| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Vascular and Interventional Radiology |
1 From the Departments of Interventional Radiology (F.J.T., T.F., F.P.M., M.J.L.) and Neurology (M.A., O.H.), Beaumont Hospital, PO Box 1297, Beaumont Rd, Dublin 9, Ireland. Received May 11, 2001; revision requested July 9; final revision received February 14, 2002; accepted March 1. Address correspondence to M.J.L. (e-mail: leebeau@iol.ie).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: During 3
years, 36 patients with ALS (19 men, 17 women; mean age, 54 years; age range, 37–69 years) underwent gastrostomy. Twenty patients were primarily referred for percutaneous endoscopic gastrostomy (PEG) and 16 for percutaneous radiologic gastrostomy (PRG). Gastrostomy method, success rate of each technique, and reason for procedure failure were reviewed in each patient. Forced vital capacity (FVC) prior to gastrostomy was recorded. PEG was performed with a pull-through technique after transillumination of the abdominal wall. PRG was performed with fluoroscopic guidance and T-fastener gastropexy. A log-rank test was used to compare survival rates after PRG and PEG, and a Wilcoxon rank sum test was used to evaluate the influence of declining FVC on PEG success. The Kaplan-Meier product limit method was used to estimate survival probabilities.
RESULTS: Of the 20 patients referred for PEG, 11 had successful results. The nine failures (45%) resulted from failure to transilluminate the abdominal wall. All 16 patients primarily referred for PRG underwent successful gastrostomy. The nine patients in whom PEG failed underwent subsequent successful PRG. In patients with diaphragmatic palsy and a high subcostal stomach, an angled subcostal approach (n = 6) or intercostal approach (n = 2) was required at PRG. One death occurred in the PEG group (9%) because of procedure-related aspiration, and a second patient from the PEG group required laparotomy for postprocedural peritonitis. One death occurred in the PRG group (4%) because of inadvertent placement of the feeding tube in the peritoneal cavity. There was no significant difference between PEG and PRG in patient survival. FVC did not have a statistically significant influence on PEG failure.
CONCLUSION: Results show PRG to be the method of choice.
© RSNA, 2002
Index terms: Amyotrophic lateral sclerosis, 72.613 • Gastrostomy, 72.1267 • Stomach, interventional procedures, 72.1267 • Stomach, surgery, 72.1267
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The bulbar, cervical, and thoracic loss of motor neurons associated with ALS predisposes patients to aspiration pneumonia, which aggravates respiratory failure and often accelerates the natural course of the disease (10,11). Management of respiratory failure involves strategies that limit aspiration, reduce secretions, and position the patient to a maximal mechanical advantage (6).
Both percutaneous endoscopic gastrostomy (PEG) and percutaneous radiologic gastrostomy (PRG) techniques have been used successfully in patients with ALS, although to our knowledge, there have been no formal clinical trials reported in the literature (12). Wollman and D'Agostino (13) noted that migration of the stomach into the thoracic cavity because of diaphragmatic weakness may preclude transillumination of the anterior abdominal wall at endoscopy and make safe gastrostomy impossible. PRG with fluoroscopic guidance and other inherent safety factors has an intrinsic versatility that surmounts this problem (14).
A comparison of both procedures when applied to the management of ALS has not been reported in the literature, to our knowledge. Thus, we retrospectively evaluated both PEG and PRG methods of gastrostomy placement in patients with ALS at our institution to determine the insertion method of choice.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
1. Patients who had moderate debilitation, bulbar symptoms, high risk of perioperative aspiration, more than 10% body weight loss, and FVC of 40%–60% were referred for PRG.
2. Patients who were less than 50 years old with more than 10% body weight loss and FVC of more than 50% and who were otherwise well with only mild bulbar symptoms and a low aspiration risk were referred for PEG. Those in whom PEG was unsuccessful were referred for PRG.
3. Patients who have FVC of more than 75% who are otherwise well, without bulbar symptoms or weight loss, are not currently offered gastrostomy.
4. Patients who have FVC of less than 40% are not routinely offered gastrostomy, since current evidence suggests that the procedure does not confer survival benefit in this population.
The criteria as outlined have since been revised, and since July 2001, all those requiring gastrostomy are referred directly for PRG. Evidence-based referral criteria have not yet been described in the literature, to our knowledge, although guidelines have been published (15). Informed consent was obtained from all patients.
PEG was performed by using a modification of the Gauderer-Ponsky pull-through method described in 1980 (16). The procedure was performed in the endoscopy unit by two endoscopists experienced in this technique, with the patient consciously sedated with 4 mg of midazolam (Rocke, Basel, Switzerland) and 50 mg of pethidine hydrochloride (Antigen Pharmaceuticals, Roscrea, Ireland). The gastroscope was passed into the patient’s stomach, and the gastroscope light source was used to transilluminate the anterior abdominal wall to select a suitable gastrostomy site. Local anesthesia was administered, and a small incision was made before puncturing the anterior wall of the stomach with a Seldinger needle. A guide wire was fed through the needle and snared through the gastroscope. The gastroscope was slowly removed, bringing the wire out through the patient’s mouth. The PEG tube was attached to the guide wire, and the abdominal end was pulled gently but firmly so that the tube passed through the mouth and esophagus and into the stomach. The proximal end of the 16-F feeding tube (Flexiflo Dura-PEG; Abbott, Sligo, Ireland) was pulled through the abdominal wall until the retaining flange juxtaposed the anterior stomach wall and the anterior abdominal wall. A fixation clip was attached externally to complete the procedure.
PRG was performed with fluoroscopic guidance, as previously described by using the Seldinger technique (14). Patients received 300 mL of dilute barium (E-Z-EM, Westbury, NY) either orally or by means of a nasogastric tube the night before the procedure to opacify the transverse colon, and a nasogastric tube was inserted, through which the stomach was insufflated with air during the procedure. All procedures were performed by one of three interventional radiologists (T.F., M.J.L., F.P.M.) experienced in the gastrostomy procedure, with the patient consciously sedated with 2 mg of midazolam and 100 µg of fentanyl citrate (Antigen Pharmaceuticals). Local anesthesia was administered before all cutaneous punctures.
Gastropexy with T-fasteners (Boston Scientific, Natick, Mass) was the primary step performed in all procedures. Gastropexy fixes the anterior wall of the stomach to the anterior abdominal wall, which stabilizes the anterior wall of the stomach and allows safe dilation of the gastrostomy tract (14,17). This step was achieved by using four T-fasteners inserted into the gastric antrum at the corners of a 2-cm square around the site of the proposed gastrostomy. It was not possible to establish gastropexy with four T-fasteners in six patients because of limited access to the anterior abdominal wall. At least two T-fasteners were used in all patients.
An 18-gauge needle was used to puncture the center point of the square-configured gastropexy, and a 0.038-inch Amplatz Super Stiff guide wire (Boston Scientific, Galway, Ireland) was coiled in the stomach. This tract was dilated as much as 16 F by using increasing dilator sizes passed over the guide wire. It was standard practice to screen patients between each successive dilation of the gastrostomy tract to check guide wire position and to ensure optimum stomach inflation with air. Finally, a gastrostomy feeding tube was inserted over the guide wire, and a test injection of 10–20 mL of contrast medium (Iohexol 300 or Omnipaque 300; Nycomed, Oslo, Norway) through the gastrostomy tube was performed to check gastrostomy tube position.
A 14-F gastrostomy tube (Cook, Bloomington, Ind) was used in 20 of 25 patients (16 primary PRG referrals and nine failed PEG cases). A 14-F button gastrostomy (MIC-KEY G Low Profile Gastrostomy Kit; Medical Innovation, Draper, Utah) was used in the last five patients (18), with one of these being an elective replacement of the standard, previously used gastrostomy tube.
Technical difficulties encountered, as well as success rate of placement, in both PEG and PRG were obtained from patient records, which were reviewed by two authors (F.J.T., M.A.). Immediate postprocedural complications were recorded. FVC prior to gastrostomy was documented. Follow-up continued in conjunction with the clinical team until the time this article was written, and the survival period was noted from the day of gastrostomy.
Statistical calculations were made to assess the effect of declining FVC on PEG performance and the ability to transilluminate the anterior abdominal wall. The nonparametric Wilcoxon rank sum test was used to determine any difference in FVC levels between groups. Survival time in both the PEG and PRG groups was also analyzed. Statistical results were defined as significant for a P value of .05 or less. The log-rank test for the equality of survivor functions among groups was used, and Kaplan-Meier product limit estimates to determine the survival probabilities at different intervals were implemented.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
In two patients referred for PRG after failed PEG, the stomach was not visible below the costal margin at fluoroscopy after distention with air because of diaphragmatic palsy. Intercostal gastrostomy tube placement was required in these two patients. The lowest intercostal space was chosen in both. Chest fluoroscopy performed in these patients at the end of each procedure did not show pneumothorax. In six patients (four patients with failed PEG and two primary PRG referrals), a small portion of the stomach descended below the costal margin after air insufflation; by using a steep cephalic track angulation, it was possible to place a gastrostomy tube. Gastropexy in these cases was also difficult because of diaphragmatic palsy, with space for only two T-fasteners instead of the usual four in these six patients.
One patient who had previously undergone attempted PEG provided a substantial technical challenge. The reason for PEG failure in this patient related to obesity and a high-lying stomach. These factors prevented transillumination of the anterior abdominal wall. At PRG, the air-distended stomach was visualized, but it was not possible to safely dilate the tract to the required 14-F size. A 10-F pigtail catheter (Boston Scientific, Galway, Ireland) was temporarily placed, and the patient was electively recalled for optimal dilation of the tract and placement of a 14-F gastrostomy tube.
One patient who underwent successful primary PEG aspirated pharyngeal secretions into the larynx during endoscopy and died during respiratory arrest 24 hours after the procedure. This patient had asthma and FVC of 17%. No aspiration events were recorded in the PRG patient group. A second patient in the PEG group, who developed acute abdomen within 24 hours after the procedure, required explorative laparotomy. A leak was discovered around the gastrostomy site on the anterior wall of the stomach and was repaired.
There was one procedure-related complication in the PRG group. This patient underwent a failed PEG attempt 2 days prior to PRG and, during the PRG procedure, had inadvertent placement of the gastrostomy tube in the peritoneal cavity. A test injection of contrast medium (Iohexol 300 or Omnipaque 300; Nycomed) was performed, as is standard, but results were misinterpreted, and the tube malposition was not detected. The patient was subsequently fed through the tube and developed peritonitis. Despite emergency laparotomy, the patient died 4 days after the procedure.
The mean FVC was 53% (range, 35%–97%) in the successful primary PEG group and 57% (range, 12%–95%) in the subset of patients in whom primary PEG failed. In the patient group referred primarily for PRG, the mean FVC was 61% (range, 43%–100%). No significant difference in FVC was observed between the patients who underwent successful PEG and those in whom PEG failed (P = .52) or between those who underwent PRG and those in whom PEG failed (P = .125). Twenty-one patients of the 36 have died at the time of preparing this article. The estimated median survival time was 13 months (range, 0–16 months) from the date of gastrostomy. There was no significant difference detected in survival rates between the PEG and PRG groups (P = .165) (Figure).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
There is an intrinsic risk of aspiration in patients with ALS because of dysfunction of the swallowing mechanism (22). Results of a meta-analysis of the literature by Wollmann et al (23) showed a significantly greater risk of aspiration associated with PEG, as compared with the risk associated with PRG (2.1% vs 0.6%; P < .001). Performance of PEG in patients with ALS will theoretically further increase the risk of aspiration during or immediately after the procedure. This proposed risk relates to the degree of conscious sedation required for PEG, the degree of pharyngeal anesthesia required, and the mechanical stress imposed during endoscopy in an already compromised swallowing and antireflux mechanism. In our study, one patient from the PEG group aspirated, which resulted in death. In the PRG group, there were no aspiration events recorded.
Safe performance of PEG depends on adequate transillumination of the anterior abdominal wall by the gastroscope light. In some patients with ALS, because of intrinsic weakness of the diaphragm, the stomach migrates high under the costal margin. Such anomalous anatomy, especially when compounded by obesity, precludes transillumination of the abdominal wall at a safe access site. Furthermore, there is a greater risk of penetrating a vital thoracic structure if the gastroscope light is the only guiding factor. Further navigational information is essential before safely proceeding to gastrostomy placement in such patients. This information is readily available when the PRG protocol is applied. Anteroposterior and lateral fluoroscopy while the stomach is insufflated with air and the colon opacified with barium provides an accurate real-time road map to plan the gastrostomy.
Use of an intercostal route or angled subcostal route for gastrostomy placement has been described (12). When necessary, we use the lowest anterior intercostal space, where the pleural reflection is less likely to be traversed. We have experienced no additional morbidity with this approach.
The 53% success rate of PEG in our institution is much lower than the 89% success rate in reported in another article (24). This discrepancy may be partly related to different selection criteria for patients who have ALS and undergo PEG. In the population of patients with ALS in our study, six of 20 referred for PEG had FVC of less than 50%. Evidence-based reviews by Miller (25) and Bradley et al (26) describe FVC of 30%–50% as moderate risk and FVC of less than 30% as high risk for performing PEG. These data have not been validated for PRG. It is of interest that in the patients with ALS in our study, there was no correlation between FVC and failure of the PEG technique (ie, raised diaphragmatic level). It is uncertain whether this relates to compensatory activity by accessory muscles of respiration, developing better in some patients than others, or some other factor. Formal clinical trials will be required to answer this question.
A further factor that favors PRG over PEG in patients with ALS is the degree of conscious sedation required (13). Less sedation is necessary for PRG, with less associated chance of respiratory suppression, which is already compromised in patients with ALS (10). In our institution, 2 mg of midazolam is administered prior to PRG, as compared with 4 mg administered for PEG. Furthermore, local anesthesia of the pharyngeal mucosa is not necessary in PRG, which eliminates another risk factor for aspiration.
T-fastener gastropexy enables routine use of larger gastrostomy tubes and ready replacement of a displaced tube even before the development of a mature tract (17). The inability to perform standard gastropexy with four T-fasteners in some patients with ALS is a potential hazard when performing PRG. The problem relates to difficulty in dilating the gastrostomy tract without gastropexy when the anterior stomach wall is pushed away by the dilator. There is a real risk of inadvertent placement of the tube in the peritoneal cavity in these patients. Where suboptimal gastropexy is achieved, additional care must be taken with fluoroscopy during dilation of the gastrostomy tract.
The incidence of tube replacement after PRG is greater than that seen in the PEG group. We believe this incidence results from the difference in retaining mechanisms of gastrostomy tubes designed for PRG and those designed for PEG. In our experience, the balloon-retaining device used in all the patients undergoing PRG in our study is less effective than the flange mechanism used in PEG tubes. The recommended 5-mL exchange of water at 2-week intervals in PRG balloons frequently results in overdistention and rupture. Some of the balloons ruptured spontaneously (27). The remaining tubes were inadvertently pulled out. The flange system used in the PEG feeding tubes has less potential to become displaced, which results in a lower need for PEG replacement, as seen in the PEG group in our study.
The death related to PEG resulted from aspiration of a pharyngeal mucous plug during passage of the gastroscope. This complication is consistent with the underlying inability of patients with ALS to efficiently swallow their oral secretions. PRG averts the risk of aspiration in these patients.
The death related to PRG resulted from placement of the feeding tube in the peritoneal cavity despite the presence of a T-fastener gastropexy. This was not recognized at the time of the procedure, despite a test injection of contrast medium (Iohexol 300 or Omnipaque 300; Nycomed). This patient had previously undergone PEG that failed due to failure to transilluminate the anterior abdominal wall. The high stomach position in these patients presents a technical challenge, since the approach for gastrostomy placement is angled rather than perpendicular to the abdominal wall. This angling increases the risk of losing guide wire position while dilating the gastrostomy tract, which probably occurred in this case. The deaths experienced in both groups highlight the technical difficulties that ALS presents.
In conclusion, patients with ALS present a challenge to both endoscopists and interventional radiologists when gastrostomy is required. PRG had a significantly higher success rate than did PEG in the patients with ALS in our study. The requirement for less sedation for PRG and the avoidance of the endoscopic route of placement lower the intrinsic risk of aspiration and establish PRG as the preferred primary gastrostomy technique in patients with ALS.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, M.J.L.; study concepts, O.H., M.J.L.; study design, F.J.T., O.H., M.J.L.; literature research, F.J.T.; clinical studies, T.F., F.P.M., M.J.L.; data acquisition, T.F., F.P.M., F.J.T., M.A.; data analysis/interpretation, M.A., T.F., F.J.T.; statistical analysis, F.J.T.; manuscript preparation, M.A., F.J.T.; manuscript definition of intellectual content, M.J.L., F.P.M., O.H.; manuscript editing, M.J.L., O.H.; manuscript revision/review, F.P.M., M.J.L.; manuscript final version approval, M.J.L.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. J McDermott and P. J Shaw Diagnosis and management of motor neurone disease BMJ, March 22, 2008; 336(7645): 658 - 662. [Full Text] [PDF]
|
|
|
|
 |
|
 A. S. Shaw, M. A. Ampong, A. Rio, J. McClure, P. N. Leigh, and P. S. Sidhu Entristar Skin-Level Gastrostomy Tube: Primary Placement with Radiologic Guidance in Patients with Amyotrophic Lateral Sclerosis Radiology, November 1, 2004; 233(2): 392 - 399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A Chio, R Galletti, C Finocchiaro, D Righi, M A Ruffino, A Calvo, N Di Vito, P Ghiglione, A A Terreni, and R Mutani Percutaneous radiological gastrostomy: a safe and effective method of nutritional tube placement in advanced ALS J. Neurol. Neurosurg. Psychiatry, April 1, 2004; 75(4): 645 - 647. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P N Leigh, S Abrahams, A Al-Chalabi, M-A Ampong, L H Goldstein, J Johnson, R Lyall, J Moxham, N Mustfa, A Rio, et al. The management of motor neurone disease J. Neurol. Neurosurg. Psychiatry, December 1, 2003; 74(90004): iv32 - 47. [Full Text] [PDF]
|
|
|
|
|
|
B J Traynor, M Alexander, B Corr, E Frost, and O Hardiman Effect of a multidisciplinary amyotrophic lateral sclerosis (ALS) clinic on ALS survival: a population based study, 1996-2000 J. Neurol. Neurosurg. Psychiatry, September 1, 2003; 74(9): 1258 - 1261. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
