Methods	After surgery, women were randomised centrally by the Finnish Cancer Registry to two groups: Both groups received pelvic radiotherapy. Group A had no additional treatment. Group B was given chemotherapy during gaps in radiation treatment. The first dose of chemotherapy was given 1 to 2 weeks after surgery and before radiotherapy. The second cycle was given after half of the radiotherapy had been given. The third treatment cycle followed 2 weeks after the radiotherapy
Participants	The study population consisted of 156 women with endometrial cancer treated in four Finnish university hospitals between April 1992 and April 1996. They were followed up until April 2001, Women had noninvasive or early myoinvasive (< 50%) grade 3 endometrial cancer or any grade cancer that was deeply myoinvasive (> 50%), involving the cervix or peritoneal washings (old FIGO stage 1c-lllA). Thirteen (18%) women allocated radiotherapy only had stage 2 disease compared to 21 (25%) women allocated chemotherapy with radiotherapy. Nine (12.5%) women allocated radiotherapy only had malignant cells in the pelvic washings compared to 10 (11.9%) of women allocated chemotherapy with radiotherapy. The rest had stage 1 disease The surgery included a total abdominal hysterectomy (TAH), bilateral salpingo-oophorectomy. 80% of women also had pelvic lymphadenectomy
Interventions	Women allocated only radiotherapy received a total dose of 56 Gray (2 Gray/fraction) delivered to the pelvis using a four-field technique. The treatment was divided in two courses 28 Gray each, separated by a pause of 3 weeks. The first course started 4 to 5 weeks following the operation Women allocated chemotherapy with the radiotherapy received three chemotherapy cycles consisting of cisplatin 50 mg/m2 Cisplatin, epirubicin 60 mg/m2 and cyclophosphamide 500mg/m2. The first cycle was given after the final histopathological report was available, or 1 to 2 weeks after the operation. The second one was given during the pause in radiotherapy. The third course was given within 2 weeks after the completion of the second radiation course
Outcomes	The primary outcome was 5 year survival rate.
Results (OS & PFS)	During the five-year follow-up time 32 patients (20.5%) had recurrent disease. The recurrence rate was 22.6% in women allocated chemotherapy with radiotherapy and 18.0% in women allocated radiotherapy only. The difference is not statistically significant (P = 0.50) As calculated from the operation data, the median time to recurrence was 20 (range 8 to 60) months in women allocated chemotherapy compared to 15 (range 6 to 37) months in women allocated radiotherapy only (P = 0.17) Of the 32 patients with recurrent disease, only 6 stayed alive for the rest of the follow-up time. Twenty-six (81.2%) died of endometrial carcinoma. The median time interval from the recurrence to death was nine (range 2 to 22) months for women allocated chemotherapy with radiotherapy and eight (range 3 to 11) months for women allocated radiotherapy only. Even among the surviving patients, only two (one in each group) with a local recurrence were free of disease at the end of the follow-up. Fifteen women allocated chemotherapy with radiotherapy died and 11 of the women allocated radiotherapy only died. Consequently the disease-specific five-year survival was 82.1% and 84.7% respectively. The median survival for women allocated chemotherapy with radiotherapy was 37 (range 13 to 50) months compared to 23 (range 15 to 44) months for women allocated radiotherapy only (P = 0.148). The median disease-free survival was 25 months (range 12 to 49) for women allocated chemotherapy with radiotherapy and 18 (range 9 to 36) months for women allocated radiotherapy only (P = 0.13)
Results (site of recurrence)	The Kuoppala 2008 used chemotherapy sandwiched in between the radiotherapy courses. This trial reported a pure locoregional recurrence rate of 3.2% evenly distributed in the two groups, 3 women treated without chemotherapy of which 2 were alive after 5 years of follow up versus 2 treated by chemotherapy, of which one was alive after 5 years. In this trial chemotherapy did affect the distal recurrences rate, 13.8% versus 20.2%. Ten women (all died) developed a distal recurrence in the group not allocated chemotherapy compared to 17 (3 were still alive after 5 years of follow up) if chemotherapy had been used
Toxicity	Eight allocated both radiotherapy and chemotherapy required laparotomy for intestinal obstruction compared to 2 women treated by radiotherapy only Two patients in both groups experienced other major complications. One woman allocated chemotherapy with her radiotherapy suffered ureteric obstruction and one developed pelvic bone necrosis and bowel fistulation. One woman allocated radiotherapy only developed multiple gastric ulcerations causing serious malnutrition and another perforated her sigmoid colon. This was not attributable to recurrence Chemotherapy was implemented in three courses. The first, second and third treatment course could be administered according to the protocol 79 (94.0%), 70 (83.3%) and 64 (76.2%) women respectively. Of these, 6, 13 and 14 women respectively experienced grade 3 to 4 leucopenia. None suffered from grade 4 infection. There were two cases of sepsis and eight had Grade 3 infections. Nausea and vomiting were assessed as grade 3/4 in 5.6%. Grade 3 and 4 diarrhoea did not occur. Alopecia was severe (grade 3/4) in 49% of women. Liver and kidney functional test results remained within normal range throughout the chemotherapy
Notes
Risk of bias
Bias	Authors’ judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Women were allocated treatment by the central Finnish Cancer registry
Allocation concealment (selection bias)	Low risk	Allocation was coded by group A or B.
Blinding (performance bias and detection bias) All outcomes	High risk	It is not possible to conceal the treatment allocation from the participants
Incomplete outcome data (attrition bias) All outcomes	Low risk	Data were analysed on an intention to treat basis and all cases are accounted for In the group allocated chemotherapy with radiotherapy, 78 women (92.9%) were treated according to the protocol during the first period. Owing to intestinal symptoms one woman received only 20 Gray and another 22 Gray. One was treated with 49 Gray and another with 50 Gray in one period of radiotherapy. Two patients were not treated at all. In addition one patient had intestinal symptoms and one had infection, but both were nevertheless treated fully. During the second period the treatment was successful in 77 of women in this group (91.6%). Three had low leucocyte levels but were nevertheless treated according to protocol. Four patients complained of intestinal symptoms. Of, these, three were treated to the full dose but one received only 14 Gray. The second radiation period was omitted for one patient, because she developed sepsis. One refused the second treatment and four received none at all for unknown reasons Not all women received the prescribed radiotherapy dose, For women allocated radiotherapy only, 68 women (94.4%) received the full dose. One patient received 36 Gray and one was treated in one treatment period only with a dose of 45 Gray. Two patients refused the treatment and received no radiotherapy at all. During the second period 66 patients (91.6%) of the women allocated radiotherapy only were treated according to the protocol. Three suffered from intestinal symptoms but were treated to the full dose. However, there were another three patients treated to a suboptimal dose They received 11, 18 and 21 Gray, respectively. One patient was treated only during the first period and two patients opted to receive no radiotherapy at all
Selective reporting (reporting bias)	Low risk	There is no suggestion of selective reporting but that trial is old and was not reported according to CONSORT criteria
Other bias	Low risk	The groups were well matched.
Main problems with the trial as perceived by this analysis	Unclear risk	The trial design split the radiotherapy to include a break in treatment after delivering 26 Gray. This may allow metastatic deposits to recover especially in the group allocated radiotherapy. The split regimens would not be used in modern oncology and this limits the relevance of the data from this study