Vaginal creation for mllerian agenesis

Carla P. Roberts, MD, PhD,a Michael J. Haber, PhD,b and John A. Rock, MDa Atlanta, Ga
OBJECTIVE: The purpose of this study was to determine the effectiveness of passive vaginal dilation and McIndoe vaginoplasty in the creation of a neovagina for patients with mllerian agenesis. STUDY DESIGN: Fifty-one patients with Mayer-Rokitansky-Kuster-Hauser syndrome were treated for vaginal agenesis at either Johns Hopkins Hospital or Emory University. These historic prospective data were obtained by a review of medical records and a current office or telephone consultation. Initial office visits dated from November 18, 1983, through June 6, 1998. Their progress towards both anatomic and functional success was followed through August 1, 2000, which was a range of 2 to 16.8 years. One-way analysis of variance, Student t test, and logistic regression analysis were performed when appropriate. RESULTS: Four patients were lost to follow-up in various stages of the treatment. Ten patients refused vaginal dilation and proceeded to a successful modified McIndoe vaginoplasty. Of the 37 remaining patients, 91.9% anatomic and functional success was achieved from the Ingram method for vaginal dilation. Passive dilation failed in 8.1% of patients, who underwent a modified McIndoe vaginoplasty; all neovaginal creations were successful. All patients who underwent McIndoe vaginoplasty were compliant with postoperative vaginal form use. None of our patients lost vaginal space through contractions or loss of skin graft. Of those patients for whom dilation failed, only 1 patient discontinued the study because of bleeding and discomfort. In addition, only 1 patient from the 3 cases of failure had undergone a previous hymenotomy. Interestingly, 6 patients for whom dilation was successful (6/34 patients; 17.6%) had also undergone a previous hymenotomy. The mean follow-up time for all patients in this study was 111.1 7.2 months, with a range of 25 to 188 months. The mean follow-up time for those patients for whom dilation failed or who refused dilation was significantly lower at 64.5 9.5 and 65.3 18.5 months, respectively (P < .005). The mean time to successful dilation was 11.8 1.6 months with a range of 3 to 33 months. Although longer, no statistically significant difference was observed for dilation time in those patients for whom there was a failure to achieve anatomic or functional success (20.5 12.5 months; range, 8-33 months). CONCLUSION: These data reveal that passive dilation with the Ingram method is capable of creating an adequate vaginal canal in patients with vaginal agenesis, with respect to both function and anatomy even in those patients with a previous hymenotomy and resultant scar formation. Our modified McIndoe procedure has proved to be an excellent option for patients for whom conservative dilation techniques failed and who refuse to attempt any dilation. Interestingly, our data indicate that patients may now be trending toward immediate surgical correction rather than diligently using dilation techniques to create a vaginal space. (Am J Obstet Gynecol 2001;185:1349-53.)

Key words: Dilation, mllerian agenesis, vaginoplasty

Patients with vaginal agenesis have been extensively examined since the 1800s when the disorder was first diagnosed by Mayer.1 Most patients have primary amenorrhea and are found to have an absent or hypoplastic vagina on physical examination. Vaginal dilation was not offered to

From the Department of Gynecology and Obstetricsa and the Department of Biostatistics, Rollins School of Public Health,b Emory University School of Medicine. Presented at the Twenty-seventh Annual Meeting of the Society of Gynecologic Surgeons, Orlando, Fla, March 5-7, 2001. Reprint requests: Carla Roberts, MD, PhD, Department of Gynecology and Obstetrics, Emory University School of Medicine, 1639 Pierce Dr, Rm 4208 WMB, Atlanta, GA 30322. Copyright 2001 by Mosby, Inc. 0002-9378/2001 $35.00 + 0 6/6/119075 doi:10.1067/mob.2001.119075

the earliest patients who were treated. Surgical creation of a vaginal space and various graft techniques were used, and the space was then maintained by hard forms, such as balsa wood. Complications occurred that ranged from graft loss to sloughing of the urethra and rectovaginal fistula formation.2 After the observation that coitus alone could create a sufficient vagina in some patients with vaginal agenesis, techniques for dilation were examined. The Frank technique3 used hand-held vaginal dilators in a squatting position. Although dilation could be accomplished, the position was awkward, and the patient was unable to do any other task during the hours of daily pressure required. The method of Ingram4 uses a patients own trunk weight to apply pressure while sitting on a stool or bicycle seat and allows for other productive tasks to be performed during this dilation procedure. 1349

1350 Roberts, Haber, and Rock

December 2001 Am J Obstet Gynecol

Table I. Treatment groups for patients with mllerian agenesis

Treatment group Attempted dilation Refused dilation Lost to follow-up Total (n) 37/51 10/51 4/51 Percentage of total 72.5 19.6 7.8

Table II. Dilation success and failure rates

Groups Successful dilation Failed dilation *P < .001. Total (n) 34/37 3/37 % 91.9* 8.1

This study contains the largest series of patients with vaginal agenesis to date for whom the Ingram method of dilation was used to create a vaginal space and to afford these patients satisfactory sexual function. Material and methods Fifty-one women with mllerian agenesis were included in this analysis. The patients were first seen in the clinic of the senior author at either Johns Hopkins Hospital or Emory University School of Medicine from November 18, 1983, to June 6, 1998, and were followed through August 1, 2000. The historic prospective analysis consisted of a review of records. The mean follow-up was 9.25 years (range, <2-16 years). Those patients lost to follow-up were not included in this value. Functional success was determined as the endpoint of satisfactorily having intercourse or accepting the largest dilator without discomfort in the clinic visit. Anatomic success was determined by a series of postoperative visits to the senior authors clinic. Those patients without functional success were given the option to proceed to surgical correction with a McIndoe vaginoplasty. All patients had a recent follow-up clinic visit or were personally called by one of the authors to update our records on their clinical status. Statistical analysis was performed with 1-way analysis of variance, Student t test, correlation, and logistic regression analysis where appropriate. The attending physician counseled both the patient and her family extensively during the initial visit. Psychiatric counseling was offered if the physician, patient, or family members requested it. If the patient was hospitalized, both the attending physician and his associates saw the patient daily. All surgical procedures were performed by or under the constant supervision of the senior author. Patients were given careful instructions for the correct use of the vaginal dilators as used by Ingram.4 Lucite dilators of gradual sizes were used to create a vaginal space both by increasing alternatively the length and width until satisfactory vaginal creation was accomplished. At the first clinic visit, the physician demonstrated the patients perineal anatomy with a mirror, and the patient was shown where to place the tip of the smallest dilator, at the introital dimple between the anus and the urethra. The patient was instructed to push until mild discomfort

was achieved; this is more from pressure than pain. The patient was then told to sit on the bicycle seat, lean slightly forward with the mold in place in a pushing manner for 20 minutes 3 times a day. Vaginal lubricant was used before dilation times. It was stressed that the patient must proceed at her own pace and that, once she was pleased with the results of the current sized dilator, she should proceed to the next length or width in an alternating manner. She was told that each graduation should take approximately 2 weeks and that mild discomfort would be experienced with effective dilation. Once the patient had proceeded comfortably to the largest dilator for 1 month, intercourse was attempted. If intercourse was infrequent, dilation was continued. The Lucite forms were cleaned in warm soapy water. The patients were followed monthly until adequate dilation was achieved. Those patients who refused a trial of vaginal dilation or who failed to create a functional vaginal space with dilation underwent a modified McIndoe vaginoplasty.1,5 Results The total number of patients is shown in Table I. Of the 37 patients who attempted vaginal dilation, 34 women (91.9%) achieved success; only 3 patients (8.1%) failed to achieve the creation of a functional vagina (Table II; P < .001 vs failed to dilate). Of those patients who attempted dilation, 6 of 34 (17.6%) had undergone a previous attempted hymenotomy. Of the 3 patients for whom dilation failed, only 1 patient (33.3%) had undergone an attempted hymenotomy. Although the trend toward failure with previously attempted hymenotomy is present, the results are not significant. Fig 1 shows the starting vaginal length before dilation for all 51 patients. Three patients in this group had a starting vaginal length of 10 cm and had a narrow vaginal canal, and coitus was infeasible. The starting vaginal length for the 3 patients for whom dilation failed was 2 cm (Fig 1). Correlation analysis (Table III) demonstrates that, although increased starting vaginal lengths had a greater trend toward success, no statistical significance is reached in this study. When the starting age was analyzed with logistic regression analysis, an evocative result was obtained in that a statistically significant failure to dilate was seen in those patients who were younger than 18 years (Table IV; P < .05). It should be stressed that, in the correlation analysis, the definition of failure included

Volume 185, Number 6 Am J Obstet Gynecol

Roberts, Haber, and Rock 1351

Fig 1. Starting vaginal length. A, The total group of patients in this study; B, the group of patients for whom dilation was successful; C, the group of patients who refused a trial of dilation; D, the group of patients for whom dilation failed.

Table III. Starting vaginal length for patients in need of dilation

Starting length (cm) <1 1-2 2-4 >4 Percentage of patients for whom dilation failed (n) 0.64 (7/11) 0.20 (3/15) 0.17 (2/12) 0.12 (1/8)

Table IV. Starting age for all patients in need of dilation

Age (y) 14-16 17-18 >19 Percentage of patients for whom dilation failed (n) 0.39 (7/18) 0.43 (6/14) 0.00 (0/14)

not only those patients who unsuccessfully attempted dilation but also those patients who refused dilation. No significant correlation was seen with the use of only those 3 patients who first attempted dilation and subsequently proceeded to a McIndoe vaginoplasty. No difference was observed in the months needed to achieve functional creation of a vagina with dilation (Fig 2; 11.8 1.6 months vs 20.5 12.5 months). Fig 3 shows the total months of follow-up for these patients in our clinic after the initiation of treatment. Interestingly, the months of follow-up were significantly less for those patients for whom dilation failed or who refused dilation altogether (all, 111.1 7.2 months; successful, 129.5 5.8 months; failed, 64.5 18.5 months; refused, 65.3 18.5 months; P < .05). Comment Mllerian agenesis is second only to gonadal dysgenesis as a cause of primary amenorrhea. There have been

multiple series to describe various surgical techniques, but these series always contain multiple developmental disorders with resultant vaginal agenesis. Only one other series, by Rock et al, has limited their patients to mllerian agenesis alone, and that series involved the Frank method of dilation.6 The reasons for failure of neovaginal creation by the Frank method have been attributed to the inability of the patient to generate adequate perineal pressure for the necessary protracted time each day, boredom of manually applying pressure for an adequate time, and total consumption of that daily time with the process of dilation. The patient can perform the Ingram method of dilation while also performing other sit-down tasks. Of additional and great importance is the ability to perform vaginal dilation while fully clothed. The emotional benefit for these already self-conscious young women is immeasurable, although obviously this is not a substitute for adequate counseling by the physician. It was initially thought that dilation would work for only patients with vaginal agenesis who had a starting

1352 Roberts, Haber, and Rock

December 2001 Am J Obstet Gynecol

Fig 2. Total months necessary to achieve successful dilation versus the length of dilation time in those patients for whom the attempt to create a functional vaginal space failed.

etal need for a quick fix or a change in motivation for todays patient is an unanswered question. This series demonstrates an extremely appealing treatment for the patient with vaginal agenesis, because our success rate with the Ingram method of dilation is so striking. For those patients for whom dilation failed or who refused a trial of dilation from the initial meeting, we have been very successful with the modified McIndoe vaginoplasty. All patients who were treated in this series ultimately achieved an anatomically and functionally successful vagina. Might we begin to think of the vagina as Oliver Wendell Holmes once viewed the mind: Mans mind, once stretched by a new idea, never returns to its original dimensions.8
REFERENCES

Fig 3. Months of follow-up in our clinic after the initiation of vaginal dilation. *P < .05.

vaginal length of at least 1 to 4 cm.7 Our series demonstrates that only a dimple or less (0.5 cm) is necessary for functional success. There is a soft spot beneath the urethra in every patient that can be the starting point for dilation, even when an obvious concavity is not present. In the dilation series by Rock et al6 who used the Frank method, the most common reason for failure was a previous, inappropriate attempt at hymenotomy. In our series, no significant difference in dilation success was seen in those patients who had undergone previous operation. This could be explained as a consequence of our small number of patients for whom the attempted dilation failed and/or who underwent a previous hymenotomy attempt. More likely, however, persistent and continuous pressure is achieved with the Ingram method of dilation, which would overcome the scar formation left by the prior operative venture. We believed that the breakdown of follow-up time was of interest for two reasons. Our average follow-up was 9.25 years after the initiation of treatment, and no patient in our series was followed for <2 years. The follow-up for patients for whom dilation failed or who refused a trial was significantly shorter than those patients for whom the procedure was successful. A shorter follow-up time indicates that a particular patient is a more recent addition to our practice. Whether this is an indication of todays soci-

1. Mayer R. Uber Verdoppelungen des Uterus and ihre Arten, nebst Bemerkungen uber Hasenscharte und Wolfsrachen. J Chir Auger 1829;13:525-64. 2. Jones HW, Wheeless CR. Salvage of the reproductive potential of women with anomalous development of the mllerian ducts: 1868-1968-2068. Am J Obstet Gynecol 1969;194:348-64. 3. Frank RT. The formation of an artificial vagina without operation. Am J Obstet Gynecol 1938;35:1053-7. 4. Ingram JM. The bicycle seat stool in the treatment of vaginal agenesis and stenosis: a preliminary report. Am J Obstet Gynecol 1981;140:867-73. 5. Garcia J, Jones HW Jr. The split thickness graft technique for vaginal agenesis. Obstet Gynecol 1977;49:328-32. 6. Rock JA, Reeves LA, Retto H, et al. Success following vaginal creation for mllerian agenesis. 1983;39:809-13. 7. Bryans FE. Management of congenital absence of the vagina. Am J Obstet Gynecol 1981;139:281-4. 8. Frank LR, editor. Quotationary. New York: Random House; 2001. p. 508.

Discussion DR DIONYSIOS K. VERONIKIS, St Louis, Mo. Drs Roberts, Haber, and Rock report on the nonsurgical creation of a neovagina in patients with mllerian agenesis. Their analysis of 51 patients with a mean follow-up of 9 years is the largest series to date to use the Ingram pressure method of neovaginal construction. The authors have applied and adhered to the Ingram technique with a high percentage of the patients successfully creating their own functional neovagina. They have reconfirmed and further delineated the benefits of the Ingram pressure method in the creation of a functional neovagina in these patients. Noteworthy from their data is the time interval, 11.8 1.6 months, required to develop a functional neovagina with dilation. Also, patients with only a vaginal dimple had equal functional success at forming a neovagina, and prior attempted hymenotomy does not appear to be a contraindication to the Ingram method. There are no reported complications with the Ingram pressure method in their study. The Frank method1 of creating a neovagina was not accepted by the patients because of the technical limitations of the method, which include fatigue of the hands and fingers, the need to be in the lithotomy or squatting position, and the loss of productive time during pressure

Volume 185, Number 6 Am J Obstet Gynecol

Roberts, Haber, and Rock 1353

treatment. As described by the authors, the Ingram technique2 eliminates these technical difficulties by using the patients trunk weight in the sitting position. As with the treatment of other congenital anomalies, construction of a neovagina has gone through many changes, with successes and failures. Finally, a method of treatment evolves that is simple, easy to execute, attended by low morbidity rates, and has a high percentage of good results. The findings of the authors clearly support that dilation should be the initial recommendation to patients with vaginal agenesis. I congratulate the authors in providing important and needed clinical data in the treatment of patients with vaginal agenesis. I have the following questions: 1. The functional success endpoint in your study was satisfactory intercourse or acceptance of the largest dilator without discomfort. Because the largest dilator in the Ingram set is 3.5 cm in diameter, in these patients was there a gaping effect to the neovaginal introitus? 2. Because normal vaginal supporting structures are absent and there is no scarring in the Ingram method, have you seen any patients with vaginal prolapse or urinary incontinence? 3. What was recommended as a vaginal lubricant during dilation? Is there a role for a topical anesthetic agent to facilitate dilation, such as 20% benzocaine? 4. The Ingram method also suggested the use of a previous patient who had successfully created her own neovagina as a volunteer. Dr Ingram felt that the successful patient model was a prime component of his technique. Do you feel that a patient model is uniformly required?

REFERENCES

1. Frank RT. The formation of an artificial vagina without operation. Am J Obstet Gynecol 1938:35:1053-5. 2. Ingram JM. The bicycle seat stool in the treatment of vaginal agenesis and stenosis: a preliminary report. Am J Obstet Gynecol 1981;140:867-73.

DR ROBERTS (Closing). Yes, we used functional success as our primary endpoint because that was ultimately what was needed to achieve an improved quality of life for our patients. We did not see any gaping effect in our patients. The introitus is anatomically normal. In addition, we do not have any instances of prolapse in our patients, and we do have quite thorough followup. If we are not regularly seeing them in our clinic, then we have contact with their referring physician. We are obviously continuing to follow our patients for these issues. For lubrication, we use either K-Y jelly or sometimes Sultrin cream (a triple sulfa cream), if any irritated tissue is present. As for the use of patient models, if a patient who has not undergone the procedure needs to talk with a patient who achieved success, we would act as a mediator to set up an appointment between them. We do not do this with every patient. As a routine, we do, however, have contact with every patient and thorough counseling sessions with both the patient and her family. During these counseling sessions, we use both visual aids and an actual set of dilators. We also provide psychiatric help if either the physician or the patient feels that it is necessary after the counseling sessions.