- Email: [email protected]
Primary staging in ovarian tumors of low malignant potential
GYNECOLOGIC ONCOLOGY 31, 402--408 (1988)
Primary Staging in Ovarian Tumors of Low Malignant Potential ROBERTO YAZIGI,
M.D.,* JULIE
SANDSTAD,
M.D., AND
ALAN K .
MuNoz, M . D . *
*Division of Gynecologic Oncology, *Department of Obstetrics and Gynecology, and Department of Pathology, University of Texas, Southwestern Medical School, Dallas, Texas 75235 Received March 10, 1987 Surgical staging, consisting of peritoneal washings for cytology, infracolic omentectomy, and biopsies of diaphragm, extrapelvic peritoneum, and pelvic and aortic lymph nodes, was performed in 29 patients with ovarian tumors of low malignant potential, presumed to be either Stage 1 (25) or Stage 11 (4), in order to determine the incidence of unsuspected metastases in patients with localized disease. Fourteen patients had all and fifteen patients had one or more of these procedures performed. Overall, in stages 1 and II, positive peritoneal cytology was found in 7%, unexpected omental metastases in 13%, diaphragmatic metastases in 7%, positive pelvic lymph nodes in 27%, and positive aortic lymph nodes in 7%. Seven out of 29 (24%) patients with presumed localized disease, were upstaged by virtue of the staging procedures. Based on our findings, we conclude that surgical-pathologic staging to search for occult metastases in ovarian tumors of low malignant potential is justified from an investigational standpoint: however, its impact on therapeutic management is far from being defined. © 1988AcademicPress. Inc.
INTRODUCTION Ovarian tumors of low malignant potential were first described by Taylor in 1929 [1]; official recognition as a pathologic entity by the International Federation of Gynecology and Obstetrics (FIGO) was given in 1971 [2] and by the World Health Organization in 1973 [3]. Tumors of low malignant potential comprise about 15% of all ovarian epithelial cancers [4,5], and tend to occur at an earlier age than do the usual ovarian carcinoma [4,6,7]. Extra ovarian manifestations do not always correlate with outcome [8], and recurrences within the pelvis or abdomen may develop after long latent intervals of up to 20 years [9-11]. These neoplasms are distinguished from the frankly malignant invasive epithelial tumors of the ovary because of their apparently more indolent biologic behavior. Fortunately an evolving set of morphologic criteria help identify these tumors, so that these patients can be followed as a group [7,12-15]. Much progress has been made on the natural history of ovarian adenocarcinoma by detecting sites of subclinical metastases of apparently early stage disease [16,17], upstaging up to one third of presumed stage I or II patients.
402 0090-8258/88 $1.50 Copyright © 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
STAGING IN LOW MALIGNANTOVARIANTUMORS
403
Previous reports of tumors of low malignant potential have retrospectively staged patients without the adequate surgical staging our present day knowledge dictates. The main purpose of this study is to evaluate the role of careful surgicalpathologic staging in patients with tumors of low malignant potential, focusing on the upstaging of cases with apparent localized disease. MATERIALS AND METHODS Forty-seven patients aged 19-75 years (mean 42, median 39) from 1972 to the present, from Parkland Memorial and St. Paul Hospitals, were diagnosed with ovarian serous or mucinous tumors of low malignant potential. All histologic sections from the original diagnostic material were reviewed by one of the authors (J.S.) with no knowledge of clinical outcome. Six of the 47 cases were excluded due to inadequate histological material for review. Two cases were excluded after review upgrading to frank carcinoma. One case was excluded after review diagnosis change to cystadenofibroma. Clinical data were reviewed on the 38 remaining cases. Thirty six of the 38 patients (95%) were presumed to have stage I or II disease at the time of primary exploration (31 stage I and 5 stage II); the remaining two patients were found to have gross intraabdominal disease at the time of primary surgery and so were initially staged as III during the procedure. These last two patients are excluded from further staging analysis. Patients were surgically staged according to FIGO classification (see Table 1). The initial surgical procedures in the 36 patients with apparent stage I or II disease consisted of hysterectomy and bilateral oophorectomy in 22 cases and unilateral salpingo-oophorectomy in 14 cases, 8 of which had a contralateral ovarian wedge resection. In three cases the contralateral ovary was surgically absent. In addition to this approach, 29 of these 36 patients (80%) had one or more of the following staging procedures performed: peritoneal washings, infracolic omentectomy, diaphragmatic biopsies, multiple peritoneal biopsies, and lymph node sampling (pelvic and aortic). Fourteen patients (48%) had all of these procedures performed and 15 (52%) had one or more performed.
TABLE 1 PATIENT CHARACTERISTICS (N = 36) Mean Age (range years); Presumed FIGO Stage IAi IAii IBi IBii IC IIA liB
42 (19-75) N
%
21 4 2 1 3 1 4
58 11 6 3 8 3 11
404
YAZIGI, SANDSTAD, AND MUNOZ
Peritoneal washings were obtained on entering the abdomen. Approximately 100 ml of saline solution was placed into the pelvis and aspirated; the fluid was immediately sent for analysis. Diaphragmatic biopsies were obtained under direct vision. Pelvic nodes sampled varied from 1 to 25, with an average of 8. Selective aortic lymph node biopsy was carried out by removing the largest palpable nodes from the bifurcation up to the inferior mesenteric artery and varied from 1 to 10 with an average of 3. The histology of the tumors proved to be serous in 25 patients (69%) and mucinous in 11 (31%). None fit recently delineated criteria for seromucinous type [15]. Morphologic criteria utilized for diagnosis of ovarian serous tumors of low malignant potential (borderline) include the formation of tumor papillary projections of varying complexity with lining cell stratification, cytologic atypia and/or mitotic figures, and tufting or detachment of cell clusters floating free among the papillary projections [7,14,15]. Morphologic criteria utilized for the diagnosis of mucinous tumors of low malignant potential (borderline) include villoglandular growth pattern of varying complexity, mitotic figures and/or cytologic atypia, and cell stratification [7,12,13,15]. These changes seen in both tumor types take place within the tumor cystic spaces, and no evidence for ovarian stromal or capsular invasion is present. RESULTS
Clinical Symptoms The most frequent complaint was abdominal pain, seen in 49% of patients, followed by: increased abdominal girth, abdominal mass, dyspareunia, and vaginal bleeding. Eleven patients (30%) were asymptomatic and tumor was found on a routine exam.
Findings At Surgical Staging Contralateral ovary. The contralateral ovary was explored in 27 patients with presumed stage I disease, either by oophorectomy (19) or wedge resection (8). Four patients (15%) had bilateral disease. In no patient with bilateral involvement was this an unexpected microscopic finding since all four had evident macroscopic disease at the time of surgery. Therefore, in 23 patients in which the tumor was apparently confined to one ovary no patient was upstaged based on contralateral ovarian biopsies. Peritoneal cytology. Peritoneal washings for cytology were obtained in 27 apparent stage I or II patients, and two were positive (7%). These were found in presumed stage IA and IB patients who were also found to have microscopic metastatic disease to the aortic and pelvic lymph nodes, respectively. Ascites was present in three stage I patients; none showed cytologic malignancy. Omentum-diaphragm-peritoneum. Omentectomy was performed in 24 patients with apparent stage I or lI disease; three patients (13%) revealed microscopic implants, all in presumed stage IIB patients. No patient considered to be stage I had positive omental findings.
405
STAGING IN LOW MALIGNANT OVARIAN TUMORS
Diaphragmatic biopsies were obtained in 15 patients with presumed stage I or I! disease, and of these, only one (7%) was positive in an apparent stage II-B patient. No patient was upstaged by virtue of unexpected positive peritoneal biopsies. Pelvic lymph nodes. Pelvic lymph node sampling was performed in 15 cases, 11 of which were apparent stage IA, 1 apparent stage IB, 1 apparent stage IC, and 2 apparent stage IIB. Four patients (27%) had metastatic lymph node disease; two stage IA patients, one stage IB patient, and one stage liB patient. Aortic lymph nodes. Aortic lymph nodes were sampled in 15 cases thought to be stage I or II and only one case (7%) had metastatic lymph node disease and was therefore upstaged. This patient had negative pelvic node sampling. Overall, on the basis of lymph node biopsies, 5 of 15 patients sampled (33%) were found to have unexpected microscopic metastases. See Table 2 for summary of findings. Upstaging. Based on the above described procedures, of 25 presumed stage I patients who underwent full or partial staging, four (16%) were upstaged to stage III by virtue of lymph node metastases (three pelvic and one aortic). Of four presumed stage II patients, three (75%) were upstaged to stage III due to omental microscopic implants (plus diaphragm and pelvic nodes, respectively, in each patient). Altogether, of the 29 staged patients with presumed stage I or II disease, 7 (24%) were upstaged by these surgical procedures (Table 3). DISCUSSION Obtaining random biopsies during surgical staging of ovarian cancer to detect sites of subclinical metastases has received special emphasis recently. This has resulted in upstaging a significant proportion of patients with presumed localized disease. The relevance of this fact is obvious regarding management and prognosis. Ovarian tumors of low malignant potential, however, are considered to be different pathologic and clinical entities, and staging procedures in order to detect early extraovarian spread may not have the same meaning as in invasive carcinoma. Despite this, we know that those patients with evidence of extraovarian disease may experience up to a 36 to 40% 10-year mortality rate [14,5]. Thus, adequate TABLE 2 STAGINGPROCEDURESANDFINDINGSINPRESUMEDSTAGESIANDII(N = 29)
Procedure Pertioneal cytology Omentectomy Diaphragm biopsy Peritoneal biopsy Pelvic lymph nodes Aortic lymph node
Stage I (=25)
Stage II (=4)
No. +/
No. +/
Total (=29)
No. +/
No. Sampled
No. Sampled
Sampled
2/25 0/20 0/11 0/11 3/13 1/12
0/2 3/4 1/4 0/4 1/2 0/3
2/27 3/24 1/15 0/15 4/15 1/15
(7%) (13%) (7%) (0%) (27%) (7%)
406
YAZIGI, SANDSTAD, AND MUNOZ TABLE 3 COMPARATIVE [N1TIAL AND FINAL STAGE
Patient 1 2 3 4 5 6 7
Initial stage IAi IAi IAi IBi liB liB IIB
Positive findings
Final stage
Pelvic L / N Pelvic L / N Aortic L/N Pelvic L / N Ornentum-diaphragm Omentum Omentum-Pel L/N
Ill II1 III III III Ill Ill
staging provides prognostic information in tumors of low malignant potential although the data suggest that, for instance, a stage III low malignant potential tumor has a consistently better outcome than a stage III, grade 2 or 3 cystadenocarcinoma. The dilemma of deciding whether to perform surgical staging in patients with tumors of low malignant potential can only be resolved by (1) determining what is the real yield of staging laparotomies in presumed early stage tumors and, (2) determining if upstaging and its concommitant change in therapeutic management exert any impact on final outcome. This study essentially addresses the first issue, and our results point to several important findings. Of patients thought to have stage I or II disease at the time of primary surgery when surgical staging procedures were performed, 13% (3/24) had microscopic omental metastasis, 7% (1/15) had occult diaphragmatic metastasis, and 33% (5/15) had positive lymph nodes. Overall, of the 29 staged patients with presumed early stage disease, 7 (24%) were upstaged to stage III as the result of these procedures. In a similar study, Nation and Krepart [18] upgraded two cases from stage I to III based on microscopic omental metastasis. Remarkable is the finding that none of our 25 patients with stage I disease were found to have histologically confirmed occult intraperitoneal spread; however, 3 of 13 patients sampled had retroperitoneal lymph node metastases. On the contrary, 75% (3/4) of our stage II patients showed occult intraabdominal extension of their tumor. The significance of omental, diaphragmatic, or peritoneal implants has been a matter of controversy, centering on the issue that these implants may indeed not be true metastases, but multicentric proliferation of coelomic mesothelium and mesenchyme, which have retained some Mullerian potential at extraovarian sites. Attempts to correlate different microscopic features with prognosis have been made [19], however unsuccessfully. The great issue here is how best to interpret the significance of extraovarian deposits when fortuitously found at a microscopic level related to the patients outcome. Lymph node involvement by tumors of low malignant potential have been previously described in pelvic, aortic, mediastinal, supraclavicular, axillary, and inguinal lymph nodes in Bergman's autopsy study [20]. Ehrmann et al. [21] raised the important issue of benign glandular inclusions in lymph nodes being confused
STAGING IN LOW MALIGNANT OVARIAN TUMORS
407
with metastatic tissue. Our five cases with lymph node metastases all showed complex cribiform or papillary structures with cytologic atypia, sometimes mitoses, and tufting and were not difficult to differentiate from benign glandular inclusions. Interestingly, the presence of extraovarian implants in 16-47% of cases [4,22,14,5,23] herald those patients who die of their disease. For stage IA patients who desire fertility, most gynecologic oncologists would agree that conservative treatment is justified and outcome after unilateral salpingooophorectomy seems to be similar to hysterectomy with bilateral salpingooophorectomy [19]. Our findings in the contralateral ovary (by oophorectomy or wedge biopsy) agree with the data from Talezaar et al. [24] None of our 23 patients with disease apparently confined to one ovary had positive microscopic findings in the contralateral ovary. In Talezaar's series, one patient developed a subsequent serous tumor of low malignant potential in a previously negative biopsied ovary. Williams [25], however, found "malignancy" in the contralateral ovary in 7% of his cases which appeared grossly normal at surgery. Despite our findings, we still believe that contralateral wedge biopsy should be performed, based on the incidence of bilateral involvement reported by other authors [26,5]. The results of our study lead us to believe that: Staging laparotomy for tumors of low malignant potential appears justified from an investigational standpoint, on the basis of a 24% incidence of upstaging in presumed stages I and II disease. However, its impact on management and survival will have to wait until some form of therapy is identified that benefits patients with early extraovarian spread. The value of adjuvant therapy in stage I disease has been negated by the one randomized study published [27]. The use of chemotherapy for advanced disease awaits :further randomized trials. The ongoing study by the Gynecologic Oncology Group may help resolve the issue. REFERENCES 1. Taylor, H. C. Malignant and semi malignant tumors of the ovary, Surg. Gynecol. Obstet. 48, 204 (1929). 2. Classification and staging of malignant tumors in the female pelvis, Acta Obstet. Gynecol. Scand. 50, 1 (1971). 3. Serov, S. F., Scully, R. E., and Sobin, L. H. Histological typing of ovarian turnours, World Health Organization, Geneva (1973). 4. Aure, J. C., Hoeg, K., and Kolstad, P. Clinical and histologic studies of ovarian carcinoma, Obstet. Gynecol. 37, 1 (1971). 5. Russell, P. Borderline epithelial tumours of the ovary: A conceptual dilemma, Clin. Obstet. Gynaecol. 11, 259 (1984). 6. Purola, E. Serous papillary ovarian tumors, Acta. Obstet. Gynecol. Scand. (Suppl.) 42, 3 (1963). 7. Colgan, T. J., and Norris, H. J. Ovarian epithelial tumors of low malignant potential: A review, lnt. J. Gynecol. Pathol. 1, 367 (1983). 8. Tasker, M., and Langley, F. A. The outlook for women with borderline epithelial tumours of the ovary, Brit. J. Obstet. Gynaecol. 92, 969 (1985). 9. Malloy, J. J., Dockerty, M. B., Welch, J. S., et al. Papillary ovarian tumors, Amer. J. Obstet. Gynecol. 93, 867 (1965). 10. Woodruff, J. D., and Julian, C. G. Histologic grading and morphologic changes of significance
408
11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
22. 23. 24. 25. 26. 27.
YAZIGI, SANDSTAD~ AND MUNOZ
in the treatment of semi malignant and malignant ovarian tumors, in Proceedings o f the Sixth National Cancer Conference, Lippincott, Philadelphia/Toronto, pp. 347-351 (1968). Minyi, T., Lijuan, L., and Tonghua, L. The characteristics of ovarian serous tumors of borderline malignancy, Chin. Med. J. 93, 459 (1980). Hart, W. R., and Norris, H. J. Borderline and malignant mucinous tumor of the ovary: Histologic criteria and clinical behavior, Cancer 31, 1031 (1973). Chaitin, B. A., Gershenson, D. M., and Evans, H. L. Mucinous tumors of the ovary: A clinicopathologic study of 70 cases, Cancer 55, 1958 (1985). Katzenstein, A., Mazur, M. T., Morgan, E. E., et al. Proliferative serous tumors of the ovary, Amer. J. Surg. Pathol. 2, 339 (1978). Bostwick, D. G., Tazelaar, H. D., Ballon, S. C., et al. Ovarian epithelial tumors of borderline malignancy, Cancer 58, 2052 (1986). Piver, M. S. Ovarian carcinoma: A decade of progress, Cancer 54, 2706 (1984). Young, R. C., Decker, D. G., Wharton, J. T., et al. Staging laparotomy in early ovarian cancer, J. Amer. Med. Assoc., 250, 3072 (1983). Nation, J. G., and Krepart, G. V. Ovarian carcinoma of low malignant potential: Staging and treatment, Amer. J. Obstet. Gynecol. 15, 290 (1986). Michael, H., and Roth, L. M. Invasive and non invasive implants in ovarian serous tumors of low malignant potential, Cancer 57, 1240 (1986). Bergman, F. Carcinoma of the ovary: A clinicopathological study of 86 autopsy cases with special references to mode of spread, Acta Obstet. Gynecol. Seand. 45, 211 (1966). Ehrmann, R. L., Federschneider, J. M., and Knapp, R. C. Distinguishing lymph node metastases from benign glandular inclusions in low-grade ovarian carcinoma, Amer. J. Obstet. Gynecol. 136, 737 (1980). Julian, C. G., and Woodruff, J. D. The biologic behavior of low-grade papillary serous carcinoma of the ovary, Obstet, Gyneeol. 40, 860 (1972). Genadry, R., Poliakoff, S., Rotmensch, J., et al. Primary, papillary peritoneal neoplasia, Obstet. Gynecol. 58, 730 (1981). Talezaar, H. D., Bostwick, D. G., Ballon, S. C., et al. Conservative treatment of borderline ovarian tumors, Obstet. Gynecol. 66, 417 (1985). Williams, T. J., Symmonds, R. E., and Litvak, O. Management of unilateral and encapsulated ovarian cancer in young women, Gynecol. Oneol. 1, 143 (1973). Nikrui, N. Survey of clinical behavior of patients with borderline epithelial tumors of the ovary, Gynecol. Oncol. 12, 107 (1985). Creasman, W. T., Park, R., Norris, H., et al. Stage I borderline ovarian tumors, Obstet. Gynecol. 59, 93 (1982),
Primary Staging in Ovarian Tumors of Low Malignant Potential ROBERTO YAZIGI,
M.D.,* JULIE
SANDSTAD,
M.D., AND
ALAN K .
MuNoz, M . D . *
*Division of Gynecologic Oncology, *Department of Obstetrics and Gynecology, and Department of Pathology, University of Texas, Southwestern Medical School, Dallas, Texas 75235 Received March 10, 1987 Surgical staging, consisting of peritoneal washings for cytology, infracolic omentectomy, and biopsies of diaphragm, extrapelvic peritoneum, and pelvic and aortic lymph nodes, was performed in 29 patients with ovarian tumors of low malignant potential, presumed to be either Stage 1 (25) or Stage 11 (4), in order to determine the incidence of unsuspected metastases in patients with localized disease. Fourteen patients had all and fifteen patients had one or more of these procedures performed. Overall, in stages 1 and II, positive peritoneal cytology was found in 7%, unexpected omental metastases in 13%, diaphragmatic metastases in 7%, positive pelvic lymph nodes in 27%, and positive aortic lymph nodes in 7%. Seven out of 29 (24%) patients with presumed localized disease, were upstaged by virtue of the staging procedures. Based on our findings, we conclude that surgical-pathologic staging to search for occult metastases in ovarian tumors of low malignant potential is justified from an investigational standpoint: however, its impact on therapeutic management is far from being defined. © 1988AcademicPress. Inc.
INTRODUCTION Ovarian tumors of low malignant potential were first described by Taylor in 1929 [1]; official recognition as a pathologic entity by the International Federation of Gynecology and Obstetrics (FIGO) was given in 1971 [2] and by the World Health Organization in 1973 [3]. Tumors of low malignant potential comprise about 15% of all ovarian epithelial cancers [4,5], and tend to occur at an earlier age than do the usual ovarian carcinoma [4,6,7]. Extra ovarian manifestations do not always correlate with outcome [8], and recurrences within the pelvis or abdomen may develop after long latent intervals of up to 20 years [9-11]. These neoplasms are distinguished from the frankly malignant invasive epithelial tumors of the ovary because of their apparently more indolent biologic behavior. Fortunately an evolving set of morphologic criteria help identify these tumors, so that these patients can be followed as a group [7,12-15]. Much progress has been made on the natural history of ovarian adenocarcinoma by detecting sites of subclinical metastases of apparently early stage disease [16,17], upstaging up to one third of presumed stage I or II patients.
402 0090-8258/88 $1.50 Copyright © 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
STAGING IN LOW MALIGNANTOVARIANTUMORS
403
Previous reports of tumors of low malignant potential have retrospectively staged patients without the adequate surgical staging our present day knowledge dictates. The main purpose of this study is to evaluate the role of careful surgicalpathologic staging in patients with tumors of low malignant potential, focusing on the upstaging of cases with apparent localized disease. MATERIALS AND METHODS Forty-seven patients aged 19-75 years (mean 42, median 39) from 1972 to the present, from Parkland Memorial and St. Paul Hospitals, were diagnosed with ovarian serous or mucinous tumors of low malignant potential. All histologic sections from the original diagnostic material were reviewed by one of the authors (J.S.) with no knowledge of clinical outcome. Six of the 47 cases were excluded due to inadequate histological material for review. Two cases were excluded after review upgrading to frank carcinoma. One case was excluded after review diagnosis change to cystadenofibroma. Clinical data were reviewed on the 38 remaining cases. Thirty six of the 38 patients (95%) were presumed to have stage I or II disease at the time of primary exploration (31 stage I and 5 stage II); the remaining two patients were found to have gross intraabdominal disease at the time of primary surgery and so were initially staged as III during the procedure. These last two patients are excluded from further staging analysis. Patients were surgically staged according to FIGO classification (see Table 1). The initial surgical procedures in the 36 patients with apparent stage I or II disease consisted of hysterectomy and bilateral oophorectomy in 22 cases and unilateral salpingo-oophorectomy in 14 cases, 8 of which had a contralateral ovarian wedge resection. In three cases the contralateral ovary was surgically absent. In addition to this approach, 29 of these 36 patients (80%) had one or more of the following staging procedures performed: peritoneal washings, infracolic omentectomy, diaphragmatic biopsies, multiple peritoneal biopsies, and lymph node sampling (pelvic and aortic). Fourteen patients (48%) had all of these procedures performed and 15 (52%) had one or more performed.
TABLE 1 PATIENT CHARACTERISTICS (N = 36) Mean Age (range years); Presumed FIGO Stage IAi IAii IBi IBii IC IIA liB
42 (19-75) N
%
21 4 2 1 3 1 4
58 11 6 3 8 3 11
404
YAZIGI, SANDSTAD, AND MUNOZ
Peritoneal washings were obtained on entering the abdomen. Approximately 100 ml of saline solution was placed into the pelvis and aspirated; the fluid was immediately sent for analysis. Diaphragmatic biopsies were obtained under direct vision. Pelvic nodes sampled varied from 1 to 25, with an average of 8. Selective aortic lymph node biopsy was carried out by removing the largest palpable nodes from the bifurcation up to the inferior mesenteric artery and varied from 1 to 10 with an average of 3. The histology of the tumors proved to be serous in 25 patients (69%) and mucinous in 11 (31%). None fit recently delineated criteria for seromucinous type [15]. Morphologic criteria utilized for diagnosis of ovarian serous tumors of low malignant potential (borderline) include the formation of tumor papillary projections of varying complexity with lining cell stratification, cytologic atypia and/or mitotic figures, and tufting or detachment of cell clusters floating free among the papillary projections [7,14,15]. Morphologic criteria utilized for the diagnosis of mucinous tumors of low malignant potential (borderline) include villoglandular growth pattern of varying complexity, mitotic figures and/or cytologic atypia, and cell stratification [7,12,13,15]. These changes seen in both tumor types take place within the tumor cystic spaces, and no evidence for ovarian stromal or capsular invasion is present. RESULTS
Clinical Symptoms The most frequent complaint was abdominal pain, seen in 49% of patients, followed by: increased abdominal girth, abdominal mass, dyspareunia, and vaginal bleeding. Eleven patients (30%) were asymptomatic and tumor was found on a routine exam.
Findings At Surgical Staging Contralateral ovary. The contralateral ovary was explored in 27 patients with presumed stage I disease, either by oophorectomy (19) or wedge resection (8). Four patients (15%) had bilateral disease. In no patient with bilateral involvement was this an unexpected microscopic finding since all four had evident macroscopic disease at the time of surgery. Therefore, in 23 patients in which the tumor was apparently confined to one ovary no patient was upstaged based on contralateral ovarian biopsies. Peritoneal cytology. Peritoneal washings for cytology were obtained in 27 apparent stage I or II patients, and two were positive (7%). These were found in presumed stage IA and IB patients who were also found to have microscopic metastatic disease to the aortic and pelvic lymph nodes, respectively. Ascites was present in three stage I patients; none showed cytologic malignancy. Omentum-diaphragm-peritoneum. Omentectomy was performed in 24 patients with apparent stage I or lI disease; three patients (13%) revealed microscopic implants, all in presumed stage IIB patients. No patient considered to be stage I had positive omental findings.
405
STAGING IN LOW MALIGNANT OVARIAN TUMORS
Diaphragmatic biopsies were obtained in 15 patients with presumed stage I or I! disease, and of these, only one (7%) was positive in an apparent stage II-B patient. No patient was upstaged by virtue of unexpected positive peritoneal biopsies. Pelvic lymph nodes. Pelvic lymph node sampling was performed in 15 cases, 11 of which were apparent stage IA, 1 apparent stage IB, 1 apparent stage IC, and 2 apparent stage IIB. Four patients (27%) had metastatic lymph node disease; two stage IA patients, one stage IB patient, and one stage liB patient. Aortic lymph nodes. Aortic lymph nodes were sampled in 15 cases thought to be stage I or II and only one case (7%) had metastatic lymph node disease and was therefore upstaged. This patient had negative pelvic node sampling. Overall, on the basis of lymph node biopsies, 5 of 15 patients sampled (33%) were found to have unexpected microscopic metastases. See Table 2 for summary of findings. Upstaging. Based on the above described procedures, of 25 presumed stage I patients who underwent full or partial staging, four (16%) were upstaged to stage III by virtue of lymph node metastases (three pelvic and one aortic). Of four presumed stage II patients, three (75%) were upstaged to stage III due to omental microscopic implants (plus diaphragm and pelvic nodes, respectively, in each patient). Altogether, of the 29 staged patients with presumed stage I or II disease, 7 (24%) were upstaged by these surgical procedures (Table 3). DISCUSSION Obtaining random biopsies during surgical staging of ovarian cancer to detect sites of subclinical metastases has received special emphasis recently. This has resulted in upstaging a significant proportion of patients with presumed localized disease. The relevance of this fact is obvious regarding management and prognosis. Ovarian tumors of low malignant potential, however, are considered to be different pathologic and clinical entities, and staging procedures in order to detect early extraovarian spread may not have the same meaning as in invasive carcinoma. Despite this, we know that those patients with evidence of extraovarian disease may experience up to a 36 to 40% 10-year mortality rate [14,5]. Thus, adequate TABLE 2 STAGINGPROCEDURESANDFINDINGSINPRESUMEDSTAGESIANDII(N = 29)
Procedure Pertioneal cytology Omentectomy Diaphragm biopsy Peritoneal biopsy Pelvic lymph nodes Aortic lymph node
Stage I (=25)
Stage II (=4)
No. +/
No. +/
Total (=29)
No. +/
No. Sampled
No. Sampled
Sampled
2/25 0/20 0/11 0/11 3/13 1/12
0/2 3/4 1/4 0/4 1/2 0/3
2/27 3/24 1/15 0/15 4/15 1/15
(7%) (13%) (7%) (0%) (27%) (7%)
406
YAZIGI, SANDSTAD, AND MUNOZ TABLE 3 COMPARATIVE [N1TIAL AND FINAL STAGE
Patient 1 2 3 4 5 6 7
Initial stage IAi IAi IAi IBi liB liB IIB
Positive findings
Final stage
Pelvic L / N Pelvic L / N Aortic L/N Pelvic L / N Ornentum-diaphragm Omentum Omentum-Pel L/N
Ill II1 III III III Ill Ill
staging provides prognostic information in tumors of low malignant potential although the data suggest that, for instance, a stage III low malignant potential tumor has a consistently better outcome than a stage III, grade 2 or 3 cystadenocarcinoma. The dilemma of deciding whether to perform surgical staging in patients with tumors of low malignant potential can only be resolved by (1) determining what is the real yield of staging laparotomies in presumed early stage tumors and, (2) determining if upstaging and its concommitant change in therapeutic management exert any impact on final outcome. This study essentially addresses the first issue, and our results point to several important findings. Of patients thought to have stage I or II disease at the time of primary surgery when surgical staging procedures were performed, 13% (3/24) had microscopic omental metastasis, 7% (1/15) had occult diaphragmatic metastasis, and 33% (5/15) had positive lymph nodes. Overall, of the 29 staged patients with presumed early stage disease, 7 (24%) were upstaged to stage III as the result of these procedures. In a similar study, Nation and Krepart [18] upgraded two cases from stage I to III based on microscopic omental metastasis. Remarkable is the finding that none of our 25 patients with stage I disease were found to have histologically confirmed occult intraperitoneal spread; however, 3 of 13 patients sampled had retroperitoneal lymph node metastases. On the contrary, 75% (3/4) of our stage II patients showed occult intraabdominal extension of their tumor. The significance of omental, diaphragmatic, or peritoneal implants has been a matter of controversy, centering on the issue that these implants may indeed not be true metastases, but multicentric proliferation of coelomic mesothelium and mesenchyme, which have retained some Mullerian potential at extraovarian sites. Attempts to correlate different microscopic features with prognosis have been made [19], however unsuccessfully. The great issue here is how best to interpret the significance of extraovarian deposits when fortuitously found at a microscopic level related to the patients outcome. Lymph node involvement by tumors of low malignant potential have been previously described in pelvic, aortic, mediastinal, supraclavicular, axillary, and inguinal lymph nodes in Bergman's autopsy study [20]. Ehrmann et al. [21] raised the important issue of benign glandular inclusions in lymph nodes being confused
STAGING IN LOW MALIGNANT OVARIAN TUMORS
407
with metastatic tissue. Our five cases with lymph node metastases all showed complex cribiform or papillary structures with cytologic atypia, sometimes mitoses, and tufting and were not difficult to differentiate from benign glandular inclusions. Interestingly, the presence of extraovarian implants in 16-47% of cases [4,22,14,5,23] herald those patients who die of their disease. For stage IA patients who desire fertility, most gynecologic oncologists would agree that conservative treatment is justified and outcome after unilateral salpingooophorectomy seems to be similar to hysterectomy with bilateral salpingooophorectomy [19]. Our findings in the contralateral ovary (by oophorectomy or wedge biopsy) agree with the data from Talezaar et al. [24] None of our 23 patients with disease apparently confined to one ovary had positive microscopic findings in the contralateral ovary. In Talezaar's series, one patient developed a subsequent serous tumor of low malignant potential in a previously negative biopsied ovary. Williams [25], however, found "malignancy" in the contralateral ovary in 7% of his cases which appeared grossly normal at surgery. Despite our findings, we still believe that contralateral wedge biopsy should be performed, based on the incidence of bilateral involvement reported by other authors [26,5]. The results of our study lead us to believe that: Staging laparotomy for tumors of low malignant potential appears justified from an investigational standpoint, on the basis of a 24% incidence of upstaging in presumed stages I and II disease. However, its impact on management and survival will have to wait until some form of therapy is identified that benefits patients with early extraovarian spread. The value of adjuvant therapy in stage I disease has been negated by the one randomized study published [27]. The use of chemotherapy for advanced disease awaits :further randomized trials. The ongoing study by the Gynecologic Oncology Group may help resolve the issue. REFERENCES 1. Taylor, H. C. Malignant and semi malignant tumors of the ovary, Surg. Gynecol. Obstet. 48, 204 (1929). 2. Classification and staging of malignant tumors in the female pelvis, Acta Obstet. Gynecol. Scand. 50, 1 (1971). 3. Serov, S. F., Scully, R. E., and Sobin, L. H. Histological typing of ovarian turnours, World Health Organization, Geneva (1973). 4. Aure, J. C., Hoeg, K., and Kolstad, P. Clinical and histologic studies of ovarian carcinoma, Obstet. Gynecol. 37, 1 (1971). 5. Russell, P. Borderline epithelial tumours of the ovary: A conceptual dilemma, Clin. Obstet. Gynaecol. 11, 259 (1984). 6. Purola, E. Serous papillary ovarian tumors, Acta. Obstet. Gynecol. Scand. (Suppl.) 42, 3 (1963). 7. Colgan, T. J., and Norris, H. J. Ovarian epithelial tumors of low malignant potential: A review, lnt. J. Gynecol. Pathol. 1, 367 (1983). 8. Tasker, M., and Langley, F. A. The outlook for women with borderline epithelial tumours of the ovary, Brit. J. Obstet. Gynaecol. 92, 969 (1985). 9. Malloy, J. J., Dockerty, M. B., Welch, J. S., et al. Papillary ovarian tumors, Amer. J. Obstet. Gynecol. 93, 867 (1965). 10. Woodruff, J. D., and Julian, C. G. Histologic grading and morphologic changes of significance
408
11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21.
22. 23. 24. 25. 26. 27.
YAZIGI, SANDSTAD~ AND MUNOZ
in the treatment of semi malignant and malignant ovarian tumors, in Proceedings o f the Sixth National Cancer Conference, Lippincott, Philadelphia/Toronto, pp. 347-351 (1968). Minyi, T., Lijuan, L., and Tonghua, L. The characteristics of ovarian serous tumors of borderline malignancy, Chin. Med. J. 93, 459 (1980). Hart, W. R., and Norris, H. J. Borderline and malignant mucinous tumor of the ovary: Histologic criteria and clinical behavior, Cancer 31, 1031 (1973). Chaitin, B. A., Gershenson, D. M., and Evans, H. L. Mucinous tumors of the ovary: A clinicopathologic study of 70 cases, Cancer 55, 1958 (1985). Katzenstein, A., Mazur, M. T., Morgan, E. E., et al. Proliferative serous tumors of the ovary, Amer. J. Surg. Pathol. 2, 339 (1978). Bostwick, D. G., Tazelaar, H. D., Ballon, S. C., et al. Ovarian epithelial tumors of borderline malignancy, Cancer 58, 2052 (1986). Piver, M. S. Ovarian carcinoma: A decade of progress, Cancer 54, 2706 (1984). Young, R. C., Decker, D. G., Wharton, J. T., et al. Staging laparotomy in early ovarian cancer, J. Amer. Med. Assoc., 250, 3072 (1983). Nation, J. G., and Krepart, G. V. Ovarian carcinoma of low malignant potential: Staging and treatment, Amer. J. Obstet. Gynecol. 15, 290 (1986). Michael, H., and Roth, L. M. Invasive and non invasive implants in ovarian serous tumors of low malignant potential, Cancer 57, 1240 (1986). Bergman, F. Carcinoma of the ovary: A clinicopathological study of 86 autopsy cases with special references to mode of spread, Acta Obstet. Gynecol. Seand. 45, 211 (1966). Ehrmann, R. L., Federschneider, J. M., and Knapp, R. C. Distinguishing lymph node metastases from benign glandular inclusions in low-grade ovarian carcinoma, Amer. J. Obstet. Gynecol. 136, 737 (1980). Julian, C. G., and Woodruff, J. D. The biologic behavior of low-grade papillary serous carcinoma of the ovary, Obstet, Gyneeol. 40, 860 (1972). Genadry, R., Poliakoff, S., Rotmensch, J., et al. Primary, papillary peritoneal neoplasia, Obstet. Gynecol. 58, 730 (1981). Talezaar, H. D., Bostwick, D. G., Ballon, S. C., et al. Conservative treatment of borderline ovarian tumors, Obstet. Gynecol. 66, 417 (1985). Williams, T. J., Symmonds, R. E., and Litvak, O. Management of unilateral and encapsulated ovarian cancer in young women, Gynecol. Oneol. 1, 143 (1973). Nikrui, N. Survey of clinical behavior of patients with borderline epithelial tumors of the ovary, Gynecol. Oncol. 12, 107 (1985). Creasman, W. T., Park, R., Norris, H., et al. Stage I borderline ovarian tumors, Obstet. Gynecol. 59, 93 (1982),