THE JOURNAL OF UROLOGY
Vol. 73, No. 5, May 1955 Pn:nted 1:n U.S.A.
SURGICAL TREATMENT OF HORSESHOE KIDNEY: COMPARISON OF RESULTS AFTER VARIOUS TYPES OF OPERATIONS ORMOND S. CULP
AND
JAMES R. WINTERRINGER
From the Section of Urology, Mayo Clinic and JVlayo Foundation, Rochester, Minn.
Successful surgical treatment of horseshoe kidney is no longer a novelty. The voluminous literature that has accumulated during the past 20 years abounds ·with accounts of gratifying postoperative results. Most of these, however, have been isolated case reports or analyses of very small groups of patients. It is generally agreed that pathologic processes in either half of a horseshoe kidney deserve and demand the same therapeutic considerations as similar lesions in nonfused kidneys. There are widely divergent opinions, however, regarding the significance of the fusion per se and the value of some of the operations that have been proposed. Although horseshoe kidneys are relatively common, most of these anomalies do not require surgical treatment. During the 15 years from 1939 through 1953, approximately one fourth of the horseshoe kidneys that were diagnosed at the Mayo Clinic were subjected to some type of operation. It is not surprising, therefore, that the personal experience of most urologists in the surgical treatment of this anomaly has been rather limited and that the literature is devoid of critical comparison of different surgical techniques . .We have reviewed all of the cases of horseshoe kidneys that have been operated on at the Mayo Clinic. Other types of renal fusion, such as unilateral fused kidney and sigmoid or S-shaped kidneys have been excluded. In the early part of the twentieth century many horseshoe kidneys ·were noted as an incidental finding during abdominal operations or they were merely explored because of a palpable mass. Two such cases in which secondary nephrectomy eventually was performed have been included in our series but all other simple explorations have been excluded. From 1912 through 1953, 106 patients with true horseshoe kidney had renal operations at this institution. It is this group which we have studied. Males predominated in a ratio of 4: 1 (table 1). The youngest patient was 5 years of age and the oldest 65, but a little more than half of the total number were in the third and fourth decades. All of the patients complained of pain, either in the flank or in the lower part of the abdomen. The indications for the initial operations performed at the clinic were: calculi (with or without hydronephrosis), 65 patients, hydronephrosis (without calculi) 27, tuberculosis 5, pain only 5, postoperative urinary fistula 2, and renal cyst 2. Eighty-seven per cent of the 106 patients had calculi or hydronephrosis or both. Only 2 patients were suspected of having a renal neoplasm and in both cases the growth proved to be a simple cyst. One patient who died 17 years after Read at annual meeting, North Central Section of American Urological Association, Detroit, Mich. October 7 to 9, 1954. 747
748 TABLE
ORMOND S. CULP AND JAMES R. WINTERRINGER
1. Age and sex distribution of 106 patients with horseshoe kidney operated on at the
Mayo Clinic from 1912 through 1953
Total ......
Age, Years
Males
Females
Total
5 10-19 20-29 30-39 40-49 50-59 60-65
1 6 18 23 18 12
0 2
7
6 3 2 1
1 8 25 29 21 14 8
85
21
106
.......
. . . ...
7
pyelolithotomy was found to have an unsuspected carcinoma of the renal pelvis. One of the kidneys removed because of hydronephrosis contained an unsuspected hemangioma. These were the only renal neoplasms encountered in the entire group. In 1 case the fusion involved the upper poles of the kidneys. In another instance the kidneys were said to be joined at the pelves. In all other cases some type of isthmus joined the lower poles. The 106 patients included in this study had a total of 160 renal operations. Seventeen of these were performed elsewhere before the patients were seen at the clinic and 6 are known to have been performed elsewhere after the patients had been dismissed. The remaining 137 operations were performed at the clinic. Eighty-one of the 106 patients had only one operation at the clinic. Eleven individuals had single but separate procedures on both sides while 14 required more than one operation on one or both kidneys. Nearly 40 per cent of all patients, therefore, eventually had to have multiple operations. Of the 137 operations at the clinic, the lumbar extraperitoneal approach was used in 128. Anterior extraperitoneal exposure was employed in 4 cases and only five operations were performed transperitoneally. Sixteen surgeons contributed cases to this series but 82 per cent of the patients were treated by 7 of the surgeons. Their respective therapeutic plans and operative techniques varied considerably. The many different surgical procedures lend themselves to a variety of groupings, but two basic categories prevailed, as shown in table 2. In 67 of the operations the isthmus was divided while in 70 it was left intact. N ephrectomy and simple removal of calculi dominated their respective categories and accounted for 74 per cent of all the operations. POSTOPERATIVE COMPLICATIONS
No attempt was made to evaluate transient febrile episodes and other minor disturbances that followed some of the operations. The significant complications appeared to be: urinary fistula, 6 cases, hemorrhage 3, pneumonia 3, incisional hernia 3, phlebothrombosis 3, fecal fistula 1, cerebral hemorrhage 1, and renal failure 1. Most of these sequelae were noted after removal of one half of a horseshoe kidney.
749
SURGERY OF HORSESHOE KIDNEY TABLE
2. Operations on 106 patients with horseshoe kidney seen at the Mayo Clinic frorn 191.'l through 1953 With Division of Isthmus
Without Division of Isthmus
Operation
Number
N ephrectomy Nephropexy N ephropexy and removal of calculus N ephropexy, pyeloplasty and removal of calculus N ephropexy and pyeloplasty Division only
52 5 3
4
1
Number
Removal of calculus only Removal of calculus and pyeloplasty Pyeloplasty only U reterolysis Ligation of vessels Excision of cyst Exploration or drainage
67 I Total.
Total ... TABLE
2
Operation
50 5
3 3 2 2
5 -70
3. Follow-i,p periods for 77 living patients operated on for horseshoe kidney at the
Mayo Clinic frorn 1912 through 1953 Follow~up Period 1 Years
Patients
1-5
24 21
6-10
11-15 16-20 21-25 26-30 31-35
9
12 3 4 4
FOLLOW-UP
Eighteen patients are known to be dead. Seventy-seven have returned for examination or have answered questionnaires. The fate of the remaining 11 is unknown. Early deaths. There were 2 deaths soon after nephrectomy for hydronephrosis. One of these patients died of pneumonia on the seventh postoperative day and the other died in uremia on the eleventh day after operation. The operative mortality rate for the entire series, based on the number of operations, was 1.5 per cent. A third patient died at home 27;2 months after nephrectomy for tuberculosis, and a fourth died of renal insufficiency 57;2 months after bilateral calculi had been removed. Late deaths. Five patients eventually died of renal insufficiency after operations for calculi but they lived for 3, 13, 15 and 29 years, respectively. In all probability the operations added to the longevity of 4 of these patients. Five additional deaths which occurred 1, 2, 4, 20 and 21 years after operation were due to cardiovascular diseases and were not related to the horseshoe kidney or the operations. Four other patients died 3, 4, 5 and 26 years after operations for calculi but no accurate details could be obtained regarding their terminal illnesses. Living patients. The average interval between operation and last report on 77 patients was 11.5 years (table 3). Fifty-six patients were in excellent health and
750
ORMOND S. CULP AND JAMES R. WINTERRINGER
21 were ailing. Nine of the ailing had chronic renal disease (calculi or pyelonephritis or both), 7 had no urologic troubles but were incapacitated by cardiovascular disease (4 cases), peptic ulcer (2 cases) or herniated intervertebral disk (1 case), and 5 complained of vague pains which were difficult to evaluate. General results. Of the 95 traced patients, 56 had no further complaints, 7 were relieved of all renal symptoms but were troubled by unrelated ailments, 5 had urologic relief but died 1 to 21 years later of cardiovascular disturbances, and 4 eventually died of renal insufficiency but survived for 13 to 29 years after the removal of calculi. It would appear, therefore, that at least 75 per cent of the patients derived significant benefit from the surgical treatment of their horseshoe kidneys. Since it is most unlikely that all of the procedures noted in table 2 have the same potential values or the same inherent risks, they have been subjected to further critical appraisal, as presented in the following sections. NEPHRECTOMY
Removal of one half of a horseshoe kidney warrants special consideration, not only because it promises to continue to be the most common operation but also because it appears to be fraught with some of the most disconcerting complications. Indications for the 52 nephrectomies in this series were: hydronephrosis alone, 21 cases, calculi (with or without hydronephrosis) 18, tuberculosis 5, postoperative urinary fistula 6, postoperative hemorrhage 1, and pain only 1. These ·were primary procedures in only 37 cases; in the others, one to three operations had been performed previously. Hemorrhage and fistula formation appear to be the greatest hazards of nephrectomy for horseshoe kidney. Thus, in this series, 3 patients had excessive postoperative bleeding. In 1 of these the responsible vessels in the lower pole of the remaining kidney were ligated 5 hours after the original operation and the patient recovered. In the other 2 the hemorrhage was complicated further by fatal uremia and fulminating pneumonia. In 2 instances there was difficulty controlling bleeding during the operation and hemostats were left on the vessels. These were removed in 48 hours in 1 case and after 5 days in the other with no subsequent hemorrhage. In 5 cases the so-called nephrectomy failed to include all of the functioning tissue on that side and persistent urinary fistulas occurred. Two of these patients had undergone the incomplete nephrectomy elsewhere and 3 were operated on at the clinic. In 2 of these cases even the "repeat nephrectomy" was too conservative, for the fistulas persisted. In the other 3, the patients were cured promptly when care was taken to remove the isthmus as well as the remaining lower pole. For example: Case 1. A man, aged 39 years, had marked left hydronephrosis and numerous small renal calculi (fig. la). A large left renal mass was removed but a portion of the lower pole was left in situ, as demonstrated by clusters of remaining calculi (fig. lb). A urinary fistula persisted even after more stones had been
SURGERY OF HORSESHOE KIDNEY
751
FIG. 1. Incomplete initial nephrectomy. a, Preoperative excretory urogram showing -advanced hydronephrosis and multiple calculi in left half of horseshoe kidney. b, Excretory urogram after removal of bulk of left kidney. Note remaining stone-bearing calyces to left of vertebrae. c, Excretory urogram after additional calculi were removed.Note functioning hydrocalyx in stump of left kidney. d, Excretory urogram after isthmus and residual portion of left kidney were excised.
evacuated. Functioning tissue was still present on the left (fig. le). After the isthmus and this residual segment of left kidney were removed the patient had no further trouble (fig. ld). Another urinary fistula that followed simple renal exploration was cured by the first attempt at nephrectomy. Two additional urinary fistulas were noted after nephrectomy for tuberculosis but both healed eventually without further surgical treatment. There was 1 fecal fistula after a left nephrectomy but this healed spontaneously. The technique for division of the isthmus has been reiterated throughout the literature with many personal modifications. Since the isthmus can vary from a fibrous band to a broad zone of functioning cortex no one method of symphysiotomy is likely to suffice. Thorough orientation regarding the nature and extent of the isthmus and careful identification of all of the anomalous blood vessels seem to be much more important than elaborate methods of suturing the remaining cortex. Accurate preoperative pyelographic studies are necessary when delineation of the separate calyceal system is attempted, for many of the pelvic ramifications in the isthmus are not apparent even after the kidney has been freed.
752
ORMOND S. CULP AND JAMES R. WINTERRINGER
To insure success and avoid troublesome complications removal of one half of a horseshoe kidney should include excision of all cortical tissue to the adjacent pole of the remaining kidney rather than simple division at a convenient level. If the isthmus cannot be freed easily, subcapsular enucleation over the great vessels will simplify the procedure. Subsequent excessive bleeding usually comes from the cut surface of the remaining kidney. It is imperative, therefore, that hemostasis be complete in this area. Simple mattress sutures tied over pads of fat usually suffice. Occasionally the isthmus has a separate blood supply which can be ligated. One must remember, however, that the entire blood supply to both segments of the horseshoe kidney may enter via the isthmus. Meticulous study of the entire vascular pattern is mandatory if cataatrophes are to be averted. In the course of two operations in this series the surgeon abandoned his original plan to divide the isthmus because of peculiarities in the blood supply. The large, ill-defined isthmus that extends well beyond the midline has probably been the greatest pitfall. There is reason to suspect that these can be managed more accurately by transperitoneal exposure despite the continued popularity of incisions confined to the flank. CALCULI
More than 60 per cent of the patients in this series had nephrolithiasis. Although 18 nephrectmnies were performed, conservative measures were employed in 47 cases. Unfortunately, the treatment of many of these patients antedated the appreciation of metabolic disturbances, ureteropelvic obstruction and some of the other factors in the etiology of calculi. (Our first case of hyperparathyroidism associated with horseshoe kidney was diagnosed in 1954.) Thirty-eight patients had 50 simple pyelolithotomies or nephrolithotomies on 45 kidneys. There were 22 known recurrences in 19 patients. Five of the recurrent calculi were removed at the clinic and 2 patients had subsequent operations elsewhere. In this group the average postoperative follow-up period was 10 years. Seventeen of the 38 patients eventually claimed to be enjoying excellent health without further recurrences. Six were alive but ureinic when last heard from. Nine patients died (4 of renal insufficiency, 2 of cardiovascular disease and 3 of unknown causes). The fate of the remaining 6 patients is unknown. On the other hand, 9 patients had 10 kidneys treated by pyelolithotomy combined with pyeloplasty or nephropexy or both. There have been 2 known recurrences. Both of these followed pyelolithotomy and pyeloplasty without symphysiotomy and nephropexy. In 1 case all of the additional stones passed spontaneously or were extracted cystoscopically during the first 2 months after operation and there have been no recurrences since. It is possible that these may have been Ininute calculi that escaped removal at the original operation. The other patient eventually underwent a pyelolithotomy elsewhere. All of the 9 patients in this group are living 1 to 16 years after operation with an average follow-up period of 4.5 years. Two patients have hypertension but no urologic complaints. The others are asymptomatic.
SURGERY OF HORSESHOE KIDNEY
753
It is regrettable that these two groups of patients are not comparable in size and have not been followed for similar periods. It is noteworthy, however, that most recurrent calculi in both groups manifested themselves within 12 months after the original operation. The striking difference in results in these two groups is more likely due to improved renal drainage than to numerical variations. Recurrence of calculi followed 44 per cent of the simple pyelolithotomies and nephrolithotomies and plagued 50 per cent of the patients who received such treatment. Recurrences were reduced to 20 per cent when more efficient drainage of the renal pelvis was provided. HYDRONEPHROSIS
Hydronephrosis undoubtedly is the greatest single menace to horseshoe kidneys. In 21 cases one of the fused kidneys had to be removed because of faradvanced hydronephrosis without calculi. In 19 cases an attempt was made to correct the obstruction. Four of the patients required operations on both kidneys. The 23 conservative procedures included: ligation of vessels only, 1 case, ureterolysis only 2, symphysiotomy and nephropexy 6, pyeloplasty without symphysiotomy 8, and symphysiotomy, pyeloplasty and nephropexy 6. Ligation of vessels. There was no follow-up in the case in which this procedure was performed. It is most unlikely that ligation of vessels alone holds much promise for success in most instances. Ureterolysis. Fibrous tissue was excised from the ureteropelvic region in 1 case and both ureters were freed from an inflammatory presacral mass in the other There was dramatic improvement in both cases but these were unique situations and probably should not be compared with more conventional problems. Nephropexy. Six patients had symphysiotomy and unilateral nephropexy. Four had excellent results and were well 3, 5, 8 and 12 years later. The fact that 2 have failed to show recurrent renal calculi is further proof of the success of the procedure. In 2 cases the hydronephrosis became worse. One of these kidneys eventually had to be removed. Pyeloplasty. In the 8 pyeloplasties performed without division of the isthmus three techniques were employed. The first technique, a Foley Y-plasty, was used in 1 case, with disappearance of all dilatation. The second technique, reimplantation of the ureter, was used in 2 patients, 1 of whom improved and the other became worse. The third technique, creation of another communication between the ureter and the dependent portion of the pelvis, was used in 5 patients, 4 of whom showed marked improvement while the fifth had disappointing results. The 6 patients with good results have been followed for 1 to 16 years. Pyeloplasty and nephropexy. Six patients had combined symphysiotomy, pyeloplasty and nephropexy. Three of these were greatly improved after a Foley Y-plasty. Two required subsequent nephrectomy after reimplantations of the ureter. The sixth patient underwent creation of a new stoma between the ureter and the pelvis and had an excellent result. The 4 patients with good results have remained asymptomatic for 2 to 4 years after operation.
754
ORMOND S. CULP AND JAMES R. WINTERRINGER
FIG. 2. a, Preoperative excretory urogram. b, Excretory urogram 3 years after symphysiotomy and left caliectomy, pyeloplasty and nephropexy.
It is noteworthy that 3 of the 4 failures subsequent to pyeloplasty followed reimplantation of the ureter and that all 4 of the patients with a Foley Y-plasty had gratifying results. Unfortunately, there are insufficient cases for adequate comparison of similar procedures with and without symphysiotomy. It may be significant, however, that calculi recurred only when the isthmus was left intact. This small group of cases lacks statistical significance but it demonstrates several cardinal facts. Correction of hydronephrosis in horseshoe kidneys need not be stereotyped but requires careful evaluation of all possible causes of obstruction. If there is any doubt about the role of the isthmus, it had best be divided or resected. If the ureteropelvic junct~re is dependent and funnel-shaped simple nephropexy will probably suffice. If there is high insertion of the ureter, some dependable type of pyeloplasty is mandatory. Furthermore, the surgical adjuncts which are employed in nonfused kidneys can be adapted advantageously to horseshoe kidneys. For example: Case 2. A man, aged 42 years, had pain in the left flank and was found to have hydronephrosis but good function in the left half of a horseshoe kidney (fig. 2a). There were calcium oxalate calculi in one of the lower calyces. The isthmus was divided, the lower calyx was removed, pyeloplasty was performed and the kidney was sutured in a more nearly normal position. All discomfort disappeared promptly. Three years later the left kidney was essentially normal and there has been no recurrence of the stones (fig. 2b). OPERATIONS FOR PAIN ONLY
Five of the 106 patients had no pathologic changes in their kidneys and were operated on solely for the relief of pain. The following operations were performed: symphysiotomy and nephropexy, 2 cases, symphysiotomy alone 1, ureterolysis 1,. and nephrectomy 1. The kidney that was removed proved to be entirely normaL
SURGERY OF HORSESHOE KIDNEY
755
Fm. 3. Excretory urogram 33 years after symphysiotomy and right nephropexy.
Nevertheless, this was the only patient who was relieved. The others continued to have the same discomfort. For example: Case 3. A woman had symphysiotomy and right nephropexy in 1917 at the age of 27 because of vague abdominal discomfort. Thirty-three years later she had similar pain and the upper part of her urinary tract appeared normal except for malrotation of the kidneys (fig. 3). Guiterrez made a notable contribution when he postulated and elaborated on the so-called horseshoe kidney syndrome. But identification of these rare individuals is a painstaking process of diagnosis by exclusion. Although much has been written regarding symptom complexes attributable to the fusion alone, there are astonishingly few well-documented examples in the literature. Most of the patients who have been relieved of pain by symphysiotomy and unilateral or bilateral nephropexy had some degree of pyelectasis, calculosis or infection. It becomes exceedingly difficult to evaluate the role of the fusion per se in the production of pain when one considers the number of patients who have obvious pyelectasis of their horseshoe kidneys but no discomfort. Most operations on horseshoe kidneys performed solely for the relief of pain seem to be doomed to failure unless the candidates for such procedures are selected with extreme caution. SUMMARY AND CONCLUSIONS
One hundred six cases of horseshoe kidney treated surgically at the Mayo Clinic from 1912 through 1953 were reviewed. The consolidated experiences at one institution, as disclosed in this review, are offered as further evidence to support the impressions and convictions of many independent authors.
756
ORMOND S. CULP AND JAMES R. WINTERRINGER
It was observed that approximately 25 per cent of the horseshoe kidneys that were diagnosed clinically required surgical treatment. Eighty-seven per cent of the patients had calculi or hydronephrosis or both, and nearly 40 per cent had to have more than one operation. A great variety of procedures was employed. More than 75 per cent of the patients eventually had significant improvement. The operative mortality rate, based on 137 procedures, was 1.5 per cent. Most of the serious complications followed nephrectomy. Postoperative hemorrhage and urinary fistula were the greatest problems. Meticulous attention to the anomalous renal blood supply and removal of the isthmus as well as the diseased half of the kidney will prevent these sequelae. Forty-four per cent of the simple pyelolithotomies and nephrolithotomies were followed by recurrence of the stones. When these operations were combined with procedures designed to improve drainage of the renal pelvis the recurrence of calculi dropped to 20 per cent. Hydronephrosis due to obstruction at or near the ureteropelvic juncture is the most common abnormality in horseshoe kidney. Many different operations are capable of relieving such obstruction but each case must be appraised carefully and thoroughly. When in doubt, the combination of symphysiotomy, pyeloplasty and nephropexy is the most dependable. Operations solely for relief of pain are usually unsuccessful and are seldom indicated.