Impact of peritoneal cytology on survival of endometrial cancer patients treated with surgery and radiotherapy

Stage IIIA endometrial cancer includes patients with serosal or adnexal invasion and patients with positive peritoneal cytology only. In this study, we assessed the impact of peritoneal cytology on endometrial cancer survival. All endometrial cancer patients receiving surgery and radiotherapy at the Geneva University Hospitals between 1980 and 1993 were included. Stage lllA cancers were categorised into ‘cytological’ stage lllA (only positive peritoneal cytology) and ‘histological’ stage lllA (serosal or adnexal infiltration). Survival rates were analysed by Kaplan–Meier method and compared using log-rank test. The prognostic importance of peritoneal cytology was analysed by multivariate regression analysis. This study included 170 endometrial cancers (112 stage I, 17 cytological stage IIIA, 18 histological stage IIIA, 9 stage lllB+). Disease-specific survival of cytological stage IIIA was not different from stage I (94 vs 88% respectively, P=0.5) but better than histological stage IIIA (94 vs 51% respectively, P<0.01). Histological stage IIIA patients were at increased risk to die from cancer compared to stage I patients (HR 2.7, 95% CI 1.0–7.7), while cytological stage IIIA patients were not (HR 0.3, 95% CI 0.3–2.0). Cytological stage lllA endometrial cancer has similar prognosis as stage l and better prognosis than histological stage IIIA. Additional research, definitively separating stage and cytology is warranted.

Endometrial cancer is the most common genital tract malignancy in Western countries (Grady and Ernster, 1996;Burke et al, 1997;Parkin et al, 1997;Greenlee et al, 2000). The long-term survival of patients with endometrial cancer is related to the stage at diagnosis. Disease-specific 5-year survival for tumours confined to the uterus is over 80%, but decreases to approximately 40% for tumours invading the serosa, the fallopian tube or the ovary (AJCC, 2002).
Before 1988, endometrial cancer was staged clinically (International Federation of Gynaecology & Obstetrics, FIGO 1971). This staging system was based on the fractional biopsy of the endometrium, the size of the uterine cavity and physical examination. The clinical system was abandoned because newly accumulated data from surgical staging reports, among others including depth of myometrial invasion, allowed stratification. A new surgical staging system was approved at the 1988 FIGO meeting (Grady and Ernster, 1996;Burke et al., 1997;FIGO, 2000;AJCC, 2002). In addition to pathological extension, the result of cytology obtained by peritoneal washings became an important determinant of staging. In particular, stage l patients with positive peritoneal cytology were upstaged to stage lllA (Konski et al, 1988). Stage IIIA therefore covers both patients with only positive peritoneal cytology and patients with macroscopic/histological invasion of serosa or adnexal tissues.
As some studies have suggested that cytology is an important prognostic factor, while others did not (Grimshaw et al, 1990;Morrow et al, 1991;Kadar et al, 1992;Obermair et al, 2001;Preyer et al, 2002;Kasamatsu et al, 2003), the value of peritoneal cytology in the staging of endometrial cancer is still controversial (Kasamatsu et al, 2003). In this study, we evaluated the impact of positive peritoneal cytology on the survival of patients operated for endometrial carcinoma and treated with adjuvant brachytherapy and/or external radiotherapy.

METHODS Patients
The current study included women who received surgery followed by adjuvant external radiotherapy and/or brachytherapy for endometrial cancer between 1980 and 1993 at the departments of Gynaecology and Radiation Oncology of the Geneva University Hospitals (n ¼ 295). We excluded patients with stage I or stage II endometrial cancer without peritoneal cytology assessment (n ¼ 111), women with previous malignancy occurring within 5 years prior to the diagnosis of endometrial cancer (n ¼ 2), or patients who were not resident in the canton (n ¼ 12). The final study population included 170 patients. According to the Geneva cancer registry, these women represented approximately 50% of all endometrial cancer patients treated in the public sector during the period considered.
Peritoneal washing was performed by either collecting liquid present in the peritoneal cavity, or by rinsing the cavity with 100 cm 3 of physiological saline. The liquid was centrifuged and assessed for the presence of malignant cells. Surgical treatment generally included hysterectomy and oophorosalpingectomy, except for three patients (o2%) who received only simple hysterectomy. A total of 12 patients (7%) underwent lymph node dissection. All surgical procedures were followed by external radiotherapy and/or brachytherapy.

Variables
Histologic type and differentiation of the endometrial tumours were coded according to the International Classification of Diseases for Oncology (World Health Organization, 1990). A patient was considered to have positive peritoneal cytology if adenocarcinoma cells were detected, regardless of the number of cancer cells. Stages were coded according to the 1988 FIGO staging system: stage I, tumour confined to uterus; stage II, tumour invading cervix; stage IIIA, tumour associated with positive peritoneal cytology or macroscopic or histological involvement of serosa or adnexa and stage IIIB þ , tumour invading vagina, mucosa of bladder/bowel, regional lymph node or distant metastases. For the purpose of the present study, stage lllA cancers were further categorised as: 'cytological' stage lllA defined as cases with only positive peritoneal cytology and 'histological' stage lllA defined as cases with histological or macroscopic infiltration of serosa or adnexal tissues.
Data on survival and follow-up were derived from the Geneva cancer registry and included vital status and date of death or departure from the canton (regularly and systematically obtained from the Cantonal Population Office) and cause of death (retrieved from medical files).

Statistical analyses
The 5-year overall and disease-specific survival rates were calculated. Disease-specific survival was calculated considering death from endometrial cancer only. Survival curves were obtained by Kaplan -Meier method and were compared by nonparametric survival analyses using log-rank test. The effect of peritoneal cytology on endometrial cancer mortality was analysed by multivariate Cox's proportional-hazards regression analysis, taking into account other variables significantly linked to survival. Analyses were performed using SPSS (Hull and Nie, 1995). Differences were considered statistically significant at Po0.05.

RESULTS
Of the 170 endometrial cancer patients included in the study, 112 were diagnosed as stage I, 14 as stage II, 17 as cytological stage IIIA, 18 as histological stage IIIA, and nine were diagnosed as stage IIIB þ . If peritoneal cytology had not been incorporated in the  staging of the endometrial cancers, 15 of the 17 cytological stage IIIA tumours would have been classified as stage I and two as stage II. Table 1 presents the characteristics of these 170 patients according to stage of endometrial cancer. The mean age of the patients was 64.7 years and was similar for the different stages. Compared to histological stage IIIA tumours, cytological stage IIIA tumours invaded the myometrium to a lesser extent, were better differentiated and received less aggressive radiotherapy.
The median duration of follow-up was 126 months and 14 patients (8%) were lost to follow-up. Table 2 presents overall and disease-specific 5-year survival rates per stage. Figure 1 presents disease-specific survival curves for stage I, cytological stage IIIA and histological stage IIIA endometrial cancer. The disease specific 5-year survival of patients with cytological stage IIIA tumours was not significantly different from that of patients with stage I tumours (94 vs 88% respectively, log-rank test: P ¼ 0.5). In contrast, the 5-year disease-specific survival of patients with histological stage IIIA tumours was significantly worse (51 vs 94% for histological vs cytological stage IIIA cancers, log-rank test: Po0.01). A similar pattern was observed when considering the overall survival rates accounting for all deaths, including death from other causes than endometrial cancer.
After adjustment for factors significantly linked to survival from endometrial cancer (i.e. age, tumour differentiation and type of radiotherapy) as shown in Table 3, the risk to die from endometrial cancer was still not significantly different in patients with cytological stage IIIA compared with patients with stage I cancer (HR 0.3, 95% CI 0.3 -2.0). Compared to stage I, patients with stage II, histological IIIA, and IIIB þ tumours were at respectively 3.2fold (95% CI 1.2 -8.6), 2.7-fold (95% CI 1.0 -7.7) and 3.5-fold (95% CI 1.2 -10.4) increased risk to die from their endometrial cancer. The results were not modified when excluding the two patients in the cytological stage IIIA category who had cervical invasion (i.e. who would have been staged as stage II in the absence of positive peritoneal cytology).

DISCUSSION
This study is the first to compare the prognosis of patients with stage I, cytological stage IIIA, and histological stage IIIA endometrial cancer in the same patient population. It suggests that positive peritoneal cytology is not an independent prognostic factor in patients with endometrial cancer treated with surgery and radiotherapy. Irradiated women with only positive cytology (cytological stage IIIA) had the same prognosis as women with endometrial cancer confined to the uterus (stage I) and a much better prognosis than women with histological stage IIIA cancer.
However, our study has several limitations: 111 women had no peritoneal cytology and were therefore excluded, patients were selected for inclusion in the study on the basis of administration of radiotherapy, and the study has a relatively low statistical power due to the limited number of patients included. Nonetheless, the lack of significant survival difference between the stage I and cytological stage IIIA groups suggest that the FIGO classification system classifies patients with positive peritoneal cytology into an inappropriately high-risk category. However, this finding might also be explained by a disproportionate benefit derived from radiotherapy in patients with positive peritoneal cytology. Table 4 shows that similarly high survival rates (88 -94% 3 -5 year survival) in women with stage I and cytological stage IIIA cancer have been observed in series where, as in our study, the large majority of cytological stage IIIA patients underwent either pelvic radiotherapy or total pelvic lymphadenectomy (Kadar et al, 1992;Kasamatsu et al, 2003). In contrast, the relatively poor results reported by other investigators (64 -67% 3 -5 year survival) in cytological stage IIIA patients might be explained by the fact that in these studies, only a minority of patients received treatment to pelvic lymph nodes (Morrow et al, 1991;Obermair et al, 2001;Preyer et al, 2002). Pelvic radiotherapy or lymphadenectomy might be especially important for patients with positive peritoneal cytology, as it has been shown by Morrow et al (1991) that the presence of positive peritoneal cytology increases the risk of paraaortic lymph node invasion (15 vs 4%), and by inference increases the likelihood of pelvic node involvement.  The proportion of cytological stage IIIA cancer patients with cervical involvement is unknown in the majority of studies, except for the study of Kasamatsu et al and ours, in which the proportions of cytological stage IIIA patients with cervical involvement were relatively low (0 and 12%, respectively). As cervical involvement is known to affect the survival negatively, a higher proportion of cytological stage IIIA patients with cervical involvement might explain unfavourable survival rates. Therefore, the unknown proportions of patients with cervical involvement complicate interpretation of outcomes and impair comparison between the different studies.
In conclusion, this study found comparable survival rates for patients with cytological IIIA and stage I endometrial cancer, which were significantly superior to that of patients with histological stage IIIA endometrial cancer. Because all studies, including ours, have important shortcomings, no definitive conclusion on the importance of positive peritoneal cytology can be drawn and therefore, additional research is warranted. We strongly recommend that stage and cytology be separated in future studies, in order to better define the impact of peritoneal cytology on outcome of endometrial cancer.