CONSERVATIVE TREATMENT OF ENDOMETRIOSIS: THE EFFECTS OF LIMITED SURGERY AND HORMONAL PSEUDOPREGNANCY*t

Transcription

1 SCIENTIFIC ARTICLES FERTILITY AND STERILITY Copyright c 1976 The American Fertility Society Vol. 27, No.7, July 1976 Printed in U.S.A. CONSERVATIVE TREATMENT OF ENDOMETRIOSIS: THE EFFECTS OF LIMITED SURGERY AND HORMONAL PSEUDOPREGNANCY*t CHARLES B. HAMMOND, M.D.J JOHN A. ROCK, M.D., AND ROY T. PARKER, M.D. Division of Gynecologic Endocrinology and Infertility, Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, North Carolina This study compares the effects of limited surgery or hormonal pseudopregnancy, or a combination of these two, upon fertility and the need for subsequent surgery with respect to the extent of the disease at the time of initial diagnosis in patients with endometriosis externa. Of the 61 patients who desired to enhance or preserve reproductive capacity, 20 patients became pregnant, for a pregnancy rate of 33%. The pregnancy rate in all categories, that is, those patients treated with pseudopregnancy, conservative surgery, and combined pseudopregnancy and surgery, was found to be in direct relationship to the initial extent of disease. In such patients, conservative surgery alone seemed to give the best results in the achievement of pregnancy. There seemed to be little difference between pseudopregnancy alone and conservative surgery in regard to the need for subsequent surgery after initial therapy, although there seemed to be a significantly greater chance for the need for subsequent surgery in patients receiving a combination of the two forms of therapy. The need for subsequent surgery after initial therapy in 80 patients increased in direct relationship to the initial extent of disease present, despite the form of therapy used. Fifty-nine other patients with endometriosis, who did not desire to preserve fertility and presented for relief of other symptoms, underwent initial "radical" therapy. Forty-six patients underwent complete operation, including removal of uterus, tubes and ovaries, and none required subsequent reoperation. Of the 13 remaining patients, who underwent incomplete surgical removal, leaving one or both ovaries in situ, 11 required subsequent reoperation for recurrent pelvic endometriosis. The medical literature is replete with retrospective studies of patients thought to be infertile as a result of endometriosis externa. These studies have contrasted the benefits and detriments of Accepted February 13, *Supported in part by a grant from the G. D. Searle Co., Chicago, Ill. tpresented at the Thirty-First Annual Meeting of The American Fertility Society, April 3 to 5, 1975, Los Angeles, Calif. +Reprint requests: Dr. Charles B. Hammond, Department of Obstetrics and Gynecology, Duke University Medical Center, Durham, N. C various forms of therapy, but, unfortunately, analyses of the results are cumbersome because of a lack of uniformity in the documentation of location and extent of endometriosis present and in the myriad of therapeutic programs utilized. While several authors have suggested classification systems for patients with endometriosis externa, none has been widely used. Acosta et al. 1 and Spangler et al. 2 pointed out that, in the absence of double-blind prospective studies, retrospective analyses of patient groups with regard to therapeutic modalities

2 Vol. 27, No.7 CONSERVATIVE TREATMENT OF ENDOMETRIOSIS 757 remain the major thrust for evaluating the treatment of endometriosis. A recent paper by Andrews and Larsen 3 does provide a comparison of the results of several different forms of treatment for patients with endometriosis. It specifically discusses the need for subsequent surgery and the pregnancy rate that follows therapy and does comment on the initial extent of disease. Their study showed that a significantly greater fertility rate could be obtained by conservative operation alone rather than by hormonal pseudopregnancy alone or by a combination of these two therapies. In fact, the combination of conservative surgery and postoperative pseudopregnancy appeared to decrease the chance of fertility when it was compared with conservative surgical treatment alone. This paper attempts to provide further insight into such questions. The present report is an analysis of the effects of several forms of therapy in patients with varying degrees of, but documented, endometriosis externa. The study investigates the resolution of infertility and the need for subsequent surgery in response to therapy with hormonal pseudopregnancy, conservative surgery, and combined surgery, and combined surgery with postoperative pseudopregnancy. Finally, we present our experience with a limited number of patients with proven endometriosis externa who did not desire further pregnancy and who underwent therapy not oriented toward preservation or enhancement of fertility. MATERIALS AND METHODS The present study is a retrospective analysis extending from January 1, 1962, through December 31, 1974, during which time more than 175 patients with proven endometriosis externa were evaluated by physicians of the Department of Ob- TABLE 1. Classification of Endometriosis" Extent of disease Mild Moderate Severe Findings 1. Scattered, superficial implants on structures other than uterus, tubes, or ovaries; no scarring 2. Rare, superficial implants on ovaries 3. No significant adhesions 1. Involvement of one or both ovaries with multiple implants or small endometriomas ( < 2 em) 2. Minimal peritubular or periovarian adhesions 3. Scattered, scarred implants on other structures 1. Large ovarian endometrium(> 2 em) 2. Significant tubal or ovarian adhesions 3. Tubal obstruction 4. Obliteration of cul-de-sac, major uterosacral involvement 5. Signficant bowel or urinary tract disease "Modified from Acosta et al. 1 stetrics and Gynecology, Duke University Medical Center. Review of operative notes and pathology reports allowed patient classification according to the number and location of implants of endometriosis. The classification system used in this study is a modification of the system presented by Acosta et al. 1 (Table 1). Of the patients reviewed, 139 provided adequate documentation for inclusion into the study (Tables 2 and 3). The 139 patients were divided into two groups TABLE 2. Patient Categorization: 139 Patients with Proven Endometriosis Externa ( ) Group A B Categorization 80 patients who desired to maintain or enhance reproductive function 61 patients presented with infertility 38 (62%) primary infertility 23 (38%) secondary infertility 19 patients presented with pain, etc. 59 patients who did not desire to maintain reproductive function; presented for evaluation of dysmenorrhea, dyspareunia, or abnormal findings on pelvic examination

3 758 HAMMOND ET AL. July 1976 TABLE 3. Patient Population Group A Group B Observation (conservative (radical treatment) treatment) No. of patients Age (yr) Mean Range Parity 0 65% 36% 1 or more 35% 64% Race Caucasian 90% 81% Other 10% 19% Extent of disease Mild 35% 36% Moderate 43% 42% Severe 22% 22% on the basis of the desire to maintain or enhance reproductive function or the lack of this desire. Patients in group A, receiving conservative therapy, were significantly younger than those who initially underwent radical therapy (group B). Infertility was more prevalent among those patients who presented for conservative therapy. The majority of the patients treated for endometriosis were white. The extent of pelvic endometriosis externa was identical in the two groups at the time of initial surgical documentation. RESULTS Conservative Therapy Eighty patients (group A) desired to maintain or enhance reproductive function. Of these 80 patients, 61 presented with the initial complaint of infertility and desired pregnancy at that time. Of the 61 patients, 38 patients (62%) had primary infertility while 23 patients (38%) had secondary infertility. Nineteen patients of the eighty presented with pelvic pain, dysmenorrhea, dyspareunia, or pelvic mass, but desired fertility to be preserved. In these 80 patients the diagnosis of endometriosis externa was confirmed by laparotomy (62%) or by direct endoscopic visualization (laparoscopy or culdoscopy) of implants. Of the 61 patients who presented for initial infertility, evaluation of other causes of infertility included: laparotomy or endoscopy with tubal lavage to document tubal status; documentation of ovulatory status; postcoital test; and analysis of husband's semen. Patients included in this study were considered infertile as a result of endometriosis; patients with other obvious causes of infertility were excluded. All had been infertile for at least 1 year prior to therapy. All 61 patients have been followed for at least 1 year since completion of treatment, and 37 have been followed for more than 2 years. The average length of followup for the entire group of 61 patients was 38 months. The remaining 19 patients did not desire fertility at the time of evaluation, but presented for other symptoms as noted. Evaluation was carried out by laparotomy or endoscopic visualization; therapy was then begun. Hormonal Pseudopregnancy. Thirtythree patients with pelvic endometriosis were intially treated with hormonal pseudopregnancy following documentation of Extent of disease TABLE 4. Effect of Hormonal Pseudopregnancy for Endometriosis Externa (Group A-Conservative Therapy) No. of patients Infertility Subsequent surgery No. infertile Conceptions Tota] no. z- operations Mild (33%) 22 2(9%) Moderate 6 1 0(0%) 6 4 (66%) Severe (25%) 5 3 (60o/o) Total (30%) 33 9 (27%)

4 Vol. 27, No.7 CONSERVATIVE TREATMENT OF ENDOMETRIOSIS 759 the presence and extent of disease (Table 4). While there was some variation in the regimens utilized, the vast majority of these patients were treated with Enovid, 10 to 30 mg/day given continuously for 6 to 9 months. No other specific infertility therapy was carried out. The average Enovid dose was 20 mg/day for an average length of therapy of 8.6 months. A few patients received injectable depo-progesterone (Depo-Provera) for a similar interval of time. Possible complications of hormonal pseudopregnancy occurred in two patients who developed superficial lower extremity thrombophlebitis, one patient who developed severe headaches, and one patient who discontinued therapy because of continued vaginal bleeding while taking 40 mg of Enovid!day. Of the 33 patients, 30 (91%) showed subjective improvement during and after hormonal pseudopregnancy. Such improvement was usually significant. Twenty-five of these patients (75.8%) showed objective responses to treatment during follow-up examinations, usually within 4 months of treatment. The most prevalent objective finding was regression of nodules of endometriosis documented at the time of endoscopy. Eleven of the thirty-three patients (33.3%) had recurrence of symptoms following completion of pseudopregnancy treatment. The average interval to recurrence from the end of pseudopregnancy was 8.4 months, with a range of 3 to 16 months. Six of the eleven recurrences were within 6 months of termination of treatment. Of these 11 patients with recurrent symptoms, 9 (27% of the total group of 33 patients) required subsequent surgery for pelvic endometriosis. When the extent of endometriosis present prior to the initial pseudopregnancy treatment was examined, it was found that, of those patients requiring reoperation, only 2 of22 patients had mild endometriosis (9% ), while 4 of 6 patients who had moderate endometriosis (66%) required subsequent reoperation. Thus, the more extensiy.e the diilea.s ~Ctile -fime-"of initiation of seudo re - nancy, the mo 1 el were symptoms ft-recur and consequently the need for subsequent surgery or these 33 patients, 22 were found to have mild endometriosis, and 18 of these had presented with an initial complaint of infertility. Of this number, six (33%) conceived following termination of the pseudopregnancy regimen. Of the remaining 11 patients with more extensive pelvic endometriosis, 5 had presented with the initial complaint of infertility. Only one of these 5 patients conceived following treatment. Thus, 7 of the 23 patients treated with pseudopregnancy and who desired pregnancy at that time conceived, a pregnancy rate of 30%. Of the seven patients who achieved pregnancy, six (85.7%) had mild endometriosis. One patient had severe disease. The average time to conception was 10.4 months from the end of hormonal treatment, with the range being 2 to 18 months. Four of the seven patients became pregnant within 1 year of termination of pseudopregnancy. Three of the seven patients who conceived had normal term deliveries, and the remaining four patients had early abortions (57%). Conservative Surgery. Twelve patients desired to preserve or enhance reproductive capacity and underwent conservative surgery alone (Table 5). Conservative surgery included laparotomy, extirpation of implants of endometriosis, lysis of adhesions, fulguration or excision of peritoneal implants, and uterine suspension. None of these patients received hormonal therapy preceding or following the surgical procedure. There were no serious complications resulting from conservative surgery. All 12 of the patients who underwent primary conservative surgery without hormonal therapy achieved remission of

5 760 HAMMOND ET AL. July 1976 Extent of disease TABLE 5. Effect of Conservative Surgery for Endometriosis Extema (Group A-Conservative Therapy) No. of patients No. infertile Mild 3 3 Moderate 7 5 Severe 2 2 Total Infertility Subsequent surgery Conceptions Total no. T operations 2 (66%) 3 0(0%) 3(60%) 7 2 (2990) 0(0%) 2 1 (50%) 5 (50%) 12 3 (25%) symptoms of endometriosis in the postoperative setting. Of these 12 patients (25% ), 3 required reoperation for recurrent symptoms, with the average time to recurrence of 40 months. This was more likely to occur in patients with more extensive disease at the time of initial operation. Ten of these twelve patients had presented with an initial complaint of infertility; following surgery, five conceived, for a pregnancy rate of 50%. When these patients were grouped according to the initial extent of disease it was found that two of three patients with mild disease conceived (66%), and three of seven patients with moderate or severe endometriosis conceived following surgery (43%). All five patients who conceived after surgery delivered normal infants at term. There were no abortions. Average time to pregnancy was 20 months from surgery, the range being 5 to 36 months. The average follow-up after initial surgery was 2.9 years, with a range of 1 to 12 years. Combined Therapy (Conservative Surgery with Hormonal Pseudopregnancy. Thirty-five patients who desired to pre- serve reproduction or enhance fertility were treated with a combination of initial conservative surgery followed by hormonal pseudopregnancy (Table 6). Conservative surgery was more often performed in patients who were shown endoscopically to have more extensive disease. Only three patients were felt to have mild endometriosis. Twentyone of these patients were found to have moderate endometriosis, while eleven had severe disease. All 35 patients treated with combined therapy achieved remission of symptoms of pelvic endometriosis. However, 19 of the 35 patients (52. 7%) had recurrence of symptoms of pelvic endometriosis following combined therapy. The average time from the end of therapy to recurrence was 18 months, with a range of 4 to 39 months. Ofthis number, 18 (51%) required subsequent reoperation. However, 10 of 21 patients with moderate endometriosis and 8 of 11 patients with severe disease required subsequent reoperation for recurrence. The more extensive the disease at the time of therapy, the more likely the need for reoperation. TABLE 6. Effect of Conservative Surgery and Hormonal Pseudopregnancy (Combined Therapy) for Endometriosis Externa (Group A-Conservative Therapy) Extent of disease No. of patients Infertility Subsequent surgery No. infertile Conceptions Total no. T operations Mild (100%) 3 0 (0%) Moderate (25%) (48%) Severe (14%) 11 8 (73%) Total (31%) (51%)

6 Vol. 27, No. 7 CONSERVATIVE TREATMENT OF ENDOMETRIOSIS 761 Type of therapy TABLE 7. Comparison of Various Therapies for Endometriosis Externa (Group A-Conservative Therapy) No. of patients Pseudopregnancy Conservative surgery Combined treatment (surgery & pseudopregnancy) Total Infertility Subsequent surgery No. infertile Pregnancies Total no. 2" operations 7 (30%) 33 9 (27%) 5 (50%) 12 3 (25%) 8(31%) 35 18(51%) 20 (34%) (37%) Of the 26 patients in this group who presented with infertility, 8 patients (31%) became pregnant. All three patients with mild endometriosis treated in this fashion became pregnant. Four ofthe sixteen patients with moderate endometriosis and one of the seven patients with severe endometriosis achieved pregnancy following combined treatment. It was noted that the more extensive the disease at the time of the initial operation, the less likely was pregnancy to occur after combined therapy. Of the eight patients in this group who achieved pregnancy, seven had normal deliveries and one had an early spontaneous abortion. The average time to pregnancy was 15.6 months from the termination of pseudopregnancy, with a range of 8 to 41 months. Two of the eight patients were pregnant within 1 year, seven of the eight patients were pregnant within 2 years, and the eighth patient was pregnant within 3.4 years. Comparison of the Various Therapies in Conservative Treatment. The results of the three therapeutic regimens in this group of 80 patients who desired to preserve or enhance reproductive capability are compared in Table 7. Thirty-seven per cent of these patients later required subsequent reoperation for symptoms of recurrent endometriosis. However, essentially all patients achieved remission of symptoms following the initial form of therapy, regardless of the type of therapy. If one analyzes the likelihood of recurrent symptoms after initial treatment, it is apparent that the combined therapy of surgery followed by hormonal pseudopregnancy seemed more likely to result in the need for subsequent reoperation than did either conservative surgery alone or pseudopregnancy therapy alone. In all of these categories, both the pregnancy rate and the need for subsequent surgery seemed to be related to the extent of the disease at the time of diagnosis; the greater the extent the more likely was recurrence and the less likely was pregnancy to be achieved. The highest pregnancy rate after therapy was found to be among patients treated with initial conservative surgery alone. There seemed to be a reduction in the conception rate in patients treated with conservative surgery if it was followed by hormonal pseudopregnancy. There seemed little difference in the pregnancy rate following pseudopregnancy alone when compared with combined conservative surgery followed by pseudopregnancy. However, one must recall that patients who underwent conservative surgery likely had more extensive pelvic endometriosis than patients treated with pseudopregnancy alone. Radical Therapy Fifty-nine patients (group B) did not desire to maintain their reproductive function and presented for evaluation of dysmenorrhea, dyspareunia, or abnormal findings on pelvic examination (Table 8). All of these patients underwent extirpa-

7 762 HAMMOND ET AL. July 1976 TABLE 8. Radical Therapy for Endometriosis Externa (Group B-59 Patients) Therapy No. of Subsequent patients surgery Complete operation, including 46 (78%) 0 both ovaries Incomplete operation leaving 13 (22%) 11 (85%) one or both ovaries Total tive surgery and the diagnosis was confirmed histologically and categorized as to location and amount. These patients were older, more likely parous, but had essentially the same extent of disease as did group A patients. Of these 59 patients, 46 underwent complete operation, including removal of the uterus, tubes, and both ovaries. The vast majority of these patients were returned to full estrogen replacement therapy prior to discharge after surgery. Thirteen of the fifty-nine patients underwent incomplete surgical removal which consisted of removal of the uterus alone, or of the uterus and only one ovary. All patients in this group have been followed for at least 1 year since the primary surgical procedure. All of these 59 patients achieved remission of the symptoms of endometriosis following surgery. Symptoms referrable to recurrent endometriosis were of sufficient severity to require reoperation in 11 of the 13 patients who had initially had incomplete operation (85% ). The range of follow-up was 1 to 5 years. None of the 46 patients who initially underwent complete operation (and none of the 11 patients who subsequently underwent complete operation) have required reoperation, despite prompt and adequate postoperative estrogen replacement. DISCUSSION Although endometriosis has been under close observation and intense study for more than 40 years, the disease often remains an enigma to those who must diagnose and treat it. The etiology is not established, the clinical and pathological manifestations are often obscure, and the methods of treatment are frequently ineffective. The natural course of endometriosis is often that of a progressive, painful, sterilizing, and crippling disease. It has been defined as "the presence of functioning endometrium outside its usual placement in the uterine cavity, but usually confined to the pelvis in the region of the ovaries, uterosacral ligaments, and uterovesicle peritoneum."4 Whether the disease process usually develops early in life and gives rise to symptoms only after years of subclinical activity or develops shortly before clinical manifestations occur is unknown. It is well known that endometriosis is an important factor in infertility. Infertility has been reported as a primar:)" complaint in 6 to 15% of patients with endometriosis, although it is estimated that 30 to 40% of patients with endometriosis are infertile. 5 The cause and effect are unclear. Carter 6 presented several items the gynecologist should consider prior to the institution of a therapeutic regimen for endometriosis. Among these are "the couple's desire for progeny, the age of the patient, the duration of infertility, and the severity of symptoms and associated pelvic pathology." Before a therapeutic modality is instituted, the lesion must be proven to be endometriosis, the location and extent of disease established and a thorough infertility survey completed, and commitment between physician and patient for long-term followup should be made. Surgical procedures have become the major therapeutic measure for the treatment of endometriosis, and pregnancy success rates of 40 to 65% have been reported ' 10 In a young woman who desires to preserve or enhance fertility,

8 -~ Vol. 27, No.7 CONSERVATIVE TREATMENT OF ENDOMETRIOSIS 763 the operative procedure should be conservative, that is, removal of the smallest portion of reproductive tissue that will eradicate or control the disease and repair of damaged tubes and/or ovaries in the hope of future pregnancy. In the woman in whom fertility is not a consideration, a more radical operation is usually performed in an effort to obtain symptomatic relief. There has been much research in an effort to develop an effective therapeutic approach to endometriosis. The use of estrogenic hormonal preparations such as diethylstilbestrop 1 in doses adequate to eliminate cyclic ovarian function, the use of potent progestins to develop and maintain a decidual state in the endometrium, and the administration of small, subvirilizing, amounts of androgen 14 for prolonged periods have been found to relieve symptoms and promote pregnancy in selected patients. The use of these types of hormonal preparations, however, is not without undesirable features and complications The pregnancy success rates reported for hormonal therapy alone are generally lower than those reported for conservative surgery alone. More recently, Danazol has afforded a new hope in medical therapy, although reservation is warranted in view of past experience with new medical modalities. 15 The classification of endometriosis externa is not a new concept. Wick and Larsen 16 presented a classification based on "grades." A criterion for histologic grading of these lesions was presented, and a plea given for a survey correlating the clinical symptomatology and the physical findings with the grading of endometriosis. Riva et al. 13 presented a classification based on the location of endometriosis recorded at the time of culdoscopy. The extent of disease was based on the number of pelvic structures involved. They noted that the majority of those patients treated with hormonal pseudopregnancy with minimal disease had no evidence of endometriosis at the time of re-evaluation after 1 to 6 months of progestin therapy. This correlation was also found after 7 to 15 months of therapy. Although the classification by Riva et al. 13 did not accurately describe the extent of endometriosis present, there was a direct relationship between the number of pelvic organs involved and the response to hormonal pseudopregnancy. Beecham 17 suggested a classification for endometriosis externa based on either palpable findings on pelvic examination or palpable-visual findings at operation. He maintained that descriptions of each stage contain a minimum of useful words, while describing accurately the extent of disease present. The descriptions of the amount of disease in each stage are very brief in his classification system, and palpable subjective findings lend some confusion. Acosta et al. 1 composed a concise classification of pelvic endometriosis based on the sites and amounts of endometriosis present at the time of surgery (Table 1). Each area of classification, that is, mild, moderate, and severe, was precisely defined. This classification system was unique in that it recognized paraovarian and peri tubular adhesions as important visual findings in differentiating moderate from mild endometriosis. These findings at the time of laparoscopy would be held by many as criteria for proceeding with conservative surgery rather than utilizing hormonal therapy in those patients with minimal disease. Most recently, Mitchell and Farber 5 recommended a less complicated staging classification for pelvic endometriosis based on the sites and amounts of endometriosis present at the time of surgery, similar to that described by BeechamY Their system was unique in that it characterized stage V as endometriosis giving rise to adenocarcinoma. Certainly all of the above classification systems have their advantages and disadvantages. The problem existing for

9 764 HAMMOND ET AL. July 1976 most practicing gynecologists is that no uniformly accepted system of documentation has been devised to assist them in their therapeutic regimens. The present study utilized the classification systems of Acosta et al. 1 for the following reasons: (1) The system presented seemed to be the most concise and complete in the literature to date. (2) A direct relationship was established between the initial extent of disease and the pregnancy success rate at the time of initial surgery. In 1974 Andrews and Larsen 3 presented one of the most complete studies to date comparing treatment modalities and suggested that conservative surgery offered the best chance for pregnancy as compared with pseudopregnancy or a combination of surgery and postoperative pseudopregnancy. The authors demonstrated that postoperative pseudopregnancy lowered the pregnancy rate expected for conservative surgery to that expected for pseudopregnancy alone. As more than one-half of conceptions after conservative surgery reportedly occurred within 1 year, it was understandable that ovulatory suppression during this critical period was detrimental. In the present study a combination of conservative surgery and postoperative psuedopregnancy was generally reserved for those patients with advanced endometriosis. Thus, 23 patients with advanced disease were so treated. The overall pregnancy rate fo 31% was similar to that obtained with pseudopregnancy alone. The average time to pregnancy was 15.6 months, differing from that previously reported. 1 3 These data appear to supportthe opinion~f AnarewsHand Larsell?l j~ that the use of postoperative pseudopregnancy did not enhance the chance for conception, but actually decreased it. Acosta et al. 1 have shown a direct correlation between the pregnancy success rate and the extent of disease present at the time of conservative surgery. Riva et al. 13 demonstrated a correlation between pelvic organ involvement and response to hormonal pseudopregnancy. The present study demonstrated not only a correlation between the pregnancy success rate and the extent of endometriosis but also a direct relationship between the need for subsequent surgery and the extent of disease. The pregnancy success rate of 50% for conservative surgery was consistent with that previously reported by several authors. The number of patients in this series is small,and therefore must be viewed with some reservation. The average follow-up was 3 years, but the average time to pregnancy after conservative surgery was 20 months. This differs from the generally accepted view that a majority of the pregnancies would occur within 1 year postsurgery. A majority of the patients with minimal disease at the time of endoscopy were treated with hormonal pseudopregnancy. The highest pregnancy rate with this type of therapy was reported in those patients with minimal disease, 33%. Of the seven patients who achieved pregnancy, six had mild endometriosis. This finding supports the generally held belief that there is no place for pseudopregnancy in advanced endometriosis. As expected, symptomatic relief was attained in essentially all (91%) of the patients treated with hormonal pseudopregnancy. In comparing the three forms of therapy with regard to the pregnancy success rate, it appears that conservative surgery offers the best chance for conception, and the addition of postoperative pseudopregnancy is not warranted. A discussion of conservative therapy for endometriosis is not complete without attention to the need for subsequent surgery. Williams 10 reported a 10 to 12% risk of subsequent surgery after conservative surgery alone. Spangler et al. 2

10 Vol. 27, No. 7 CONSERVATIVE TREATMENT OF ENDOMETRIOSIS 765 reported that 13 of 101 patients treated with conservative surgery alone required reoperation. Andrews and Larsen 3 noted that the addition of pseudopregnancy did not decrease the incidence of subsequent reoperation for endometriosis. Comparison of the various modalities in our studies revealed little difference in conservative surgery or pseudopregnancy alone, but a combination of these two treatments seemed more likely to result in the need for future reoperation. The average length of time from initial therapy to recurrence was longer (40 months) in those patients treated with conservative surgery alone, whereas in patients treated with pseudopregnancy alone, or in patients treated with a combination of pseudopregnancy and conservative surgery, the time interval to recurrence was 8.4 months and 15.6 months, respectively. It appears that conservative surgery affords a longer interval of time between initial operation and time of recurrence. Gray, 7 in his discussion of the surgical treatment of endometriosis, stated that "when hysterectomy is necessary it is not good judgement to leave a very small portion of ovarian tissue if doing so would almost certainly lead to a necessary second operation." Ranney 18 presented certain criteria which he believed would aid the surgeon in deciding whether to remove ovarian tissue. He felt that one should not attempt the preservation of ovarian tissue if (1) the hilar region of both ovaries contained deep plaques of endometriosis; (2) ifthere was extensive endometriosis deeply involving bowel or ureters which could not be carefully and thoroughly resected; and (3) if the operation was an exploratory emergency operation due to hematoperitoneum caused by spontaneous rupture of endometriotic implants into the pelvis during the secretory decidual phase of the cycle or secondary to progestin therapy, and if hysterectomy and bilateral salpingo-oophorectomy were indicated because of associated or dissociated pelvic pathology. Andrews and Larsen 3 reported a relatively low rate of recurrence (3.5%) following hysterectomy with one or both ovaries remaining. A recurrence rate of 4.1% was reported by Rogers and Jacobs9 in similarly treated cases. This would suggest the advisability of ovarian conservation in younger women with endometriosis. In our series, 46 of 59 patients did not desire to maintain their reproductive capacity and underwent complete operation, including removal of the uterus, tubes, and ovaries, and did not require subsequent reoperation. The vast majority of these patients were returned to full estrogen replacement therapy prior to discharge after surgery. Of 13 patients who had initially undergone an incomplete operation consisting of the removal of the uterus alone or removal of the uterus and only one ovary, 11 required reoperation for recurrent symptoms or pelvic mass found to be due to recurrent endometriosis. Again, none of the 46 patients who initially underwent complete operation and none of the 11 patients who subsequently underwent complete operation have required reoperation despite prompt and adequate estrogen therapy. In our experience complete removal of the uterus, tubes, and both ovaries was a beneficial procedure for those patients who did not desire further reproduction. However, leaving one or both ovaries in place was associated with an 85% chance for the later need for reoperation for recurrent or persistent disease. Analysis of results from retrospective studies is cumbersome because of the lack of uniformity in documentation of the location and extent of endometriosis present. The present study is an attempt to classify the sites and amounts of endometriosis present at the time of

11 766 HAMMOND ET AL. July 1976 endoscopy and laparotomy and to correlate these findings with the pregnancy success rate and the need for subsequent surgery. This study is not unlike those previously reported in that it is a nonrandomized, retrospective analysis of a small patient group. This study points to the need for more randomly assigned, larger patient groups to allow better comparison of the several methods of therapy, and we strongly urge the adoption of a uniform diagnostic and staging classification. REFERENCES 1. Acosta AA, Buttram VC, Besch PK, Malinak LR, Franklin RR, Vanderheyden JD: A proposed classification of pelvic endometriosis. Obstet Gynecol 42:19, Spangler DB, Jones CS, Jones HW: Infertility due to endometriosis. Am J Obstet Gynecol 109:850, Andrews WC, Larsen DC: Endometriosis: treatment with hormonal pseudopregnancy and/or operation. Am J Obstet Gynecol 118:643, Kistner RW: The treatment of endometriosis by inducing pseudopregnancy with ovarian hormones. Fertil Steril 10:539, Mitchell GW, Farber M: Medical vs. surgical management of endometriosis. In Controversy in Obstetrics and Gynecology, Vol 2, Edited by DE Reid, CD Christian. Philadelphia, WB Saunders Co, 1974, p Carter BC: Treatment of endometriosis. J Obstet Gynaecol Br Commonw 69:783, Gray LA: Endometriosis. Clin Obstet Gynecol 3:472, Ranney B: Endometriosis. I. Conservative operations. Am J Obstet Gynecol 107:743, Rogers SF, Jacobs WM: Infertility and endometriosis: conservative surgical approach. Fertil Steril 19:529, Williams TJ: The role of surgery in the management of endometriosis. Mayo Clin Proc 50:198, Karnaky KJ: The use of stilbestrol for endometriosis: preliminary report. South Med J 41:1109, Andrews MC, Andrews WC, Strauss AF: Effects of progestin-induced pseudopregnancy on endometriosis. Am J Obstet Gynecol 78:776, Riva HL, Kawasaki DM, Messinger AJ: Further experience with norethynodrel in treatment of endometriosis. Obstet Gynecol 19:111, Preston SN, Campbell HB: Pelvic endometriosis-treatment with methyltestosterone. Obstet Gynecol12:152, Ansbacher R: Treatment of endometriosis with Danazol. Am J Obstet Gynecol 121:283, Wick MJ, Larsen CD: Histologic criteria for evaluating endometriosis. Northwest Med 48:611, Beecham CT: Classification of endometriosis (editorial). Obstet Gynecol 28:437, Ranney B: Endometriosis. III. Complete operations. Am J Obstet Gynecol 109:1137, 1971