I. Introduction

Endometrial cancer is the most common gynecological cancer in the western world. In the United States every year 36,100 new cases are diagnosed and approximately 6,500 women die of this malignancy (Greenlee et al, 2000). Of the 6,088 endometrial cancer patients assessed by the International Federation of Gynecology and Obstetrics (FIGO) Annual Report n. 24, 2.8% were younger than 40 years of age, 9.9% were 40-49 years old, 25.6% were 50-59 years old, 33.3% were 60-69 years old, and 28.4% were 70 years or older (Creasman et al, 2001).

Two different clinico-pathological and biological types of endometrial cancer can be considered (Sherman et al, 1995; Matias-Guiu et al, 2001; Prat, 2002). Type-1 endometrioid carcinomas are estrogen-dependent, are often associated with endometrial atypical hyperplasia, and generally have a good prognosis, whereas type-2 non-endometrioid carcinomas (i.e. serous papillary and clear cell carcinomas) are estrogen- independent, arise in an atrophic endometrium, affect older women, and display a high biological aggressiveness with poor clinical outcome. The recurrence rate of these latter is extremely high and the most frequent extra-pelvic sites of relapse are the upper abdomen, lungs and liver (Trope et al, 2001). Molecular pathogenesis is quite different for the two variants (Sherman et al, 1995; Matias-Guiu et al, 2001; Prat, 2002). Four different genetic abnormalities may occur in endometrioid carcinomas (microsatellite instability and mutations in the PTEN, k-RAS and beta-catenin genes), whereas nonendometrioid carcinomas often display p53 gene mutations and loss of heterozygosity on several chromosomes. Occasionally a nonendometrioid carcinoma may develop as a result of dedifferentiation of a preexisting endometrioid carcinoma; in such a case, the tumor exhibits overlapping clinical, morphologic, immunohistochemical and molecular features of the two types (Matias-Guiu et al, 2001).

Endometrial cancer spreads by direct extension (myometrium, cervix, vagina, adnexa, and more rarely, bladder and rectum), lymphatic spread, and hematogeneous spread (Oram, 1990). Lymphatic trunks drain mainly into the pelvic and para-aortic nodes, and less frequently into the groin nodes through vessels lying in the round ligament (Boronow et al, 1984; Oram, 1990; Ayhan et al, 1994; Benedetti Panici et al, 1998). In a prospective trial performed by Ayan et al (1994) on 183 patients with clinical stage I disease submitted to systematic lymphadenectomy, pelvic and para-aortic metastases were found in 15.3% and 9.3% of the cases, respectively. In a study of the Gynecologic Oncology Group (GOG), para-aortic metastases were detected in 2% of patients with negative pelvic nodes compared to 67% of those with positive pelvic nodes (Boronow et al, 1984).

In 1988 the FIGO Committee on Gynecologic Oncology changed the staging of endometrial cancer from a clinical to a surgical-pathologic staging that requires an accurate histopathologic examination of the surgical sample aimed to assess the true extension of the disease and to define a series of prognostic variables (Mikuta, 1993) (Table 1). Each case is graded histologically as G₁, G₂ or G₃, according to whether nonsquamous or nonmorular solid growth pattern involves <5%, 6-50%, or >50% of the glandular component (Mikuta, 1993; Creasman et al, 2001). Significant nuclear atypia increases the histologic grade one step. In serous papillary, clear cell and adenosquamous carcinomas, the nuclear grade takes precedence. Adenocarcinoma with squamous differentiation is graded according to the nuclear grade of the glandular component.

An accurate surgical staging shows that approximately 15-20 % of patients with tumor apparently confined to the uterine body have a subclinical extension of the disease to the cervix, adnexa, lymph nodes or peritoneum (De Palo et al, 1993). The high operability rate of endometrial cancer makes this staging system a viable one, which provides information about the need for additional treatment. Patients who are treated primarily with radiation therapy should be staged according to the clinical staging system adopted by FIGO in 1971.

Table 2 reports the most important surgical-pathological prognostic variables, most of which are included in the 1988 FIGO staging system (Mikuta, 1993). Among the 5,694 surgically staged patients reported in the FIGO Annual Report n. 24, 5-year survival was 88.9% for stage Ia, 90.0% for stage Ib, 80.7% for stage Ic, 79.9% for stage IIa, 72.3% for stage IIb, 63.4% for stage IIIa, 38.8% for stage IIIb, 51.1% for stage IIIc, 19.9% for stage IVa, and 17.2% for stage IVb (Creasman et al, 2001). Survival was related to the degree of differentiation for any stage and to histological type. For instance 5-year survival ranged from 92.0% for grade G₁to 74.0% for grade G₃ among stage I patients_,from 78.6% for grade G₁ to 44.4% for grade G₃among stage IIIa patients, and from 61.2%for grade G₁ to 44.0% for grade G₃among stage IIIc patients. In the same series 5-year survival was 79.7% for endometrioid carcinoma, 79.1% for adenosquamous carcinoma, 72.9% for mucinous carcinoma, 54.3% for serous papillary carcinoma, and 63.2 % for clear cell carcinoma. The aggressive biological behaviour, and in particular the higher frequency to extrauterine spreading, helps to explain the poorer survival of serous papillary and clear cell carcinomas (Cirisano et al, 2000; Trope et al, 2001). Several authors reported that uterine papillary serous carcinoma can relapse also in patients with tumor confined to the endometrium or to an endometrial polyp (Silva and Jenkins, 1990; Lee and Belinson, 1991; Carcangiu and Chambers, 1992; Suzuki et al, 1999).

Lymph node involvement is related to several pathological variables such as histological grade, myometrial invasion, histological type, lymph-vascular space involvement, tumor size, cervical extension, peritoneal cytology, and adnexal metastases (Creasman et al, 1987; Feuer and Calanog, 1987; Morrow et al, 1991; Bell et al, 1997). In a GOG study including 621 stage I endometrial cancer patients positive pelvic and para-aortic node rate ranged from 3% and 2% for grade G₁ to 18% and 11% respectively for grade G₃, and from 1% and 1% for tumors confined to endometrium to 25% and 17% respectively for tumors invading myometrium deeply (Creasman et al, 1987).

The prognostic relevance of peritoneal cytology in patients with endometrial cancer apparently confined to the uterine body, but that are classified in stage IIIa only for the presence of neoplastic cells in peritoneal fluid or washing is still debated (Turner et al, 1989; Grimshaw et al, 1990; Grigsby et al, 1992; Kadar et al, 1992; Vecek et al, 1993; Kennedy et al, 1993; Kashimura et al, 1997; Ebina et al, 1997; Obermair et al, 2001; Takeshima et al, 2001; Luo et al, 2001; Kasamatsu et al, 2003) (Table 3). Conversely positive peritoneal cytology is a poor prognostic factor in patients with extrauterine disease (Kadar et al, 1992; Ebina et al, 1997; Takeshima et al, 2001). For instance according to Kadar et al (1992) if the tumor had spread to the adnexa, lymph nodes or peritoneum, positive peritoneal cytology had a detrimental effect on 5-year survival, decreasing it from 73 to 13%. Similarly Ebina et al (1997) reported that in stages IIIc and IV the prognosis was significantly poorer for patients with positive than for those with negative peritoneal cytology.

Hirai et al (2001) investigated the malignant potential of endometrial cancer cells in peritoneal washing in 50 patients with clinical stage I–II disease in whom positive peritoneal cytology was found at surgery. A tube for cytologic analyses was inserted into the peritoneal cavity when closing the abdomen, and afterwards the peritoneal cavity was irrigated with physiologic saline and washings were obtained through the tube seven and fourteen days after surgery. Persistence of positive peritoneal cytology was observed in 4 out of 7 patients with adnexal metastases, none of the 9 patients with nodal disease, and 1 of 34 patients with disease confined to the uterus, for a total of 10% (5 of 50).

Table 1. FIGO surgical staging of endometrial cancer (Rio de Janeiro, 1988)

Stage
Ia	Tumor limited to the endometrium (G₁, G₂, G₃)
Ib	Invasion to < half of myometrium (G₁, G₂, G₃)
Ic	Invasion to > half of myometrium (G₁, G₂, G₃)
IIa	Endocervical glandular involvement only (G₁, G₂, G₃)
IIb	Cervical stromal involvement (G₁, G₂, G₃)
IIIa	Involvement of uterine serosa and/or adnexae and/or positive peritoneal cytology (G₁, G₂, G₃)
IIIb	Vaginal involvement (G₁, G₂, G₃)
IIIc	Metastases to pelvic and/or para-aortic lymph nodes (G₁, G₂, G₃)
IVa	Involvement of bladder and/or rectal mucosa (G₁, G₂, G₃)
IVb	Distant metastases (including intra-abdominal and groin lymph node metastases) (G₁, G₂, G₃)

G₁, grade 1; G₂, grade 2; G₃, grade 3

Table 2. Surgical-pathological risk factors of endometrial cancer

Intra-uterine risk factors

1. Depth of myometrial invasion

2. Histologic grade

3. Histology

4. Cervical extension

5. Lymph vascular space involvement

Extra-uterine risk factors

1. Pelvic and para-aortic node metastases

2. Adnexal involvement

3. Penetration of uterine serosa

4. Positive peritoneal cytology

Table 3. Independent prognostic significance of positive peritoneal cytology in endometrial cancer confined to the uterine body

NO	YES
Grimshaw et al, 1990	Turner et al, 1989
Lurain et al, 1991	Grigsby et al, 1992
Vecek et al, 1993	Kennedy et al, 1993
Ebina et al, 1997	Kashimura et al, 1997
Takeshima et al, 2001	Luo et al, 2001
Kasamatsu et al, 2003	Obermair et al, 2001

In the other 45 (90%) patients no malignant cells were found in any of the washings. These data seem to suggest that endometrial cancer cells found in the peritoneal cavity usually disappear within a short time and have a low malignant potential. Only malignant cells from particular cases, such as adnexal metastases, appear to be capable of independent growth and could represent a risk factor for recurrence.

II. Surgery

Whenever possible surgery is the initial treatment for both early and advanced endometrial cancer (Bremond et al, 2001). The standard surgical approach consists of laparotomy, peritoneal washing, extrafascial hysterectomy, bilateral salpingo-oophorectomy, partial colpectomy, and pelvic and para-aortic lymph node dissection (Oram, 1990; Boronow et al, 1984; De Palo et al, 1993; la Vecchia et al, 1983; Chen, 1989; Vardi et al, 1992; Lanciano et al, 1993; Gretz et al, 1996; Boronow, 1997). However the lack of resection of the upper vagina does not seem to have a strong impact on survival (la Vecchia et al, 1983). Different surgical procedures can be performed to assess retroperitoneum, ranging from biopsies of enlarged nodes only to selective node sampling from multiple sites to systematic pelvic and para-aortic lymphadenectomy (Chuang et al, 1995; Kilgore et al, 1995; Onda et al, 1997). In the experience of Kilgore et al (1995) the chance to detect microscopic metastases was related to the number of sampled pelvic sites and to the number of removed nodes.

Maximal surgical cytoreduction should be recommended for patients with advanced endometrial cancer, since the achievement of a residual disease <1 cm appears to be an independent prognostic variable for survival (Goff et al, 1994; Chi et al, 1997; Bristow et al, 2000, 2001; Ayhan et al, 2002). For instance in the series of Bristow et al (2000), including 65 patients submitted to primary surgery for stage IVb endometrial cancer, median survival was 34.3 months for patients with residual disease <1 cm compared to 11.0 months (p=0.0001) for those with larger residum). As for the former, the patients with microscopic residual disease had a significant better survival compared to those with macroscopic residuum. Similarly, among the 37 patients with stage IVb endometrial cancer treated by Ayhan et al (2002), median survival was 25 months for the patients with residuum <1cm compared to 10 months (p=0.01) for those with bulkier residual disease.

Intensive surgical staging (also including omentectomy and peritoneal biopsies) similar to that required for ovarian cancer is warranted for serous papillary and clear cell carcinoma of the endometrium (Gallion et al, 1989; Bancher-Todesca et al, 1998; Cirisano et al, 2000; Chan et al, 2003).

It is uncertain whether lymph node dissection itself gives a clinical benefit (Candiani et al, 1990; Chuang et al, 1995; Kilgore et al, 1995; Fanning et al, 1996; Massi et al, 1996; Onda et al, 1997; Orr et al, 1997; Bar-Am et al, 1998; Larson et al, 1998; Podratz et al, 1998; Mohan et al, 1998; Trimble et al, 1998; Mariani et al, 2000; Seago et al, 2001; Rittenberg et al, 2003). Cumulative data from the literature showed the development of recurrent disease in 20 (6.6%) out of 305 patients who had been submitted to systematic lymphadenectomy for medium risk endometrial cancer, who had negative nodes and who had not undergone postoperative pelvic irradiation (Fanning et al, 1996; Orr et al, 1997; Podratz et al, 1998; Larson et al, 1998; Mohan et al, 1998). Only 5 of these 20 recurrences were loco-regional. Subsequent studies confirmed that whole pelvis irradiation can be safely omitted in patients with FIGO stage Ic or stage I grade G₃ endometrial cancer if nodal status is known (Seago et al, 2001; Rittenberg et al, 2003). Therefore systematic lymphadenectomy could avoid adjuvant external irradiation in medium risk patients with histologically proven negative lymph nodes.

The therapeutic relevance of lymphadenectomy has been long debated (Candiani et al, 1990; Chuang et al, 1995; Kilgore et al, 1995; Massi et al, 1996; Onda et al, 1997; Larson et al, 1998; Mohan et al, 1998; Podratz et al, 1998; Mariani et al, 2000). For instance, both Candiani et al (1990) and Massi et al (1996) failed to detect a difference in survival between patients who had undergone pelvic lymphadenectomy and those who had not. Conversely, Kilgore et al (1995) showed a survival benefit for patients undergoing multiple site pelvic lymph node sampling compared with those not receiving such procedure. Mariani et al (2000) assessed 137 endometrial cancer patients with an high risk of para-aortic metastases due to deep myometrial invasion or palpable pelvic nodes or adnexal involvement, and 51 patients who had positive pelvic or para-aortic nodes. Among the former 5-year survival was 85% for patients who had undergone para-aortic lymphadenectomy compared to 71% (p=0.06) for those who had not, and among the latter 5-year survival was 77% for patients who had undergone this surgical procedure compared to 42% (p=0.05) for those who had not.

Piver II-III radical hysterectomy is sometimes performed in endometrial cancer patients, and particularly in those with macroscopic involvement of the uterine cervix (Rutledge, 1974; Boothby et al, 1989; Boente et al, 1995, Mariani et al, 2001; Sartori et al, 2001). The rationale for this operation is that the removal of parametria provides more adequate free surgical margins when endometrial cancer has spread to the cervix (Sartori et al, 2001). In 1995 a study performed on specific diagnostic and therapeutic options in endometrial cancer by means of a questionnaire sent to several leading centres for gynecologic oncology in Western Europe, revealed that Piver II-III hysterectomy was routinely performed in 6.2% of 81 institutions, never used in 11.1%, and adopted for specific conditions (such as FIGO stage >I, younger age, poorly differentiated tumor) in 82.7% of the centres (Maggino et al, 1995). As for the elective surgical management in stage II disease, radical hysterectomy was considered to be the treatment of choice in 79.5% of the institutions.

Rutledge (1974) reported that 5-year survival was not significantly different between endometrial cancer patients who underwent extended hysterectomy and those treated with simple hysterectomy, even though, among stage I patients, the local recurrence rate was lower in the radical surgery group. In the study of Bonte et al (1995) on patients with stage II disease, pelvic recurrence rate was 6% among the 33 patients submitted to radical hysterectomy and 14% among the 37 submitted to extrafascial hysterectomy. The retrospective analysis of 203 stage II endometrial cancer cases treated in five different Italian gynecologic oncology centres detected that 5-year and 10-year survivals were significantly better in the 68 patients who had undergone radical hysterectomy compared to the 135 who had undergone simple hysterectomy (94% versus 79%, and, respectively, 94% versus 74%, p=0.03) (Sartori et al, 2001). Vaginal hysterectomy may be considered a reasonable alternative to the abdominal approach in specific clinical conditions, such as obesity, old age, uterine prolapse, poor performance status, and high anesthesiological risk (Maggino et al, 1995; Massi et al, 1996; Chan et al, 2001). In the series of Massi et al (1996) including stage I endometrial cancer patients, 5-year and 10-year-survivals were superimposible in the 180 patients who underwent vaginal hysterectomy (90% and 87%, respectively) and in the 147 who underwent abdominal hysterectomy (91% and 90%, respectively). Vaginal operation can be complemented by extraperitoneal pelvic lymphadenectomy according to a modification of Mitra’s technique, that is a fast procedure applicable in high-risk surgical patients under spinal anesthesia (Massi et al, 2000), as well as by a laparoscopic staging (Gemignani et al, 1999; Eltabbakh et al, 2001; Malur et al, 2001; Eltabbakh, 2002; Fram, 2002; Holub et al, 2002; Langebrekke et al, 2002; Occelli et al, 2003).

In particular laparoscopic assisted vaginal hysterectomy has been recently considered as a technically acceptable and safe surgical procedure for early endometrial cancer in the hands of experienced operators. It is associated with a longer operating time, but with a less blood loss, a shorter hospitalisation and equivalent clinical outcomes when compared to abdominal hysterectomy (Gemignani et al, 1999; Eltabbakh et al, 2001; Malur et al, 2001; Eltabbakh, 2002; Holub et al, 2002; Fram, 2002; Occelli et al, 2003). For instance a retrospective review of Eltabbakh (2002) on women with clinical stage I endometrial cancer showed that the 100 patients who underwent laparoscopy and the 86 who underwent laparotomy had similar 2-year and 5-year recurrence-free survivals (93% versus 94%, and 90% versus 92%, respectively), as well as similar 2-year and 5-year overall survivals (98% versus 96% and 92% versus 92%, respectively). Both groups were similar with regard to age, lymphadenectomy, surgical stage, tumor grade, histology, and postoperative radiation therapy, and moreover there was no apparent difference with regard to the sites of recurrence between the two groups. However in low-risk endometrial cancer patients treated by laparoscopic assisted vaginal hysterectomy an higher incidence of vaginal cuff recurrences or positive peritoneal cytology has been sometimes reported, and some cases of port-site recurrences have been described, probably due to the intraoperative dissemination of tumor cells enhanced by the use of intrauterine manipulator (Wang et al, 1997; Zayyan and Kazmi, 1998; Muntz et al, 1999; Sonoda et al, 2001; Vergote et al, 2002; Chu et al, 2003). For instance, in a large series of low risk endometrial cancer patients Sonoda et al (2001) found a positive peritoneal cytology in 10.3% of the 131 patients treated with laparoscopic assisted vaginal hysterectomy compared to 2.8% of the 246 patients who underwent abdominal hysterectomy. Therefore, the routine use of laparoscopic assisted vaginal hysterectomy should be undertaken with caution, as the long-term risks for recurrence and survival have yet to be defined in large, randomised controlled trials (Chu et al, 2003).

As for recurrent endometrial cancer, intensive surgery has been sometimes suggested for women with large pelvic or abdominal relapse (Scarabelli et al, 1998). In carefully selected cases of isolated central recurrences pelvic exenteration is the only potential option for cure (Morris et al, 1996; Barakat et al, 1999; Chi and Barakat, 2001). In the series of Morris et al (1996), including 20 patients with recurrent endometrial cancer who underwent exenteration with curative intent, the estimated 5-year disease-free survival was 45%. Twelve (60%) patients experienced major complications, the most common of which was neovaginal flap necrosis, and 1 (5%) patient died of surgical complications. Barakat et al (1999) reassessed 44 patients with central recurrence after surgery with or without radiotherapy who underwent pelvic exenteration. Major postoperative complications occurred in 35 (80%) patients and included urinary/intestinal tract fistulas, pelvic abscess, septicemia, pulmonary embolism, and stroke. Median survival for the entire group of patients was 10.2 months, but 9 (20%) achieved long-term survival (>5 years). Further investigation into the techniques of intraoperative radiotherapy could increase the pool of patients to whom a salvage surgery may be sometimes offered (Chi and Barakat, 2001).

III. Radiotherapy

Exclusive radiotherapy consisting of both external beam irradiation and uterovaginal brachytherapy is undertaken only in selected cases unsuitable for surgical management due to medical contraindications or advanced age (Rose et al, 1993; Rouanet et al, 1993; Fishman et al, 1996; Thomas et al, 2001; Peiffert et al, 2003). In patients with early tumor this treatment modality can obtain a local control rate of about 80-90% (Peiffert et al, 2003). Fishman et al (1996) assessed 54 patients with clinical stage I and II endometrioid carcinoma who were deemed medically inoperable and exclusively received radiation therapy, and a cohort of 108 operable patients adjusted for age, clinical stage, and grade as a control group. The 5-year actuarial cancer-specific survivals for patients with stage I inoperable, stage II inoperable, stage I operable, and stage II operable disease were 80, 85, 98, and 100%, respectively.

External pelvic irradiation is the most common adjuvant treatment in endometrial cancer, able to reduce loco-regional recurrences without improving overall survival (Aalders et al, 1980; Roberts et al, 1999; Creutzberg et al, 2000; Straugh et al, 2003; Creutzberg et al, 2003). In the randomised study of Aalders et al (1980) enrolling 540 stage I patients, vaginal-pelvic recurrence rate was 1.9% in patients who received adjuvant intravaginal irradiation plus external pelvic irradiation compared to 6.9% (p<0.01) in those who received adjuvant intravaginal irradiation alone, whereas distant metastases developed in 9.9% of the former and in 5.4% of the latter. Thus, the 5-year survival was not improved by external irradiation. A more detailed analysis led to the conclusion that only patients with poorly differentiated tumors which infiltrate more than half of the myometrial thickness, might benefit from additional external radiotherapy. In the GOG 99 trial comparing surgery versus surgery plus external irradiation in patients with intermediate-risk stage endometrial cancer, loco-regional recurrence rate was 1.6% in the patients who had radiotherapy compared to 8.5% in those who had no further treatment, whereas survivals were not significantly different (Roberts et al, 1999). In the multicenter randomised PORTEC [Post Operative Radiation Therapy in Endometrial Cancer] study enrolling 715 patients with intermediate-risk stage I disease, adjuvant external pelvic irradiation reduced loco-regional recurrence rate (4% versus 14%, p<0.001) and increased treatment complication rate (25% versus 6%, p<0.0001) compared to surgery alone, whereas deaths for cancer (9% versus 6%), 5–year survivals (81% versus 85%), and 8-year actuarial survivals (71% versus 77%) were unchanged (Creutzberg et al, 2000; Creutzberg et al, 2003). As pelvic radiotherapy appears to improve loco-regional control without a survival benefit, its use should be limited to those patients at sufficiently high risk (15% or over) for recurrence in order to maximize local control and relapse-free survival (Creutzberg et al, 2003).

Vaginal vault brachytherapy alone has been used as postoperative treatment in patients with low risk (stage Ib G₁, G₂) disease (Alektiar et al, 2002) or in patients with medium risk (stage I G₃ or stage Ic) disease with histologically proven negative lymph nodes (Seago et al, 2001; Rittenberg et al, 2003), but the clinical benefit of this adjuvant therapy versus surgery alone has not yet been defined. Moreover, there is no definitive suggestion that the addition of a vaginal vault brachytherapy boost to external beam irradiation is beneficial for pelvic control or disease-free survival of stage I or II endometrial cancer patients and prospective randomised trials designed to study external beam irradiation alone versus external beam irradiation plus vaginal brachytherapy are unlike to show a positive result (Greven et al, 1998).

Little data are currently available about extended-field irradiation in patients with positive para-aortic nodes (Feuer and Calanog, 1987; Corn et al, 1992; Rose et al, 1992; De Palo et al, 1993; Hicks et al, 1993; Rotman et al, 1993). In the study of Feuer and Calanog (1987) this treatment obtained a 5-year survival of 66.7% in patients with microscopic aortic metastases compared to 16.7% in those with macroscopic aortic metastases. Rose et al (1992) gave extended field irradiation to 17 out of 26 patients with para-aortic involvement, and found that 9 (53%) of them were alive with no evidence of disease with a median survival of 27 months, whereas 87.5% of the 8 patients treated with chemotherapy or progestins died after a median time of 13 months. By assessing 19 patients with positive para-aortic nodes, Hicks et al (1993) detected a 5-year disease-free survival of 27% in patients treated with pelvic plus para-aortic irradiation compared to 0% in those who received pelvic irradiation plus hormonotherapy. Para-aortic irradiation could be effective particularly in the control of microscopic disease even after surgical debulking; however, this procedure is sporadically used in the clinical practice for both the risk of severe bowel complications and the suggestion that para-aortic node involvement is associated with systemic spread of disease (Corn et al, 1992; Rotman et al, 1993).

The role of whole abdomen irradiation in selected patients is still debated (Martinez et al, 1988; Small et al, 2000; Smith et al, 2000; Lee et al, 2003; Martinez et al, 2003). Recently Lee et al (2003) reported that whole abdomen irradiation with a cumulative dose of 3000 cGy plus supplementary doses to partial abdominal volumes can eradicate small peritoneal deposits after surgical cytoreduction. This treatment modality has been evaluated also in patients with uterine serous papillary and clear cell carcinoma (Smith et al, 2000; Martinez et al, 2003). Smith et al (2000) assessed 22 patients with FIGO Stage III-IV endometrioid carcinoma and 26 patients with FIGO Stage I-IV serous papillary or clear cell carcinoma treated postoperatively with whole abdomen irradiation (median dose: 3000 cGy to the upper abdomen and 4980 cGy to the pelvis, respectively). Patients with endometrioid carcinoma had 3-year disease-free and 3-year overall survival of 79% and 89%, respectively, compared with 47% and 68% in the group of patients with serous papillary or clear cell carcinoma. Stage I-II patients with serous papillary or clear cell carcinoma had 3-year disease-free survival and overall survival of 87% compared with 32% and 61% in those with stage III and IV disease. The 3-year actuarial major complication rate was 7%, with no treatment-related deaths. These data appear to suggest that whole abdomen irradiation is a safe, effective treatment for patients with optimally debulked advanced stage endometrioid or early stage serous papillary or clear cell carcinoma. Martinez et al (2003) gave postoperative whole abdomen irradiation with a pelvic/vaginal boost to 132 patients with stage I-III endometrial carcinoma at high risk for abdomen-pelvic recurrence, including serous-papillary and clear cell histology. The 5- and 10-year cause-specific survival was 77% and 72% for the entire group, 75% and 70% for adenocarcinoma, and 80% and 74% for serous papillary and clear cell carcinoma, and the long-term complication rate was acceptable (chronic grade 3/4 gastrointestinal toxicity in 14% and grade 3 renal toxici

Review Article

Received: 18 December 2003; Accepted: 29 December 2003; electronically published: December 2003

I. Introduction

II. Surgery

III. Radiotherapy