not occur (141 ± 367.8 days; p = 0.61). Conclusions VPEG tubes can be safely placed in patients with obstructive symptoms due to inoperable malignancy, with complete relief in the majority of patients. Ascites was a risk factor for post-procedural infectious AEs. Drainage of ascites prior to VPEG tube placement may decrease this risk, although this requires further study. Keywords Cancer Endoscopic retrograde cholangiopancreatography (ERCP) Quality of life Therapeutic/palliation Endoscopy
Refractory nausea and vomiting are common symptoms in patients with advanced malignancy and lead to diminished quality of life [1–4]. The most common etiologies of nausea and vomiting in these patients are impaired gastric emptying, medication side effects, and gastrointestinal obstruction [5]. Although medications may improve some of the symptoms in mechanical obstruction, they are inadequate for complete relief [6]. Focal obstruction is best addressed either by endoscopic stenting or surgical bypass [7]. However, there is a subset of patients in whom gastrointestinal obstruction is especially challenging due to multifocal obstruction. In these patients, surgical bypass or endoscopic stenting are rarely viable options [8, 9]. Refractory nausea and vomiting leads to decreased quality of life and frequent hospitalizations for management of this syndrome, which often includes insertion of nasogastric
123
Surg Endosc Table 1 Summary of current literature Study
Study size
Primary malignancy (n)
Ascites present [n (%)]
Symptom relief [n (%)]
Brooksbank et al. [11]
51
Colon and rectum (27)
NA
47 (92)
Complication rate [n (%)] 7 (13.7)
Campagnutta et al. [20]
34
Ovarian cancer (29)
NA
27 (84)
2 (6.2)
Herman et al. [19]
46
Ovarian cancer (26)
NA
41 (89)
2 (4.3)
Pothuri et al. [14]
94
Ovarian cancer (94)
59 (63)
86 (91)
18 (19.1)
Vashi et al. [17]
73
Ovarian and cervical (49)
NA
73 (100)
10 (13.7)
This study
96
Colon and rectum (26)
45 (46.9)
81 (91)
9 (9.4)
NA not available
tubes (NGT) for decompression [4, 10]. However, NGT placement is an inadequate long-term solution as patients must remain in hospital until the symptoms subside. In these patients, a venting percutaneous endoscopic gastrostomy (VPEG) tube may be a preferred option, allowing patients to achieve palliation at home [11–14]. The percutaneous endoscopic gastrostomy (PEG) tube was originally described in 1979 [15]. Its most common indication remains long-term enteral feeding in those unable to maintain sufficient oral intake due to impaired swallowing [16]. Alternatively, PEG tubes may be used for venting purposes in those patients with advanced malignancy and chronic gastrointestinal obstruction. However, there are currently no practice guidelines that exist for this purpose, and data is sparse regarding the clinical outcomes in this setting [11, 12, 14, 17, 18], with the majority of data in the gynecological literature [12–14, 18–20]. The largest study to date in a heterogeneous cancer population, described 51 patients who received a VPEG for obstructive symptoms as a palliative measure, with the majority of patients symptomatically improving [11]. However, the study did not detail the characteristics of these patients, leaving little guidance as to which patients benefit most from VPEG tubes and whether placement is safe in end-stage disease, such as those patients with malignant ascites. Current existing data regarding this topic are summarized in Table 1. We hypothesize that placement of VPEG in this patient population is safe and feasible and may decrease the recurrence of obstructive symptoms in addition to serving as palliation for otherwise inoperable malignancies. Herein, we present the largest study to date examining the outcomes of patients who received VPEG for refractory nausea and vomiting secondary to obstruction from unresectable malignancies. We also aim to identify predictors of those patients who would benefit most from VPEG placement and the risk factors for procedural adverse events (AEs) [21].
123
Patients and methods Study design and patient population We retrospectively analyzed the data of all patients in whom a VPEG tube was attempted for a malignant indication over a 13-year period (1998–2010) at a single tertiary care academic center (Northwestern Memorial Hospital). All patients who had a VPEG attempted due to malignant obstruction over this time period were identified through a query of the Northwestern Medical Enterprise Data Warehouse, an integrated database of clinical and research data from patients receiving treatment through Northwestern medical affiliates. Selection for VPEG placement was not standardized and was at the discretion of the primary gastroenterologist following consultation by the primary medical team; patients who were not offered an attempt at VPEG are not included in this study. Further demographic and clinical information was obtained from manual review of the electronic medical record. All manual data were collected by two physicians (RI and DS) and verified by the study lead investigator (RK). Patients who received VPEG for any other indication other than malignant obstructive symptoms were excluded. The Institutional Review Board at Northwestern University approved this study. The indication for placement in all cases was persistent nausea and vomiting due to malignant gastrointestinal obstruction. Clinical information, including malignancy type, hospitalization history, procedure details, NGT use, and symptom resolution, was extracted. We also reviewed radiographic studies to determine the presence or absence of visible gastric metastatic disease, herein referred to as gastric disease, abdominal carcinomatosis, and ascites. All endoscopically placed tubes were placed using the ‘pull’ method. Tube size was not standardized. T-fasteners were not used for endoscopic tubes. We determined whether ascites was drained, and the method of drainage prior to the procedure. The decision to
Surg Endosc Table 2 Baseline characteristics (n = 96)
Table 3 Primary malignancy type (n = 96) [n (%)]
Age ± SD (years)
57 ± 14
Colorectal
26 (27.1)
Female gender [n (%)]
57 (59.4)
Pancreas
18 (18.8)
Inpatient [n (%)]
91 (94.8)
Gynecologic
17 (17.7)
Body mass index ± SD
24.2 ± 5.6
Gastric
Abdominal carcinomatosis [n (%)]
64 (66.6)
Cholangiocarcinoma
5 (5.2)
Gastric disease [n (%)]
49 (51.0)
Adenocarcinoma of unknown primary
5 (5.2)
Ascites [n (%)]
45 (46.9)
Transitional cell carcinoma
3 (3.1)
Appendiceal Other
drain ascites, time, method, and amount drained was not standardized and occurred at the discretion of the team of clinicians; it typically occurred 1–3 days prior to VPEG placement via large-volume paracentesis. It is standard practice for ultrasounds to be used in bedside and interventional radiology-performed paracentesis at our institution. Patients were maintained on parenteral nutrition and advised to limit oral intake. We reviewed the medical chart to identify re-hospitalizations for obstructive symptoms and VPEG malfunction. Immediate and delayed AEs were noted. The Social Security Death Index Interactive Search was used to determine the date of death.
6 (6.3)
3 (3.1) 13 (13.5)
(66.6 %) and gastric disease was noted in 49 patients (51.0 %). Ascites was also present in 46.9 % of patients, with drainage performed in only a minority of these patients (35 %, n = 16). In all patients, the etiology of ascites was presumed to be peritoneal metastatic disease. Drainage was performed by paracentesis (n = 11) or permanent drainage catheter (n = 5). Broad-spectrum gramnegative pre-procedural antibiotics were administered to all patients. Gastroenterology trainees assisted in 57.3 % (n = 55) of all cases. VPEG technical success
Statistics Patient demographic information and clinical characteristics are described by means and ranges for continuous variables, and by presence or absence for categorical variables. Fisher’s exact test was used to compare outcomes in the patients who suffered AEs versus those who did not, as well as determine the significance of post-procedural survival time. p-Values \0.05 were considered statistically significant and all tests were two-sided. Two separate reviewers completed the data analysis to ensure congruency in the final reported values.
Results
Venting percutaneous endoscopic gastrostomy was successfully placed by gastroenterologists in 89 patients (92.7 %), 88 on the first attempt. Of the remaining seven patients, VPEG could not be placed because of failure to transilluminate the anterior abdominal wall. In six patients with unsuccessful VPEG placement (85.7 %), failure was attributed to peritoneal metastases resulting in no appropriate window for placement. The etiology of unsuccessful placement was not detailed in the remaining patient. In those patients who failed endoscopic placement, venting gastrostomy was placed by interventional radiology in four patients, and surgically in two patients. Placement was unsuccessful in one patient despite attempts by endoscopy, interventional radiology, and surgery (Fig. 1).
Patient demographics VPEG clinical success We identified 96 patients in whom an endoscopic VPEG placement was attempted over the 13-year period. However, we were unable to determine the total volume of VPEGs placed during this time. Overall, 95 % (n = 91) of the procedures were performed during an inpatient hospitalization. The median age at the time of the procedure was 57 years (range 21–90 years) (Table 2). The most common malignancies were colorectal (27.1 %), pancreatic (18.8 %), and gynecologic (17.7 %) (Table 3). Abdominal carcinomatosis was present in the majority of patients
Complete relief of nausea and vomiting was observed in the majority of patients (n = 81, 91.0 %). Four of these patients were re-admitted for recurrent symptoms due to VPEG tube dysfunction caused by clogging, and all were successfully replaced. Seven patients (7.9 %) were readmitted for obstructive symptoms despite a functioning VPEG tube, and were considered treatment failures. A single patient expired before symptom relief could be assessed.
123
Surg Endosc Fig. 1 VPEG placement data. VPEG venting percutaneous endoscopic gastrostomy
Table 4 Procedure adverse events Patient no.
Gender
Ascites
Abdominal carcinomatosis
Gastric disease
Adverse events
1
M
No
Yes
No
PEG occlusion
2
F
Yes
Yes
No
PEG occlusion
3
F
Yes
Yes
No
PEG occlusion
4
F
Yes
Yes
No
Infection
5
M
Yes
Yes
No
PEG occlusion/infection
6
M
No
No
No
PEG migration
7
F
Yes
Yes
Yes
Infection
8
M
Yes
Yes
No
Infection
9
M
Yes
Yes
No
Infection/sepsis/death
M male, F female, PEG percutaneous endoscopic gastrostomy
An NGT was present prior to VPEG placement in 73 % of patients (n = 65). Documentation on clinical response to NGT was available in only 36 patients, with 32 (89 %) patients reporting improvement in symptoms with NGT. Of those patients reporting clinical improvement with NGT, 93 % (n = 30) had no re-hospitalizations for obstructive symptoms following VPEG tube placement. In the four patients without symptom relief from NGT, resolution of obstructive symptoms occurred in only a single patient. Procedural adverse events Ten post-procedural AEs were noted in nine patients. The most common observed AE was infection (n = 5), with one death due to suspected intra-abdominal sepsis (Table 4). Ascites was more likely to be present in patients who had post-procedural infectious AEs (12.2 vs. 0 %; p = 0.02) (Fig. 2). There was a trend towards decreased infectious AEs when ascites was drained prior to endoscopic VPEG placement (7.1 vs. 14.8 %; p = 0.64). When positive ascitic fluid cultures were obtained, organisms
123
were typical gastrointestinal bacteria, including Enterobacter cloacae and Klebsiella pneumonia. Gastric involvement of the tumor was associated with a decreased rate of AEs compared with patients without gastric disease (22.9 vs. 2.2 %; p = 0.013). In the patients in whom follow-up was documented, overall mean survival after VPEG placement was 135 ± 347.9 days (range 5–2,772 days). We observed a trend towards decreased survival when AEs occurred (73 ± 47.8 days) compared with when they did not (141 ± 367.8 days; p = 0.61). Thirty-day all-cause mortality rate within our cohort was 22.9 %.
Discussion Patients with inoperable obstructive gastrointestinal malignancies can be especially difficult to manage given recurrent nausea and vomiting that is not always alleviated by medications. There is emerging evidence suggesting that, in these patients, placement of VPEG tubes may improve symptoms and lead to more hospital-free days [7,
Surg Endosc
Fig. 2 VPEG adverse events. Infectious adverse events were increased in patients with ascites who underwent VPEG placement. Drainage of ascites prior to VPEG placement appeared to reduce the
risk of infectious adverse events. Overall adverse events were decreased in patients who had gastric involvement of tumor. VPEG venting percutaneous endoscopic gastrostomy
11–14, 22]. However, the majority of these studies are smaller case series and thus provide limited clinical guidance on how to approach more complex patients such as those with ascites or abdominal carcinomatosis who require VPEG placement. Based on the results of our single-center experience, the largest study to date, we conclude that VPEG tubes can be successfully placed endoscopically in nearly all patients. Furthermore, most patients with inoperable malignancy and obstructive symptoms had complete relief of symptoms and did not require re-hospitalization for these symptoms after VPEG placement. We also observed that a clinical response to NGT placement could serve as a predictor of those in whom endoscopic VPEG placement might be most successful. Thus, placement should be performed if the etiology of symptoms is unclear prior to placement of VPEG tubes. Based on our findings, the presence of ascites and abdominal carcinomatosis are not absolute contraindications to successful VPEG placement; however, extensive abdominal carcinomatosis was common in failed endoscopic VPEG placement. Additionally, while endoscopic VPEG placement is successful in patients with malignant ascites, we noted its presence as a risk factor for postprocedural infectious AEs, suggesting that this may be a useful predictor when attempting to risk-stratify patients. This risk is possibly decreased by drainage of fluid prior to VPEG tube placement, and we advocate that this should be performed whenever possible. However, given the
multifactorial issues that may be present in patients with ascites from underlying malignancies, this warrants further study with larger cohorts. Interestingly, we found a decreased rate of overall complications in patients with gastric involvement of the tumor. The reasons for this remain unclear but it may be that larger tumors adhere to the anterior abdominal wall and thus may facilitate VPEG placement. Additionally, patients with visible gastric disease could present with symptoms earlier and may be an overall healthier cohort of patients than those with abdominal carcinomatosis; however, this finding requires further validation. Reassuringly, our study also revealed that the majority of those who received VPEG tubes were rarely re-admitted for recurrent obstructive symptoms. The mean patient survival in this population (135 ± 347.9 days) was significantly greater than previously reported and suggests that VPEG placement is a durable palliative solution [11, 22]. As expected, the length of survival was increased when procedural AEs did not occur.
Conclusions In those patients with malignant obstruction not amenable to enteral stenting or surgical bypass, VPEG placement has a high technical and clinical success rate with an acceptable AEs rate. Although ascites and malignant abdominal carcinomatosis appeared to increase the risk of AEs, the
123
Surg Endosc
majority of patients still achieved clinical success. Thus, in patients whose therapeutic options are limited, endoscopic placement of venting gastrostomy tube should be considered for palliation.
11.
12. Disclosures Rachel Issaka, David Shapiro, Neehar Parikh, Mary Mulcahy, Srinadh Komanduri, John Martin, and Rajesh Keswani have no conflicts of interest to declare.
References 1. Bentley A, Boyd K (2001) Use of clinical pictures in the management of nausea and vomiting: a prospective audit. Palliat Med 15:247–253 2. Coates A, Abraham S, Kaye SB, Sowerbutts T, Frewin C, Fox RM, Tattersall MH (1983) On the receiving end: patient perception of the side-effects of cancer chemotherapy. Eur J Cancer Clin Oncol 19:203–208 3. Griffin AM, Butow PN, Coates AS, Childs AM, Ellis PM, Dunn SM, Tattersall MH (1996) On the receiving end. V: patient perceptions of the side effects of cancer chemotherapy in 1993. Ann Oncol 7:189–195 4. Osoba D, Zee B, Warr D, Latreille J, Kaizer L, Pater J (1997) Effect of postchemotherapy nausea and vomiting on healthrelated quality of life. The Quality of Life and Symptom Control Committees of the National Cancer Institute of Canada Clinical Trials Group. Support Care Cancer 5:307–313 5. Stephenson J, Davies A (2006) An assessment of aetiology-based guidelines for the management of nausea and vomiting in patients with advanced cancer. Support Care Cancer 14:348–353 6. Ang SK, Shoemaker LK, Davis MP (2010) Nausea and vomiting in advanced cancer. Am J Hosp Palliat Care 27:219–225 7. Dolan EA (2011) Malignant bowel obstruction: a review of current treatment strategies. Am J Hosp Palliat Care 28:576–582 8. Keswani RN, Azar RR, Edmundowicz SA, Zhang Q, Ammar T, Banerjee B, Early DS, Jonnalagadda SS (2009) Stenting for malignant colonic obstruction: a comparison of efficacy and complications in colonic versus extracolonic malignancy. Gastrointest Endosc 69:675–680 9. Jung MK, Park SY, Jeon SW, Cho CM, Tak WY, Kweon YO, Kim SK, Choi YH, Kim GC, Ryeom HK (2010) Factors associated with the long-term outcome of a self-expandable colon stent used for palliation of malignant colorectal obstruction. Surg Endosc 24:525–530 10. Ripamonti C, Mercadante S, Groff L, Zecca E, De Conno F, Casuccio A (2000) Role of octreotide, scopolamine butylbromide,
123
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
and hydration in symptom control of patients with inoperable bowel obstruction and nasogastric tubes: a prospective randomized trial. J Pain Symptom Manage 19:23–34 Brooksbank MA, Game PA, Ashby MA (2002) Palliative venting gastrostomy in malignant intestinal obstruction. Palliat Med 16:520–526 Campagnutta E, Cannizzaro R (2000) Percutaneous endoscopic gastrostomy (PEG) in palliative treatment of non-operable intestinal obstruction due to gynecologic cancer: a review. Eur J Gynaecol Oncol 21:397–402 Meyer L, Pothuri B (2006) Decompressive percutaneous gastrostomy tube use in gynecologic malignancies. Curr Treat Options Oncol 7:111–120 Pothuri B, Montemarano M, Gerardi M, Shike M, Ben-Porat L, Sabbatini P, Barakat RR (2005) Percutaneous endoscopic gastrostomy tube placement in patients with malignant bowel obstruction due to ovarian carcinoma. Gynecol Oncol 96:330–334 Gauderer MW, Ponsky JL, Izant RJ Jr (1980) Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 15:872–875 Eisen GM, Baron TH, Dominitz JA, Faigel DO, Goldstein JL, Johanson JF, Mallery JS, Raddawi HM, Vargo JJ 2nd, Waring JP, Fanelli RD, Wheeler-Harbough J (2002) Role of endoscopy in enteral feeding. Gastrointest Endosc 55:794–797 Vashi PG, Dahlk S, Vashi RP, Gupta D (2011) Percutaneous endoscopic gastrostomy tube occlusion in malignant peritoneal carcinomatosis-induced bowel obstruction. Eur J Gastroenterol Hepatol 23:1069–1073 Campagnutta E, Cannizzaro R, De Cicco M, De Piero G, Giorda G, Sopracordevole F, Parin A, Scarabelli C (1998) Percutaneous endoscopic gastrostomy (PEG) in upper gastrointestinal tract occlusion in gynecologic oncology. Minerva Ginecol 50:305–311 in Italian Herman LL, Hoskins WJ, Shike M (1992) Percutaneous endoscopic gastrostomy for decompression of the stomach and small bowel. Gastrointest Endosc 38:314–318 Campagnutta E, Cannizzaro R, Gallo A, Zarrelli A, Valentini M, De Cicco M, Scarabelli C (1996) Palliative treatment of upper intestinal obstruction by gynecological malignancy: the usefulness of percutaneous endoscopic gastrostomy. Gynecol Oncol 62:103–105 Cotton PB, Eisen GM, Aabakken L, Baron TH, Hutter MM, Jacobson BC, Mergener K, Nemcek A Jr, Petersen BT, Petrini JL, Pike IM, Rabeneck L, Romagnuolo J, Vargo JJ (2010) A lexicon for endoscopic adverse events: report of an ASGE workshop. Gastrointest Endosc 71:446–454 Teriaky A, Gregor J, Chande N (2012) Percutaneous endoscopic gastrostomy tube placement for end-stage palliation of malignant gastrointestinal obstructions. Saudi J Gastroenterol 18:95–98
and Other Interventional Techniques
Palliative venting percutaneous endoscopic gastrostomy tube is safe and effective in patients with malignant obstruction Rachel B. Issaka • David M. Shapiro • Neehar D. Parikh • Mary F. Mulcahy • Srinadh Komanduri • John A. Martin Rajesh N. Keswani
•
Received: 14 June 2013 / Accepted: 1 December 2013 Ó Springer Science+Business Media New York 2013
Abstract Background and study aims Obstructive symptoms are common in advanced malignancies. Venting percutaneous endoscopic gastrostomy (VPEG) tubes can be placed for palliation. The aim of this study was to determine the outcomes of VPEG placement in patients with advanced malignancy. Methods We retrospectively reviewed patients in whom a VPEG tube was attempted for a malignant indication from 1998 to 2010 at a tertiary care center. Clinical information, procedure details, and adverse events (AEs) were recorded. Results Ninety-six patients meeting the inclusion criteria were identified. Colorectal (27 %), pancreas (18.8 %), and gynecologic (17.7 %) malignancies were most common. Overall, 46.9 % of patients had ascites, with 35.6 % undergoing drainage prior to VPEG placement. VPEG placement was successful in 89 patients (92.7 %), and relief of obstructive symptoms was observed in 91.0 % of patients. Seven patients had refractory symptoms despite functioning VPEG tube. Ten post-procedural AEs were noted in nine patients, with one death. Infectious
R. B. Issaka (&) Division of Internal Medicine, Feinberg School of Medicine, Northwestern University, 251 E. Huron, Galter 3-150, Chicago, IL 60611, USA e-mail: [email protected]; [email protected] D. M. Shapiro N. D. Parikh S. Komanduri J. A. Martin R. N. Keswani Division of Gastroenterology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA M. F. Mulcahy Division of Hematology/Oncology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA
complications were more common in patients with ascites (12.2 %) versus those without (0 %; p = 0.02). There was a trend towards decreased infectious AEs when ascites was drained prior to VPEG (14.8 vs. 7.1 %; p = 0.64) in our patient cohort. We observed a decreased survival when AEs occurred (73 ± 47.8 days) compared with when they did not occur (141 ± 367.8 days; p = 0.61). Conclusions VPEG tubes can be safely placed in patients with obstructive symptoms due to inoperable malignancy, with complete relief in the majority of patients. Ascites was a risk factor for post-procedural infectious AEs. Drainage of ascites prior to VPEG tube placement may decrease this risk, although this requires further study. Keywords Cancer Endoscopic retrograde cholangiopancreatography (ERCP) Quality of life Therapeutic/palliation Endoscopy
Refractory nausea and vomiting are common symptoms in patients with advanced malignancy and lead to diminished quality of life [1–4]. The most common etiologies of nausea and vomiting in these patients are impaired gastric emptying, medication side effects, and gastrointestinal obstruction [5]. Although medications may improve some of the symptoms in mechanical obstruction, they are inadequate for complete relief [6]. Focal obstruction is best addressed either by endoscopic stenting or surgical bypass [7]. However, there is a subset of patients in whom gastrointestinal obstruction is especially challenging due to multifocal obstruction. In these patients, surgical bypass or endoscopic stenting are rarely viable options [8, 9]. Refractory nausea and vomiting leads to decreased quality of life and frequent hospitalizations for management of this syndrome, which often includes insertion of nasogastric
123
Surg Endosc Table 1 Summary of current literature Study
Study size
Primary malignancy (n)
Ascites present [n (%)]
Symptom relief [n (%)]
Brooksbank et al. [11]
51
Colon and rectum (27)
NA
47 (92)
Complication rate [n (%)] 7 (13.7)
Campagnutta et al. [20]
34
Ovarian cancer (29)
NA
27 (84)
2 (6.2)
Herman et al. [19]
46
Ovarian cancer (26)
NA
41 (89)
2 (4.3)
Pothuri et al. [14]
94
Ovarian cancer (94)
59 (63)
86 (91)
18 (19.1)
Vashi et al. [17]
73
Ovarian and cervical (49)
NA
73 (100)
10 (13.7)
This study
96
Colon and rectum (26)
45 (46.9)
81 (91)
9 (9.4)
NA not available
tubes (NGT) for decompression [4, 10]. However, NGT placement is an inadequate long-term solution as patients must remain in hospital until the symptoms subside. In these patients, a venting percutaneous endoscopic gastrostomy (VPEG) tube may be a preferred option, allowing patients to achieve palliation at home [11–14]. The percutaneous endoscopic gastrostomy (PEG) tube was originally described in 1979 [15]. Its most common indication remains long-term enteral feeding in those unable to maintain sufficient oral intake due to impaired swallowing [16]. Alternatively, PEG tubes may be used for venting purposes in those patients with advanced malignancy and chronic gastrointestinal obstruction. However, there are currently no practice guidelines that exist for this purpose, and data is sparse regarding the clinical outcomes in this setting [11, 12, 14, 17, 18], with the majority of data in the gynecological literature [12–14, 18–20]. The largest study to date in a heterogeneous cancer population, described 51 patients who received a VPEG for obstructive symptoms as a palliative measure, with the majority of patients symptomatically improving [11]. However, the study did not detail the characteristics of these patients, leaving little guidance as to which patients benefit most from VPEG tubes and whether placement is safe in end-stage disease, such as those patients with malignant ascites. Current existing data regarding this topic are summarized in Table 1. We hypothesize that placement of VPEG in this patient population is safe and feasible and may decrease the recurrence of obstructive symptoms in addition to serving as palliation for otherwise inoperable malignancies. Herein, we present the largest study to date examining the outcomes of patients who received VPEG for refractory nausea and vomiting secondary to obstruction from unresectable malignancies. We also aim to identify predictors of those patients who would benefit most from VPEG placement and the risk factors for procedural adverse events (AEs) [21].
123
Patients and methods Study design and patient population We retrospectively analyzed the data of all patients in whom a VPEG tube was attempted for a malignant indication over a 13-year period (1998–2010) at a single tertiary care academic center (Northwestern Memorial Hospital). All patients who had a VPEG attempted due to malignant obstruction over this time period were identified through a query of the Northwestern Medical Enterprise Data Warehouse, an integrated database of clinical and research data from patients receiving treatment through Northwestern medical affiliates. Selection for VPEG placement was not standardized and was at the discretion of the primary gastroenterologist following consultation by the primary medical team; patients who were not offered an attempt at VPEG are not included in this study. Further demographic and clinical information was obtained from manual review of the electronic medical record. All manual data were collected by two physicians (RI and DS) and verified by the study lead investigator (RK). Patients who received VPEG for any other indication other than malignant obstructive symptoms were excluded. The Institutional Review Board at Northwestern University approved this study. The indication for placement in all cases was persistent nausea and vomiting due to malignant gastrointestinal obstruction. Clinical information, including malignancy type, hospitalization history, procedure details, NGT use, and symptom resolution, was extracted. We also reviewed radiographic studies to determine the presence or absence of visible gastric metastatic disease, herein referred to as gastric disease, abdominal carcinomatosis, and ascites. All endoscopically placed tubes were placed using the ‘pull’ method. Tube size was not standardized. T-fasteners were not used for endoscopic tubes. We determined whether ascites was drained, and the method of drainage prior to the procedure. The decision to
Surg Endosc Table 2 Baseline characteristics (n = 96)
Table 3 Primary malignancy type (n = 96) [n (%)]
Age ± SD (years)
57 ± 14
Colorectal
26 (27.1)
Female gender [n (%)]
57 (59.4)
Pancreas
18 (18.8)
Inpatient [n (%)]
91 (94.8)
Gynecologic
17 (17.7)
Body mass index ± SD
24.2 ± 5.6
Gastric
Abdominal carcinomatosis [n (%)]
64 (66.6)
Cholangiocarcinoma
5 (5.2)
Gastric disease [n (%)]
49 (51.0)
Adenocarcinoma of unknown primary
5 (5.2)
Ascites [n (%)]
45 (46.9)
Transitional cell carcinoma
3 (3.1)
Appendiceal Other
drain ascites, time, method, and amount drained was not standardized and occurred at the discretion of the team of clinicians; it typically occurred 1–3 days prior to VPEG placement via large-volume paracentesis. It is standard practice for ultrasounds to be used in bedside and interventional radiology-performed paracentesis at our institution. Patients were maintained on parenteral nutrition and advised to limit oral intake. We reviewed the medical chart to identify re-hospitalizations for obstructive symptoms and VPEG malfunction. Immediate and delayed AEs were noted. The Social Security Death Index Interactive Search was used to determine the date of death.
6 (6.3)
3 (3.1) 13 (13.5)
(66.6 %) and gastric disease was noted in 49 patients (51.0 %). Ascites was also present in 46.9 % of patients, with drainage performed in only a minority of these patients (35 %, n = 16). In all patients, the etiology of ascites was presumed to be peritoneal metastatic disease. Drainage was performed by paracentesis (n = 11) or permanent drainage catheter (n = 5). Broad-spectrum gramnegative pre-procedural antibiotics were administered to all patients. Gastroenterology trainees assisted in 57.3 % (n = 55) of all cases. VPEG technical success
Statistics Patient demographic information and clinical characteristics are described by means and ranges for continuous variables, and by presence or absence for categorical variables. Fisher’s exact test was used to compare outcomes in the patients who suffered AEs versus those who did not, as well as determine the significance of post-procedural survival time. p-Values \0.05 were considered statistically significant and all tests were two-sided. Two separate reviewers completed the data analysis to ensure congruency in the final reported values.
Results
Venting percutaneous endoscopic gastrostomy was successfully placed by gastroenterologists in 89 patients (92.7 %), 88 on the first attempt. Of the remaining seven patients, VPEG could not be placed because of failure to transilluminate the anterior abdominal wall. In six patients with unsuccessful VPEG placement (85.7 %), failure was attributed to peritoneal metastases resulting in no appropriate window for placement. The etiology of unsuccessful placement was not detailed in the remaining patient. In those patients who failed endoscopic placement, venting gastrostomy was placed by interventional radiology in four patients, and surgically in two patients. Placement was unsuccessful in one patient despite attempts by endoscopy, interventional radiology, and surgery (Fig. 1).
Patient demographics VPEG clinical success We identified 96 patients in whom an endoscopic VPEG placement was attempted over the 13-year period. However, we were unable to determine the total volume of VPEGs placed during this time. Overall, 95 % (n = 91) of the procedures were performed during an inpatient hospitalization. The median age at the time of the procedure was 57 years (range 21–90 years) (Table 2). The most common malignancies were colorectal (27.1 %), pancreatic (18.8 %), and gynecologic (17.7 %) (Table 3). Abdominal carcinomatosis was present in the majority of patients
Complete relief of nausea and vomiting was observed in the majority of patients (n = 81, 91.0 %). Four of these patients were re-admitted for recurrent symptoms due to VPEG tube dysfunction caused by clogging, and all were successfully replaced. Seven patients (7.9 %) were readmitted for obstructive symptoms despite a functioning VPEG tube, and were considered treatment failures. A single patient expired before symptom relief could be assessed.
123
Surg Endosc Fig. 1 VPEG placement data. VPEG venting percutaneous endoscopic gastrostomy
Table 4 Procedure adverse events Patient no.
Gender
Ascites
Abdominal carcinomatosis
Gastric disease
Adverse events
1
M
No
Yes
No
PEG occlusion
2
F
Yes
Yes
No
PEG occlusion
3
F
Yes
Yes
No
PEG occlusion
4
F
Yes
Yes
No
Infection
5
M
Yes
Yes
No
PEG occlusion/infection
6
M
No
No
No
PEG migration
7
F
Yes
Yes
Yes
Infection
8
M
Yes
Yes
No
Infection
9
M
Yes
Yes
No
Infection/sepsis/death
M male, F female, PEG percutaneous endoscopic gastrostomy
An NGT was present prior to VPEG placement in 73 % of patients (n = 65). Documentation on clinical response to NGT was available in only 36 patients, with 32 (89 %) patients reporting improvement in symptoms with NGT. Of those patients reporting clinical improvement with NGT, 93 % (n = 30) had no re-hospitalizations for obstructive symptoms following VPEG tube placement. In the four patients without symptom relief from NGT, resolution of obstructive symptoms occurred in only a single patient. Procedural adverse events Ten post-procedural AEs were noted in nine patients. The most common observed AE was infection (n = 5), with one death due to suspected intra-abdominal sepsis (Table 4). Ascites was more likely to be present in patients who had post-procedural infectious AEs (12.2 vs. 0 %; p = 0.02) (Fig. 2). There was a trend towards decreased infectious AEs when ascites was drained prior to endoscopic VPEG placement (7.1 vs. 14.8 %; p = 0.64). When positive ascitic fluid cultures were obtained, organisms
123
were typical gastrointestinal bacteria, including Enterobacter cloacae and Klebsiella pneumonia. Gastric involvement of the tumor was associated with a decreased rate of AEs compared with patients without gastric disease (22.9 vs. 2.2 %; p = 0.013). In the patients in whom follow-up was documented, overall mean survival after VPEG placement was 135 ± 347.9 days (range 5–2,772 days). We observed a trend towards decreased survival when AEs occurred (73 ± 47.8 days) compared with when they did not (141 ± 367.8 days; p = 0.61). Thirty-day all-cause mortality rate within our cohort was 22.9 %.
Discussion Patients with inoperable obstructive gastrointestinal malignancies can be especially difficult to manage given recurrent nausea and vomiting that is not always alleviated by medications. There is emerging evidence suggesting that, in these patients, placement of VPEG tubes may improve symptoms and lead to more hospital-free days [7,
Surg Endosc
Fig. 2 VPEG adverse events. Infectious adverse events were increased in patients with ascites who underwent VPEG placement. Drainage of ascites prior to VPEG placement appeared to reduce the
risk of infectious adverse events. Overall adverse events were decreased in patients who had gastric involvement of tumor. VPEG venting percutaneous endoscopic gastrostomy
11–14, 22]. However, the majority of these studies are smaller case series and thus provide limited clinical guidance on how to approach more complex patients such as those with ascites or abdominal carcinomatosis who require VPEG placement. Based on the results of our single-center experience, the largest study to date, we conclude that VPEG tubes can be successfully placed endoscopically in nearly all patients. Furthermore, most patients with inoperable malignancy and obstructive symptoms had complete relief of symptoms and did not require re-hospitalization for these symptoms after VPEG placement. We also observed that a clinical response to NGT placement could serve as a predictor of those in whom endoscopic VPEG placement might be most successful. Thus, placement should be performed if the etiology of symptoms is unclear prior to placement of VPEG tubes. Based on our findings, the presence of ascites and abdominal carcinomatosis are not absolute contraindications to successful VPEG placement; however, extensive abdominal carcinomatosis was common in failed endoscopic VPEG placement. Additionally, while endoscopic VPEG placement is successful in patients with malignant ascites, we noted its presence as a risk factor for postprocedural infectious AEs, suggesting that this may be a useful predictor when attempting to risk-stratify patients. This risk is possibly decreased by drainage of fluid prior to VPEG tube placement, and we advocate that this should be performed whenever possible. However, given the
multifactorial issues that may be present in patients with ascites from underlying malignancies, this warrants further study with larger cohorts. Interestingly, we found a decreased rate of overall complications in patients with gastric involvement of the tumor. The reasons for this remain unclear but it may be that larger tumors adhere to the anterior abdominal wall and thus may facilitate VPEG placement. Additionally, patients with visible gastric disease could present with symptoms earlier and may be an overall healthier cohort of patients than those with abdominal carcinomatosis; however, this finding requires further validation. Reassuringly, our study also revealed that the majority of those who received VPEG tubes were rarely re-admitted for recurrent obstructive symptoms. The mean patient survival in this population (135 ± 347.9 days) was significantly greater than previously reported and suggests that VPEG placement is a durable palliative solution [11, 22]. As expected, the length of survival was increased when procedural AEs did not occur.
Conclusions In those patients with malignant obstruction not amenable to enteral stenting or surgical bypass, VPEG placement has a high technical and clinical success rate with an acceptable AEs rate. Although ascites and malignant abdominal carcinomatosis appeared to increase the risk of AEs, the
123
Surg Endosc
majority of patients still achieved clinical success. Thus, in patients whose therapeutic options are limited, endoscopic placement of venting gastrostomy tube should be considered for palliation.
11.
12. Disclosures Rachel Issaka, David Shapiro, Neehar Parikh, Mary Mulcahy, Srinadh Komanduri, John Martin, and Rajesh Keswani have no conflicts of interest to declare.
References 1. Bentley A, Boyd K (2001) Use of clinical pictures in the management of nausea and vomiting: a prospective audit. Palliat Med 15:247–253 2. Coates A, Abraham S, Kaye SB, Sowerbutts T, Frewin C, Fox RM, Tattersall MH (1983) On the receiving end: patient perception of the side-effects of cancer chemotherapy. Eur J Cancer Clin Oncol 19:203–208 3. Griffin AM, Butow PN, Coates AS, Childs AM, Ellis PM, Dunn SM, Tattersall MH (1996) On the receiving end. V: patient perceptions of the side effects of cancer chemotherapy in 1993. Ann Oncol 7:189–195 4. Osoba D, Zee B, Warr D, Latreille J, Kaizer L, Pater J (1997) Effect of postchemotherapy nausea and vomiting on healthrelated quality of life. The Quality of Life and Symptom Control Committees of the National Cancer Institute of Canada Clinical Trials Group. Support Care Cancer 5:307–313 5. Stephenson J, Davies A (2006) An assessment of aetiology-based guidelines for the management of nausea and vomiting in patients with advanced cancer. Support Care Cancer 14:348–353 6. Ang SK, Shoemaker LK, Davis MP (2010) Nausea and vomiting in advanced cancer. Am J Hosp Palliat Care 27:219–225 7. Dolan EA (2011) Malignant bowel obstruction: a review of current treatment strategies. Am J Hosp Palliat Care 28:576–582 8. Keswani RN, Azar RR, Edmundowicz SA, Zhang Q, Ammar T, Banerjee B, Early DS, Jonnalagadda SS (2009) Stenting for malignant colonic obstruction: a comparison of efficacy and complications in colonic versus extracolonic malignancy. Gastrointest Endosc 69:675–680 9. Jung MK, Park SY, Jeon SW, Cho CM, Tak WY, Kweon YO, Kim SK, Choi YH, Kim GC, Ryeom HK (2010) Factors associated with the long-term outcome of a self-expandable colon stent used for palliation of malignant colorectal obstruction. Surg Endosc 24:525–530 10. Ripamonti C, Mercadante S, Groff L, Zecca E, De Conno F, Casuccio A (2000) Role of octreotide, scopolamine butylbromide,
123
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
and hydration in symptom control of patients with inoperable bowel obstruction and nasogastric tubes: a prospective randomized trial. J Pain Symptom Manage 19:23–34 Brooksbank MA, Game PA, Ashby MA (2002) Palliative venting gastrostomy in malignant intestinal obstruction. Palliat Med 16:520–526 Campagnutta E, Cannizzaro R (2000) Percutaneous endoscopic gastrostomy (PEG) in palliative treatment of non-operable intestinal obstruction due to gynecologic cancer: a review. Eur J Gynaecol Oncol 21:397–402 Meyer L, Pothuri B (2006) Decompressive percutaneous gastrostomy tube use in gynecologic malignancies. Curr Treat Options Oncol 7:111–120 Pothuri B, Montemarano M, Gerardi M, Shike M, Ben-Porat L, Sabbatini P, Barakat RR (2005) Percutaneous endoscopic gastrostomy tube placement in patients with malignant bowel obstruction due to ovarian carcinoma. Gynecol Oncol 96:330–334 Gauderer MW, Ponsky JL, Izant RJ Jr (1980) Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 15:872–875 Eisen GM, Baron TH, Dominitz JA, Faigel DO, Goldstein JL, Johanson JF, Mallery JS, Raddawi HM, Vargo JJ 2nd, Waring JP, Fanelli RD, Wheeler-Harbough J (2002) Role of endoscopy in enteral feeding. Gastrointest Endosc 55:794–797 Vashi PG, Dahlk S, Vashi RP, Gupta D (2011) Percutaneous endoscopic gastrostomy tube occlusion in malignant peritoneal carcinomatosis-induced bowel obstruction. Eur J Gastroenterol Hepatol 23:1069–1073 Campagnutta E, Cannizzaro R, De Cicco M, De Piero G, Giorda G, Sopracordevole F, Parin A, Scarabelli C (1998) Percutaneous endoscopic gastrostomy (PEG) in upper gastrointestinal tract occlusion in gynecologic oncology. Minerva Ginecol 50:305–311 in Italian Herman LL, Hoskins WJ, Shike M (1992) Percutaneous endoscopic gastrostomy for decompression of the stomach and small bowel. Gastrointest Endosc 38:314–318 Campagnutta E, Cannizzaro R, Gallo A, Zarrelli A, Valentini M, De Cicco M, Scarabelli C (1996) Palliative treatment of upper intestinal obstruction by gynecological malignancy: the usefulness of percutaneous endoscopic gastrostomy. Gynecol Oncol 62:103–105 Cotton PB, Eisen GM, Aabakken L, Baron TH, Hutter MM, Jacobson BC, Mergener K, Nemcek A Jr, Petersen BT, Petrini JL, Pike IM, Rabeneck L, Romagnuolo J, Vargo JJ (2010) A lexicon for endoscopic adverse events: report of an ASGE workshop. Gastrointest Endosc 71:446–454 Teriaky A, Gregor J, Chande N (2012) Percutaneous endoscopic gastrostomy tube placement for end-stage palliation of malignant gastrointestinal obstructions. Saudi J Gastroenterol 18:95–98
