1 Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif, 2 Centre Oscar Lambret, 3 rue Frédéric Combemale, 59020 Lille, 3 Centre Henri Becquerel, Rue d'Amiens, 76038 Rouen, 4 Institut Claudius Regaud, 2024 rue du Pont Saint Pierre, 31052 Toulouse, 5 Centre Eugene Marquis, Rue de la Bataille de Flandre Dunkerque, 35042 Rennes, 6 Groupe des Chirurgiens de Centre de Lutte Contre le Cancer and 7 Société Française d'Oncologie Gynécologique, France
8 To whom correspondence should be addressed at: Service de Chirurgie, Institut Gustave Roussy, 39 rue Camille Desmoulins, 94805 Villejuif Cedex, France. Email: morice{at}igr.fr
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: conservative surgery/epithelial ovarian tumour/pregnancy/recurrence/restaging surgery
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The FIGO stage (International Federation of Gynecology and Obstetrics, 1987) was defined using the macroscopic description during the surgical procedure(s) and histological analysis of specimens removed during initial and restaging surgery (if the initial procedure was incomplete).
Patient follow-up included a clinical examination, blood marker measurements and an ultrasound scan at least every 6 months during the first year following the procedure, then yearly afterwards.
The recurrence rate, overall survival (OS), disease-free survival (DFS), fertility outcome and the number of pregnancies were studied. The characteristics of patients who developed a recurrence were studied. OS and DFS probabilities were estimated from the first surgical procedure to the last visit, using the KaplanMeier method.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Patient characteristics
The median age of the 34 patients evaluated at the time of the surgical procedure was 27 years (range 1436). Twenty-three patients were nulliparous. Four patients had a history of infertility. Two of them had undergone ovulation induction therapy. Two patients had a previous history of borderline ovarian tumour (BOT). Patients presented with disease-related symptoms in 68% of the cases (23/34). Abdominal distension (41%, 14/34) and pain (32%, 11/34) were the most frequent presenting symptoms. EOC was diagnosed in six (18%) asymptomatic patients during the abdomino-pelvic ultrasound examination. In five patients, the presenting symptom was unknown. In one patient, the EOC was diagnosed during pregnancy (at 16 weeks) on routine pelvic ultrasound.
The histological types were: serous (39%), mucinous (2162%), endometrioid (515%), small-cell (25%) and mixed (mucinous and serous) EOC (39%). Tumour differentiation was: grade 1 (1544%), grade 2 (1544%) and grade 3 (412%). The distribution of these 34 patients according to FIGO staging was: 30 stage IA (grade 1, n=13; grade 2, n=14; grade 3, n=3); three stage IC (grade 1, n=2; grade 3 n=1) and one stage II (grade 2).
Complete initial staging surgery was performed in two patients and a restaging surgical procedure in 32 patients over a median interval of 45 (range 7149) days since the first operation. A laparoscopic approach was used for restaging surgery in four cases. Twenty-six and 24 patients underwent a para-aortic and a pelvic lymphadenectomy respectively. None of them was upstaged following the histological analysis of lymph nodes. In one patient with a previous history of BOT (case no. 11 in Table I), mixed endosalpingiosis components and non-invasive implants were found in nodes. An appendectomy was performed in 16 patients and uterine curettage in nine. Of 23 patients with mucinous or mixed histological subtype, 13 underwent an appendectomy. Twenty-five patients had a routine biopsy of the contralateral ovary (normal in all cases). Among the 32 patients who underwent restaging surgery, two (6%) were upstaged (peritoneal implants were observed on the Fallopian tube serosa in one case and peritoneal cytology was positive in the other patient). Thirty-three patients had stage I and one had stage II disease.
|
Recurrence and survival
The recurrence and survival rates according to disease stage were studied in these 34 patients treated conservatively after restaging surgery. The median duration of follow-up was 47 months (range 11224). Twenty-four patients were alive and disease-free. Two patients underwent a laparoscopic or a laparotomic cystectomy for a suspicious lesion on the remaining ovary. In these two cases the histology was benign. Eleven patients developed a recurrence over an average period of 16 months (range 248). The characteristics of those 11 patients are detailed in Table I. Among seven patients with mucinous or mixed subtype, three patients underwent lymphadenectomy during the surgical management of their initial disease. Ten patients had recurrent invasive disease and one patient (with stage IA grade 1 disease) had a borderline recurrence on the preserved ovary (case no. 11 in Table I). Among these 11 patients, the first recurrence was observed on the contralateral remaining ovary in nine cases (isolated at this site in five patients). One patient (case no. 10) treated for a stage IA grade 2 tumour (with a normal routine biopsy specimen of the remaining ovary at restaging surgery) rapidly developed a contralateral tumour 2 months after restaging surgery. Ten patients underwent removal of the remaining ovary at surgical treatment of the recurrence and eight had a hysterectomy.
One patient developed an unusual recurrence. This case was previously published as a case report by the team who treated her (Baron et al., 2002). This young patient was initially treated for a large ovarian tumour by laparotomy. She underwent a transverse low incision laparotomy with a right salpingo-oophorectomy but with peroperative rupture of the tumour. The histological analysis demonstrated the presence of a mucinous BOT together with small foci of invasive well-differentiated carcinoma. Three months later she underwent a restaging laparotomy including complete peritoneal staging, lymphadenectomy and appendectomy. As all specimens were negative, no adjuvant chemotherapy was administered but this lesion was considered as a stage IC tumour due to peroperative rupture of the lesion during initial surgery. Four years later she developed an isolated wound recurrence (without peritoneal or ovarian disease) which was treated by surgical resection followed by chemotherapy (platinum-based and paclitaxel). A contralateral wound recurrence was observed 2 years later. This patient is currently under treatment for pleural effusion.
Seven patients with stage IA disease and all patients with stage IC developed a recurrence. Among 10 patients with an invasive recurrence, stages and grades were: stage IA G1, n=1; stage IA G2, n=4; stage IA G3, n=1; stage
IC, n=4. OS rates at 3 and 5 years (CI) for all patients were respectively 96% (7999) and 84% (6295). The DFS rates at 3 and 5 years (CI) for all patients were respectively 74% (5787) and 63% (4380) (Figure 1). OS and DFS rates at 5 years (CI) for patients with stage IA disease were respectively: 87% (6296) and 78% (5990) (Figure 2). According to the tumour grade, DFS at 5 years (CI) for patients with stage IA disease was 92% (6799) in patients with grade 1 and 70% (4388) in patients with grade 2 disease (Figure 3).
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Three years ago, however, we published a preliminary series about the results of conservative surgery of EOC in a group of patients treated in a single institution (Morice et al., 2001). These preliminary results were disturbing because all patients with a disease stage > IA developed a recurrence. In an attempt to explain such a difference with the Italian and American series, we decided to conduct a retrospective multicentre study in our country. Very strict inclusion criteria were defined in order to obtain the most accurate results. The definition of these inclusion criteria is a key to explain the differences observed. Among the 59 potentially eligible patients reviewed in the present series, 34 fulfilled all inclusion criteria. In these inclusion criteria, only patients with complete peritoneal staging were enrolled so that the definition of the tumour stage would be as precise as possible. A recent randomized study demonstrated the therapeutic implications of complete peritoneal staging in patients with early stage disease (Trimbos et al., 2003
). This staging procedure is particularly important in patients with apparently early-stage disease, as adjuvant therapy may be avoided in patients with stage IA or IB disease (and grade 1 or 2 tumours). Staging should include peritoneal cytology or washing, an omentectomy and multiple peritoneal biopsies (Di Saia, 1989
; Morice et al., 2003
). Nodal staging was deemed optional because the benefit of this procedure is still controversial in patients with EOC.
The second inclusion criterion concerned the use of chemotherapy. As one of the major differences between published series concerns clinical outcomes in patients with stage IC disease, we preferred to include only patients who received adjuvant chemotherapy for stage
IC disease. The interest of adjuvant treatment in patients with stage IC disease continues to be debated in the literature. However, in the present series, we decided to enrol only patients with stage
IC disease treated by conservative surgery followed by adjuvant therapy in order to ensure that a subsequent recurrence would not be related to the absence of adjuvant treatment. Only one patient in the present series was treated exclusively by surgery for stage IC disease (case no. 9 in Table I). This case is amply detailed in the Results section. Although this lesion could have been classified initially as stage IA or IC disease, it was considered by the team who treated her as a stage IC tumour due to its peroperative rupture (absence of peritoneal cytology during initial surgery). This is the only case of stage IC disease which was not treated with adjuvant chemotherapy in the present series. Furthermore, the characteristics of the recurrence (wound recurrence at 48 months) observed in this case were probably not related to the absence of adjuvant treatment during the initial management. In the Italian series, 11 out of 19 patients with stage IC disease did not receive adjuvant treatment (Colombo et al., 1994
). In Schilder et al.'s (2002)
series, 2/10 patients with stage IC disease were treated exclusively by surgery. The type of chemotherapy is also important. The standard treatment is a platinum-based chemotherapy. In the study published by Schilder et al. (2002)
, 5/19 patients received adjuvant treatment containing melphalan alone. In the present series, patients who received regimens other than platinum-based chemotherapy were excluded.
The last important inclusion criterion, and probably one of the most important, concerns the centralized histological review of the ovarian tumour by the same pathologist (patients were excluded if histological slides of the initial tumour were not available). In Schilder et al.'s (2002) series, the histological slides were reviewed by one pathologist in each of the eight centres which participated in this multicentre study. The final centralized review was performed only in dubious cases concerning the histological subtype or tumour grade (Schilder et al., 2002
). In the present series, four patients reviewed were reclassified as having a borderline tumour and the tumour grade was reclassified in one patient. A routinely centralized pathological review is probably a key point.
Our results confirm, as do those in the Italian and American series, that conservative management could be safely proposed in patients with stage IA grade 1 disease. However, unlike the other series, and even if the number of patients with stage IC is low, the fact that all of them developed a recurrence seems to suggest that such management should not be proposed to patients with a stage
IC tumour. This result could not be explained by the absence of adjuvant treatment, the absence of complete surgical staging (misdiagnosis of a more advanced stage disease) or inadequate chemotherapy, all of which were among our inclusion criteria. Furthermore, most of the recurrent lesions were on the remaining ovary (8/10 invasive recurrences). Consequently, the spared ovary is the site of the first recurrence in the majority of patients. This point was not so evident in the other two series: only 2/5 recurrent lesions were located on the preserved ovary in the Italian series and 3/5 in the American series (Colombo et al., 1994
; Zanetta et al., 1997
; Schilder et al., 2002
).
We also observed that the DFS of stage IA grade 2 disease was quite different from that reported in the literature following radical surgery, even if the number of such patients is small in the present study. In the literature, the survival of patients with this type of tumour varies from 77 to 94%, whereas in our study the DFS in patients with a stage IA grade 2 tumour was only 70% (Dembo et al., 1990; Young et al., 1990
; Baiocchi et al., 1998
; Villa et al., 1998
). In the Italian and American series, respectively 2/8 and 2/6 patients with stage IA grade 2 disease developed a recurrence (Colombo et al., 1994
; Zanetta et al., 1997
; Schilder et al., 2002
). Thus, in three main series about conservative surgery, the risk of recurrence appears to be increased following conservative surgery in patients with stage IA grade 2 disease. Systematic adjuvant chemotherapy is not routinely indicated in patients who undergo radical treatment, but could such treatment reduce the risk of recurrence after conservative surgery? In the present series, 2/4 patients with recurrent stage IA grade 2 disease received adjuvant treatment and in the Italian series both patients with recurrent lesions received six courses of platinum-based chemotherapy (Zanetta et al., 1997
).
In our retrospective study, only 1/3 patients with stage IA grade 3 disease developed a recurrence. But the number of patients with stage IA grade 3 disease is too small to draw a definitive conclusion about this subgroup of patients. Furthermore, two of them had a very small grade 3 borderline or endometrioid tumour (with a clear-cell component in one of them). This point (small tumour size) explained why some patients were treated exclusively by surgery for stage IA grade 3 disease whereas these patients are treated routinely with adjuvant treatment in our institution. With a follow-up of 23 and 29 months, neither of the two above-mentioned patients developed a recurrence. However, if the question is whether conservative management of stage IA grade 3 disease is appropriate, such surgery should not be proposed to patients with a bulky stage IA tumour and grade 3 disease. The histological subtype can also be an important prognostic factor. Schilder et al. (2002) did not observe a higher recurrence rate in five patients with a clear-cell tumour. However, in general, clear-cell tumours seem to have a worse prognosis than their serous, mucinous and endometrioid counterparts (Dembo et al., 1990
; Young et al., 1990
; Baiocchi et al., 1998
; Villa et al., 1998
). In this series, 1/2 patients with this kind of tumour developed a recurrence (the other patient had a very small tumour <1 cm within a large endometrioid cyst). In our opinion, conservative management could be considered in patients with a serous, mucinous or an endometrioid tumour but should not be performed in patients with more aggressive tumours such as clear-cell or anaplastic carcinoma. How can we explain the higher recurrence rate observed in our series after conservative management of stage IC or IA grade 2 disease? One of the dangers of the conservative treatment of EOC is not to remove a microscopic invasive cancer in the remaining ovary. Munnell (1969)
estimated the risk of microscopic cancer on the contralateral macroscopically normal ovary to be 12%. These findings explain why routine biopsies on the contralateral ovary were recommended (Munnell, 1969
; Di Saia, 1989
). However, such a procedure could lead to adhesions and subsequent infertility. In the study by Zanetta et al. and in our series, none of the patients with a macroscopically normal ovary had microscopic metastasis found on routine biopsies (Colombo et al., 1994
; Zanetta et al., 1997
). This procedure should be performed only in patients with suspicious lesions of the remaining ovary.
Two recent series focus on the evaluation of the risk of occult contralateral involvement in patients with unilateral macroscopic stage I EOC (Benjamin et al., 1999; Amsalem et al., 2003
). In the series by Benjamin et al. (1999)
, only 3/118 patients (2.5%) with stage I EOC had occult ovarian involvement. None of these three patients had stage IA disease: two of them had stage IC (grade 1 and 3) and the last one had stage IB grade 3 disease (Benjamin et al., 1999
). In the series reported by Amsalem et al. (2003)
, 3/25 patients with surgical stage IA or IC disease had microscopic occult involvement of the contralateral ovary: two had stage IB grade 3 disease and the last patient had histological stage III grade 1 disease. These findings suggest that microscopic contralateral involvement could be found in patients with stage IC and/or in patients with early stage disease but a grade 3 tumour. Such findings were not observed in patients with a stage IA grade 1 or 2 tumour.
Recently an interesting paper was published about the incidence of premalignant lesions in the contralateral ovary in patients with unilateral ovarian cancer (Kaur et al., 2004). Premalignant lesions (inclusion cysts, nuclear atypia and epithelial stratification) and Bcl-2 overexpression were seen more frequently in patients with contralateral ovarian cancer compared to control groups (Kaur et al., 2004
). The histological subtype and disease stage were detailed but the tumour grade was not given. Perhaps these premalignant lesions were mainly observed in a specific subgroup of patients with stage IA and could so explain a higher risk of recurrences if conservative surgery were performed.
According to these results, indications for conservative surgery in patients treated for an EOC should be rigorously and wisely selected. The main objective for conservative surgery is to preserve fertility. Zanetta et al. (1997) and Schilder et al. (2002)
reported 27 pregnancies in 20 patients and 26 full-term deliveries in 17 patients treated conservatively. In our series, 10 pregnancies were observed in nine patients. Perhaps these results could be improved using a laparoscopic approach for the restaging procedure, thereby reducing the risk of postoperative adhesions that could impact on subsequent fertility (Leblanc et al., 1999
).
One unsolved question is whether radical surgery ought to be recommended after the completion of pregnancy(ies) in order to reduce the risk of recurrence. Di Saia (1989) recommended performing radical surgery when the desire for reproduction is no longer being considered. Moore et al. (1999)
recently published a case report of an infertile patient with an early-stage poorly differentiated tumour who developed a recurrence 10 years following the treatment of the ovarian tumour. As the recurrence rate reported in our series is relatively high, and as late recurrences could occur (48 months in this series), we think that such completion surgery should be proposed to patients whose pregnancy(ies) was(were) successful.
In conclusion, our results confirm that conservative surgery could be safely performed in young patients treated for a stage IA grade 1 EOC who wish to preserve their fertility. However, this procedure should be evaluated in patients with stage IA grade 2 disease but should not be performed in patients with FIGO stage > IA.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Baiocchi G, Raspagliesi F, Grosso G, Fontanelli R, Cobellis L, Di Re E and Di Re F (1998) Early ovarian cancer: is there a role for systematic pelvic and para-aortic lymphadenectomy? Int J Gynecol Cancer 8, 103108.[CrossRef][ISI]
Baron M, Ladonne JM and Resch B (2002) Abdominal wall metastasis from ovarian cancer after laparotomy. A case report. Eur J Gynaecol Oncol 23, 561562.[ISI][Medline]
Benjamin I, Morgan MA and Rubin SC (1999) Occult bilateral involvement in stage I epithelial ovarian cancer. Gynecol Oncol 72, 288291.[CrossRef][ISI][Medline]
Colombo N, Chiari S, Maggioni A, Bocciolone L, Torri V and Mangioni C (1994) Controversial issues in the management of early epithelial ovarian cancer: conservative surgery and role of adjuvant therapy. Gynecol Oncol 55, S47S51.[CrossRef][ISI][Medline]
Dembo AJ, Davy M, Stenwig AE, Berle EJ, Bush RS and Kjorstad K (1990) Prognostic factors in patients with stage I epithelial ovarian cancer. Obstet Gynecol 75, 263273.[Abstract]
Di Saia PJ (1989) Conservative management of the patient with early gynecologic cancer. CA Cancer J Clin 39, 135154.[ISI][Medline]
International Federation of Gynecology and Obstetrics (1987) Changes in definitions of clinical staging for carcinoma of the cervix and ovary. Am J Obstet Gynecol 156, 263264.[Medline]
Kaur TB, Shen T, Gaughan J, Thompson J, Thompson J, Houck KL and Hernandez E (2004) Premalignant lesions in the contralateral ovary of women with unilateral ovarian carcinoma. Gynecol Oncol 93, 6977.[CrossRef][ISI][Medline]
Leblanc E, Querleu D and Lanvin D (1999) Laparoscopic staging of early invasive adnexal tumors. Int J Gynecol Cancer (Suppl 1), A125.
Moore MM, Tewari K, Rose GS, Frauehauf JP and Di Saia PJ (1999) Long-term consequences following conservative management of epithelial ovarian cancer in an infertile patient. Gynecol Oncol 73, 452454.[CrossRef][ISI][Medline]
Morice P, Camatte S, Wicart-Poque F, El-Hassan J, Pautier P, Lhomme C, De Crevoisier R, Haie-Meder C, Duvillard P and Castaigne D (2001) Results of conservative treatment in epithelial ovarian carcinoma. Cancer 92, 24122418.[CrossRef][ISI][Medline]
Morice P, Camatte S, Wicart-Poque F, Atallah D, Rouzier R, Pautier P, Pomel C, Lhomme C, Duvillard P and Castaigne D (2003) Results of conservative management of epithelial malignant and borderline ovarian tumours. Hum Reprod Update 9, 185192.
Munnell EW (1969) Is conservative therapy ever justified in stage I (IA) cancer of the ovary? Am J Obstet Gynecol 103, 641650.[ISI][Medline]
Raspagliesi F, Fontanelli R, Paladini D and Di Re EM (1997) Conservative surgery in high-risk epithelial ovarian carcinoma. J Am Coll Surg 185, 457460.[CrossRef][ISI][Medline]
Schilder JM, Thompson AM, DePriest PD, Ueland FR, Cibull ML et al. (2002) Outcome of reproductive age women with stage IA or IC invasive epithelial ovarian cancer treated with fertility-sparing therapy. Gynecol Oncol 87, 17.[CrossRef][ISI][Medline]
Silverberg SG (2000) Histopathologic grading of ovarian carcinoma: a review and proposal. Int J Gynecol Pathol 19, 715.[ISI][Medline]
Trimbos JB, Vergote I, Bolis G, Vermorken JB, Mangioni C et al. (2003) Impact of adjuvant chemotherapy and surgical staging in early-stage ovarian carcinoma: European Organisation for Research and Treatment of Cancer-Adjuvant ChemoTherapy in Ovarian Neoplasm trial. J Natl Cancer Inst 95, 113125.
Villa A, Parazzini F, Acerboni S, Guarniero P and Bolis G (1998) Survival and prognostic factors of early ovarian cancer. Br J Cancer 77, 123124.[ISI][Medline]
Young RC, Walton LA, Ellenberg SS, Homesley HD, Wilbanks GD, Decker DG, Miller A, Park R and Major F Jr (1990) Adjuvant therapy in stage I and stage II epithelial ovarian cancer. Results of two prospective randomized trials. N Engl J Med 322, 10211027.[Abstract]
Zanetta G, Chiari S, Rota S, Bratina G, Maneo A, Torri V and Mangioni C (1997) Conservative surgery for stage I ovarian carcinoma in women of childbearing age. Br J Obstet Gynaecol 104, 10301035.[ISI][Medline]
Submitted on September 6, 2004; accepted on January 10, 2005.