Comparison of Complications of Percutaneous Endoscopic Versus Surgically Placed Gastrostomy Tubes in 42 Dogs and 52 Cats

Introduction

A variety of methods exist for maintaining enteral nutritional support in animals, including nasoesophageal, pharyngostomy, esophagostomy, gastrostomy, and jejunal feeding tubes.^1–5 Gastrostomy feeding tubes are preferred for long-term enteral nutritional support in dogs and cats, because they are well tolerated, provide better patient comfort, allow feeding of a blended commercial diet, and are relatively easy to place.²

Gastrostomy tubes may be inserted by several different surgical methods.^4,6 Advantages of surgically placed gastrostomy (SPG) tubes include direct visualization of the site of placement and the ability to ensure a good adhesion between the stomach and body wall (by suturing them together).⁷ Disadvantages of these tubes include greater insertion time, need for a deeper plane of anesthesia, higher cost, and more tissue trauma caused by laparotomy.^8–10

The introduction of the percutaneous endoscopic gastrostomy (PEG) tube in 1986 offered a less invasive alternative to the SPG tube. Since their introduction, PEG tubes have become the most common method of assisted feeding in dogs and cats.^3,7,11 Percutaneous endoscopic gastrostomy tubes can be placed more rapidly and under a lighter plane of anesthesia than the SPG tube, and placement of the PEG tube does not require a laparotomy.^4,6,9–12 Despite these advantages, the method used to place PEG tubes does not allow a gastropexy that would ensure an early and permanent adhesion between the stomach and body wall.⁵ Inadvertent removal of a PEG tube prior to adhesion formation may expose the animal to the risk of peritonitis and sepsis.^13–15 Percutaneous endoscopic gastrostomy tubes should not be removed until a gastric adhesion to the body wall has formed, usually at least 8 to 10 days after placement.¹³ Although adhesions have been reported to form in 3 days in healthy dogs, it may take longer in immunosuppressed or debilitated animals and in those that did not have a gastropexy.^5,9

The complications associated with PEG tubes and SPG tubes have been described, but a direct comparison between the two has not been performed.^8,16 The purposes of this retrospective study were to document the complications associated with PEG and SPG tubes and to directly compare the rates and severity of complications from both methods.

Materials and Methods

The medical records of 122 animals having PEG tubes or SPG tubes inserted at the University of Illinois College of Veterinary Medicine and the Kansas State University Veterinary Medical Teaching Hospital between June 1993 and June 2003 were evaluated. The criteria required for inclusion in the study were use of a Pezzer mushroom-tip catheter,^a discharge of the animal from the hospital with the feeding tube in place, and return of the animal to the hospital for at least one recheck appointment in the first 2 weeks following tube placement. Data retrieved from the medical records included age, sex, breed, primary diagnosis (i.e., reason for tube), method of tube placement (i.e., surgical or endoscopic), and any complications that occurred.

Percutaneous endoscopic gastrostomy tubes were placed as described by Armstrong and Hardie.¹² Briefly, PEG tubes were placed by using a needle to advance suture through the body wall into the stomach, retrieving the suture endoscopically, pulling the suture out the mouth, attaching the suture to the feeding tube, and pulling on the suture to draw the feeding tube out through the body wall. Rubber flanges were used internally to secure the stomach to the body wall, and externally to secure the tube to the body wall. Surgically placed gastrostomy tubes were placed as described by Crowe.¹³ Briefly, pursestring sutures were placed around the tube where it entered the stomach, the stomach was sutured to the abdominal fascia using preplaced mattress sutures, and a Chinese finger trap suture was used to secure the tube to the skin. All animals had Pezzer mushroom-tip catheters^a inserted by a faculty member or a resident supervised by a faculty member.

Minor complications were defined as those that did not require medical attention for correction and did not pose a threat to the health of the animal (e.g., peristomal inflammation). Moderate complications were defined as those that posed a non-life-threatening health risk to the animal, such as peristomal infection, fistula formation after tube removal, and tube removal by the animal that did not require intervention. Severe complications were defined as those that posed a life-threatening health risk to the animal, such as tube removal requiring intervention and septic peritonitis. Animals that did not develop complications were assigned a severity score of one for statistical analysis, while those that developed mild, moderate, or severe complications were assigned severity scores of two, three, or four, respectively.

The Fisher’s exact test^b was used to compare the complication rates between PEG tubes and SPG tubes in cats, PEG tubes and SPG tubes in dogs, and PEG tubes and SPG tubes in dogs and cats combined. The Wilcoxon’s rank sum test^c was utilized to compare the complication severity scores between PEG tubes and SPG tubes in cats, PEG tubes and SPG tubes in dogs, and PEG tubes and SPG tubes in dogs and cats combined. Statistical significance was set at P ≤ 0.05.

Results

Ninety-four animals met the criteria for inclusion in the study. Forty-three cats had PEG tubes inserted; 36 were domestic shorthair cats, one was a domestic longhair, and six were purebred cats. The age range of these cats was 1 to 16 years (mean age 8.9 years). There were 21 spayed females, two intact females, 19 castrated males, and one intact male. Nine cats had SPG tubes inserted; three of these were domestic shorthair cats, two were domestic longhair cats, and four were purebred cats. Ages of cats in the SPG tube group ranged from 1 to 16 years (mean age 9.8 years). In the SPG tube group, there were five spayed females, one intact female, and three castrated males.

Thirty dogs of various breeds had PEG tubes placed. There were 11 spayed females, four intact females, six castrated males, and nine intact males. Ages ranged from 2 months to 11 years (mean age 5.4 years). Twelve dogs of various breeds had SPG tubes inserted. Their ages ranged from 2 months to 11 years (mean age 6.8 years). In the SPG tube group, there were three spayed females, two intact females, five castrated males, and two intact males.

The most common indications for PEG tube placement in cats were hepatic lipidosis (30.2%) and oral neoplasia (27.9%) [Table 1]. The most common indications for SPG tube placement in cats were hepatic lipidosis (22.2%), intestinal neoplasia (22.2%), and anorexia of unknown origin (22.2%) [Table 1]. The most common indications for PEG tube placement in dogs were esophageal stricture (20%) and oral neoplasia (20%) [Table 2]. The most common indication for SPG tube placement in dogs was esophageal disease (33.3%) [Table 2].

Of the 43 cats that had PEG tubes placed, 25 (58.1%) had no complications, two (4.6%) had mild complications, 14 (32.6%) had moderate complications, and two (4.6%) had severe complications [Table 3]. Of the nine cats that had gastrostomy tubes surgically inserted, seven (77.8%) had no complications, one (11.1%) had a minor complication, one (11.1%) had a moderate complication, and none had severe complications [Table 3]. Of the 30 dogs that had PEG tubes placed, 17 (56.7%) had no complications, three (10%) had minor complications, five (16.6%) had moderate complications, and five (16.6%) had severe complications. One dog in the PEG tube group that developed septic peritonitis after inadvertent tube removal was euthanized [Table 3]. Of the 12 dogs that had gastrostomy tubes surgically placed, 10 (83.3%) had no complications, one (8.3%) had a minor complication, one (8.3%) had a moderate complication, and none had severe complications [Table 3].

Of the 73 animals that had PEG tubes placed, 42 (57.5%) had no complications, five (6.8%) had minor complications, 19 (26.0%) had moderate complications, and seven (9.6%) had severe complications. Of the 21 animals that had gastrostomy tubes placed surgically, 17 (81.0%) had no complications, two (9.5%) had minor complications, two (9.5%) had moderate complications, and none had severe complications. Of the 11 animals (six dogs, five cats) that inadvertently removed their PEG tubes, two of the six dogs developed septic peritonitis, but none of the cats did.

No significant differences were noted when complication rates between PEG tubes and SPG tubes in cats (P=0.24) or in dogs (P=0.10) were compared. However, when the dogs and cats were combined, PEG tubes had a significantly higher complication rate than SPG tubes (P=0.04). No significant differences were noted when complication severity scores between PEG tubes and SPG tubes in cats (P=0.20) or in dogs (P=0.08) were evaluated. However, when the dogs and cats were combined, PEG tubes had significantly higher complication severity scores than SPG tubes (P=0.03).

Discussion

The rate of PEG tube complications in this study (42.5%) was slightly lower than rates documented by Armstrong et al. (56%) and by Ireland et al. (57%).^8,17 These complication rates were difficult to compare because of the varying criteria used to classify the complications. The complication rate in the study reported here was lower than the rates reported in other studies, because gulping and scratching at the tube site were not included as complications.^8,17 Prior reports classified inadvertent tube removal by the animal <14 days postplacement as a severe or major complication (whether or not an adverse consequence resulted). In the study reported here, tube removal without an adverse effect was classified as a moderate complication, and tube removal requiring tube replacement or resulting in peritonitis was classified as a severe complication.^8,17

The rate of tube removal by the animals in this study (13.7%) was higher than the rate in the study by Armstrong et al. (dogs and cats, 4%) and was comparable to rates reported by Ireland et al. (cats 9.5%) and Elliot (dogs 19.6%).^8,16,17 Adhesions between the stomach and peritoneum are necessary to prevent leakage of stomach contents when the tube is withdrawn; therefore, the tube is not usually removed until these adhesions have had time to develop. Recommendations have been made that the tube remain in place for at least 8 to 10 days for an adhesion to form.^8,13 A study of healthy dogs reported that adhesions formed 3 days after PEG tube placement, but studies examining adhesion formation in debilitated animals are lacking.¹¹ In one report, a debilitated cat developed peritonitis when the tube was removed 23 days after placement.⁸ In the current study, five cats removed their tubes prematurely. Two of these cats had their tubes replaced surgically, and none suffered septic peritonitis as a result of inadvertent PEG tube removal. Six dogs removed their PEG tubes prematurely. Of these dogs, one had the tube surgically replaced, and two suffered septic peritonitis as a consequence. Tube removal requiring intervention and resulting in septic peritonitis were classified as severe complications in this study. Cats and dogs in the PEG tube group had severe complication rates of 4.7% and 17%, respectively, while cats and dogs in the SPG tube group had severe complication rates of 0% and 0%, respectively. The timing of tube removal by the animal did not seem to affect the severity of complications. Three dogs removed their PEG tubes and did not require intervention at 6, 14, and 38 days after placement; one dog removed the PEG tube 6 days after placement, and the tube was replaced surgically without complication; and two dogs developed septic peritonitis after PEG tube removal 4 and 28 days after placement.

The higher incidence of peristomal infection in animals of the PEG tube group (19.2%) versus animals of the SPG tube group (4.8%) may have been affected by the method of placement and the use of an internal flange of tubing between the mushroom-tipped catheter and the gastric mucosa, as well as the use of a rubber flange on the skin to secure the tube externally. In this study, these flanges were used in animals of the PEG tube group and not in animals of the SPG tube group. Some authors advocate the use of the inner and outer flange, while others believe they are unnecessary.^10,11 If the outer flange is not used or is too loose, the animal may be predisposed to leakage of gastric contents into the abdomen, and peritonitis may result. If the outer flange is too tight, it may lead to necrosis.¹⁸ The presence of the flange itself, even if placed properly, can cause the animal discomfort by producing a moist dermatitis at the rubber-to-skin contact site.¹⁰

The use of prophylactic antibiotics has been advocated with the use of PEG tubes and SPG tubes in both animals and humans.^9,10,19,20 In one study, the use of prophylactic antibiotics reduced the wound infection rate associated with PEG tube placement from 30% to 7%.²⁰ The present study did not assess the effect of antibiotics on the rate of wound infection; however, antibiotics may have had an effect on peristomal infection. It is possible that more animals with SPG tubes would have received perioperative antibiotics since they may have had other procedures performed at the same time, such as intestinal biopsies.²¹

One of the difficulties in evaluating complication rates in a retrospective study involving two different institutions is the inability to standardize the technique for tube placement. However, at both institutions, identical procedures previously described in the Materials and Methods section were used. DeBowes et al. demonstrated that Pezzer mushroom-tip catheters had a lower complication rate than Malecot catheters in cats.²² At the two institutions involved in this study, it was standard practice to use Pezzer mushroom-tip catheters. Complication rates may differ depending on the level of expertise of the operator. However, in this study, PEG tubes and SPG tubes were placed by faculty members or residents under the supervision of a faculty member, which should have minimized complications associated with inexperience. In this study, no trends were apparent between operator experience and the rate or severity of complications.

The most common indications for placement of PEG or SPG tubes in the cats of this study were hepatic lipidosis and oral neoplasia. The most common indication for placement of PEG or SPG tubes in dogs was esophageal disease. The indications for PEG and SPG tubes in this study were similar to other studies. In a study by Armstrong et al., gastrostomy tubes were placed most commonly in cats with hepatic disease and in dogs with esophageal disease.⁸ It is possible that animals with more serious diseases were prone to more complications. In a study by Elliot et al., 11% of dogs with renal failure died or were euthanized after development of complications associated with the gastrostomy tube, suggesting that the type of disease may affect the severity of complications.¹⁶ A correlation between disease severity and complication rate or complication severity was not apparent in this study. The disease processes of the seven animals that suffered severe PEG tube complications included esophageal disease (n=3), anorexia of unknown origin (n=1), hepatic disease (n=1), intestinal neoplasia (n=1), and oral neoplasia (n=1). These disease processes were also present in animals that did not have complications associated with PEG tube placement. It is also possible that age influenced the complication severity, although a correlation between age and complication severity was not apparent in this study. The ages of the seven animals that suffered severe PEG tube complications were 2, 4, 5 (n=2), 8 (n=2), and 11 years. The mean age of the animals with severe PEG tube complications was 6.1 years. This is similar to the mean age (7.4 years) of all the animals in this study that had PEG tubes inserted.

In this study, a comparison of dogs with PEG tubes versus SPG tubes failed to demonstrate a significant difference in complication rates and severity scores. Likewise, a comparison of cats with PEG tubes versus SPG tubes failed to demonstrate a significant difference in complication rates and severity scores. Only when the dogs and cats were combined were significantly higher complication rates and severity scores detected in the animals with SPG tubes. Because of the small population of animals with SPG tubes, the importance of this statistical result is uncertain. Based on the numbers, PEG tubes were preferred over SPG tubes at both institutions. This choice may be influenced by cost, procedure time, and client perception—all of which favor PEG tube placement.

Conclusion

The complication rates and complication severity scores were compared in dogs and cats with either PEG tubes or SPG tubes inserted. No statistical differences were found between the two tube groups until the data from cats and dogs were combined. The combined data revealed a significantly higher complication rate with PEG tubes. The absence of severe complications in the SPG group may indicate that SPG tubes should be placed preferentially over PEG tubes when there are no contraindications to laparotomy. However, based on the findings of this study, both PEG tubes and SPG tubes are viable options for enteric support in dogs and cats.