HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Pickhardt, P. J. Articles by Cossentino, M. J. Search for Related Content PubMed PubMed Citation Articles by Pickhardt, P. J. Articles by Cossentino, M. J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?

Stomal Metastases Complicating Percutaneous Endoscopic Gastrostomy: CT Findings and the Argument for Radiologic Tube Placement

¹ Department of Radiology, National Naval Medical Center, 8901 Wisconsin Ave., Bethesda, MD 20889-5600.
² Department of Radiology and Nuclear Medicine, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Rd., Bethesda, MD 20814.
³ Department of Radiology, University of Washington, 1959 N.E. Pacific, Box 357115, Seattle, WA 98195-7115.
⁴ Gastroenterology Service, Walter Reed Army Medical Center, Bldg. 2, 7F, 6900 Georgia Ave., N.W., Washington, DC 20307.

Received November 12, 2001; accepted after revision March 4, 2002.

 
The opinions and assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Departments of the Navy, Army, Air Force, or Defense.

Address correspondence to P.J. Pickhardt.

Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
OBJECTIVE. This article describes the CT appearance of metastatic implantation at the percutaneous endoscopic gastrostomy (PEG) tract in patients with malignancy of the upper aerodigestive tract. Cumulative data from previous case reports are also considered for insight into causes of metastasis and the implications for gastrostomy placement in these patients.

CONCLUSION. CT showed lobulated soft tissue involving the entire abdominal wall PEG tract in all proven cases. CT is an effective method for evaluation because the tumor burden lies predominately in the abdominal wall and not at the entry or exit site. The stomal implant is often the only site of metastatic disease at presentation. In general, CT findings of mildly increased soft tissue along the PEG tract are nonspecific, but a lobulated mass is highly suspicious for tumor implantation, especially if the one-sided thickness exceeds 1 cm. The preponderance of evidence from the existing literature points to direct tumor implantation during endoscopic placement as the likely cause (rather than hematogenous spread). This conclusion would support the alternative of radiologic tube placement in these patients.

Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
Percutaneous endoscopic gastrostomy (PEG) to provide nutritional support is a commonly performed procedure in patients with head and neck or esophageal malignancy [1]. Tumor growth about the PEG stoma is an uncommon complication that has received considerable attention in recent years in the gastroenterology and oncology literature [2,3,4,5,6]. Clinical presentation of a PEG site metastasis is typically delayed until the mass has grown large enough to cause cutaneous manifestations (e.g., fungating mass, hemorrhage, necrosis, or superinfection) or gastric complication (e.g., gastrointestinal bleeding) [2,3,4,5,6]. The PEG tube itself usually continues to function normally. We undertook this study because CT is an effective method for evaluating the extent of these stomal metastases. Because these cancer patients often undergo CT for tumor staging and various other indications, a mechanism to monitor the PEG tube and its tract already exists in many cases. To our knowledge, this article constitutes the largest single series of PEG site metastases and is the first description of the CT findings beyond isolated case examples [7]. Anecdotal evidence will be presented that strongly favors direct stomal seeding from PEG tube contamination during endoscopic placement as the most likely cause. This conclusion would support the alternative of percutaneous radiologic gastrostomy in this patient population.

Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Four patients with malignancies of the upper aerodigestive tract had biopsy-proven metastatic involvement at their PEG site identified over a 7-year period. This corresponds to an incidence of approximately 3% for PEG tube placement for this indication at our institutions. CT scans obtained before or at the time of clinical diagnosis of stomal metastasis were available in all four patients. All CT was performed using helical HiSpeed scanners (General Electric Medical Systems, Milwaukee, WI) with 7.0-mm collimation and the administration of IV contrast material.

CT review was focused primarily on the PEG site, which was evaluated for the presence of soft tissue around the gastrostomy tube. The degree of involvement at the cutaneous exit site, abdominal wall tract, and gastric entry site was subjectively assessed. Estimated dimensions of the abnormal soft tissue were recorded. CT evidence of additional sites of abdominal metastatic disease was evaluated.

The available clinical and pathologic data were reviewed for these four patients. Portions of their clinical information have been previously described in medical case reports or letters, but the CT findings have not been emphasized [4,5,6]. The CT findings were correlated with images from endoscopic examination in two patients and with photographs of the PEG exit site in three patients.

In addition, 25 CT scans from 20 cancer patients with PEG tubes in place but with no stomal metastatic disease were reviewed in a similar fashion to provide data on the specificity of CT findings. These studies were performed an average of 3.4 months after PEG tube placement (range, 2 days-13 months). No obvious concern existed for stomal infection or mass in any of these patients. None of these patients has subsequently developed PEG stomal metastasis on clinical follow-up.

Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
All four patients with stomal metastatic disease were men who were 45-76 years old (mean, 62 years). Squamous cell carcinoma was the underlying malignancy in all four patients (primary sites: tongue in two, palatine tonsillar fossa in one, and mid esophagus in one). PEG tubes were placed before radiation therapy using the "pull" technique in all four patients. The time from PEG tube placement to clinical diagnosis of stomal metastatic disease ranged from 3 to 9 months (mean, 7.3 months). Clinical presentations evolved slowly over a period of weeks to months and included various combinations of peristomal drainage, bleeding, super-infection, and friable soft tissue at the exit site. Visible peristomal soft tissue was initially interpreted in three patients as irritated or infected granulation tissue; the fourth patient presented late with a large fungating mass at the PEG exit site. None of the patients presented with signs or symptoms related to involvement at the gastric entry site. The PEG tubes continued to function normally in all four patients. Pathologic diagnosis of metastatic squamous cell carcinoma was based on biopsy of the exit site in all four patients and also on endoscopic biopsy of the entry site in two patients.

CT revealed lobulated soft tissue surrounding the PEG tubes in all four patients (Figs. 1A,1B,1C,1D,2A,2B,3). The maximum width of soft tissue perpendicular to the PEG tube (measured on one side only and excluding the tube itself) averaged 2.1 cm (range, 1.0-2.8 cm). The soft tissue extended along the entire PEG tract within the abdominal wall, which measured an average of 5.9 cm in length (range, 5.6-6.0 cm). The left rectus abdominis muscle was asymmetrically enlarged where it was traversed by the PEG tube in all four patients, and direct tumor extension into the muscle was apparent in two patients. The tumor made contact with the gastric entry site and the cutaneous exit site in all four patients. Gastric entry site involvement was more prominent than cutaneous exit site involvement in two patients (Fig. 1A,1B,1C,1D), whereas exit site involvement predominated over entry site involvement in the other two patients. No other evidence of intraabdominal metastatic disease was seen in any patient. The PEG lesion represented a solitary metastasis in two patients; the other two patients had limited extraabdominal metastatic disease.

View larger version (144K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1A. —76-year-old man with squamous cell carcinoma of tongue who presented with persistent drainage around percutaneous endoscopic gastrostomy (PEG) tube 9 months after placement. Axial contrast-enhanced CT scan shows irregular soft-tissue mass involving PEG tract (arrow). Note eccentric component at gastric entry site (arrowhead) but relatively little mass effect at cutaneous exit site. Lesion was solitary metastasis.

View larger version (143K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1B. —76-year-old man with squamous cell carcinoma of tongue who presented with persistent drainage around percutaneous endoscopic gastrostomy (PEG) tube 9 months after placement. Axial contrast-enhanced CT scan obtained 3 months before A shows clear interval progression of soft tissue around PEG tube. Minimal soft tissue (arrow) present at this point is nonspecific and appears similar to some cases with no tumor that we reviewed.

View larger version (142K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1C. —76-year-old man with squamous cell carcinoma of tongue who presented with persistent drainage around percutaneous endoscopic gastrostomy (PEG) tube 9 months after placement. Photograph from upper endoscopy shows lobulated gastric entry site mass (arrowheads) adjacent to PEG bumper, corresponding to CT finding.

View larger version (156K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 1D. —76-year-old man with squamous cell carcinoma of tongue who presented with persistent drainage around percutaneous endoscopic gastrostomy (PEG) tube 9 months after placement. Photograph of PEG exit site shows mild induration of surrounding skin and small area of friable soft tissue (arrow) centrally.

View larger version (185K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2A. —64-year-old man with squamous cell carcinoma of tonsillar fossa. Axial contrast-enhanced CT scans obtained for tumor restaging 3 months before clinical presentation of stomal metastasis show lobulated soft tissue around percutaneous endoscopic gastrostomy tube (arrowheads). Associated thickening of musculus rectus abdominis (arrows) is less specific and was also seen in most patients with no stomal metastases. Coarse pancreatic calcifications are present from chronic pancreatitis, but no other sites of abdominal metastatic disease were seen. Patient presented 3 months later with peristomal bleeding from presumed granulation tissue that was later proven at pathology to be metastatic disease.

View larger version (184K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 2B. —64-year-old man with squamous cell carcinoma of tonsillar fossa. Axial contrast-enhanced CT scans obtained for tumor restaging 3 months before clinical presentation of stomal metastasis show lobulated soft tissue around percutaneous endoscopic gastrostomy tube (arrowheads). Associated thickening of musculus rectus abdominis (arrows) is less specific and was also seen in most patients with no stomal metastases. Coarse pancreatic calcifications are present from chronic pancreatitis, but no other sites of abdominal metastatic disease were seen. Patient presented 3 months later with peristomal bleeding from presumed granulation tissue that was later proven at pathology to be metastatic disease.

View larger version (143K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 3. —45-year-old man with squamous cell carcinoma of mid esophagus who presented with fungating mass at percutaneous endoscopic gastrostomy (PEG) exit site only 3 months after placement. Axial contrast-enhanced CT scan shows increased abdominal wall soft tissue that encases PEG tube (arrow).

Three of the four CT examinations were performed at or shortly before the time of clinical diagnosis of the implantation metastasis but after symptoms had developed. CT of the fourth patient (Fig. 2A,2B) was performed at a subclinical stage 3 months before symptoms developed, at which time bleeding from the PEG stoma was presumed to be due to granulation tissue. CT of two patients performed before the index CT (but after PEG placement) showed that the soft tissue had clearly progressed in the interval. The CT findings of mildly increased soft tissue along the PEG tract on these earlier scans (Fig. 1B) were nonspecific and comparable to the appearance seen in some patients without implantation metastases.

The degree of involvement at the cutaneous exit site on CT correlated well with photographs of the abdomen that were available in three patients (Fig. 1A,1B,1C,1D). Likewise, the gastric entry finding seen on CT correlated well with photographs obtained from upper endoscopy in two patients (Fig. 1A,1B,1C,1D).

The 25 CT scans of the 20 patients with no PEG site metastases showed, at most, only minimal hazy soft-tissue infiltration adjacent to the PEG tube in the abdominal wall (Fig. 4). No patient showed a well-defined mass or lobulated contour to the soft tissue. The one-sided width of the ill-defined soft tissue adjacent to the tube measured less than 2-3 mm in most patients and never exceeded 4 mm. Mild or moderate fusiform thickening of the left rectus abdominis muscle was noted in 11 (55%) of 20 patients and appears to correspond to a normal or nonspecific finding.

View larger version (157K): [in this window] [in a new window] [as a PowerPoint slide]   Fig. 4. —80-year-old man with gastric non-Hodgkin's lymphoma and no evidence of stomal metastatic disease. Axial contrast-enhanced CT scan shows expected appearance of normal percutaneous endoscopic gastrostomy tube tract, with only minimal hazy soft tissue adjacent to tube (arrows). Lymphomatous involvement of stomach, right adrenal gland, and perirenal spaces is present, in addition to gastrohepatic and retrocrural lymphadenopathy.

Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

 
Gastrostomy tube placement for nutritional support is a frequently performed procedure that can be accomplished through fluoroscopic, endoscopic, or open surgical approaches [1, 8]. Advantages of percutaneous gastrostomy (endoscopic or radiologic) over surgical placement include shorter procedure and recovery times, avoidance of general anesthesia, and cost-effectiveness [1, 8]. Reported advantages of radiologic gastrostomy over endoscopic gastrostomy include a greater success rate and less need for conscious sedation [8]. Some of the well-recognized complications of percutaneous gastrostomy include aspiration, peritonitis, tube migration, perforation, peristomal infection, hemorrhage, and leakage around the tube [9].

Malignancy of the upper aerodigestive tract is a common clinical indication for feeding gastrostomy [1]. Metastatic seeding of the gastrostomy stoma is an unusual complication that can be seen in this subset of patients. The overall incidence is unknown, but case reports have been increasing in frequency [1,2,3,4,5,6]. On the basis of our experience, the incidence may be as high as 3%. All 27 reported cases of stomal metastases after percutaneous gastrostomy have involved tubes placed endoscopically via the os [1,2,3,4,5,6]. To our knowledge, no case has been reported after percutaneous radiologic gastrostomy. This absence is notable given the rather extensive clinical experience with percutaneous radiologic gastrostomy. To wit, one academic radiology department that places 500-700 gastrostomy tubes each year under fluoroscopy has had no case of stomal metastasis over a period of more than 15 years (Picus D, personal communication).

Our study shows that CT is an effective method for evaluating implantation metastases at the PEG site because the bulk of the tumor usually lies in the abdominal wall. Early disease is often clinically silent because the tube itself usually continues to function normally, even after the surrounding soft tissue has developed into a bulky mass. Furthermore, the early CT findings of hazy or mildly increased soft tissue are nonspecific and cannot be reliably distinguished from irritated or even normal granulation tissue [7]. Fusiform thickening of the ipsilateral rectus abdominis relative to the contralateral side is also a nonspecific finding that may be seen in most healthy patients with no tumor. Lobulated or increasing soft tissue seen along the PEG tube tract, however, should be viewed with suspicion, especially if the onesided thickness exceeds 1 cm. Extension to the cutaneous exit site (or, less frequently, the gastric entry site) is generally what leads to clinical presentation but is often delayed until the tumor has grown quite large or becomes complicated by bleeding, infection, or irritation [2,3,4,5,6]. Even when such cutaneous manifestations are present, a malignant cause is often not suspected initially [2]. These cancer patients may undergo CT for restaging or other indications, which affords an opportunity to examine the PEG region for abnormal soft tissue. Depending on the timing, CT would presumably detect most unsuspected implantation metastases before clinical manifestations develop, as seen in one of the four patients we present (Fig. 2A,2B).

The two leading hypotheses as to the mechanism of tumor spread to the PEG site are direct stomal seeding from tube contamination during placement and hematogenous spread. Each of these theories will be discussed in more detail because this issue directly affects which route of gastrostomy placement is most appropriate.

Seeding of the PEG stoma via tube contamination from direct tumor contact during endoscopic placement is a logical theory. Both the "pull" and "push" techniques entail passage of the gastrostomy tube through the oropharynx and esophagus, which results in direct tumor contact by the tube [1]. Translocation of dislodged tumor cells could result from this direct contact with the primary tumor. Substantial anecdotal evidence supports this theory, including the fact noted previously that all reported cases after percutaneous gastrostomy have involved PEG tubes placed via the os. The absence of any reported cases involving either radiologically placed gastrostomy tubes or PEG tubes placed by the "introducer" method (neither of which entails oral tube passage) further supports direct implantation as the mechanism of tumor spreading.

Port site tumor recurrences after laparoscopic and thoracoscopic tumor manipulations are well-documented phenomena [10]. Wound recurrences from open surgical procedures, however, are exceedingly rare, which lends further credence to the idea that the port site recurrences are the result of direct implantation and not hematogenous spread [10]. Tumor implantation of the percutaneous tract is a rare complication of radiologic procedures, including fine-needle biopsies and more invasive studies such as biliary interventions [11, 12]. Hematogenous spread to a fine-needle biopsy tract would seem highly unlikely given the minimal trauma involved.

One study examined the issue of metastatic spread to PEG stomata using predictions from tumor kinetics [4]. Using tumor-doubling times based on currently available biologic data, the authors concluded that direct tumor implantation seemed more feasible than hematogenous spread. The use of tumor-doubling times was particularly effective for explaining the rapid development of some stomal metastases that presented as early as 3 months after tube placement.

One argument in support of a hematogenous route is that coexistent hematogenous metastases are present in as many as half of all reported cases [5, 6]. However, coexistent hematogenous metastases are not present in a significant number of remaining cases, which seems to favor direct implantation even more. Nonetheless, hematogenous spread may have accounted for at least some of the reported cases. One case report of a stomal metastasis in a surgically placed gastrostomy tube also supports hematogenous spread [13]. Furthermore, some evidence exists from animal models that healing wounds are at increased risk for metastatic involvement by circulating tumor cells [14].

Although mostly anecdotal, these existing data strongly favor direct implantation over hematogenous spread as a cause of metastasis. This conclusion would support the practice of radiologic tube placement in this patient population to avoid direct contact of the tube with the primary tumor. Because stomal metastases tend to occur in patients with aggressive tumors and often carry a dismal prognosis, the route of tube placement is inconsequential. However, the stomal metastasis represents a solitary metastasis in approximately half of all cases [5, 6]. It remains unproven but conceivable that patient outcome would be improved by preventing a solitary implantation metastasis from occurring.

The small number of patients in this series is a limitation of our study that precludes firm conclusions. However, it is unlikely that any single institution will accumulate significant experience with PEG site metastases, so we reviewed the existing data from the many individual case reports in the literature. Another limitation of our study is that in none of the CT studies did patients with no stomal metastases have a history of stomal irritation or infection, which would have more fairly tested the specificity of CT findings.

In conclusion, CT is effective for evaluating symptomatic implantation metastases that complicate PEG placement in patients with malignancy of the upper aerodigestive tract. Although the CT findings of early asymptomatic stomal metastases are nonspecific, the observation of increased soft tissue along the PEG tract may warrant further investigation or close follow-up. Because ample anecdotal evidence supports direct tumor implantation from endoscopic tube placement as the predominate cause, percutaneous radiologic gastrostomy would seem more appropriate in these patients to prevent this complication.

References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. Safadi BY, Marks JM, Ponsky JL. Percutaneous endoscopic gastrostomy. Gastrointest Endosc Clin N Am 1998;8:551 -568[Medline]
2. Sinclair JJ, Scolapio JS, Stark ME, Hinder RA. Metastasis of head and neck carcinoma to the site of percutaneous endoscopic gastrostomy: case report and literature review. J Parenter Enteral Nutr 2001;25:282 -285[Abstract/Free Full Text]
3. Peghini PL, Guaouguaou N, Salcedo JA, Al-Kawas FH. Implantation metastasis after PEG: case report and review. Gastrointest Endosc 2000;51:480 -482[Medline]
4. Douglas JG, Koh W, Laramore GE. Metastasis to a percutaneous gastrostomy site from head and neck cancer: radiobiologic considerations. Head Neck 2000;22:826 -330[Medline]
5. Brown MC. Cancer metastasis at percutaneous endoscopic gastrostomy stomata is related to the hematogenous or lymphatic spread of circulating tumor cells. Am J Gastroenterol 2000;95:3288 -3291[Medline]
6. Cossentino MJ, Fukuda MM, Butler JA, Sanders JW. Cancer metastasis to a percutaneous gastrostomy site. (letter) Head Neck 2001;23:1080[Medline]
7. Levine CD, Handler B, Baker SR, et al. Imaging of percutaneous tube gastrostomies: spectrum of normal and abnormal findings. AJR 1995;164:347 -351[Abstract/Free Full Text]
8. Wollman B, D'Agostino HB. Percutaneous radiologic and endoscopic gastrostomy: a 3-year institutional analysis of procedure performance. AJR 1997;169:1551 -1553[Abstract/Free Full Text]
9. Schapiro GD, Edmundowicz SA. Complications of percutaneous endoscopic gastrostomy. Gastrointest Endosc Clin N Am 1996;6:409 -422[Medline]
10. Johnstone PA, Rohde DC, Swartz SE, Fetter JE, Wexner SD. Port site recurrences after laparoscopic and thoracoscopic procedures in malignancy. J Clin Oncol 1996;14:1950 -1956[Abstract/Free Full Text]
11. Soyer P, Pelage JP, Dufresne AC, et al. CT of abdominal wall implantation metastases after abdominal percutaneous procedures. J Comput Assist Tomogr 1998;22:889 -893[Medline]
12. Lundstedt C, Stridbeck H, Andersson R, Tranberg KG, Andren-Sandberg A. Tumor seeding occurring after fine-needle biopsy of abdominal malignancies. Acta Radiol 1991;32:518 -520[Medline]
13. Alagaratnam TT, Ong GB. Wound implantation: a surgical hazard. Br J Surg 1977;64:872 -875[Medline]
14. Murthy SM, Goldschmidt RA, Rao LN, Ammirati M, Buchmann T, Scanlon EF. The influence of surgical trauma on experimental metastasis. Cancer 1989;64:2035 -2044[Medline]

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This Article Abstract Figures Only Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Pickhardt, P. J. Articles by Cossentino, M. J. Search for Related Content PubMed PubMed Citation Articles by Pickhardt, P. J. Articles by Cossentino, M. J. Social Bookmarking                      What's this? Hotlight (NEW!) What's Hotlight?