- Email: [email protected]
Timing of percutaneous endoscopic gastrostomy tube placement in head and neck cancer patients
Timing of percutaneous endoscopic gastrostomy tube placement in head and neck cancer patients EILEEN M. RAYNOR, MD, MARK F. WILLIAMS, MD, ROBERT G. MARTINDALE, MD, PhD, and EDWARD S. PORUBSKY, MD,
Augusta Georgia
Percutaneous endoscopic gastrostomy (PEG) is an effective method for providing alimentation in patients with upper aerodigestive tract carcinoma. Multiple complications of this procedure have been reported, ranging from leakage around the tube to tumor seeding of the abdominal cavity. This study was undertaken to determine whether the timing of PEG tube placement with respect to primary tumor extirpation led to a difference in the number and severity of observed complications. The medical records of 43 patients with head and neck carcinoma who had PEG tubes placed from 1995 to 1996 were retrospectively reviewed. Comparisons of timing of PEG tube placement, complication, location, and stage of the primary tumor were performed. In addition, the use of adjuvant therapy with respect to the time of PEG tube placement and complications was evaluated. Of these, 23% were done before and 30% during surgery at the time of primary tumor resection (9 of 13 were after primary removal). One patient had an intraabdominal abscess. Minor complications occurred in 15 of 43 patients (35%) and included granulation tissue at the PEG site, leakage, and tube displacement. Eight of the 9 patients who underwent intraoperative PEG after tumor resection had no complications. Patients who underwent PEG during or after surgery had significantly fewer complications than those who underwent preoperative PEG or had unresectable tumors (P = 0.038). The largest number of complications occurred in patients who underwent preoperative PEG (57%) followed by patients whose tumors were unresectable (31%). There was no sta-
From the Department of Surgery, Division of Otolaryngology (Drs Raynor, Williams, and Porubsky) and the Division of Gastrointestinal Surgery (Dr Martindale), Medical College of Georgia. Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, San Francisco, CA, September 7-10, 1997. Reprint requests: Edward S. Porubsky, MD, Department of Surgery, Division of Otolaryngology, Medical College of Georgia, 1120 15th St, Augusta, GA 30912. Copyright © 1999 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/90/$8.00 + 0 23/1/91408
tistical difference with regard to tumor location or postoperative x-ray therapy in PEG complications. This study demonstrates that PEG tube placement after tumor resection has the lowest incidence of postoperative complications. Performing PEGs intraoperatively after tumor resection can prevent the need for additional anesthesia to provide alimentation in patients with upper aerodigestive tract carcinoma. (Otolaryngol Head Neck Surg 1999;120: 479-82.)
Patients with head and neck cancer often require an extraoral route for nutritional support after tumor resection and while undergoing adjuvant therapy because of altered pharyngeal anatomy and impaired deglutition. These patients are commonly nutritionally depleted and cannot meet their caloric needs without nutritional support. Traditionally nasogastric tube (NGT) placement has been the preferred approach by which these patients receive their nutrients.1 The use of NGTs can be associated with multiple problems, including laryngeal irritation, persistent gastroesophageal reflux, nasal alar necrosis, sinusitis, and inadvertent removal, as well as patient dissatisfaction. Recently, percutaneous endoscopic gastrostomy (PEG) tube placement has been used extensively for providing nutritional support during recovery from head and neck cancer extirpation and has proved to be safe with relatively few complications.2,3 Stomal seeding at the PEG site from passage through a cancerous oral cavity has recently been described in the literature.4,5 These reports have led us to investigate whether timing of PEG tube placement with respect to primary tumor extirpation led to a difference in the number and severity of complications. METHODS AND MATERIAL Of 50 patients with primary head and neck carcinomas who had PEG tubes placed at the Medical College of Georgia from 1995 to 1996, 43 had records that were available for review. The remaining 7 patients had incomplete data recorded in their medical records. PEG tube placement was separated into 4 categories. Preoperative PEG tubes were placed before tumor removal; intraoperative PEG tubes were placed 479
480
Otolaryngology– Head and Neck Surgery April 1999
RAYNOR et al
Table 1. List of all complications and frequency of occurrence Complication
None External leak Granulation at site Pain Extrusion Abscess Abdominal metastases Other*
No. of patients
27 7 5 2 2 1 0 5
*Other
complications included bleeding (2), cracked tube (1), and clogged tube requiring changing (2).
Table 2. List of all tumor locations and frequency of occurrence Tumor site
Larynx Hypopharynx Tonsil Floor of mouth Base of tongue Oral tongue, buccal mucosa Supraglottic Larynx Neck Multiple* Other (head and neck)†
No. of patients (n = 43)
7 7 6 5 5 2 2 1 4 4
*Multiple
sites: Tonsil and larynx (1), base of tongue and pyriform (1), floor of mouth and lung primary (1), and base of tongue and parotid gland (1). †Other head and neck: Parotid gland (1), nasopharynx (1), and temporal bone (2).
during primary tumor resection, either before or immediately after tumor removal; and postoperative PEG tubes were placed from 7 days to several months after tumor resection. The final category included patients who had unresectable tumors and underwent palliative therapy. In patients who received PEG tubes after surgery, initial attempts at oral feeding generally failed, or the patient had significant difficulty with deglutition during radiation therapy, thereby requiring an alternative route for alimentation. PEGs were further classified as pull, the standard Ponsky technique,6 in which the catheter is pulled through the oral cavity and esophagus and out the abdominal wall; or, the push method7 whereby a guide wire is placed into the stomach under direct visualization and serially dilated with the Seldinger technique until the catheter is inserted through the abdomen directly into the stomach. Intraoperative PEGs were performed by introducing the flexible endoscope directly through the opened pharynx. A guide wire was placed percutaneously into the stomach and snared; the PEG catheter was
then fed transesophageally over the guide wire through the pharynx into the stomach. Insufflation of the stomach in these intraoperative PEGs was facilitated by gentle approximation of the mucosa around the endoscope with Babcock forceps. Patient age, sex, tumor site and stage, surgical procedure, use of additional tumor-directed therapy, date and type of PEG, and type of complications were recorded. Comparisons of time of PEG tube placement; the number, type, and severity of complications; and the location and stage of the tumor were performed with χ2 analysis. The use of adjuvant radiation therapy with respect to the time of PEG and complications was also evaluated. Two-tailed Fisher’s exact tests were used when 2 groups were compared with each other. Statistical significance was determined at the P ≤ 0.05 level. RESULTS
Thirty-six of the patients were male (84%), and the average age was 58 years. The majority of patients had stage IV tumor at the time of their PEG (79%). Time of PEG tube placement was initially divided into 5 categories: preoperative (10 patients), intraoperative before tumor extraction (4 patients), intraoperative after tumor removal (9 patients), postoperative (12 patients), and unresectable (8 patients). For purposes of statistical analysis, the preoperative and intraoperative groups before tumor removal were added together (n = 14). Of the 7 women, 1 had an intraoperative PEG, 4 received their PEG tube before surgery, and 2 had PEG tubes placed after surgery. Their tumor stages were as follows: preoperative—stage II (1), stage IV (3); intraoperative—stage IV (1); and postoperative—stage II (1), stage III (1). Nine patients had tumors from stages I through III (T1N0-T3N1). One patient had a T1N0 base of tongue tumor and received a preoperative PEG tube. Stage II cancer patients had only preoperative (2) and postoperative (2) PEG tubes placed. Patients with stage III tumors got PEG tubes at all 3 times: preoperative (1), intraoperative (1), and postoperative (2). Comparisons were made between timing of placement and the total number of complications. Minor and major complications are listed in Table 1. The most common reported complications were external leakage from the base of the PEG site (7) followed by granulation tissue (5). One major complication occurred in this series. This patient received his PEG tube before tumor resection, and subsequently an intraabdominal abscess developed requiring a laparotomy. Only 3 patients had the push PEG by the Russell method; therefore no statistical comparisons could be made with regard to method of PEG tube placement and complications. At the time of manuscript preparation, no patients in this series had evidence of tumor at the PEG site. PEG tubes were removed in 10 patients because of sole reliance on oral intake for nutri-
Otolaryngology– Head and Neck Surgery Volume 120 Number 4
RAYNOR et al
481
Fig 1. Graph of time of PEG tube placement versus number of patients for no complication and any complication.
tion. Four patients died before PEG tube removal of causes unrelated to their PEGs. None of them had any PEG complication before death. Patients who underwent PEG after tumor removal, either during or after surgery, had a statistically significantly lower rate of complications than the preoperative groups (27% vs 57%; P = 0.027). Those patients who underwent intraoperative PEG after tumor removal had significantly fewer complications than those who underwent preoperative PEGs (11% vs 57%; P = 0.04. Patients who received preoperative versus postoperative PEGs had significantly more complications (57% vs 17%; P = 0.05). Patients who had unresectable tumor also had a higher rate of complications than patients with postoperative PEGs (31% vs 17%). These comparisons are demonstrated in Fig 1. Tumor sites are listed in Table 2. There was no statistically significant difference between tumor location and PEG complications. Twelve patients had recurrent squamous cell carcinoma (28%), 29 of the 35 surgical patients received postoperative radiation therapy (83%), and 6 of 35 underwent preoperative x-ray therapy (17%). Neither of these groups had a statistically significant increase in complications. DISCUSSION
Patients with head and neck malignancies are often malnourished because of impaired deglutition from either the tumor itself or decreased oral competence,
cranial nerve deficiencies, adynamic pharyngeal segments from reconstructive flaps, or radiation-induced xerostomia. These patients often have anorexia, general debilitation, and nausea in addition to any swallowing difficulties. Many of these problems can be overcome by providing an alternative route for enteral nutrition through either NGTs or gastrostomy tubes. Open gastrostomies require a laparotomy and have a significant complication rate ranging from 3% to 50%.8 In addition, open gastrostomies may prolong reaching optimal feeding because of significant postoperative pain and ileus.3 Nasoenteral tubes, while being easy to insert and use, also have marked disadvantages including the potential for sinusitis, gastroesophageal reflux, possible suture line irritation, and social unacceptability.9,10 Use of PEG tubes may decrease hospital stay, as demonstrated by Gibson and Wenig.11 In this study patients with pharyngeal and laryngeal cancers had 61% shorter hospital stays after PEG tube placement than similar patients who had NGTs. PEG tubes are technically simple to place and carry a low rate of complications (9% to 15%) with serious complications less than 1%. Most PEG complications are minor, consisting of superficial wound infections, peristomal leakage, and chronic granulation.3,6 A serious but rare complication that has been reported in head and neck cancers is seeding of the tumor at the PEG site. This is probably the result of implantation of cancer cells from the primary tumor into the abdominal
482
Otolaryngology– Head and Neck Surgery April 1999
RAYNOR et al
wall by pulling the tumor through with the catheter.4,5 This is a rare complication, with only a few cases reported in the literature and an unknown prevalence.11 It seems likely that this complication can be minimized either by use of the Russell technique or placement of the PEG tube after the primary has been resected. Because of our desire to limit complications, especially tumor seeding, we designed this retrospective study in an attempt to determine whether timing of PEG tube placement with respect to tumor removal correlated to incidence of complications. We had no abdominal metastatic lesions in this series. We did find that there was a significant decrease in overall complications when the PEG was done after tumor extirpation as opposed to preoperative placement. In general, we found that the patients who underwent preoperative PEG tube placement tended to have a greater weight loss and a lower preoperative weight than the other groups. All 4 groups of patients had significant alcohol and tobacco use. The groups were equivalent in relation to comorbidities (hypertension, chronic obstructive pulmonary disease). Because patients whose tumors were unresectable had a higher rate of complications than the 2 postoperative groups, it would be beneficial to consider further work in this area to determine whether the method of PEG tube placement would lower the complication rate in this subset of patients. We also found that placing the PEG tube during surgery after tumor removal had the lowest complication rate and allowed for excellent visualization of the upper aerodigestive tract during the procedure, thereby making the PEG even easier and more efficiently performed. Intraoperative PEG tube placement is a straightforward procedure and can prevent the patient from requiring further procedures to achieve enteral feeding access. There is no significant delay by placing the PEG tube during surgery; this can often be done while awaiting the results of frozen sections or acquiring a dermal graft. Further studies comparing the push technique ver-
sus the pull method in preoperative head and neck carcinoma are warranted. CONCLUSIONS
Intraoperative PEG tube placement is associated with statistically fewer complications than PEG tube placement before tumor removal. Intraoperative PEGs can be readily performed without the need for an additional procedure to attain a route for enteral nutrition. We recommend intraoperative PEG tube placement, after tumor removal, in those patients with head and neck carcinoma who will require an alternative to oral feeding during their recovery period. REFERENCES 1. Selz PA, Santos PM. Percutaneous endoscopic gastrostomy. Arch Otolaryngol Head Neck Surg 1995;121:1249-52. 2. Wilson WR, Hariri SM. Experience with percutaneous endoscopic gastrostomy on an otolaryngology service. Ear Nose Throat J 1995;74:760-2. 3. Urban KG, Terris DJ. Percutaneous endoscopic gastrostomy by head and neck surgeons. Otolaryngol Head Neck Surg 1997;116: 489-92. 4. Sharma P, Berry SM, Wilson K, et al. Metastatic implantation of an oral squamous-cell carcinoma at a percutaneous endoscopic gastrostomy site. Surg Endosc 1994;8:1232-5. 5. van Erpecum KJ, Akkerdijk WL. Metastasis of hypopharyngeal carcinoma into the gastrostomy tract after placement of a percutaneous endoscopic gastrostomy catheter. Endoscopy 1995;27:124-7. 6. Gauderer MWL, Ponsky JL, Izant RJ. Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 1980;15:872-5. 7. Russell TR, Brotman M, Norris F. Percutaneous gastrostomy: a new simplified and cost effective technique. Am J Surg 1984; 142:132-7. 8. Connar R, Sealy W. Gastrostomy and its complications. Ann Surg 1956;143:245-50. 9. Park RH, Allison MC, Lang J, et al. Randomized comparison of percutaneous endoscopic gastrostomy and nasogastric tube feeding in patients with persisting neurological dysphagia. BMJ 1992;304:1406-9. 10. Hunter JG, Laurentano L, Shelito PC. Percutaneous endoscopic gastrostomy in head and neck cancer patients. Ann Surg 1989; 210:42-6. 11. Gibson S, Wenig BL. Percutaneous endoscopic gastrostomy in the management of head and neck carcinoma. Laryngoscope 1992:102:977-80.
Clinical MRI
This conference, sponsored by the University of Pennsylvania Medical Center, will be held June 7-11, 1999, in Hamilton Princess, Bermuda. Course directors are J. Bruce Kneeland, MD, and Robert Lenkinski, PhD. Credits: 24 hours in category 1. For further information, contact Janice Ford Benner, Hospital of the University of Pennsylvania, MRI Center, 3400 Spruce St, 1 Founders Building, Philadelphia, PA 19104; phone, 215-662-6904; fax, 215-349-5925.
Augusta Georgia
Percutaneous endoscopic gastrostomy (PEG) is an effective method for providing alimentation in patients with upper aerodigestive tract carcinoma. Multiple complications of this procedure have been reported, ranging from leakage around the tube to tumor seeding of the abdominal cavity. This study was undertaken to determine whether the timing of PEG tube placement with respect to primary tumor extirpation led to a difference in the number and severity of observed complications. The medical records of 43 patients with head and neck carcinoma who had PEG tubes placed from 1995 to 1996 were retrospectively reviewed. Comparisons of timing of PEG tube placement, complication, location, and stage of the primary tumor were performed. In addition, the use of adjuvant therapy with respect to the time of PEG tube placement and complications was evaluated. Of these, 23% were done before and 30% during surgery at the time of primary tumor resection (9 of 13 were after primary removal). One patient had an intraabdominal abscess. Minor complications occurred in 15 of 43 patients (35%) and included granulation tissue at the PEG site, leakage, and tube displacement. Eight of the 9 patients who underwent intraoperative PEG after tumor resection had no complications. Patients who underwent PEG during or after surgery had significantly fewer complications than those who underwent preoperative PEG or had unresectable tumors (P = 0.038). The largest number of complications occurred in patients who underwent preoperative PEG (57%) followed by patients whose tumors were unresectable (31%). There was no sta-
From the Department of Surgery, Division of Otolaryngology (Drs Raynor, Williams, and Porubsky) and the Division of Gastrointestinal Surgery (Dr Martindale), Medical College of Georgia. Presented at the Annual Meeting of the American Academy of Otolaryngology–Head and Neck Surgery, San Francisco, CA, September 7-10, 1997. Reprint requests: Edward S. Porubsky, MD, Department of Surgery, Division of Otolaryngology, Medical College of Georgia, 1120 15th St, Augusta, GA 30912. Copyright © 1999 by the American Academy of Otolaryngology– Head and Neck Surgery Foundation, Inc. 0194-5998/90/$8.00 + 0 23/1/91408
tistical difference with regard to tumor location or postoperative x-ray therapy in PEG complications. This study demonstrates that PEG tube placement after tumor resection has the lowest incidence of postoperative complications. Performing PEGs intraoperatively after tumor resection can prevent the need for additional anesthesia to provide alimentation in patients with upper aerodigestive tract carcinoma. (Otolaryngol Head Neck Surg 1999;120: 479-82.)
Patients with head and neck cancer often require an extraoral route for nutritional support after tumor resection and while undergoing adjuvant therapy because of altered pharyngeal anatomy and impaired deglutition. These patients are commonly nutritionally depleted and cannot meet their caloric needs without nutritional support. Traditionally nasogastric tube (NGT) placement has been the preferred approach by which these patients receive their nutrients.1 The use of NGTs can be associated with multiple problems, including laryngeal irritation, persistent gastroesophageal reflux, nasal alar necrosis, sinusitis, and inadvertent removal, as well as patient dissatisfaction. Recently, percutaneous endoscopic gastrostomy (PEG) tube placement has been used extensively for providing nutritional support during recovery from head and neck cancer extirpation and has proved to be safe with relatively few complications.2,3 Stomal seeding at the PEG site from passage through a cancerous oral cavity has recently been described in the literature.4,5 These reports have led us to investigate whether timing of PEG tube placement with respect to primary tumor extirpation led to a difference in the number and severity of complications. METHODS AND MATERIAL Of 50 patients with primary head and neck carcinomas who had PEG tubes placed at the Medical College of Georgia from 1995 to 1996, 43 had records that were available for review. The remaining 7 patients had incomplete data recorded in their medical records. PEG tube placement was separated into 4 categories. Preoperative PEG tubes were placed before tumor removal; intraoperative PEG tubes were placed 479
480
Otolaryngology– Head and Neck Surgery April 1999
RAYNOR et al
Table 1. List of all complications and frequency of occurrence Complication
None External leak Granulation at site Pain Extrusion Abscess Abdominal metastases Other*
No. of patients
27 7 5 2 2 1 0 5
*Other
complications included bleeding (2), cracked tube (1), and clogged tube requiring changing (2).
Table 2. List of all tumor locations and frequency of occurrence Tumor site
Larynx Hypopharynx Tonsil Floor of mouth Base of tongue Oral tongue, buccal mucosa Supraglottic Larynx Neck Multiple* Other (head and neck)†
No. of patients (n = 43)
7 7 6 5 5 2 2 1 4 4
*Multiple
sites: Tonsil and larynx (1), base of tongue and pyriform (1), floor of mouth and lung primary (1), and base of tongue and parotid gland (1). †Other head and neck: Parotid gland (1), nasopharynx (1), and temporal bone (2).
during primary tumor resection, either before or immediately after tumor removal; and postoperative PEG tubes were placed from 7 days to several months after tumor resection. The final category included patients who had unresectable tumors and underwent palliative therapy. In patients who received PEG tubes after surgery, initial attempts at oral feeding generally failed, or the patient had significant difficulty with deglutition during radiation therapy, thereby requiring an alternative route for alimentation. PEGs were further classified as pull, the standard Ponsky technique,6 in which the catheter is pulled through the oral cavity and esophagus and out the abdominal wall; or, the push method7 whereby a guide wire is placed into the stomach under direct visualization and serially dilated with the Seldinger technique until the catheter is inserted through the abdomen directly into the stomach. Intraoperative PEGs were performed by introducing the flexible endoscope directly through the opened pharynx. A guide wire was placed percutaneously into the stomach and snared; the PEG catheter was
then fed transesophageally over the guide wire through the pharynx into the stomach. Insufflation of the stomach in these intraoperative PEGs was facilitated by gentle approximation of the mucosa around the endoscope with Babcock forceps. Patient age, sex, tumor site and stage, surgical procedure, use of additional tumor-directed therapy, date and type of PEG, and type of complications were recorded. Comparisons of time of PEG tube placement; the number, type, and severity of complications; and the location and stage of the tumor were performed with χ2 analysis. The use of adjuvant radiation therapy with respect to the time of PEG and complications was also evaluated. Two-tailed Fisher’s exact tests were used when 2 groups were compared with each other. Statistical significance was determined at the P ≤ 0.05 level. RESULTS
Thirty-six of the patients were male (84%), and the average age was 58 years. The majority of patients had stage IV tumor at the time of their PEG (79%). Time of PEG tube placement was initially divided into 5 categories: preoperative (10 patients), intraoperative before tumor extraction (4 patients), intraoperative after tumor removal (9 patients), postoperative (12 patients), and unresectable (8 patients). For purposes of statistical analysis, the preoperative and intraoperative groups before tumor removal were added together (n = 14). Of the 7 women, 1 had an intraoperative PEG, 4 received their PEG tube before surgery, and 2 had PEG tubes placed after surgery. Their tumor stages were as follows: preoperative—stage II (1), stage IV (3); intraoperative—stage IV (1); and postoperative—stage II (1), stage III (1). Nine patients had tumors from stages I through III (T1N0-T3N1). One patient had a T1N0 base of tongue tumor and received a preoperative PEG tube. Stage II cancer patients had only preoperative (2) and postoperative (2) PEG tubes placed. Patients with stage III tumors got PEG tubes at all 3 times: preoperative (1), intraoperative (1), and postoperative (2). Comparisons were made between timing of placement and the total number of complications. Minor and major complications are listed in Table 1. The most common reported complications were external leakage from the base of the PEG site (7) followed by granulation tissue (5). One major complication occurred in this series. This patient received his PEG tube before tumor resection, and subsequently an intraabdominal abscess developed requiring a laparotomy. Only 3 patients had the push PEG by the Russell method; therefore no statistical comparisons could be made with regard to method of PEG tube placement and complications. At the time of manuscript preparation, no patients in this series had evidence of tumor at the PEG site. PEG tubes were removed in 10 patients because of sole reliance on oral intake for nutri-
Otolaryngology– Head and Neck Surgery Volume 120 Number 4
RAYNOR et al
481
Fig 1. Graph of time of PEG tube placement versus number of patients for no complication and any complication.
tion. Four patients died before PEG tube removal of causes unrelated to their PEGs. None of them had any PEG complication before death. Patients who underwent PEG after tumor removal, either during or after surgery, had a statistically significantly lower rate of complications than the preoperative groups (27% vs 57%; P = 0.027). Those patients who underwent intraoperative PEG after tumor removal had significantly fewer complications than those who underwent preoperative PEGs (11% vs 57%; P = 0.04. Patients who received preoperative versus postoperative PEGs had significantly more complications (57% vs 17%; P = 0.05). Patients who had unresectable tumor also had a higher rate of complications than patients with postoperative PEGs (31% vs 17%). These comparisons are demonstrated in Fig 1. Tumor sites are listed in Table 2. There was no statistically significant difference between tumor location and PEG complications. Twelve patients had recurrent squamous cell carcinoma (28%), 29 of the 35 surgical patients received postoperative radiation therapy (83%), and 6 of 35 underwent preoperative x-ray therapy (17%). Neither of these groups had a statistically significant increase in complications. DISCUSSION
Patients with head and neck malignancies are often malnourished because of impaired deglutition from either the tumor itself or decreased oral competence,
cranial nerve deficiencies, adynamic pharyngeal segments from reconstructive flaps, or radiation-induced xerostomia. These patients often have anorexia, general debilitation, and nausea in addition to any swallowing difficulties. Many of these problems can be overcome by providing an alternative route for enteral nutrition through either NGTs or gastrostomy tubes. Open gastrostomies require a laparotomy and have a significant complication rate ranging from 3% to 50%.8 In addition, open gastrostomies may prolong reaching optimal feeding because of significant postoperative pain and ileus.3 Nasoenteral tubes, while being easy to insert and use, also have marked disadvantages including the potential for sinusitis, gastroesophageal reflux, possible suture line irritation, and social unacceptability.9,10 Use of PEG tubes may decrease hospital stay, as demonstrated by Gibson and Wenig.11 In this study patients with pharyngeal and laryngeal cancers had 61% shorter hospital stays after PEG tube placement than similar patients who had NGTs. PEG tubes are technically simple to place and carry a low rate of complications (9% to 15%) with serious complications less than 1%. Most PEG complications are minor, consisting of superficial wound infections, peristomal leakage, and chronic granulation.3,6 A serious but rare complication that has been reported in head and neck cancers is seeding of the tumor at the PEG site. This is probably the result of implantation of cancer cells from the primary tumor into the abdominal
482
Otolaryngology– Head and Neck Surgery April 1999
RAYNOR et al
wall by pulling the tumor through with the catheter.4,5 This is a rare complication, with only a few cases reported in the literature and an unknown prevalence.11 It seems likely that this complication can be minimized either by use of the Russell technique or placement of the PEG tube after the primary has been resected. Because of our desire to limit complications, especially tumor seeding, we designed this retrospective study in an attempt to determine whether timing of PEG tube placement with respect to tumor removal correlated to incidence of complications. We had no abdominal metastatic lesions in this series. We did find that there was a significant decrease in overall complications when the PEG was done after tumor extirpation as opposed to preoperative placement. In general, we found that the patients who underwent preoperative PEG tube placement tended to have a greater weight loss and a lower preoperative weight than the other groups. All 4 groups of patients had significant alcohol and tobacco use. The groups were equivalent in relation to comorbidities (hypertension, chronic obstructive pulmonary disease). Because patients whose tumors were unresectable had a higher rate of complications than the 2 postoperative groups, it would be beneficial to consider further work in this area to determine whether the method of PEG tube placement would lower the complication rate in this subset of patients. We also found that placing the PEG tube during surgery after tumor removal had the lowest complication rate and allowed for excellent visualization of the upper aerodigestive tract during the procedure, thereby making the PEG even easier and more efficiently performed. Intraoperative PEG tube placement is a straightforward procedure and can prevent the patient from requiring further procedures to achieve enteral feeding access. There is no significant delay by placing the PEG tube during surgery; this can often be done while awaiting the results of frozen sections or acquiring a dermal graft. Further studies comparing the push technique ver-
sus the pull method in preoperative head and neck carcinoma are warranted. CONCLUSIONS
Intraoperative PEG tube placement is associated with statistically fewer complications than PEG tube placement before tumor removal. Intraoperative PEGs can be readily performed without the need for an additional procedure to attain a route for enteral nutrition. We recommend intraoperative PEG tube placement, after tumor removal, in those patients with head and neck carcinoma who will require an alternative to oral feeding during their recovery period. REFERENCES 1. Selz PA, Santos PM. Percutaneous endoscopic gastrostomy. Arch Otolaryngol Head Neck Surg 1995;121:1249-52. 2. Wilson WR, Hariri SM. Experience with percutaneous endoscopic gastrostomy on an otolaryngology service. Ear Nose Throat J 1995;74:760-2. 3. Urban KG, Terris DJ. Percutaneous endoscopic gastrostomy by head and neck surgeons. Otolaryngol Head Neck Surg 1997;116: 489-92. 4. Sharma P, Berry SM, Wilson K, et al. Metastatic implantation of an oral squamous-cell carcinoma at a percutaneous endoscopic gastrostomy site. Surg Endosc 1994;8:1232-5. 5. van Erpecum KJ, Akkerdijk WL. Metastasis of hypopharyngeal carcinoma into the gastrostomy tract after placement of a percutaneous endoscopic gastrostomy catheter. Endoscopy 1995;27:124-7. 6. Gauderer MWL, Ponsky JL, Izant RJ. Gastrostomy without laparotomy: a percutaneous endoscopic technique. J Pediatr Surg 1980;15:872-5. 7. Russell TR, Brotman M, Norris F. Percutaneous gastrostomy: a new simplified and cost effective technique. Am J Surg 1984; 142:132-7. 8. Connar R, Sealy W. Gastrostomy and its complications. Ann Surg 1956;143:245-50. 9. Park RH, Allison MC, Lang J, et al. Randomized comparison of percutaneous endoscopic gastrostomy and nasogastric tube feeding in patients with persisting neurological dysphagia. BMJ 1992;304:1406-9. 10. Hunter JG, Laurentano L, Shelito PC. Percutaneous endoscopic gastrostomy in head and neck cancer patients. Ann Surg 1989; 210:42-6. 11. Gibson S, Wenig BL. Percutaneous endoscopic gastrostomy in the management of head and neck carcinoma. Laryngoscope 1992:102:977-80.
Clinical MRI
This conference, sponsored by the University of Pennsylvania Medical Center, will be held June 7-11, 1999, in Hamilton Princess, Bermuda. Course directors are J. Bruce Kneeland, MD, and Robert Lenkinski, PhD. Credits: 24 hours in category 1. For further information, contact Janice Ford Benner, Hospital of the University of Pennsylvania, MRI Center, 3400 Spruce St, 1 Founders Building, Philadelphia, PA 19104; phone, 215-662-6904; fax, 215-349-5925.