A Hard Look at Colonoscopy Jerome S. Abrams, MD, Burlington, Vermont
Colonoscopy utilizing the modern fiberoptic endoscope was introduced into clinical practice in 1971; four years later, the report by Work Group X of the American Gastroenterological Association stated, “colonoscopy may well prove to be the single most important recent advance in the management of colonic disease” [I]. Or, more picturesquely, “almost overnight the lights of colonoscopists shown from the rectum to the cecum in ever-increasing numbers” [Z]. In the many presentations and publications that have appeared since 1972, colonoscopy has been described as simple, rapid, thorough, reliable, inexpensive, and
seen but not able to be snared; and three with polypectomy in whom the polyp was “lost.” There were five complications (2.6 per cent). Perforation followed colonoscopic biopsy at the splenic flexure in a patient with Crohn’s colitis; the site of perforation was not seen at surgery and he underwent uncomplicated colectomy. Four patients bled after polypectomy; two required transfusion, but none came to laparotomy.
safe.
Considerable experience has now been obtained throughout the world as evidenced by a recent collective review by Smith and Nivatvongs [3] consisting of almost 8,000 procedures. Several clinics have reported personal series exceeding 4,000 patients. The experience at the Medical Center Hospital of Vermont includes 196 procedures (colonscopy, 63; colonoscopy and biopsy, 68; and colonoscopy and polypectomy, 65). Indications for colonoscopy in the 196 patients were abnormal barium enema in fortyseven (24 per cent), inflammatory bowel disease in twenty-five (13 per cent), polyp(s) in ninety-six (49 per cent), status post colon resection in twelve (6 per cent), and miscellaneous in sixteen (8 per cent). The location of the suspected lesion is shown in Figure 1 and the site reached by the instrument is shown in Figure 2. All procedures were performed with a 100 cm scope. Three of sixty-five polyps (4.6 per cent) were never retrieved. Invasive carcinoma was found in only one pedunculated polyp. There were twenty-seven (14 per cent) “failures” in this series: thirteen in whom objective was not reached; four (2 per cent) in whom a lesion was present but not seen (2 polyps, 2 cancers); two in whom a lesion was seen but not able to be biopsied; five in whom a polyp was
From the Department of Surgery, College of Medicine, the University of Vermont, Burlington, Vermont. Reprint requests should be addressed to Jerome S. Abram% MD, Department of Surgery, Given Building E-307, College of Medicine, the University of Vermont, Burlington, Vermont 05401. Presented at the Seventeenth Annual Meeting of the Society for Surgery of the Alimentary Tract, Miami Beach, Florida, May 25-26, 1976.
Volume 133, January 1977
Figure 7. Location of suspected /esion( s) in 196 patients. Seventy-nine per cent of lesions lie distal to splenic flexwe, 94 per cent distal to hepatic flexure.
111
Abrams
before polypectomy is attempted and that “solo” colonoscopy should follow the completion of forty procedures supervised by an “experienced” endoscopist. To some, diagnostic colonoscopy and even polypectomy is appropriately performed as an outpatient procedure in the hospital or in a physician’s office (without fluoroscopy or x-ray examination). On the other hand, Sugarbaker et al [4] and Farringer [5] perform all colonoscopy in the operating room and most authors believe that fluoroscopy and/or overhead x-ray examination is extremely helpful in passing the scope in a difficult patient and mandatory to confirm the location of a lesion that may subsequently require laparotomy and resection. How Rapid While the cecum may be reached in as little as 5 to 7 minutes, a complete and thorough examination usually requires at least 1 hour [6] and some examinations take as long as 4 hours [4]. Thus, time is a limiting factor in the number of patients who can be examined by any one colonoscope and any one endoscopist. How Thorough
Figure 2. Site reached by 100 cm colonoscope in 196 patients.
Now that colonoscopy has become an established diagnostic and therapeutic procedure, it seems appropriate to take a “hard look” and see just how simple, how rapid, how thorough, how reliable, how inexpensive, and how safe it really is. How Simple Despite reports by some experienced endoscopists regarding the simplicity of the procedure, others have suggested that endoscopic examination of the colon is considerably more difficult than upper gastrointestinal endoscopy and rates with endoscopic retrograde cholangiopancreatography in terms of difficulty. “It [colonoscopy] should be more widely used than sigmoidoscopy, but the expertise and time required to pass the colonoscope to the cecum seriously limits its wide application” [I]. Some have suggested that a minimum of 200 colonoscopies be performed
112
Early reports stated that the cecum could be reached in only 30 to 50 per cent of cases; more recently, using 160 to 180 cm instruments, endoscopists are reaching the cecum in up to 95 per cent of cases [6]. However, even recent reports note that total colonoscopy is not always achieved [ 71 and despite very extensive experience, Williams and Teague [6] reported that “when there are fixed loops or bends as it curves with the adhesions following previous surgery or diverticular disease, it may be impossible to pass the colonoscope.” I was unable to pass the scope through the sigmoid colon at the time of laparotomy in a patient with adhesions from a remote hysterectomy! Just how important is total colonoscopy? As seen in Figure 1,79 per cent of all potential lesions in our patients occurred distal to the splenic flexure whereas 94 per cent were distal to the hepatic flexure. In their large series of more than 2,000 colonoscopies (including 499 polypectomies), Wolff and Shinya [8] noted that 35 per cent of all polyps were less than 50 cm and 91 per cent less than 100 cm from the anus, or that 87 per cent of all polyps were distal to the splenic flexure and 95 per cent distal to the ascending colon, and they were achieving “total” colonoscopy in almost all patients. Of major importance in terms of potential cancer, no villous polyps were found
The American Journal of Surgery
Colonoscopy
proximal to the sigmoid colon. Twelve of fourteen (86 per cent) malignant polyps reported by Knutson, Schrock, and Polk [9] were distal to the splenic flexure. Thus, it is evident that about 90 per cent of all colonic lesions, benign and malignant, fall within the range of the 100 cm colonoscope. How Reliable
Is colonoscopy more reliable than a barium enema? There are many reports in which lesions seen on colonoscopy had been missed by the barium enema. However, Williams [IO], in his review of 800 patients, found a 98 per cent correlation between colonoscopy and air contrast barium enema for polyps greater than 1 cm in diameter. On the other hand, in a study of seventy-eight patients with x-ray evidence of polyp, six polyps were missed on the initial endoscopy, demonstrated on a repeat barium examination, and subsequently seen on a second colonoscopy [1I]. Loose and Williams [12] performed colonoscopy on ninety-nine patients with rectal bleeding and a normal barium enema; twelve had a benign polyp, two had cancer, and four had a miscellaneous finding. Thus eighty-two of ninety-nine symptomatic patients underwent colonoscopy to confirm a negative barium enema. Colonoscopic biopsy, limited technically by size and depth of bite, may not reflect the true histologic nature of a lesion, particularly in villous adenomas, carcinomas, and inflammatory bowel disease. We have had two patients with infiltrating carcinomas where multiple biopsies showed benign adenoma. Of even greater concern is the problem of the “lost” polyp. In our own series, 5 per cent of all tissue removed was never retrieved and Wolff and Shinya [13] reported that forty-two of 616 polyps (7 per cent) were never found. Similar figures have been noted by others [4]. Approximately 2 per cent of pedunculated polyps contain invasive cancer [14]; how many may be lost along with the opportunity for curative resection? How Inexpensive
Reliable estimates suggest that there are currently 3,000 modern fiberoptic colonoscopes in the United States. The cost of equipping a colonoscopy unit may reach $25,000 [15], bringing the total initial expenditures to as much as $75,000,000. In addition to the cost of setting up a unit, Williams and Teague [6] believed that the average life of an instrument was about 200 examinations and that repairs would equal approximately one third of the original cost, a total investment of approximately
Volume 133, January 1977
TABLE
I
Comparison
of Annual
Costs of Colonic
Polypectomy-Laparotomy
vs Colonoscopy
at Medical
of Vermont
Center
Hospital
Laparotomy Hospital costs Surgeon’s fees Total No. of patients Annual cost
$ 2,744 528 B 3.272 ,22,90:
Colonoscopy $
241 495 736 B 30 $22,080
$10,000 or $50 per examination exclusive of physician’s fees and hospital costs. The hospital expenses for colonoscopic polypectomy are roughly one third that of transabdominal colotomy [9,16,17] and considerably less when outpatient services are used. However, the surgeon’s fee for laparotomy and polypectomy, including pre- and postoperative care and seven to fourteen days in the hospital is listed as sixteen relative value units by Blue Cross and Blue Shield, whereas polypectomy via colonoscopy (0 to 2 days in the hospital) earns a fee of fifteen relative value units. To determine the financial impact of colonoscopy on total health costs, some additional data must be examined. It has been suggested that all endoscopists perform a minimum of 200 diagnostic procedures before attempting polypectomy [13]. The cost of the necessary 600,000 diagnostic colonoscopies (at $400 per procedure) would be $240,000,000. It is particularly discouraging to realize that even the presence of symptoms and/or signs results in a low yield at a high cost. As noted, Loose and Williams [12] performed colonoscopy on ninety-nine patients who presented with rectal bleeding and a negative barium enema-a primary indication for colonoscopy-and found two cancers. The average cost (based on charges at the Medical Center Hospital of Vermont) for each cancer detected was more than $25,000. Recently, Coller, Corman, and Veidenheimer [18] recommended preoperative total colonoscopy for all patients scheduled for colectomy for colorectal cancer to detect a synchronous malignancy. Approximately 100,000 new colorectal cancers will occur in 1976 and 3.6 per cent, or 3,600 patients, will have a second colon primary [19]. The endoscopy expenses alone would equal $40,000,000 or approximately $11,000 per synchronous cancer, assuming none could be detected by sigmoidoscopy or barium enema. Finally, it must be recognized that the indications for colonoscopic polypectomy exceed those for laparotomy. As noted in Table I, we went from an average of seven transabdominal polypectomies per year to thirty in 1975 at an equal annual cost. Thus,
113
Abrams
potential savings per procedure may be overwhelmed by the increasing volume and indications for colonoscopy. How Safe Smith and Nivatvongs [3] found sixteen perforations (1.2 per cent) and thirty-one hemorrhages (2.5 per cent) in a series of 1,244 polypectomies. There were seven perforations (2.6 per cent) and six hemorrhages (2.3 per cent) in 265 patients with sessile polyps. Spencer, Coates, and Anderson [ZO]reported a perforation rate of 1.9 per cent after polypectomy. Perforation has occurred after removal of a 3 mm polyp [21] and fulgeration of a 2 mm polyp [22]. Serosal tears, colonic perforations, and even perforation of the terminal ileum has been described during diagnostic colonoscopy. The true complication rate for colonoscopy/polypectomy is unknown. The data we have come from published series and a few surveys, most of which fall far short of a 100 per cent response rate. It has been suggested that responders generally include those with the best results. Although there are no collective figures available regarding mortality, deaths are known. Summary
The fiberoptic colonoscope represents a significant diagnostic and therapeutic achievement in the management of colonic disease. Despite some claims to the contrary, it is not universally simple, rapid, thorough, reliable, and inexpensive or without morbidity and mortality. Until these goals are achieved through future refinements, modifications, and new instrumentation, it is strongly recommended that the indications, for colonoscopy be sharply defined to include the following: (1) Investigation of colonic lesions seen on two successive or one air-contrast barium enema examination. Pedunculated polyps less than 1 cm in diameter can be observed. (2) Preoperative investigation of patients with demonstrable colorectal cancer (to rule out synchronous lesions). (3) Postoperative evaluation after colectomy for carcinoma where the anastomosis is above the reach of a sigmoidoscope. (4) Evaluation of patients with inflammatory bowel disease of the colon. .(5) Investigation of persistent occult or gross rectal bleeding when anoscopy, sigmoidoscopy, and barium studies are negative. Patients who represent prohibitive surgical risks because of age or systemic disease should not be considered as candidates for colonoscopy without compelling indications. The decision to biopsy rather
114
than remove or fulgerate a detected lesion must be left to the judgment and experience of the endoscopist. The morbidity that follows polypectomy or fulgeration of sessile lesions less than 0.5 cm in diameter probably exceeds the incidence of carcinoma. References I. Rubin CE, Auth D, Barrett B, et al: Work group X. Instruments for diagnosis, investigation and treatment of digestive diseases. Gastroenterology 69: 115 1, 1975. 2. Ovetilt BF: Colonoscopy. Review.Gastroentero/ogy 66: 1306, 1975. 3. Smith. LE. Nivatvongs S: Complication in colonoscopy. Dis Colon Rectum 16: 214, 1975. 4. Sugarbaker PH, Vineyard GC, Lewicki AM et al: Colonoscopy in the management of diseases of the colon and rectum. Surg Gynecol Obstet 139: 341, 1974. 5. Farringer GL Jr: Discussion of [ 91. 6. Williams C. Teaaue R: Colonoscoov. Gut 14: 990. 1973. 7. Dagrodi Ai, Alaama A, Ruiz R: Clinical experiences with colonoscopy. Am J Gastroenterol63: 406, 1975. a. Wolff WI, Shinya H: A new approach to colonic polyps. Ann Surg 178: 367, 1973. 9. Knutson CO, Schrock LG. Polk HC Jr: Polypoid lesions of the proximal colon: comparison of experiences with the removal at laparotomy and by colonoscopy. Ann Surg 179: 657, 1974. 10. Williams C: Evaluation of the colonoscopic examination. Dis Colon Rectum 16: 366, 1975. Il. Laufer I, Smith NCW, Mullens JE: The radiological demonstration of colorectal polyps undetected by endoscopy. Gastroenterology 70: 167. 1976. 12. Loose HWC, Williams CB: Barium enema vs. colonoscopy. Proc R Sot Med 67: 1033, 1974. 13. Wolff WI, Shinya H: Modern endoscopy of the alimentary tract. Curr Probl Surg I, January 1974. 14. Shatney CH, Lober PH. Gilbertsen VA, Sosin H: The treatment of pedunculated adenomatous colorectal polyps with focal cancer. Surg Gynecol Obstet 139: 645, 1974. 15. Sivak MV Jr, Sullivan BH Jr, Rankin GB: Colonoscopy: the report of 644 cases and review of the literature. Am J Surg 126: 351,1974. 16. Bloom B.S. Goldhaber SZ, Sugarbaker PH. O’Conner NE: Fiberoptics: morbidity and cost. Editorial. N Engl J Med 266: 368.1973. 17. Goldhaber SZ, Bloom BS, Sugarbaker PH. D’Conner NE: Effects of the fiberoptic laparoscope and colonoscope on morbidity and cost. Ann Surg 179: 160, 1974. ia. Coller JA, Corman ML, Veidenheimer MC: Colonic polypoid disease: need for total colonoscopy. Am J Surg 13 1: 490, 1976. 19. Winawer SJ, Sherlock P, Schottenfeld D, Miller DG: Screening for colon cancer. Gastroenterology 70: 763. 1976. 20. Spencer RJ, Coates HL, Anderson MJ Jr: Colonoscopic polypectomies. A&y0 Clin Proc 49: 40. 1974. 21. Meyers MA, Ghahremani GG: Complications of fiberoptic endoscopy. II. Colonoscopy. Radiology 115: 301, 1975. 22. Nivatvongs S, Goldberg SM: The results of 100 consecutive polypectomies with a fiberoptic colonoscope. Am J Surg 128: 347.1974.
Discussion Henry Colcher (Boston, MA): I read Doctor Abram’s manuscript and I would like to answer each point as rapidly as I can.
The American Journalof Surgery
Colonoscopy
Easy? No, it is not an easy procedure. We do accept the fact that there are failures and that we cannot reach the cecum in all patients. Duration? It is not a rapid procedure. We assign 1 hour for each endoscopic examination. Safe? Yes and no. It is not a safe procedure in the hands of people who have not learned how to do it. In 1973, when I was President of the ASGE, we conducted a survey and were shocked by the number of perforations and other complications. We realized that these were often seen by people who did not have enough training. With the development of training programs the procedure has definitely become safe and the benefits justify the morbidity that it involves. Reliable? Yes, if we consider again a well trained endoscopist who knows that there are blind areas in a certain examination and repeats the examination six months or a year later. Is it worth it? Yes. The results of the study conducted by Doctors Coller, Corman, and Veidenheimer [18] show that most of the polyps located on barium enema study were on the left side and a few were on the right side. Of 146 patients thirty-six had none; they were spared surgery and 110 presented 120 polyps. In addition, however, there were numerous unsuspected lesions on the right side. What does it represent in terms of disease? At the suspected site there were 103 benign lesions but also eight polyps with invasive carcinoma and nine cancers (polypoid or ulcerating adenocarcinoma) that were not suspected on x-ray examinations. At the unsuspected sites, 120 benign polyps, two polypoid adenomas with invasive carcinoma, and four adenccarcinomas-thus, a total of twenty-three invasive carcinomas-were found in twenty-two patients. Pedunculated polyps less than 1 cm in diameter may have invasive carcinoma. Total colonosopy is worthwhile because we find a large number of cancers that can be removed at a very early phase either by polypectomy or by surgery when it is needed. We are left now with two more questions concerning cost. I look at cost in terms of effort on the physician’s part. I believe it is worth it. Cost in terms of money: I believe it can be justified when we offer the patient a polypectomy or a resection for cancer at an earlier phase. Forget the figures and the money. If we can improve the results in colon cancer by polypectomy and by follow-up for synchronous and metachronous lesions, it is worth it. William I. Woolf (New York, NY): One has to go back to the days of laparotomy and colotomy and colectomy to realize what the alternatives are. I will briefly mention two appropriate cases. A fifty-five year old lady was referred to us with a polyp of the transverse colon. She had had a polyp in the descending colon removed fifteen months previously. This was followed by a wound infection, by repeated admissions for intestinal obstruction, and by a bout of homologous
Volume 133, January 1977
serum jaundice. The transverse colonic polyp was removed endoscopically without difficulty. A small polyp of the descending colon was found in an elderly obese businessman in 1971. He was asked to return in six months, at which time the polyp had not changed in size. Removal was suggested but not pressed. We saw him in 1975. Obviously the lesion was larger. The lesion was obviously a cancer, and it was removed. Colostomy was necessary as well to protect the anastomosis. It is appropriate that we know what the complication rate is. But, as we have insisted all along, it is equally important to know who is doing each procedure and under what circumstance. It is important that physicians performing these procedures are trained, competent, and have a good basic knowledge of gastrointestinal disease. At the Beth Israel Medical Center we have performed more than 10,000 diagnostic colonoscopies now and have removed more than 2,500 colonic polyps without a single mortality. I do take issue with the statement that cancer in polyps under 1 cm in diameter does not occur. It does occur. It occurs frequently, and I am referring to invasive cancer, not carcinoma in-situ. Gerald 0. Strauch (Stamford CT): I work in a community hospital. Our colonoscopies are performed by surgeons, and I have run the Surgical Audit Committee since colonoscopy came into existence in the hospital. One of the problems that we have seen is the difficulty with the recovery of polyps which have been snipped off at the time of colonoscopy. Hard data in this regard seem to be very difficult to gather. Doctor Abrams quoted a figure of 7 per cent. In our hospital that rate is substantially higher. Any polyp that is worth removing is certainly worth looking at under a microscope, and failure to be able to do so represents a serious shortcoming. Jerome S. Abrams (closing): In answer to Doctor Colcher’s points, I would remind you again that, based on autopsy data, approximately 15 to 45 per cent or almost one third of patients more than forty-five years old have at least one polyp, and therefore we should not be surprised that we can randomly put the scope in anybody and find one or more polyps. Secondly, I am impressed by the high incidence of carcinoma in Doctor Colcher’s particular group, and I have absolutely no explanation for it, since as best I can tell from a very large series of cases the incidence of carcinoma has been significantly lower. Finally, although I am impressed philosophically with the idea that you cannot place dollar values on cancers, I am also impressed that there is fear in this country and the world about the overall cost of medical care. In answer to Doctor Wolff, I think he may have misinterpreted my statements. We are certainly not questioning the value of colonoscopy, and we are indebted to his unit for pioneering this work. But I do wonder how we can control an instrument that can be used in a doctor’s office and is available at a reasonably low cost.
115
Colonoscopy utilizing the modern fiberoptic endoscope was introduced into clinical practice in 1971; four years later, the report by Work Group X of the American Gastroenterological Association stated, “colonoscopy may well prove to be the single most important recent advance in the management of colonic disease” [I]. Or, more picturesquely, “almost overnight the lights of colonoscopists shown from the rectum to the cecum in ever-increasing numbers” [Z]. In the many presentations and publications that have appeared since 1972, colonoscopy has been described as simple, rapid, thorough, reliable, inexpensive, and
seen but not able to be snared; and three with polypectomy in whom the polyp was “lost.” There were five complications (2.6 per cent). Perforation followed colonoscopic biopsy at the splenic flexure in a patient with Crohn’s colitis; the site of perforation was not seen at surgery and he underwent uncomplicated colectomy. Four patients bled after polypectomy; two required transfusion, but none came to laparotomy.
safe.
Considerable experience has now been obtained throughout the world as evidenced by a recent collective review by Smith and Nivatvongs [3] consisting of almost 8,000 procedures. Several clinics have reported personal series exceeding 4,000 patients. The experience at the Medical Center Hospital of Vermont includes 196 procedures (colonscopy, 63; colonoscopy and biopsy, 68; and colonoscopy and polypectomy, 65). Indications for colonoscopy in the 196 patients were abnormal barium enema in fortyseven (24 per cent), inflammatory bowel disease in twenty-five (13 per cent), polyp(s) in ninety-six (49 per cent), status post colon resection in twelve (6 per cent), and miscellaneous in sixteen (8 per cent). The location of the suspected lesion is shown in Figure 1 and the site reached by the instrument is shown in Figure 2. All procedures were performed with a 100 cm scope. Three of sixty-five polyps (4.6 per cent) were never retrieved. Invasive carcinoma was found in only one pedunculated polyp. There were twenty-seven (14 per cent) “failures” in this series: thirteen in whom objective was not reached; four (2 per cent) in whom a lesion was present but not seen (2 polyps, 2 cancers); two in whom a lesion was seen but not able to be biopsied; five in whom a polyp was
From the Department of Surgery, College of Medicine, the University of Vermont, Burlington, Vermont. Reprint requests should be addressed to Jerome S. Abram% MD, Department of Surgery, Given Building E-307, College of Medicine, the University of Vermont, Burlington, Vermont 05401. Presented at the Seventeenth Annual Meeting of the Society for Surgery of the Alimentary Tract, Miami Beach, Florida, May 25-26, 1976.
Volume 133, January 1977
Figure 7. Location of suspected /esion( s) in 196 patients. Seventy-nine per cent of lesions lie distal to splenic flexwe, 94 per cent distal to hepatic flexure.
111
Abrams
before polypectomy is attempted and that “solo” colonoscopy should follow the completion of forty procedures supervised by an “experienced” endoscopist. To some, diagnostic colonoscopy and even polypectomy is appropriately performed as an outpatient procedure in the hospital or in a physician’s office (without fluoroscopy or x-ray examination). On the other hand, Sugarbaker et al [4] and Farringer [5] perform all colonoscopy in the operating room and most authors believe that fluoroscopy and/or overhead x-ray examination is extremely helpful in passing the scope in a difficult patient and mandatory to confirm the location of a lesion that may subsequently require laparotomy and resection. How Rapid While the cecum may be reached in as little as 5 to 7 minutes, a complete and thorough examination usually requires at least 1 hour [6] and some examinations take as long as 4 hours [4]. Thus, time is a limiting factor in the number of patients who can be examined by any one colonoscope and any one endoscopist. How Thorough
Figure 2. Site reached by 100 cm colonoscope in 196 patients.
Now that colonoscopy has become an established diagnostic and therapeutic procedure, it seems appropriate to take a “hard look” and see just how simple, how rapid, how thorough, how reliable, how inexpensive, and how safe it really is. How Simple Despite reports by some experienced endoscopists regarding the simplicity of the procedure, others have suggested that endoscopic examination of the colon is considerably more difficult than upper gastrointestinal endoscopy and rates with endoscopic retrograde cholangiopancreatography in terms of difficulty. “It [colonoscopy] should be more widely used than sigmoidoscopy, but the expertise and time required to pass the colonoscope to the cecum seriously limits its wide application” [I]. Some have suggested that a minimum of 200 colonoscopies be performed
112
Early reports stated that the cecum could be reached in only 30 to 50 per cent of cases; more recently, using 160 to 180 cm instruments, endoscopists are reaching the cecum in up to 95 per cent of cases [6]. However, even recent reports note that total colonoscopy is not always achieved [ 71 and despite very extensive experience, Williams and Teague [6] reported that “when there are fixed loops or bends as it curves with the adhesions following previous surgery or diverticular disease, it may be impossible to pass the colonoscope.” I was unable to pass the scope through the sigmoid colon at the time of laparotomy in a patient with adhesions from a remote hysterectomy! Just how important is total colonoscopy? As seen in Figure 1,79 per cent of all potential lesions in our patients occurred distal to the splenic flexure whereas 94 per cent were distal to the hepatic flexure. In their large series of more than 2,000 colonoscopies (including 499 polypectomies), Wolff and Shinya [8] noted that 35 per cent of all polyps were less than 50 cm and 91 per cent less than 100 cm from the anus, or that 87 per cent of all polyps were distal to the splenic flexure and 95 per cent distal to the ascending colon, and they were achieving “total” colonoscopy in almost all patients. Of major importance in terms of potential cancer, no villous polyps were found
The American Journal of Surgery
Colonoscopy
proximal to the sigmoid colon. Twelve of fourteen (86 per cent) malignant polyps reported by Knutson, Schrock, and Polk [9] were distal to the splenic flexure. Thus, it is evident that about 90 per cent of all colonic lesions, benign and malignant, fall within the range of the 100 cm colonoscope. How Reliable
Is colonoscopy more reliable than a barium enema? There are many reports in which lesions seen on colonoscopy had been missed by the barium enema. However, Williams [IO], in his review of 800 patients, found a 98 per cent correlation between colonoscopy and air contrast barium enema for polyps greater than 1 cm in diameter. On the other hand, in a study of seventy-eight patients with x-ray evidence of polyp, six polyps were missed on the initial endoscopy, demonstrated on a repeat barium examination, and subsequently seen on a second colonoscopy [1I]. Loose and Williams [12] performed colonoscopy on ninety-nine patients with rectal bleeding and a normal barium enema; twelve had a benign polyp, two had cancer, and four had a miscellaneous finding. Thus eighty-two of ninety-nine symptomatic patients underwent colonoscopy to confirm a negative barium enema. Colonoscopic biopsy, limited technically by size and depth of bite, may not reflect the true histologic nature of a lesion, particularly in villous adenomas, carcinomas, and inflammatory bowel disease. We have had two patients with infiltrating carcinomas where multiple biopsies showed benign adenoma. Of even greater concern is the problem of the “lost” polyp. In our own series, 5 per cent of all tissue removed was never retrieved and Wolff and Shinya [13] reported that forty-two of 616 polyps (7 per cent) were never found. Similar figures have been noted by others [4]. Approximately 2 per cent of pedunculated polyps contain invasive cancer [14]; how many may be lost along with the opportunity for curative resection? How Inexpensive
Reliable estimates suggest that there are currently 3,000 modern fiberoptic colonoscopes in the United States. The cost of equipping a colonoscopy unit may reach $25,000 [15], bringing the total initial expenditures to as much as $75,000,000. In addition to the cost of setting up a unit, Williams and Teague [6] believed that the average life of an instrument was about 200 examinations and that repairs would equal approximately one third of the original cost, a total investment of approximately
Volume 133, January 1977
TABLE
I
Comparison
of Annual
Costs of Colonic
Polypectomy-Laparotomy
vs Colonoscopy
at Medical
of Vermont
Center
Hospital
Laparotomy Hospital costs Surgeon’s fees Total No. of patients Annual cost
$ 2,744 528 B 3.272 ,22,90:
Colonoscopy $
241 495 736 B 30 $22,080
$10,000 or $50 per examination exclusive of physician’s fees and hospital costs. The hospital expenses for colonoscopic polypectomy are roughly one third that of transabdominal colotomy [9,16,17] and considerably less when outpatient services are used. However, the surgeon’s fee for laparotomy and polypectomy, including pre- and postoperative care and seven to fourteen days in the hospital is listed as sixteen relative value units by Blue Cross and Blue Shield, whereas polypectomy via colonoscopy (0 to 2 days in the hospital) earns a fee of fifteen relative value units. To determine the financial impact of colonoscopy on total health costs, some additional data must be examined. It has been suggested that all endoscopists perform a minimum of 200 diagnostic procedures before attempting polypectomy [13]. The cost of the necessary 600,000 diagnostic colonoscopies (at $400 per procedure) would be $240,000,000. It is particularly discouraging to realize that even the presence of symptoms and/or signs results in a low yield at a high cost. As noted, Loose and Williams [12] performed colonoscopy on ninety-nine patients who presented with rectal bleeding and a negative barium enema-a primary indication for colonoscopy-and found two cancers. The average cost (based on charges at the Medical Center Hospital of Vermont) for each cancer detected was more than $25,000. Recently, Coller, Corman, and Veidenheimer [18] recommended preoperative total colonoscopy for all patients scheduled for colectomy for colorectal cancer to detect a synchronous malignancy. Approximately 100,000 new colorectal cancers will occur in 1976 and 3.6 per cent, or 3,600 patients, will have a second colon primary [19]. The endoscopy expenses alone would equal $40,000,000 or approximately $11,000 per synchronous cancer, assuming none could be detected by sigmoidoscopy or barium enema. Finally, it must be recognized that the indications for colonoscopic polypectomy exceed those for laparotomy. As noted in Table I, we went from an average of seven transabdominal polypectomies per year to thirty in 1975 at an equal annual cost. Thus,
113
Abrams
potential savings per procedure may be overwhelmed by the increasing volume and indications for colonoscopy. How Safe Smith and Nivatvongs [3] found sixteen perforations (1.2 per cent) and thirty-one hemorrhages (2.5 per cent) in a series of 1,244 polypectomies. There were seven perforations (2.6 per cent) and six hemorrhages (2.3 per cent) in 265 patients with sessile polyps. Spencer, Coates, and Anderson [ZO]reported a perforation rate of 1.9 per cent after polypectomy. Perforation has occurred after removal of a 3 mm polyp [21] and fulgeration of a 2 mm polyp [22]. Serosal tears, colonic perforations, and even perforation of the terminal ileum has been described during diagnostic colonoscopy. The true complication rate for colonoscopy/polypectomy is unknown. The data we have come from published series and a few surveys, most of which fall far short of a 100 per cent response rate. It has been suggested that responders generally include those with the best results. Although there are no collective figures available regarding mortality, deaths are known. Summary
The fiberoptic colonoscope represents a significant diagnostic and therapeutic achievement in the management of colonic disease. Despite some claims to the contrary, it is not universally simple, rapid, thorough, reliable, and inexpensive or without morbidity and mortality. Until these goals are achieved through future refinements, modifications, and new instrumentation, it is strongly recommended that the indications, for colonoscopy be sharply defined to include the following: (1) Investigation of colonic lesions seen on two successive or one air-contrast barium enema examination. Pedunculated polyps less than 1 cm in diameter can be observed. (2) Preoperative investigation of patients with demonstrable colorectal cancer (to rule out synchronous lesions). (3) Postoperative evaluation after colectomy for carcinoma where the anastomosis is above the reach of a sigmoidoscope. (4) Evaluation of patients with inflammatory bowel disease of the colon. .(5) Investigation of persistent occult or gross rectal bleeding when anoscopy, sigmoidoscopy, and barium studies are negative. Patients who represent prohibitive surgical risks because of age or systemic disease should not be considered as candidates for colonoscopy without compelling indications. The decision to biopsy rather
114
than remove or fulgerate a detected lesion must be left to the judgment and experience of the endoscopist. The morbidity that follows polypectomy or fulgeration of sessile lesions less than 0.5 cm in diameter probably exceeds the incidence of carcinoma. References I. Rubin CE, Auth D, Barrett B, et al: Work group X. Instruments for diagnosis, investigation and treatment of digestive diseases. Gastroenterology 69: 115 1, 1975. 2. Ovetilt BF: Colonoscopy. Review.Gastroentero/ogy 66: 1306, 1975. 3. Smith. LE. Nivatvongs S: Complication in colonoscopy. Dis Colon Rectum 16: 214, 1975. 4. Sugarbaker PH, Vineyard GC, Lewicki AM et al: Colonoscopy in the management of diseases of the colon and rectum. Surg Gynecol Obstet 139: 341, 1974. 5. Farringer GL Jr: Discussion of [ 91. 6. Williams C. Teaaue R: Colonoscoov. Gut 14: 990. 1973. 7. Dagrodi Ai, Alaama A, Ruiz R: Clinical experiences with colonoscopy. Am J Gastroenterol63: 406, 1975. a. Wolff WI, Shinya H: A new approach to colonic polyps. Ann Surg 178: 367, 1973. 9. Knutson CO, Schrock LG. Polk HC Jr: Polypoid lesions of the proximal colon: comparison of experiences with the removal at laparotomy and by colonoscopy. Ann Surg 179: 657, 1974. 10. Williams C: Evaluation of the colonoscopic examination. Dis Colon Rectum 16: 366, 1975. Il. Laufer I, Smith NCW, Mullens JE: The radiological demonstration of colorectal polyps undetected by endoscopy. Gastroenterology 70: 167. 1976. 12. Loose HWC, Williams CB: Barium enema vs. colonoscopy. Proc R Sot Med 67: 1033, 1974. 13. Wolff WI, Shinya H: Modern endoscopy of the alimentary tract. Curr Probl Surg I, January 1974. 14. Shatney CH, Lober PH. Gilbertsen VA, Sosin H: The treatment of pedunculated adenomatous colorectal polyps with focal cancer. Surg Gynecol Obstet 139: 645, 1974. 15. Sivak MV Jr, Sullivan BH Jr, Rankin GB: Colonoscopy: the report of 644 cases and review of the literature. Am J Surg 126: 351,1974. 16. Bloom B.S. Goldhaber SZ, Sugarbaker PH. O’Conner NE: Fiberoptics: morbidity and cost. Editorial. N Engl J Med 266: 368.1973. 17. Goldhaber SZ, Bloom BS, Sugarbaker PH. D’Conner NE: Effects of the fiberoptic laparoscope and colonoscope on morbidity and cost. Ann Surg 179: 160, 1974. ia. Coller JA, Corman ML, Veidenheimer MC: Colonic polypoid disease: need for total colonoscopy. Am J Surg 13 1: 490, 1976. 19. Winawer SJ, Sherlock P, Schottenfeld D, Miller DG: Screening for colon cancer. Gastroenterology 70: 763. 1976. 20. Spencer RJ, Coates HL, Anderson MJ Jr: Colonoscopic polypectomies. A&y0 Clin Proc 49: 40. 1974. 21. Meyers MA, Ghahremani GG: Complications of fiberoptic endoscopy. II. Colonoscopy. Radiology 115: 301, 1975. 22. Nivatvongs S, Goldberg SM: The results of 100 consecutive polypectomies with a fiberoptic colonoscope. Am J Surg 128: 347.1974.
Discussion Henry Colcher (Boston, MA): I read Doctor Abram’s manuscript and I would like to answer each point as rapidly as I can.
The American Journalof Surgery
Colonoscopy
Easy? No, it is not an easy procedure. We do accept the fact that there are failures and that we cannot reach the cecum in all patients. Duration? It is not a rapid procedure. We assign 1 hour for each endoscopic examination. Safe? Yes and no. It is not a safe procedure in the hands of people who have not learned how to do it. In 1973, when I was President of the ASGE, we conducted a survey and were shocked by the number of perforations and other complications. We realized that these were often seen by people who did not have enough training. With the development of training programs the procedure has definitely become safe and the benefits justify the morbidity that it involves. Reliable? Yes, if we consider again a well trained endoscopist who knows that there are blind areas in a certain examination and repeats the examination six months or a year later. Is it worth it? Yes. The results of the study conducted by Doctors Coller, Corman, and Veidenheimer [18] show that most of the polyps located on barium enema study were on the left side and a few were on the right side. Of 146 patients thirty-six had none; they were spared surgery and 110 presented 120 polyps. In addition, however, there were numerous unsuspected lesions on the right side. What does it represent in terms of disease? At the suspected site there were 103 benign lesions but also eight polyps with invasive carcinoma and nine cancers (polypoid or ulcerating adenocarcinoma) that were not suspected on x-ray examinations. At the unsuspected sites, 120 benign polyps, two polypoid adenomas with invasive carcinoma, and four adenccarcinomas-thus, a total of twenty-three invasive carcinomas-were found in twenty-two patients. Pedunculated polyps less than 1 cm in diameter may have invasive carcinoma. Total colonosopy is worthwhile because we find a large number of cancers that can be removed at a very early phase either by polypectomy or by surgery when it is needed. We are left now with two more questions concerning cost. I look at cost in terms of effort on the physician’s part. I believe it is worth it. Cost in terms of money: I believe it can be justified when we offer the patient a polypectomy or a resection for cancer at an earlier phase. Forget the figures and the money. If we can improve the results in colon cancer by polypectomy and by follow-up for synchronous and metachronous lesions, it is worth it. William I. Woolf (New York, NY): One has to go back to the days of laparotomy and colotomy and colectomy to realize what the alternatives are. I will briefly mention two appropriate cases. A fifty-five year old lady was referred to us with a polyp of the transverse colon. She had had a polyp in the descending colon removed fifteen months previously. This was followed by a wound infection, by repeated admissions for intestinal obstruction, and by a bout of homologous
Volume 133, January 1977
serum jaundice. The transverse colonic polyp was removed endoscopically without difficulty. A small polyp of the descending colon was found in an elderly obese businessman in 1971. He was asked to return in six months, at which time the polyp had not changed in size. Removal was suggested but not pressed. We saw him in 1975. Obviously the lesion was larger. The lesion was obviously a cancer, and it was removed. Colostomy was necessary as well to protect the anastomosis. It is appropriate that we know what the complication rate is. But, as we have insisted all along, it is equally important to know who is doing each procedure and under what circumstance. It is important that physicians performing these procedures are trained, competent, and have a good basic knowledge of gastrointestinal disease. At the Beth Israel Medical Center we have performed more than 10,000 diagnostic colonoscopies now and have removed more than 2,500 colonic polyps without a single mortality. I do take issue with the statement that cancer in polyps under 1 cm in diameter does not occur. It does occur. It occurs frequently, and I am referring to invasive cancer, not carcinoma in-situ. Gerald 0. Strauch (Stamford CT): I work in a community hospital. Our colonoscopies are performed by surgeons, and I have run the Surgical Audit Committee since colonoscopy came into existence in the hospital. One of the problems that we have seen is the difficulty with the recovery of polyps which have been snipped off at the time of colonoscopy. Hard data in this regard seem to be very difficult to gather. Doctor Abrams quoted a figure of 7 per cent. In our hospital that rate is substantially higher. Any polyp that is worth removing is certainly worth looking at under a microscope, and failure to be able to do so represents a serious shortcoming. Jerome S. Abrams (closing): In answer to Doctor Colcher’s points, I would remind you again that, based on autopsy data, approximately 15 to 45 per cent or almost one third of patients more than forty-five years old have at least one polyp, and therefore we should not be surprised that we can randomly put the scope in anybody and find one or more polyps. Secondly, I am impressed by the high incidence of carcinoma in Doctor Colcher’s particular group, and I have absolutely no explanation for it, since as best I can tell from a very large series of cases the incidence of carcinoma has been significantly lower. Finally, although I am impressed philosophically with the idea that you cannot place dollar values on cancers, I am also impressed that there is fear in this country and the world about the overall cost of medical care. In answer to Doctor Wolff, I think he may have misinterpreted my statements. We are certainly not questioning the value of colonoscopy, and we are indebted to his unit for pioneering this work. But I do wonder how we can control an instrument that can be used in a doctor’s office and is available at a reasonably low cost.
115
