Diseases of the COLON ¢Y

VOL. 11 july-August, 1968 No. 4

Anorectal :

Current Concepts * .RAYMOND J. JACKMAN, N[.D.'~ Rochester, Minnesota

~)IA(;NOSIS and treatment of anorectal and secondary opening, since the principles fistulas at times presents the simplest and of dealing with such a lesion have been at other times the most challenging prob- well established. lems encountered in surgery, is this sinus actually an anorectal ? Or does it Etiology originate elsewhere, such as in a region of Anorectal fistulas comprise 85 per cent sigmoidal or ileocolitis? Or of sinuses opening onto the perineum. The does it connect with an infected precoccyg- other 15 per cent result from hidradenitis eal cyst? What keeps the sinus or fistula suppurativa, , and other active when the primary source has healed less common causes. There are many types over spontaneously or has been erased by of anorectal fstulas, but the one feature a previous operation? In what situations that they have in common is that they might the surgeon anticipate fecal inconti- originated inside the anus or rectum. (I nence after fistulectomy, and what can be use the past tense because the anorectal done about it? site of origin, the primary lesion, may have The purpose of this presentation is to healed over, but at one time it was present.) attempt to answer these and similar ques- All anorectal fistulas and most perianal tions that, in my opinion, have been inade- sinuses originate as . Frequently, quately answered in medical literature. I but not always, the starts as infec- shall not dwell on the simple uncompli- tion in a crypt at the mucocutaneous junc- cated fistula with a definite primary source doz~. Fistu!as also may originate in a break in the continuity of the anal skin, such as =~ Presidential Address at the meeting of the Amer- ican Proctologic Society, Denver, Colorado, June 10 a fissure, or in a break in the intestinal to 13, 1968. mucosa, such as an ulcer occurring in ile- j-Mayo Clinic and Mayo Foundation: Section of Proctology. itis, ileocolitis, or ulcerative . 247 248 JACKMAN

Differential Diagnosis on unusual or less common causes of peri- Sometimes determination of the exact anal sinuses, but the point I wish to make nature and course of the fistulous tract has here is that a perianal sinus, ostensibly an anal fistula, for which the patient previ- to be made when the patient is under anes- ously has been operated on, should alert thesia. The exploring finger is probably of more value than any other device in one immediately to some unusual source- determining the source and course of the sometimes an infected precoccygeal cyst fistula. Frequently, the primary source can and, more frequently, regional o1 be palpated with the tip of the index finger sigmoidal diverticulitis. as a depression or an area of induration. Association of Perianal Abscesses and Bidigital examination, not uncommonly, Fistulas with Other Diseaes: The tendency will reveal a cord-like channel of indura- toward fistulization, not only anorectal but tion leading to the secondary opening. enteroenteric and so on, in ileocolitis has Anoscopic examination may reveal a drop been noted by many authors. Rawls and of exuding from the primary source or co-workers,5 in a study of 1,000 patients a tuft of granulation tissue. The use of with anorectal fistulas, found that 17 per the probe, until the time of operation, is cent of them had an underlying or asso- probably the least effective means of deter- ciated systemic disease. The three disorders mining the nature of the fistula and should occurring most frequently were chronic not be resorted to if it causes the patient , diabetes mellitus, and discomfort. The differential diagnosis of regional ileitis, in a study of 114 patients draining sinuses in the nether parts has with regional , Smith and 1 s have been the subject of a previous presentation 2 found that 1 per cent either had anal ab-

•

~AYO CLINIC ..

Fro. 1. a and b, Conversion of fistulous channel or tunnel into open ditch. ANORECTAL FISTULAS 249

scess or fistula at the time of examination or gave a history of having had such a S e lesion in the preceding three years. Seven oF 8 _ _. per cent of the patients in this group had a presenting complaint of anal fistula or abscess. The abdominal or gastrointestinal symptoms were so mild that the patients ignored them. Any young adult whose pre- senting complaint is o[ anal abscess or ano- rectal fistula should be investigated [or an underlying systemic disease, particularly [or U enteritis, (Crohn's disease) or ulcerative colitis. • Dd_:rn__]ing__lo ..... Association of Anal Fistula with Tuber- culosis: Not too many years ago, the stu- Fro. 2. Traction applied to site of secondary dent was taught that if a patient was found opening and consequent retraction at primary to have an anorectal fistula, one should source. This procedure usually gives a clue to the location of the primary lesion. look for a focus of tuberculosis elsewhere in the body, either pulmonary or gastro- intestinal. This teaching no longer holds true. In fact,, many patients previously Association of Anal Fistula with Cancer: considered ito have intestinal tuberculosis Cancer developing in or near chronic drain- probably had gramdomatous ileocolitis. ing sinuses, although uncommon, has been Since the advent of chemotherapeutic drugs reported many times. In such instances the and --para-aminosalicylic acid, fistulas usually will have been present for isonicotinic acid, streptomycin, and so on- many years, it is generally contended that the incidence of tuberculosis, while still a irritation plays an important part in the public health problem, has declined mark- development of certain cancers. Perhaps edly. This is particularly true of anorectal the carcinogenic mechanism in the case of tuberculosis and tuberculous fistula. In an anal fistula is the long-continued drainage. article published in 1939, Buie, Smith, and Kline and co-workers4 reported five cases Ia reported the results of an intensive study of long-term anal fistula in which a carci- of 600 patients with anal fistula. O[ these noma developed later. They collected 20 patients, 11.5 per cent were found to have other similar cases from medical literature, tuberculosis some place in the body. Using and concluded that the difficuhy in diag- guinea pig inoculation and histologic study nosis of this complication of anal fistula of the excised fistulous tracts, we deter- resuhed mainly from the insidious nature mined that about 8 per cent o£ anal fistulas of the tumor and the masking effect of the were tuberculous, and that in most of these symptoms of the fistula on the early symp- cases there was an active or healed pulmo- toms of cancer. Or the 25 tumors, 44 per nary focus. I am certain that the per- cent were mucus-producing adenocarci- centage of tuberculous fistulas would be nomas of low-grade malignancy. In one of much iess ~oday. During the past two their five cases from this ciinic, the paden~ years ] have diagnosed definitely only one had had a fistula-in-ano for 47 years before case of tuberculous fistula. "The diagnosis an adenocarcinoma was detected. A radical was confirmed by guinea pig inoculation surgical operation is indicated for carci- and histopathologic study. noma superimposed on a fistula. 250 JACKMAN

io/~ ir~to incLl

".. , :/ ., " .'.';i' ,,i::

. ,z:. ,: /'v " .r' \,7'?'¢< ~ " .'<" " "", ~ ' ~" ?.5",

'~ aC,_ _-

lhc. 3. Operation for rectovaginal listula. Undercutting rectovaginaI scptum mobilizes it sufficiently so that it can bc pulled down and sutured to the inner border of the external sphincter after d~min has been placed beneath the septum ani.l out through the fistulous tract. (Fr(;m Jackman, R. J.: Rectovaginal and anovaginal [istulas: A surgical pl'oceduxe for treatment of certain types. J. Iowa Med. Soc. 42:435-4-10 [Sept.] 1952. With permission of the Iowa Medical Society.)

General Aspects of Treatment mary out through the secondary opening; The use of antibiotics in the treatment the overlying tissue is incised, converting of anorectal ab:~cesses and fistulas is not the fistulous channel or tunnel into an open advisable because usually incision and ditch; and debridement o1! the wound mar- gins is accomplished (Fig. 1). drainage has to be carried out anyway. Ahhough Escherichia coli, the usual offend- Obviously there are many deviations ing organism, is vulnerable to tetracycline, trom the "straightforward" fistula, and com- administration of this will not binations of two or more fistulas may resolve the problem and operation will be occnr, but the basic aim of surgical treat- necessary regardless. If the abscess is the ment in all cases is to establish adequate initial sign of a true fistula, which it usu- drainage. I)uring the operative procedure, ally is, a primary source inside the anus or the surgeon must ask himself, "vVill the rectum continues to exist and is not affected drainage be satisfactory when the tone of by any antibiotic. the muscle returns or will there be over- The basic principle involved in all oper- hanging ledges or pockets that will inter- ations for anorectal ab~cess or fistula is that fere with drainage? if the latter appears satisfactory drainage must be established. more likely, steps must be taken to correct On the [ace of it this sounds simple, but this by debriding the margins more exten- the operations for fistula may be extremely sively or by incising muscle. d~bficult and the convalescence prolonged. Finding the Primary Source of the ~.~s- On the other hand, such operations may tula: Sometimes this may be difficult. H" be very simple. If the problem is uncompli- the lesion is a fistula, there is, or at one cated--a primary opening within the anal time was, a primary or internal opening. canal and a secondary opening in the peri- ][ the patient has had a previous operation, anal skin--a probe is passed from the pri- the landmarks may be distorted and the ANORECTAL FISTULAS 251

source of the fistula difficult to find. A to incise the muscle at the site where you fistula may remain active even though the think the primary source should be. In primary source has healed over, being per- my experience, this is particularly true for petuated by the inflammatory granulation fistulas presumed to originate posteriorly, tissue in the channel or abscess cavity. less true for anterior and lateral fistulas. Use o[ dye: Many surgeons use dye to I do not have a completely adequate ex- determine the course of the channel and planation of why this is necessary; but I tO find the primary opening. My own ex- do know that time and again when I have perience with this method has been quite not done it, patients have had further unsatisfactory; furthermore, in nay opinion, trouble. Perhaps the intact anal muscle it is unnecessary, since the contrast between prevents adequate drainage during healing. the channel or tract itself with its infected granulating wall and the fresh adjacent nor- Specific Surgical Problems mal tissue is quite striking. The use of dye Fistulotomy Versus Fistulectomy: In my disguises this picture and tends to obscure opinion a fistulectomy is preferable to a the whole fiekl. fistulotomy for the following reasons: 1) Bidigital examination: "vVhile tbe pa- ~vVhen the fistulous tract is excised (fistulec- tient's muscles are relaxed during anes- tomy), one occasionally will discover a hid- thesia, bidigital palpation of the regions of den channel or ollshoot from the main tract, induration probably is the best method which otherwise would not have been for finding the source and course of the found. A spot of granulation tissue remain- tract. ing along the course of the excised fistulous Traction on the secondary opening: This tract should alert one to the possibility of procedure (Fig. 2) will sometimes cause an offshoot or abscess cavity. 2) By doing retraction or dimpling at the primary a fistulectomy, one removes the occasionally- source of the fistula. epithelized tract or the infected granulating Probing the tract: Ideally, one should tract, leaving a fresh-wound sulcus that attempt to probe from the primary opening seems to heal faster and to form a narrower out through the secondary opening. How- scar. When a wide block or scar tissue ever, probing from the secondary opening forms, the anal muscle becomes mechani- sometimes will reveal the internal or pri- cally less efficient and there may be "leak- mary source. One may need to bend and age" or difficulty in holding gas. shape the malleable probe to accommodate I realize it is sometimes poor judgment it to the course of the channel. even to attempt a fistulectomy. In a very What Does the Surgeon Do When It deep-lying channel, not only may excision Seems Impossible to Find a Primary of the scarred granulating tract be impos- Source? The patient may or may not have sible, but attempted excision may be muti- been operated on previously, the anus is lating and may create smooth and scarred, landmarks are absent, or an inefficient anus. If a fistulectomy and the probe can be inserted into the cannot be done, or if it is thought inad- sinus up to or near the site where you visable, granulations in the channel should think the primary source should be. As- be scraped away with the blade ok the suming that usuai causes for the sinus have scalpel or wiped away with gauze. been ruled out, is it adequate to lay all the tissue overlying the probe open, curette the Simultaneous Abscess Drainage and F~s- tract, and pack it? Although this may tulotomy: Are drainage of the abscess and work in some instances, I think it is best a definitive operation for the resulting fis- 252 JACKMAN tula feasible simultaneously? The answer the prone "jackknife" position--that is, with varies with the circumstances. the hips slightly elevated. 2) YVith a Smith If the abscess is relatively superficial and operating anoscope in place, a curved in- the primary source readily discernible, cision is made in the anterior anal wall, just distal or external to the source or anal drainage and fistulotomy may be carried out opening of the fistula° 3) The rectal mu- at the same time. This has obvious advan- cosa, , and muscularis (about tages when it can be done. Only one hos- half of the thickness of the rectovaginal pitalization period and one anesthetic are septum) are undercut for 2 to 3 cm proxi- necessary, and convalescence is shortened. mal to the fistula (Fig. 3). 4) The anal If the abscess is deeply situated, very skin (with , if any) overlying large, or of questionable extent, it is better the anterior anal wall is excised, exposing initially to establish drainage and subse- the anal muscles. 5) One end of a small quently to perform a delayed fistulectomy rubber drain is sutured beneath the under- or fistulotomy. It has been my experience cut septum with catgut, the free end ex- that when t have tried the complete proce- tending through the fistulous opening into dure under such circumstances, I have over- the vagina (Fig. 3). 6) About a third of looked a locu/ation of the abscess that would the thickness of the anal musculature on require another operation later. Also it the anterior wall is incised to eliminate an seems that a wider scar forms from the overhanging ledge which conld block drain- acutely infected wound, resulting in leak- age. 7) The undercut septum or flap is age and inability to control flatus. sutured with interrupted silk sutures to the Treatment of Anorectal ~istulas Pre- inner edge of the external sphincter; to senting Special Problems: Low-lying ante- prevent cutting through the tissue it is rior fistulas: Certain tistulas--such as ano- important that the flap have little or no vaginal, rectovaginat, and some anterior tension on it and that sutures not be tied perineal fistulas-sometimes require special too tightly (Fig. 4). surgical procedures. Most of these low- The advantage of this procedure is lying anterior fistulas resemble other fistu- that with the two-way drainage--primarily las in this region--that is, they originate through the drain under the septal flap from in a crypt and they involve anti secondarily by way of the partial in- abscess formation; they differ in that the cision of the sphincter nmscles--the possi- secondary opening occurs inside the vagina bilities of infection and accumulation of or near the vagina on the perineum. The serum, blood, or pus beneath the flap are usual unroofing or conversion of the tunnel eliminated. The partial severance of the into an open ditch might result in fecal anal musculature dist:d to the suture line, incontinence in certain cases and require in addition to contributing to the drainage, a secondary repair. Low-lying anterior fis- eliminates any overhanging ledge, thus pre- tulas that underlie little muscle are treated venting tension or pressure on the snture the same as any other fistula, by simple line. After two or three weeks, the severed unroofing. However, when the opening at ends of the musculature are pulled together the source of tb.e anorectovaginal fistula is by scar formation, and continence is re- ~i~s~ proxh~al or distai ~o the dentate mar- stored. gin and the vaginal opening is just inside A minimal-residue diet is prescribed post- the posterior fourchette, my colleagues and operatively, and measures are instituted to I have used the following procedure with prevent bowel movements for about a week. good results: 1) The patient is placed in The rubber drain is removed after three ANORECTAL FISTULAS 253

:/' l/

z' ,-.\ )Tt

Fro. 4, Completed operation for rectovaginal fistula. The mobilized septat llap has been sutured m place. (From Jackman, R. J.: Rectovaginal and anovaginal fistulas: A surgical procedure for treatment of certain types. J. Iowa Med. Soc. 42:435--t.t0 [Sept.] 1952. \Vith permission of the Iowa Medical Society.) days, and the sutures are removed after drainage shouhl be generous, in the hope eight days. that the wound will stay open longer. In selected cases o[ anorectovaginal fis- I)uring a remission or an inactive phase tula this procedure has a high degree of of the colitis, operation sometimes is advis- success, even in the presence of quiescent able to prevent a simple fistuht from becom- colitis or enteritis. If the tract through the ing more complicated, but even then it is perineal body is tortuous or is draining pus frequently difhcult to decide to operate actively, unroofing with subsequent peri- because wound healing may be very slow. neal repair probably is best. The tendency to fistulization in granulo- Anorectal fistula associated with ulcera- matous enterocolitis is well known and, as tive colitis and regional enterocolitis: In indicated previously, the anal tistula may be general, elective anorectal surgery is un- the main complaint of some individuals wise in cases of active ulcerative colitis. The whose intestinal symptoms are mild or mini- resulting wounds are slow to heal and very real. Operation for anorectal listula is not painful, and postoperative care is unsatis- always contraindicated in cases of enteritis, factory because of frequent rectal discharge. but usually the wound heals slowly. Surgical operations on these individuals Fecal Incontinence, Leakage, and Diffi- actually seem to make the colitis worse in culty in Holding Flatus after Operation some instances. and fis- for Fistula: Difficulty in controlling bowel tula are serious complications of ulcerative movements seldom presents a long-term colitis, which often start a chain of events problem after operation for anorectal fis- that may lead to ileostomy and colectomy. tula. On the other hand, such difiqculty is They comprise one of the most commoa frequently a temporary probiem v.ntil the complications of ulcerative colitis, about wound has had a chance to heal and the 8 per cent to I0 per cent of patients being wound margins have pulled together. The afflicted. 1[ a patient with ulcerative colitis individual who has an irritable bowel or has an anorectal abscess, the incision for erratic bowel habits with bouts of 254 JACKMAN

Use of the sewn: The seton is used widely in European countries, and to a lesser ex- tent in America, principally on patients whom the surgeon thinks may have diffi- culty with fecal incontinence if orthodox fistulectomy is performed. Its proponents think the gradual process of cutting through the sphincter with the seton permits the scar to form back of the ligature, thus hold- ing the ends of the muscle fibers together. < Personal/y, I have never used the seton in operations for fistula. It is true that a small percentage of patients have some temporary difficulty with fecal control. It FIc. 5. F, xcisions of zone of mucosa at the proxi- is also true that a very small percentage mal end of the fistulous wound. This helps obviate need a secondary repair such as a perineor- leakage of mucus when healing is complete. rhaphy, but it has been my observation that this is sometimes necessary whether or not and is apt to have the most a seton is used. trouble of this sort; consequently, one usu- Preoperative, operative, a~d postopera- ally can predict preoperatively those pa- tive measures to avert or ameliorate irwon- tients who may and those who may not tinence: Preoperatively, it is well to inform experience leakage. Certain deep-lying patients having an irritable bowel or healed anterior fistulas are more apt to resuit in ulcerative colitis that there may be some incontinence than are posterior or lateral temporary difficulty with control after oper- fistulas. For some of these individuals, par- ation and that a secondary repair may be ticutarly those having an irritable bowel necessary later. At the time of operation, syndrome aml episodes of diarrhea, I have or earlier, the surgeon must determine eradicated the lesion by denuding the area whether to excise the fistulous tract (fistu- of the primary source, covering the denuded lectomy). Especially when there is much area with a flap, and coning out the fistu- scar tissue, fistulectomy tends to create a lous tract down to the primary source. In wound that can unite or pull together more general, however, the surgeon must keep in satisfactorily than can the wound produced mind that to cure the fistula he must con- by a lesser procedure, thus reducing the vert the tunnel into an open ditch regard- likelihood of incontinence. In certain other less of how much muscle must be incised cases, especially where the fistula is deep- to accomplish this. It is also important to lying, the decision to perform a fistulot- find out and record preoperatively if the omy rather than a fistulectomy may reflect patient has any difficulty with control. better judgment. Many times I have observed patients who During either operation, i frequently are completely incontinent who have never excise a piece of rectal mucosa shaped like had an anorectal operation but have had a half-moon (Fig. 5) proximal to the pri- a long-standing fistula with many recurring mary so~>cce, i think this enables that part abscesses. The repeated insults of the ab- of the wound near the rectum to pull scesses and resulting scar formation have together better; also, there is less apt to be immobilized and destroyed the anal-sphinc- leakage of mucus after healing is complete. ter mechanism completely. Postoperatively, any packing placed in ANORECTAL FISTULAS 255

the wound should be removed within 24 5. Anai fistulas may remain active even to 48 hours. Long-continued packing of though the primary source has been oblit- the wound is undesirable since it creates a erated if drainage is inadequate or if in- wide block of scar which makes the sphinc- fected granulation tissue remains in the ter mechanism tess efficient. channel. When fecal incontinence is a problem 6. One should be able to predict pre- postoperatively, some patients may need to operatively, with a fair degree of accuracy, perform anal muscle exercises, take anti- those individuals who are apt to have diffi- cholinergic drugs, and eat minimal-residue cuhy with fecal control after operation for diets. fistula. For such individuals, those having irritable-bowel syndrome or a diarrheal Comment tendency, an alternative procedure or a Certain rules, not completely inviolable, secondary repair should be planned in may be laid clown in regard to the diag- advance. nosis and treatment of anorectal fistulas. 1. Anorectal listulas and abscesses occur- References ring in young adults should arouse suspi- I. Buie. I.. A.. N. D. Smith. and R. J. Jackman: cion of an associated enteritis or entero- The role of tul~crculosis in anal fistula. Surg. G~nec. Obstct. 68: 19l, 1939. colitis. 2. Jackman, P,. J.: I'erianal sinuses: Some unusual 2. The "much-operated-on" persistent causes. Dis. Colon Rectum. 8: 115, 1965. anorectal fistula should make one think of 3. Jackman, R..tl. and N. D. Smith: Some mani- festations of regional ileitis observed sig- an unusua] cause such as an infected pre moidcscopically. Surg. Gynec. Obstet. 76: coccygeal cyst. t.t.t, 19-[3. 3. Antibiotic therapy does not cure anal 1. Klinc, P,..]., R. j. Spencer. and E. C;. Harrison: Carcinoma associated with listula-in-ano. fistulas. .\rch. Surg. (Chicago). 89: 989, 196-t. 4. The basic principle of all operations 5. Rawls. W. E., W. J. Martin, R. J. Spenccr, and H. N. ]~o[]lllall: PeriallaI :abscess alld SilO- for anorectal fistula is adequate drainage. rectal fistula. Minnesota Meal. 46: 327, 1063.

Memoir

Bt"LT, ROBERT LI~RoY, Glendale, California; born Wellsville, Tennessee, March 3, 1899; Rush Medical College, 1924: internshil) Los Angeles General Hospital 192-1-1926; served with U. S. Armed Forces in YVorld War I. Dr. Belt joined the American Proctologic Society in 19tl and was elected to Fellowship in 1948. He was a Fellow o{ the American College of Surgeons anti a member of the American Medical Association, California Medical Associa- tion and Los Angeles County ~'Jedicai Association: on tb.e stalls o~ Me:-norial and Behrens Hospitals, Glendale, and St. Joseph's Hospital, Burbank: Emeritus Associate Clinical Professor of Surgery, U.S.C. Medical School. Dr. Belt died February 95, 1968.