The cure rate for endometrial cancer is high, with an overall 5-year expected survival of at least 77% amongst the 7045 cases reported by the UK’s Office for National Statistics (ONS) in 2006.1 Most women with the commonest histological subtype (endometrioid) present with stage-1 disease and have an intrinsically good prognosis. In 2006 the International Federation of Gynecology and Obstetrics (FIGO) reported an all stage 5-year survival of 83% for endometrioid cancer, compared with 62.5% for those with clear cell tumours and only 52.6% for serous histology.2 In light of the growing evidence base in uterine cancer care, this commentary proposes a revised management strategy for this disease.
The vast majority of women with a well- or moderately-differentiated endometrioid cancer can be cured by a simple hysterectomy with a bilateral salpingo oophorectomy. This has traditionally been performed by laparotomy but many surgeons are now performing the procedure laparoscopically. The advantages of laparoscopic surgery were evident in the Gynecologic Oncology Group’s (GOG) LAP-2 trial, in which 2616 women with stage-I–IIA endometrial cancer were assigned randomly to laparoscopy or open laparotomy for their surgery,3 which followed the traditional pattern of comprehensive surgical staging, with hysterectomy and salpingo-oophorectomy, together with pelvic cytology, and pelvic and para-aortic lymphadenectomy. As expected, laparoscopy resulted in fewer post-operative moderate or severe adverse events and a shorter hospital stay, but a longer operative time. Surprisingly, 26% of those randomised to laparoscopy were converted to open laparotomy. Laparoscopic failure was mainly associated with older age and higher body mass index (BMI). However, difficulties with lymphadenectomy, which is now considered unnecessary by many authorities in endometroid disease, may have played a part, because pelvic and para-aortic nodes were not removed in 8% of laparoscopy patients and in only 4% of laparotomy patients. Similar advantages for laparoscopic surgery were shown in a smaller single-centre Italian trial (226 subjects), in which no patients were converted to open laparotomy and a BMI of up to 37 did not preclude laparoscopic surgery.4 Further follow up of patients in both trials will give relapse-free and overall survival rates for patients treated laparoscopically, compared with the open operation.
A potential hazard with laparoscopy is port site tumours, but this has been reported to occur in less than 1% of cases.5 Traditionally, FIGO staging included peritoneal cytology, but for the latest classification this is not required; nonetheless, surgeons should consider performing peritoneal cytology for the purposes of audit.
Patients should be referred to the gynaecological oncologist if they have poorly differentiated endometrioid carcinomas (including type-II subtypes), or carcinosarcomas (including serous or clear cell subtypes). Type-I endometrial carcinoma is estrogen-dependent, appears mostly in pre- and perimenopausal women and is well differentiated, and therefore has a better prognosis. Type-II endometrial carcinoma is estrogen independent, and is diagnosed mostly in postmenopausal women, thin and fertile women, or in women with normal menstrual cycles. It is aggressive and has a poorer prognosis. Type-II subtypes are more likely to spread to the pelvic lymph nodes and to metastasize.6 Prior to surgery such patients should have a magnetic resonance imaging (MRI) scan of the pelvis and a computed tomography (CT) scan of the chest, and of the abdomen as well. MRI is the best way to demonstrate deep myometrial invasion and cervical involvement, both of which are associated with nodal spread. CT is the best modality to demonstrate peritoneal spread, omental disease and para-aortic and other sites of distant disease, and should be considered in preference to MRI in type-II disease. Currently, the use of fluorodeoxyglucose positron emission tomography (FDG-PET)/contrast-enhanced CT has not yet been shown to add value to preoperative staging, although limited reports raise this as a future possibility.7,8 Poorly differentiated endometrioid carcinomas are more likely to spread to the pelvic lymph nodes, and serous tumours tend to behave like serous ovarian cancers with trans-peritoneal spread, especially to the omentum.6 Following imaging, surgery can be planned appropriately. The recently published Medical Research Council’s (MRC) ASTEC trial and the EORTC trial have not shown any survival advantage following routine pelvic lymphadenectomy.9 However, in our opinion enlarged nodes, as demonstrated in MRI or CT scanning, should be removed. Tumour in the lymph nodes is an adverse prognostic finding, and an absolute indication for adjuvant treatment. Para-aortic lymphadenectomy may give additional prognostic information, and a recent large retrospective study suggested a survival advantage if both pelvic and para-aortic systematic lymphadencetomy are performed,10 although there is no evidence from randomised controlled trials of a therapeutic benefit.11 Optimal surgery for serous tumours is inspection of the whole peritoneal cavity, omentectomy or omental biopsy, and removal of any extrauterine disease.12,13 The role of pelvic lymphadenectomy in this group of patients has not been evaluated. Our current practice is to remove nodes that have been shown to be enlarged after imaging.
Only 20–30% of patients with endometrial cancer require adjuvant therapy. Traditionally, patients have been selected using the well-tried clinicopathological criteria of tumour histological type, grade, degree of myometrial invasion and the presence or absence of lymph nodes metastases. At present, molecular markers have not been applied routinely in diagnostic practice, but in future they may extend pathological precision.6
The days of all patients having routine postoperative radiotherapy are over. Both the PORTEC1 and the recently published combined ASTEC and EN.5 trials have shown no survival advantage for routine postoperative radiotherapy.14,15 In this commentary, we have used the new FIGO (2009)16 staging system rather than the older staging listed in the studies described below. In the PORTEC-1 trial patients with intermediate risk tumours (grade-1 FIGO stage 1B; G2 1B; G3 1A) were randomised against external beam radiotherapy versus observation alone.14,15 External beam radiotherapy produced a statistically significant reduction in local regional recurrence (14% no Radiotherapy, 4% RT, P ≥ 0.001). However, the 5-year statistical survival was similar to the 5-year survival rates of 85 and 81%, respectively (P = 0.31). One of the reasons that there was no difference in survival is that salvage radiotherapy is particularly effective amongst the small number of PORTEC patients who developed a recurrence in the vagina, which was the most frequent site of relapse: the 3-year survival rate after localised vaginal relapse was 69%.14
The interpretation of the MRC ASTEC/EN.5 trial is somewhat more difficult owing to the fact that more than half the patients in both the observation arm and the pelvic external beam radiotherapy arm received vaginal brachytherapy, and this must have had some impact upon the local recurrence rates.15 The 5-year survival was identical in both patients receiving ‘observation’ (half had brachytherapy) and those who had received pelvic external beam radiotherapy, at 84%. The role of locoregional recurrence was 6.1% in the observation group against 3.2% in those receiving pelvic external beam radiotherapy.15 There remains a case for giving external beam radiotherapy for patients with poorly differentiated tumours penetrating more than half way through the myometrium (stage 1B, G3), as a meta-analysis and systematic review carried out by Johnson and Cornes showed that pelvic radiotherapy gave a 10% survival advantage for this group of patients.16,17
The use of vaginal brachytherapy may well increase throughout the UK following the presentation of the PORTEC-2 results.18 In this study intermediate-risk endometrial cancer patients were randomised to external beam radiotherapy or vaginal brachytherapy. The vaginal relapse rate was 0.9% of patients treated with vaginal brachytherapy, with a pelvic relapse rate of 3.6%. Of those receiving pelvic external beam radiotherapy, 2.0% had a vaginal relapse and 0.7% had a pelvic relapse. Progression-free survival at 3 years was identical in both arms at 89%, but the incidence of radiation-induced side effects was significantly less in the brachytherapy group, who had a better quality of life.18,19 Brachytherapy is easy to administer via a vaginal cylinder that can be inserted without an anaesthetic. Our policy is to offer patients suitable for vaginal brachytherapy a choice between immediate treatment or a regimen of close observation, as in the PORTEC-1 study 73% of all recurrences occurred in the vagina, and the salvage rate was high at 69%.17
Until recently few patients with endometrial cancer received chemotherapy, but following the publication of the Nordic Society of Gynaecological Oncology (NSGO) trial, patients with high-risk tumours may be given chemotherapy to try and sterilise occult distant metastases.20,21 In the NSGO study, 382 patients were randomised to receive pelvic radiotherapy or a combination of chemotherapy and radiotherapy. After a median follow up of 4.3 years, the 5-year progression-free survival was 82% for patients receiving both chemotherapy and radiotherapy, compared with 74% for those receiving radiotherapy alone (HR, 0.62; 95% CI, 0.40–0.97; P = 0.03). In this study 382 patients were randomised between 1996 and 2007 and chemotherapy practice changed during this time. In the early part of the trial patients were generally treated with cisplatin (≥50 mg/m2) and either doxorubicin (50 mg/m2) or epirubicin (75 mg/m2), but after 2004 they received paclitaxel (175 mg/m2) and carboplatin (AUC 5–6), with or without epirubicin (60 mg/m2). There were compliance problems as 27% of patients did not complete the four cycles of chemotherapy. In 42% of cases the clinician decided to discontinue treatment because of unacceptable toxicity, and in 38% of cases the patient decided to stop chemotherapy. By comparison, radiotherapy was completed in over 90% of patients in both arms of the study.19,20
Ideally chemotherapy should be given to high-risk patients in the context of a clinical trial. The possible use of combined chemoradiotherapy is being tested further in the Anglo-Dutch PORTEC-3 trial.21 There are four major end points in this study: overall survival, recurrence-free survival, toxicity and quality of life. Eligible patients for this study include those with: (i) stage-1A, grade-3 disease with lymphovascular space invasion; (ii) stage-1B, G3; (iii) stage-2 G3; and (iv) stages 3A or 3C.16,21 Patients with serous or clear cell histology are also eligible. Patients are randomised to receive external beam pelvic radiotherapy with a brachytherapy boost if there was spread to the cervix, or pelvic radiotherapy with concurrent chemotherapy. In the chemoradiotherapy arm patients will receive two pulses of cisplatin chemotherapy (50 mg/m2) during their external beam radiotherapy (weeks 1 and 4), followed by four cycles of carboplatin (AUC 5) and paclitaxel (175 mg/m2, at 3-weekly intervals).21 This protocol has received some criticism, particularly about the dose of cisplatin during radiotherapy and the timing of adjuvant chemotherapy. It has been argued that compliance would be higher if the chemotherapy was given first. The study coordinators have replied that the timing of radiotherapy should be the same in both arms of the study, and point to the fact that all components of the study have been piloted and have been shown to be tolerable. In parts of the USA it is now standard care for patients with high-grade tumours to receive adjuvant chemotherapy, especially those with serous or clear cell components. As yet there is insufficient evidence to support this conclusion. It is important that the gynaecological community support the PORTEC-3 study, as this may give definitive information on the role of chemotherapy. There are no other competing studies in the UK.
In the last decade randomised controlled trials have cast doubt upon conventional beliefs in the treatment of endometrial cancer, particularly about the role of lymphadenectomy and postoperative radiotherapy. Perhaps patients with an intrinsically good prognosis should receive less treatment, and those at risk of developing metastases should receive more therapy (systematic pelvic and para-aortic lymphadenectomy, chemoradiotherapy and adjuvant chemotherapy), but this must be tested rigorously in prospective randomised trials. In the future there is a real prospect of individualised treatment using molecular-targeted therapies, perhaps in combination with radiation or conventional chemotherapeutic agents, which may improve prognosis and reduce toxicity in the small group of high-risk cases requiring additional treatment after surgery.