Repeat colonoscopy after endoscopic polypectomy

Repeat Colonoscopy after Endoscopic Polypectomy R O B E R T HOLTZMAN, M . D . , JEAN-BERNARD POULARD, M.D., SIMMY BANK, M.D., LEROY R. LEVIN, M.D., GEORGE W. FLINT, M.D., RICHARD J. STRAUSS, M.D., IRVING B. MARGOLIS, M . D .

Holtzman R, Poulard J-B, Bank S, Levin LR, Flint GW, Strauss RJ, Margolis IB. Repeat colonoscopy alter endoscopic polypectomy. Dis Colon Rectum 1987;30:185-188. T h e records of all patients undergoing endoscopic polypectomy

From the Department o/ Surgery, Queens Hospital Center AHiliation o] the Long Island Jewish Medical Center, Jamaica, New York, and the State University ol New York at Stony Brook, New York

between December 1979 and December 1982 were reviewed. One hundred seventy-two patients underwent colonoscopic polypectomy in the absence of carcinoma or inflammatory bowel disease. Of these, the polyp could not be retrieved in 4, and 19 were lost to follow-up. One hundred forty-nine patients underwent subsequent endoscopy from one to four years after the initial polylx~ctomy. Seventy-five (50.3 percent) of the patients developed new polyps. Although 61 of the 75 patients with new polyps were identified in the first two years, new polyps were noted throughout all four years. The presence of multiple polyps on the initial examination was statistically significant in predicting new polyps. The age and sex of the patients, size of the polyps, and the presence of atypia did not identify patients at higher risk for

diagnosis is desirable. Although it is known from previous studies that these patients are at risk of developing new polyps, the time required for the growth of new polyps and, thus, the optimal frequency of colonoscopic examinations, has not been established. Two recent studies, one from Denmark 6 and one from New York, 7 found that the rate of development of new polyps in patients after polypectomy was similar to rates found in the precolonoscopic period studies. This study analyzes another group of patients retrospectively in order to determine the incidence of new polyps, and to identify the time intervals at which they developed and their probability for malignant degeneration.

new polyps. T h e data indicate that new polyps are more likely to develop in patients w h o had a previous polyp. It would appear that

annual examinations should be performed until two successive examinations are negative. Following a second negative examination, reexamination at two- or three-year intervals, unless symptomatic, would appear to be adequate. [Key words: Endoscopy, Polypectomy, Colonoscopy]

Methods COLONOSCOPY HAS INCREASED o u r ability to diagnose the presence of a polyp and, with simple polypectomy, the patient is often adequately treated without the need for laparotomy.1 Several long-term studies of colonic polyps in the p recolonoscopic period have indicated an increased rate for the development of new polyps in patients already diagnosed as having a polyp3-5 As it is simpler to remove a polyp of smaller size, early

All outpatient colonoscopies (1184) performed at Long Island Jewish Medical Center from December 1979 to December 1982 were reviewed. Three hundred fifty-two patients had colonoscopic polypectomies. Of these, 137 patients were excluded due to a previous history of colonic carcinoma or polyps incidentally removed at the time of colonoscopy for the concurrent diagnosis of carcinoma of the colon or rectum. Another 43 patients were excluded due to a diagnosis of inflammatory bowel disease or familial polyposis, or if the final pathology of tissue submitted did not reveal an adenomatous polyp (i.e., mucosal excrescences, hyperplastic polyp, etc.). In four patients, the polyp was not retrieved by the endo-

Read at the meeting of the American Society of Colon and Rectal Surgeons, Houston, Texas, May 11 to 15, 1986. Address reprint requests and correspondence to Dr. Margolis: Director, Department of Surgery, Queens Hospital Center, 82-68 164 Street, Jamacia, New York 11432.

185

186

Dis. Col. ~r Rect. March 1987

HOLTZMANET AL.

scopist. There were 19 patients who were followed outside the institution for whom follow-up data could not be obtained. The remaining 149 underwent repeat colonoscopy from one to four years after the initial examination and formed the data base of this study. Only 12 patients were symptomatic at the time of repeat endoscopy. The endoscopies were performed under intravenous sedation in an outpatient suite with the patients having prepped themselves with a standard two-day bowel prep. Examinations were performed with an Olympus CF type 1T 10 colonoscope using electrocautery, snare, and biopsy forceps as needed. Many patients with multiple polyps underwent repeat colonoscopy soon after the initial examination (usually within six months) to guarantee a colon free of polyps. If a polyp was found on early second examination, it was not clear whether this was a missed or a new polyp. For purposes of analyzing the data, a polyp found on endoscopy prior to ten months after a previous endoscopy was classified as a missed polyp. The time of entry into the study of such patients was set after the repeat endoscopy when the operator believed the colon to be polyp free. The time interval between initial colonoscopy and repeat examination varied from one to four years and was left to the discretion of the endoscopist. Several factors, including the age and sex of the patients, size of the polyps, the presence of atypia, and the number of polyps found on initial, examination, were collected to determine their ability to predict which patients would develop new polyps. Results One hundred forty-nine patients underwent a total of 389 colonoscopic examinations. Seventy-five (50.3 percent) patients were found to have developed a new polyp, while the remaining 74 (49.6 percent) patients had nor30,

mal follow-up examinations. Of the 75 patients with new polyps, 40 underwent subsequent endoscopic examinations. Twenty-one (52 percent) of these 40 had another new polyp (third polyp in total). Four patients had new polyps on each of four consecutive year colonoscopies. Figure 1 depicts the year of follow-up in which the new polyps were discovered. The majority of patients who developed new polyps, 61 of 75, were identified in the first two years after the initial polypectomy. All but four patients with new polyps were discovered by the end of the third year of follow-up and despite similar numbers of colonoscopies, 76, 68, 53, in the first, second, and third year of follow-up, significantly smaller numbers of patients were found to have new polyps; 39, 22, and 10, respectively. Of the 74 patients with normal follow-up examinations, only 20 underwent subsequent examination. This is indicative of the endoscopist's willingness to prolong the interval between examinations after one normal colonoscopy. Seven patients developed a new polyp after having a normal repeat examination (Table 1). The new polyps in patients 2, 6, and 7 were conspicuously close to the areas where the original polyps were found. When comparing the site of all new polyps to the site of the index polyp, a similar trend was noted with 34 percent being in the same colonic segment (Table 2). This suggests that some new polyps may be recurrent rather than arising de novo. Several variables were examined for their ability to predict which patients would develop new polyps. Only the presence of multiple polyps on the initial examination was statistically significant in predicting new polyps (Table 3). The age and sex of the patients, size of the polyp, and the presence of atypia did not identify patients at risk for new polyps. During the follow-up period, two patients developed

'76 68

70.

60,

53

~ T o t a l # of

50

# of pts.

40, 30.

ZO

tO

o!

i !

1

|

endoscopies [] tst new polyp I []2nd new polyp t IB3rd new polyp J

2

3

Year of Follow-Up

FIG. 1. Follow-upafter indexpolypectomy.

Volume30

187

COLONOSCOPY AFTER POLYPECTOMY

Number 3

TABLE 1. New Polyps Developing after a Normal Colonoscopic Examination Patient

Normal Examination

Age

Index Polypectomy

1

68

6/80:

2

62

3 crn (L) colon-VA 2.3 cm (L) colon-VA 2.5 (L) colon-VA, CIS 12/80:1.2 cm sigmoid-VTA

New Polyp

3

62

1/83:

4

52

5

37

12/79:1.3 cm sigmoid-VTA

7/82

7 / 8 4 : 0 . 5 cm (L) colon-AP

6

66

1/80: 9/80:

3.0 cm rectum-VA 0.6 cm rectum-VA

4/83

9 / 8 4 : 0 . 7 cm rectum-VA

7

58

8/82:

2.0 cm hepatic flexure-VTA

9/83

8/81

8 / 8 4 : 1 . 5 cm hepatic-VA with invasive Ca

1/83

3 / 8 5 : 0 . 8 cm sigmoid-VTA, CIS

2.4 cm sigmoid-AP

11/83

9 / 8 5 : 0 . 5 cm hepatic flexure-AP

12/79:1.4 cm rectum-TA 0.8 cm transverse colon-TA

5/81

2 / 8 3 : 0 . 6 cm sigmoid-VA

10/84:1.3 cm cecum-TA 0.7 cm hepatic flexure-VTA

VA =-- villous adenoma; V T A = villotubular adenoma; A P = a d e n o m a t o u s polyp; T A = tubular adenoma; CIS = carcinoma in situ.

invasive carcinoma. T h e first (patient I in Table 1) refused colonoscopy between 1981 and 1984 because he was asymptomatic. When he developed guaiac-positive stool and agreed to endoscopy, he was found to have a 1.5 o n villous adenoma of the hepatic flexure with carcinoma invading through the muscularis mucosa and the stalk margins free of tumor. A Dukes' A lesion was resected. T h e other patient, a 57-year-old man with guaiacpositive stool, was thought to have a cecal polyp on barium enema. On endoscopy, a 1 cm polyp at 46 cm was found but a cecal polyp was not seen. Seventeen months late, after having guaiac-positive stool again, he was endoscoped and found to have a cecal tumor. In retrospect, this probably was missed on the initial examination. T w o patients developed polyps with carcinoma in situ during the study period. One is listed as patient 2 in Table 1. T h e other is a 60-year-old man who initially had a 1.5 a n villotubular adenoma at 7 cm. Routine sigmoidoscopy in the office to 25 cm after six months was reported as normal. T h e following year, a 0.7 cm sessile villous adenoma at 8 cm with carcinoma in situ was found. Subsequently he has had three colonoscopies at yearly intervals, all of which have been normal. There were no mortalities from colonoscopy during TABLE 2. Litigation o] New vs. Original Polyps Site of New Polyps

Site of index polyps Right Transverse Left Sigmoid Rectum

Right

Transverse

Left

Sigrnoid

Rectum

4 3 10 7 4

2 3 6 -

3 3 6 7 3

1 2 5 17 5

0 1 3 5

the three-year study period or during follow-up. In total, of the 1184 colonoscopies reviewed, there were five complications. These included one perforation and one episode of bleeding, both requiring surgery. Three episodes of minor bleeding required endoscopic coagulation and admission to the hospital. Discussion Most clinicians agree that an adenoma can undergo malignant changes and much has been published supporting this theory, s-n T h e chance for malignancy increases exponentially with size. Morson s reported malignancy rates of 1.3, 9.5, and 46 percent for adenomas less than I cm, 1 to 2 cm, and over 2 cm, respectively. Therefore, an aggressive approach to cleanse the colon of all polyps at the initial endoscopy and close follow-up to remove polyps while they are still small are recommended. Studies in the precolonoscopic period have demonstrated new polyp formation rates from 21 to 41 TABLE 3. Risk Factors/or Development o/New Polyps

Median age Sex Atypia Size of index polyp Under 1 cm 1 to 2 cm Over 2 cm N u m b e r of polyps o n initial exam:* one two three *X ~ = 9.696; P < 0.01.

Normal Follow-up (n = 74)

New Polyps o n Follow-up (n = 75)

60.2 (23-78) M:F, 48:26 10

61.5 (30-84) M:F, 47:28 11

17 42 15

23 37 15

53 13 8

35 23 17

188

Dis. Col. & Rect.

HOLTZMAN, ET AL.

TABLE4. Size o[ Polyps (Index vs. New) Size Under 1 cm 1 to 2 cm Over 2 cm

Index Polypectomy (n = t49)

New Polyps (n = 100)

40 67 42

66 24 10

X~ = 38.29; P < 0.01.

percent. Henry et al. z reported a 30 percent new polyp formation rate in patients who were followed for 20 years with semiannual sigmoidoscopy and barium-enema examinations. T h e majority of patients developed new polyps within 48 months of the original examination. This is 16 times the incidence of polyps in the population as a whole. After the fourth year, new polyps occurred at a rate only slightly higher than that of the normal population. Four recent studies have used colonoscopy for initial examination and follow-up.6, 7,12,x~ These demonstrate new polyp formation rates from 30 to 61 percent and reflect a greater accuracy in the diagnosis of polyps with the colonoscope. Fowler and Hedberg ~2found 60 percent of patients had developed a new polyp, with 50 percent of them occurring in the first 12 months. Kronborg et al. 6 reported a prospective randomized study of two groups of polypectomy patients who were to undergo repeat examinations at either 6-month or 24m o n t h intervals for a total of four years. T h e y found similar rates of new polyp formation in each group, 5 percent at each of the 6-month intervals in the first group and 20 percent in the group examined at two years. Neugut eta]. 7 report on 118 patients followed over a 45-month period found a 30 percent new polyp formation rate with the majority occurring in the first year. T h e present study showed an overall 51 percent new polyp formation rate over a 53-month follow-up period. Similar to previous studies, the majority of polyps were found in the first year (ten to 23 months) of follow-up and 81 percent developed the first new polyp within two years (up to 35 months) of the initial examination. An explanation that these represent missed polyps from the initial examination is not likely. T h e smaller sizes of the new polyps as compared with the sizes of polyps on index polypectomy (Table 4) suggest that these were not overlooked as they would have had time to grow to larger sizes and would be comparable to the sizes found on index polypectomy. T h e important question remaining is to find the appropriate balance between time intervals for colonoscopy and m a x i m u m yield of colonic polyps without chancing malignant degeneration, in Of 149 patients followed for over four years, 51 percent developed new polyps. From this information, it appears that patients who have polyps have an increased incidence of having

March 1987

another polyp during their lifetime. Further, these patients can be broken into two groups. Of those patients w h o had a second polyp, 21 (28 percent) of 75 patients went on to develop another new polyp. In comparison, only 7 of 74 patients (9.5 percent) having a normal second examination went on to develop a new polyp. Those patients who were found to have a second polyp were more likely to develop a third polyp than those who had a normal second colonoscopic examination. These patients should be followed closely and have yearly colonoscopic examinations for the first three years after the most recent polypectomy during which recurrence rates are at their highest. Of the patients who had normal second colonoscopic examinations, six of seven patients who went on to develop a new polyp did so within two years. Therefore, it is important to follow this group of patients closely also; however, they will not require as frequent examinations as the other group. A reasonable suggestion would be to perform yearly colonoscopic examinations until two consecutive examinations are negative. At this point, reexamination at two-to-three-year intervals, unless symptomatic, appear to be adequate. With this policy, the incidence of future polyp formation may be only slightly higher than that of the general population.

References 1. Shinya H, Wolff WF. Colonoscopic polypectomy: technique and safety. Hosp Pract [Off] 1975;34:71. 2. Henry LG, Condon RE, Schulte WJ, Aprahamian C, DeCosse JJ. Risk of recurrence of colon polyps. Ann Surg 1975;182:511-5. 3. Brahme F, Ekelund GR, Nord~n JG, Wenckert A. Metachronous colorectal polyps: comparison of development of coiorectal polyps and carcinomas in persons with and without histories of polyps. Dis Colon Rectum 1974;17:166-71. 4. Prager ED, Swinton NW, Young JL, Veidenheimer MC, Corman ML. Follow-up study of patients with benign mucosal polyps discovered by proctosigmoidoscopy. Dis Colon Rectum 1974; 17:322-4. 5. Weakley FL, Swinton NW. Follow-up study of patients with benign mucosal polyps of the rectum. Dis Colon Rectum 1962;5:345-8. 6. Kronborg O, Hage E, Adamsen S, Deichgraeber E. Follow-up after colorectal polypectomy. Scand J Gastroentero11983;18:1089-93. 7. Neugut AI, Johnsen CM, Forde KA, Treat MR. Recurrence rates for colorectal polyps. Cancer 1985;55:1586-9. 8. Morson B. The polyp-cancer sequence in the large bowel. Proc R Soc Med 1974;67:451-7. 9. Hermanek P, Fr/thmorgen P, Guggenmoos-Holzmann I, Ahendoff A, Matek W. The malignant potential of colorectal polyps--a new statistical approach. Endoscopy 1983;15:16-20. 10. Kozuka S. Premalignancy of the mucosal polyp in the large intestine. I. Histologic gradation of the polyp on the basis of epithelial pseudostratification and glandular branching. Dis Colon Rectum 1975;18:483-9. 11. Kozuka S, Taki T, Kubota K, Yokoyama Y. Simultaneous transition of multiple polyps into carcinoma in the large intestine. Dis Colon Rectum 1976;19:655-9. 12. Fowler DL, Hedberg SE. Followup colonoscopy after polypectomy (abstr). Gastrointest Endosc 1980;26:67. 13. Waye JD, Braunfeld S. Surveillance intervals after colonoscopic polypectomy. Endoscopy 1982;14:79-81.