Intestinal Anastomosis, with Special Reference to

the Murphy Button.

LEONARD C. SANFORD, M.D., New Haven.

In selecting the title to this paper, the writer has chosen a wellworn subject which has filled surgical literature for the past fifteen years and has left at the present time, at least, a few interesting features which are still unsettled. The advocates of the various methods of intestinal anastomosis have regularly selected some one operation which has been advanced for general use as the best, failing in many instances to appreciate the limitations which apply to any one operation, however good, and also overlooking the fact that what is often simple and speedy in the hands of a surgeon of special experience, is not as easy for the average surgeon deprived of it. It will be, then, the endeavor of this paper to treat the subject impartially and from a practical standpoint.

In

The surgery of the intestines dates back to ancient times. testinal suture was described by Celus in A. D. 20. There is definite literature on this subject by the Italian surgeons of the middle ages; de Saliceto, in about 1500, sutured the intestines, using the trachea of a goose to keep the lumen of the bowel open, the edges of the intestine being approximated by four interrupted sutures known as the suture of the four masters. In the thirteenth century four monks saved from death a case of intestinal injury by inserting the trachea of a goose into the lumen of the bowel. Ramdohr, in 1730, reported a case of recovery from an operation on the intestine in which a single suture held in apposition one cut end of bowel invaginated into the other. Ambroise Betrandi, in 1769, sutured the intestine over a piece of dried calves' trachea, softened in alcohol, which was covered with balsam of Peru. The tracheal rims were found in the fæces on the fortieth day. The patient recovered. Sentiment from 1700 until early 1800 was in favor of

leaving wounds of the small intestine open. Early in 1800, most of the methods were directed toward the formation of an artificial anus through an abdominal wound and various sutures and loops were suggested with this end in view. Schacher treated a successful case in this manner in 1720. In 1824 came the first revolution in intestinal surgery when Lembert described the stitch which united the adjacent serous coats of the intestine and which is known today as the Lembert stitch. From this time until 1883, the progress of intestinal surgery can be well appreciated from Reichel's research of the subject in a collection of reports on one hundred and twenty-one known cases. It was in 1887 that the first great stimulus was given to the subject by Senn, who suggested decalcified bone plates as a means for intestinal anastomosis, with a report of successful cases. The Senn plates were flat, round, or oval, less than two inches in diameter, with a lumen of about an inch. They were equipped in each quadrant with catgut loops and were inserted through adjacent incisions in the intestine and tied in place by the loops, the union of intestines being completed by Lembert's sutures. This method was well adapted for lateral anastomoses of large intestine. The enthusiasm that started from Senn's method, resulted in the use of a large number of different mechanical aids in intestinal surgery and really marks the development of this principle. Many were more ingenious than useful. Abbe's catgut rings were used in a number of instances, Abbe reporting successful cases by this method soon after Senn. Among the various other mechanical contrivances suggested between 1887 and 1892, were vegetable plates, potato and turnip, by Von Baracz and Dawbarn; segmented rubber rings by Brokaw; raw hide plates by Robinson; cartilage plates and gelatine plates by Schrively; bone tubes by Paul. In 1892, Murphy offered the button as a means for intestinal anastomosis. It is a matter of historic interest in this connection, that Denan's rings, suggested in 1847, are strikingly similar to the Murphy buttons. At the outset, intestinal anastomosis by means of the Murphy button received much unfavorable comment in this country and especially in England where Hutchinson, before the clinical society of London, reported fifteen cases of primary resection for gangrenous intestine

in which the Murphy button was employed with only one recovery. In the following discussion the button was generally condemned. In America, the hesitation which first marked the use of the Murphy button changed to general popularity and our surgical literature from 1893 to 1900 abounds in reports of favorable cases. Murphy, June 6, 1900, collected sixteen hundred and twenty cases, out of which in only three instances could a fatal result be traced to the button. In this list of cases, the button was found retained in the stomach twenty-two times; in the colon five times; in the ileum two times; in the jejunum once; in the cæcum two times; in the rectum four times; a truly remarkable record of cases. Since 1900, the popularity of the button as a means of intestinal anastomosis may be said to have waned; due in part, perhaps, to the perfection of the suture and various other methods of anastomosis after resection.

In this connection a brief history of the intestinal suture is important and I will begin with the Lembert stitch. This was suggested by Lembert in 1827 and was based on the principle of uniting serous surfaces to serous surfaces. This method was open to two objections. First, it afforded a lumen of insufficient strength; second, it resulted in a curling in of the mocosa; yet the stitch itself remains an important factor in most of the suture methods at the present time. The objection to a suture that pierced the mucous membrane lasted a long time. It was finally overcome, and the fact demonstrated that a secure union of intestinal surfaces required a stitch that included the submucosa. As the submucosa in places is of a smaller diameter than a fine needle, the necessity of including at least a portion of the mucosa was soon made apparent and the Czerny modification of the Lembert suture was one of the means suggested. Czerny used a deep layer of interrupted sutures reinforced by Lembert's stitches. At the time. of the introduction of this method there was a strong prejudice against the use of a continuous suture in intestinal anastomosis, on account of the unequal cutting out of stitches and a resulting weakness in the line of union. This objection was proved to be theoretical only. In 1892, Maunsell offered the method of anastomosis by means of invaginating one end of intestine into the

other. The two surfaces of cut intestine were drawn through a longitudinal incision in the intestine about one inch from the cut edges. When in position, union was effected by a continuous running suture through all the coats. The sutured intestine was drawn back and the longitudinal incision closed. Lembert's stitches were employed to cover in the first continuous suture line. The advantage gained in this way is apparent. The operation is secure and quickly performed. The objection to it is the extra longitudinal cut. The Maunsell method undoubtedly suggested. the Connell suture, described by N. E. Connell in 1893. Connell sutured one-half of the cut intestinal ends after they had been brought into position side by side by a continuous mattress stitch piercing all the coats, changing in the last half to an external mattress stitch leaving the loops long until the suture was complete then drawing them tight and reinforcing the two external knots by means of Lembert's stitches. In 1903, F. G. Connell modified the method by using an interrupted mattress suture and tying his knots internally. He reported sixty-four cases operated on by this method with twenty-one deaths, one of which was attributed to faulty union. The advantages claimed for this method are less danger of leakage, strength at the mesenteric border, less liability to necrosis and adhesion. The principles embodied in the Connell suture certainly hold good at the present time although the suture itself has been variously modified. These principles are, in the first place, a maximum of security. In a continuous suture, the cut surfaces of the bowel are subjected to a uniform pressure. There is no bulging between the stitches as a result of distension from gas or fæcal matter. This condition would not hold true in the case of interrupted sutures. Secondly, a maximum of speed. A continuous suture requires much less time than any form of interrupted. It is by no means possible in this short paper to enter into a detailed discussion of the various intestinal sutures. The method, however, of McGraw suggested in 1901 should be mentioned. He uses the elastic ligature as a means of approximation. McGraw tied the ligature interruptedly, passing it while stretched. He pierced all three coats tying the knots on the outside and closed the first suture line with external Lembert stitches..

The pressure of the elastic ligature soon caused the knot to slough into the intestine. A possible objection to the elastic ligature, that when tied and stretched it might not completely fill the needle puncture, McGraw claimed to obviate by passing the stitch, as described, on the stretch. By some surgeons twine has been used as a substitute for the elastic ligature. In the British Medical Journal of 1903 is an exhaustive article on the subject of the best general means for intra-abdominal anastomoses. Here Wolfler's method of suture is advocated. This consists of one continuous line of stitches through the mucosa reinforced by a sero-muscular

suture.

In addition to mechanical and suture means for intestinal union, a number of ingenious methods have been brought forward in the last four years. In 1902, Harrington suggested the segmented rings. The adjacent edges of intestine are approximated by purse string sutures over the ring which is held in convenient position by a specially constructed handle while anastomosis is completed by means of a Cushing's continuous stitch. The handle is then removed and the opening closed. The advantages claimed for the Harrington rings are: the weak spot at the mesenteric border is avoided by a mattress stitch; the lumen of the ring is sufficiently large; the ring can be broken by light pressure after the suture is complete if so desired.

Several intestinal forceps have been devised; one by O'Hara of Philadelphia. Each end of intestine is clamped and trimmed close to the forceps which are then approximated and the gut united. over the forceps by means of a Halstead suture. The forceps are then withdrawn and the closure completed. These methods have not been generally accepted with satisfaction.

Jaboulay has recently suggested a modification of the Murphy button by which he claims the following advantage-introduction through a small opening and no necessity for suture. The objections to Jaboulay's method are laceration of the intestine and liability of retention.