Role of the Martius Procedure in the Management of

UrinaryVaginal Fistulas
NP Rangnekar, MS, MCh (Urol), DNB (Urol), N Imdad Ali, MS, SA Kaul, MS, DNB,
HR Pathak, MS, MCh (Urol), DNB (Urol)

Background: Urinaryvaginal fistula is one of the most

common and dreaded complications of obstetric
trauma in developing countries. Management of these
fistulas is complicated by the presence of substantial
urethral loss and the tendency of the repair to break
down.

pared with six failures out of nine treated with anatomic repair. None of the patients having primary
treatment with the Martius flap had postoperative recurrence, compared with 3 of 18 having anatomic repair (16.67%). Only 1 of 12 patients with recurrent
fistulas undergoing Martius flap repair had failure
(8.33%), compared with 4 of 7 undergoing anatomic
repair (57.14%). None of the patients treated with the
Martius procedure experienced dyspareunia postoperatively, compared with 33.33% of the patients treated
with anatomic repair.

Study Design: We retrospectively studied 46 patients

with urinaryvaginal fistulas operated on in our institution over 5 years. Most of the patients had obstetric
trauma as the causative factor. Twelve patients had
urethrovaginal and 34 had vesicovaginal fistulas. Of
the 12 patients with urethrovaginal fistulas, 8 underwent a Martius procedure and 4 were treated with
simple anatomic repair. Of the 34 patients with vesicovaginal fistulas, 13 underwent a Martius procedure
and 21 were treated with anatomic repair. Nineteen
patients had recurrent fistulas and 17 had multiple
fistulas.

Conclusions: The overall success rate was far better

and the complication rate (especially incontinence and
dyspareunia) was considerably less with the Martius
procedure. We recommend the Martius procedure for
urethrovaginal and vesicovaginal fistulas, especially
those that are recurrent or multiple. (J Am Coll Surg
2000;191:259263. 2000 by the American College of Surgeons)

Results: Only one patient with a urethrovaginal fistula

treated with a Martius procedure had recurrence, compared with three of four of the patients having anatomic repair. None of the patients with vesicovaginal
fistulas treated with a Martius flap had recurrence,
compared with 4 of 21 in the anatomic-repair group
(19.05%). Thirteen patients with single fistulas (7 urethrovaginal and 6 vesicovaginal) treated with a Martius
procedure healed well without failure, compared with
1 failure among 16 fistulas (1 urethrovaginal and 15
vesicovaginal) in the anatomic-repair group. In the
group of patients with multiple fistulas, the Martius
flap also showed a definite advantage. Eight patients
with multiple fistulas were offered the Martius flap.
The procedures were successful in all but one, com-

Although uncommon in developed countries,

urinaryvaginal fistula is a common complication
of obstetric trauma in developing countries such as
India.1,2 Other causes of this dreaded condition are
operative procedures on the vagina and urethra (eg,
urethral diverticulectomy or anterior vaginal repair); pelvic fracture; vaginal or urethral neoplasm,
especially after treatment with radiation; and radical operations such as Wertheims hysterectomy.3-5
Symptoms depend on the position of the fistula in
relation to the sphincter mechanism and the size of
the fistula.
Management of these fistulas depends on a variety of factors, namely the presence or absence of
symptoms; cause, size, and location of the fistula;
and other local factors such as the extent of associated fibrosis or radiation reaction. Various surgical
procedures have been described, with varying success. We compared simple anatomic repair reinforced with vaginal-flap closure and the labial fat-

No competing interests declared.

Received December 9, 1999; Revised February 28, 2000; Accepted March 27,
2000.
From the Department of Urology, Topiwala National Medical College and
B. Y. Nair Charitable Hospital, Mumbai, India.
Correspondence address: Nilesh P Rangnekar, MS, MCh (Urol), DNB
(Urol), 5, Radha Bhuvan, 176, L.J. Rd, Shivaji Park, Dadar, Mumbai 400
028, India.
2000 by the American College of Surgeons
Published by Elsevier Science Inc.

259

ISSN 1072-7515/00/$21.00
PII S1072-7515(00)00351-3

260

Rangnekar et al

Martius Procedure for UrinaryVaginal Fistula

pad repair, which is popularly known as the Martius

procedure.6-9
METHODS
We retrospectively studied 46 women with
urinaryvaginal fistulas between 18 and 52 years of
age undergoing operative treatment in our institution from January 1994 to January 1999, with a
followup ranging from 6 months to 5.5 years. There
were 12 patients with urethrovaginal fistulas involving the bladder neck and 34 patients with vesicovaginal fistulas. Those vesicovaginal fistulas situated
high on the posterior wall of the bladder (ie, more
than 3cm away from the bladder neck) and with a
size greater than 1.5cm were excluded from the
study. Those vesicovaginal fistulas involving the
ureteric orifice on cystoscopy and intravenous urography were also excluded from the study.10 Twentyone patients underwent a Martius flap repair (8
urethrovaginal and 13 vesicovaginal) and 25 patients underwent anatomic vaginal-flap closure (4
urethrovaginal and 21 vesicovaginal). Nineteen patients had recurrent fistulas (9 urethrovaginal and
10 vesicovaginal), of whom only 11 patients could
give details about the previous operation. Six patients were operated on by a gynecologist, and three
were operated on by referring urologists. Two patients had undergone previous repair in our department. All of the patients had undergone anatomic
closure of the fistula during the previous intervention. Seventeen patients had multiple fistulas. Sizes
of the fistulas ranged from 2mm to 1.5cm. The
causes of the fistulas were hysterectomy in 12, obstetric trauma in 32, and anterior vaginal repair in 2
patients.
Identification of the urinaryvaginal fistula was
based on history, speculum examination of the vagina,
intravenous urography, and cystourethroscopy.3,4,11-13
Some patients underwent intravesical methylene
bluevaginal tampon test.11,13-15
Of the 19 patients who suffered from recurrent
fistulas after initial repair, 12 patients were treated
with the Martius procedure and 7 underwent anatomic vaginal-flap closure.
The Martius procedure involved freshening of
the edges of the fistula after raising a U-shaped vaginal flap. The fistula was closed in two layers with
urethral-wall approximation and Lamberts sutures

J Am Coll Surg

of the periurethral tissue over it. This suture line was

reinforced by the interposition of a mobilized, wellvascularized fat pad from the labia majora. A 10cm-long pedicle could be raised on the blood supply
from the branches of the pudendal artery. The fat
pad was tunneled beneath the labia minora into the
vaginal lumen, where it was sutured to the bladder
or urethral wall at the site of fistula repair. The final
layer of closure was the anterior vaginal-wall flap.
There was no cosmetic deformity in the perineum.6-9,13 The bladder was drained with suprapubic and urethral catheters for at least 3 weeks. No
sexual intercourse was allowed for at least 3 months
postoperatively.
Anatomic vaginal-flap repair was performed by
developing a plane between the vagina and the bladder or urethra facilitated by saline and epinephrine
instillation, raising a U-shaped vaginal flap. The
fistulous tract was not excised, but the edges were
freshened to prevent an iatrogenic increase in the
fistula diameter. The freshened edges of the fistula
were closed in two layers with suture lines perpendicular to each other, and the vaginal-wall flap was
closed over the repair.4,12,16 Postoperatively, the
bladder was drained with suprapubic and urethral
catheters for a minimum of 3 weeks, with avoidance
of sexual intercourse for 3 months.
After 3 weeks postoperatively, the urethral catheter was removed and the suprapubic catheter was
clamped to allow a voiding trial.2 Failure of the fistula repair was diagnosed by continuous vaginal
leak, confirmed with methylene blue dye instillation through the suprapubic catheter. Postoperative
incontinence without recurrence of the fistula was
diagnosed on urodynamic study.17
Fishers exact test of probability was applied to
determine statistical significance, with 95% confidence limits.
RESULTS
Of the 12 patients with urethrovaginal fistulas (involving the bladder neck), 8 underwent the Martius
procedure and 4 underwent anatomic repair. Of the
eight Martius flaps, only one had recurrence
(12.5%), but three of four of the anatomic repairs
(75%) had failure. One patient who did not have
recurrence with anatomic repair experienced postoperative stress incontinence, confirmed on urody-

Vol. 191, No. 3, September 2000 Rangnekar et al

261

Martius Procedure for UrinaryVaginal Fistula

Table 1. Comparison of Postoperative Outcomes Between Urethrovaginal and Vesicovaginal Fistulas

Martius flap (n21)
Type of fistula

Anatomic repair (n25)

Healed

Incontinent

Failure

Healed

Incontinent

Failure

13

17

Urethrovaginal fistula
involving bladder
neck (n12)
Vesicovaginal fistula
(n34)

namic study with low Valsalva leak point pressure

suggestive of intrinsic sphincter deficiency. None of
the patients undergoing a Martius repair had stress
incontinence. Of 34 patients with vesicovaginal fistulas, 13 underwent Martius repair and none had
failure, but of the 21 patients undergoing anatomic
repair, 4 had recurrence (19.05%). One patient experienced postoperative incontinence due to detrusor instability, confirmed on urodynamic study, but
did not have fistula recurrence (Table 1).
Of the 29 patients with a single fistula, 8 had
urethrovaginal fistulas and 21 had vesicovaginal fistulas. The only patient with a urethrovaginal fistula
in this group who underwent anatomic repair had
failure of the repair. Of the 17 patients with multiple fistulas, 4 had urethrovaginal and 13 had vesicovaginal fistulas. Of the multiple urethrovaginal
fistulas treated by anatomic repair, the fistula healed
in only one patient, who unfortunately developed
postoperative incontinence due to intrinsic sphincter deficiency. The cumulative cure rate in this
group was 0%.
Of the 19 patients who had previously failed
fistula treatment, 9 had urethrovaginal and 10 had
vesicovaginal fistulas. Nine patients undergoing
primary treatment were offered a Martius flap repair (2 urethrovaginal and 7 vesicovaginal), and 18
patients undergoing primary treatment were of-

fered anatomic repair (1 urethrovaginal and 17 vesicovaginal). None of the patients having primary
treatment with a Martius flap had postoperative recurrence, compared with 3 of 18 patients having
anatomic repair (16.67%). Table 2 compares the
postoperative outcomes of the fistula repairs depending on multiplicity and recurrence of the
fistulas.
The overall success rate of the Martius procedure in our series was 95.24%, compared with 72%
with anatomic repair (p0.038 by Fishers exact
test). One patient who underwent anatomic repair
experienced postoperative stress incontinence,
compared with none of the patients undergoing the
Martius procedure. None of the patients among the
20 cured with the Martius flap repair experienced
dyspareunia, compared with 6 of 18 cured patients
(33.33%) who underwent anatomic repair (p
0.017 by Fishers exact test) (Table 3).
DISCUSSION
Our results indicate that interposition of the labial
fat pad (Martius procedure) adds to the success of
urinaryvaginal fistula repair. We recommend this
procedure for the repair of both types of fistulas (ie,
urethrovaginal and vesicovaginal) of moderate size
situated within 3cm of the bladder neck. The labial

Table 2. Comparison of Postoperative Outcomes Depending on Recurrence and Multiplicity of the Fistulas
Martius procedure
Single

Multiple

Anatomic repair
Recurrent

Single

Multiple

Recurrent

Fistula

Urethrovaginal
Vesicovaginal

7
6

0
0

0
7

1
0

5
6

1
0

0
15

1
0

1*
2(1)

2
4

0
3

3
1

*Fistula repair healed well, but the patient developed incontinence due to intrinsic sphincter deficiency.

(1) suggests number of patients who developed postoperative incontinence due to detrusor instability.
F, failed repair of fistula; H, healed fistula.

262

Rangnekar et al

Martius Procedure for UrinaryVaginal Fistula

J Am Coll Surg

Table 3. Postoperative Complications of UrinaryVaginal Fistula Repair

Postoperative
complication
Failure of the repair
Incontinence
Voiding dysfunction
Dyspareunia

Martius flap (n21)

Anatomic repair (n25)

UV fistula

VV fistula

UV fistula

VV fistula

1
0
0
0

0
0
0
0

3
1
0
1

4
1
0
5

UV, urethrovaginal; VV, vesicovaginal.

fat pad provides the area with an additional blood

supply, lymphatic drainage, and a surface for epithelialization; it also prevents overlapping of the
urinary and vaginal suture lines.
The role of simple anatomic vaginal-flap closure is limited to vesicovaginal fistulas well away
from the sphincter zone, which are usually small
and single. Comparing overall postoperative outcomes of urinaryvaginal fistulas between anatomic
vaginal-flap closure and the Martius procedure reveals an unacceptable recurrence rate of 28% and a
high percentage of postoperative dyspareunia
(33.33%) in the former group. Only one patient
showed complete urethrovaginal fistula healing
with anatomic repair, but the same patient developed intrinsic sphincter deficiency with a low Valsalva leak point pressure.17
The results of fistula repair were particularly
poor when the patient had already undergone multiple operations. This may be due to local tissue
ischemia and fibrosis resulting from the previous
intervention.9 Failure of anatomic repair in 57.14%
of the patients with multiple operations may have
contributed to this finding. At the same time, only
8.33% of the patients undergoing the Martius procedure after multiple previous interventions had
failure, suggesting this to be the best modality for
the management of recurrent fistulas.
Gracilis muscle interposition has been used for
difficult urethrovaginal and vesicovaginal fistulas.13
This procedure has a considerably greater morbidity rate than the Martius procedure and should be
reserved for patients in whom the labial fat pad is
unavailable or insufficient, such as in a very large
fistula.9
The only urethrovaginal fistula that recurred
after the Martius procedure probably had an early
intervention after the initial surgical failure, ie,

within 3 months.2,11,12 The timing of operation is

important because of its influence on the local tissues to be repaired. We recommend an interval of at
least 3 to 6 months between the initial injury or
operation and definitive fistula repair.
The postoperative dyspareunia seen in six patients with anatomic repair could be explained on
the basis of increased scarring in the vagina compared with those treated with the Martius flap. The
lesser amount of scarring after the Martius procedure could be due to the increased blood supply and
lymphatic drainage provided by the flap.
The Martius labial fat-pad flap is an excellent
treatment for urinaryvaginal fistulas, especially for
urethrovaginal and recurrent vesicovaginal fistulas
of moderate size.
Acknowledgment: We sincerely thank Dr Shriram
S Joshi, Consultant Urologist at Jaslok Hospital and
Research Center, for editorial assistance.
References
1. Hamlin RH, Nicholson EC. Reconstruction of urethra totally
destroyed in labour. B M J 1969;1:147150.
2. Gerber GS, Schoenberg HW. Female urinary tract fistulas.
J Urol 1993;149:229236.
3. Wein AJ, Malloy TR, Carpiniello VL, et al. Repair of vesicovaginal fistula by a suprapubic transvesical approach. Surg Gynecol
Obstet 1980;150:5760.
4. Goodwin WE, Scardino PT. Vesicovaginal and ureterovaginal
fistula: a summary of 25 years of experience. J Urol
1980;123:370374.
5. Wein AJ. Vesico vaginal fistula. In: Current therapy in genitourinary surgery. Resnick MI, Kursh E, eds. Philadelphia: BC
Decker, Inc; 1987:209213.
6. Leach GE. Urethrovaginal fistula repair with Martius labial fat
pad graft. Urol Clin North Am 1991;18:409413.
7. Wang YU, Hadley HR. The use of rotated vascularized pedicle
flaps for complex transvaginal procedures. J Urol
1993;149:590592.
8. Martius H. Die Operative Wiederherstellungder Vollkommen
Fehlenden Harnrohre und des Schiessmuskels Derselbel. Zentralbl Gynakol 1928;52:480.
9. Webster GD, Sihelnik SA, Stone AR. Urethrovaginal fistula: a
review of surgical management. J Urol 1984;132:460462.

Vol. 191, No. 3, September 2000 Rangnekar et al

10. Shalmo R. Early versus late repair of vesico-vaginal fistulas by

vaginal and abdominal approaches [editorial comment]. J Urol
1995;153:11121113.
11. OConor VJ. Review of experience with vesico-vaginal fistula
repair. J Urol 1980;123:367369.
12. Keettel WC, Sehring FG, de Prosse CA, Scott JR. Surgical management of urethro-vaginal and vesico-vaginal fistulas. Am J
Obstet Gynecol 1978;131:425429.
13. Patil U, Waterhouse K, Laungani G. Management of 18 difficult
vesico-vaginal and urethro-vaginal fistulas with modified

Martius Procedure for UrinaryVaginal Fistula

14.
15.
16.
17.

263

Ingelman-Sundeberg and Martius operations. J Urol

1980;123:653656.
OConor VJ, Sokol JK, Bulkley GJ, et al. Suprapubic closure of
vesico-vaginal fistula. J Urol 1973;109:5154.
Moir JC. Vesico-vaginal fistula as seen in Britain. J Obstet
Gynaecol Br Commonw 1973;80:598.
Barnes R, Hadley H, Johnston O. Transvaginal repair of vesicovaginal fistulas. Urology 1977;10:258260.
McGuire EJ. Urodynamic evaluation of stress incontinence.
Urol Clin North Am 1995;22:551555.