The objective was to assess the long-term survival (5-15 years) in 463 women, with stages IIb-IV epithelial carcinoma of the ovary and to compare their survival with that of a normal population matched for age and sex. Statistical analysis of 463 women, with stages IIb-IV epithelial cancer of the ovary, who were participants in two consecutive North Thames Ovary Group randomized trials, which took place between 1985 and 1994, was performed. The median follow-up period was 10.5 years. The women were treated with debulking surgery, where possible, and adjuvant platinum chemotherapy. One of the randomized groups in the first North Thames trial also received total abdominal radiotherapy. Survival rates at 5, 10, and 15 years were assessed. Prognostic factors for long-term survival were determined using a mathematical model to separate early effects from late effects. The ratio of observed to expected deaths compared to the normal population was calculated. Overall survival at 5 years was 21% (95% confidence intervals 17.5-25%), at 10 years was 13.5% (95% confidence intervals 10.5-17%), and at 15 years was 12% (95% confidence intervals 9-16%). The important prognostic factors for long-term survival were disease-free or minimal residual disease (a single remaining deposit <2 cm) at initial surgery with tumor grade 1 and good performance status. Compared with the normal population (1995 data), the ratio of observed to expected deaths after start of chemotherapy at 5 years was 14.1 (P < 0.001 Fisher's exact test), at 9-10 years 4.9 (P = 0.0033, Fisher's exact test), while in the 11- to 15-year period it had dropped to 2.75 (P = 0.090, Fisher's exact test), which was not significantly different. Patients with advanced cancer of the ovary, who survive 11 years or longer, have a life expectancy which is very similar to that of a normal population of women of the same age. Women with advanced ovarian cancer have an improved chance of long-term survival following treatment if they present with minimal residual disease after primary surgical debulking, grade 1 tumors, and good performance status.

We evaluated the sequential use of carboplatin, paclitaxel and topotecan in patients with advanced, previously untreated ovarian cancer. In total, 43 patients with advanced ovarian cancer and >1 cm residual disease were treated with sequential carboplatin (area-under-the-curve (AUC) 5 days 1 and 22), paclitaxel (175 mg m(-2) days 43 and 64) and topotecan (1.5 mg m(-2) daily for 5 days from days 85, 106, 127 and 148). Median age of patients was 6 1 years. Median follow-up was 22.2 months (range 0.76-50.6 months). In all, 34 (79%) patients received all eight cycles of treatment and nine (21%) withdrew. Of the 29 evaluable patients, 19 (66%) responded according to WHO and 30 of 36 (83%) patients according to CA125. The best overall response (CA 125 and/or WHO) was 77% (33 of 43 patients). The response rates to sequential drugs based on >50% fall in CA 125 were as follows: carboplatin, 77% (30 of 39 patients); paclitaxel, 65% (15 of 23 patients); topotecan, 38% (five of 13 patients). Two patients responded to paclitaxel and one to topotecan after failure to respond to preceding chemotherapy. Median survival and time to progression was 22.24 and 10.61 months, respectively. This study demonstrates that sequential chemotherapy with just two initial courses of carboplatin is a reasonable way to introduce new agents into first-line therapy for poor prognostic ovarian cancer patients.

PURPOSE:To determine an accurate definition for progression of ovarian cancer in patients with a persistently elevated serum CA-125.PATIENTS AND METHODS:A retrospective analysis was performed on 300 patients with epithelial ovarian carcinoma with at least one measurement of CA-125. The date of progression according to clinical or radiologic criteria was ascertained in the 88 patients with persistently elevated CA-125 levels (> 23 U/mL). This was compared with the date of progression according to CA-125, defined as the date on which the CA-125 level first increased to >or= twice its nadir level, confirmed by a second sample also >or= twice the nadir.RESULTS:Eighty of the 88 patients had evidence of progression by both standard and CA-125 criteria, giving a sensitivity of 94%. In six of these patients, no sample was taken to confirm CA-125 doubling. In 13 patients, CA-125 doubling occurred after the date of clinical progression. Only one patient had a false-positive prediction of progression according to CA-125; the patient died as a result of a myocardial infarct before evidence of clinical progression.CONCLUSION:In patients whose CA-125 level decreases to normal after chemotherapy, a doubling from the upper limit of normal has been shown to predict progression. In those with persistently elevated levels, doubling of CA-125 from its nadir level has now been shown to accurately define progression. If confirmed, these CA-125 criteria should be used as additional end points in clinical trials.

PURPOSE:To determine whether precise definitions of response based on serial CA-125 levels can predict the activity of drugs in phase II trials for ovarian cancer as accurately as standard criteria.PATIENTS AND METHODS:Fourteen different drugs for relapsed ovarian cancer were analyzed in 25 treatment groups in 19 clinical trials. Response rates were estimated in 1,457 assessable patients according to standard criteria and in 1,092 assessable patients according to CA-125. For each drug trial, the observed response rates acted as the input data for an evaluation of how the two criteria would perform in a hypothetical Gehan two-stage phase II trial, accepting a target drug efficacy rate of 20% and a rejection error of 5%.RESULTS:CA-125 and clinical response criteria were concordant in 20 of the 25 groups, with less than 5% chance of rejecting the drug in nine groups and greater than 5% in 11 groups. In four groups, the drug had less than 5% chance of being rejected by CA-125 but greater than 5% chance of being rejected by standard criteria. The difference in the classification of drugs by the standard and CA-125 response criteria was not statistically significant (P =.38, McNemar's test). CA-125 response rates were slightly higher than standard response rates by a factor of 1.11.CONCLUSION:Definitions based on a 50% or 75% decrease of CA-125 levels accurately predicted which drugs in phase II trials for relapsed ovarian cancer were active and justified further investigation.

Despite the availability of serial data on CA 125 in ovarian cancer, the problem of interpreting a change over time is still unsolved. Changes in marker concentrations are due not only to patients improving or deteriorating but also to analytical imprecision and normal intra-individual biological variation. The aim of this study was to assess the analytical imprecision (CV(A)) and the intra- and inter-individual biological variation (CV(I) and CV(G), respectively) of CA 125 in a group of 26 patients with clinically stable ovarian cancer. Furthermore, the critical difference for a change between two consecutive CA 125 concentrations calculated as square root(2) x Z x (CV(A)2 + CV(I)2)(1/2) (Z =1.65 for unidirectional and 1.96 for bidirectional changes, p < or = 0.05) and the index of individuality calculated as ((CV(A)2+CV(I)2)/CV(G)2)(1/2) were estimated. After the exclusion of outliers, CV(A) and the average CV(I) and CV(G) were 12.1%, 24.0%, and 43.1%, respectively. The index of individuality was 0.62 and the critical difference calculated for unidirectional changes was 62.6%. CV(A) and CV(I) contribute considerably to the variation in serial results and should, therefore, be included in the criteria for serum tumor marker assessment during monitoring of patients with ovarian cancer. The cut-off value of CA 125 is of minor value in detecting unusual results for an individual subject, when previous measurements from an individual are available. These measurements should be preferred as reference for interpretation of new results.

BACKGROUND:Response rates to chemotherapy in relapsed, platinum resistant epithelial ovarian cancer remain poor. We have explored the effectiveness of weekly cisplatin combined with prolonged oral etoposide in this patient group.PATIENTS AND METHODS:Forty-two women with relapsed, advanced ovarian cancer were treated with cisplatin 60 mg/m2 on days 1, 8, 15, 29, 36 and 43 and oral etoposide 50 mg given from day 1-14 and day 29-43. In those who were responding and tolerating treatment (n = 13) oral etoposide 50 mg was continued for two further cycles (days 1-21 repeated every 28 days). The interval since last platinum containing chemotherapy was > 6 months in 28 patients and < 6 months in 16 patients.RESULTS:Thirty-six patients were evaluable for response according to CA 125 criteria giving an overall response rate of 44%. The response rate in evaluable patients declined with increasing numbers of previous treatments: 57% with one prior treatment, 42% with two, 40% with three or more. The response rate in patients who had received platinum chemotherapy within six months prior to treatment was 46%. The only significant non-haematological toxicity was nausea and vomiting in 4 patients who experienced greater than grade 2 toxicity. The number of patients experiencing haematological toxicity more than grade 2 was as follows: haemoglobin 3, white blood count 12, platelets 6. Sixteen patients had dose delays and two had dose reductions.CONCLUSION:We conclude that this short but intensive regimen provides worthwhile response rates, even in those patients who would ordinarily be considered refractory to platinum, and has an acceptable toxicity profile.

We retrospectively investigated the outcome of ovarian cancer in women aged less than 40 years treated in three randomised phase III studies of platinum-based chemotherapy. 624 patients had invasive epithelial ovarian cancer. A Cox proportional hazard model was used to study prognostic variables. 29 women (5%) were under 40 years of age. Stage, histological grade and amount of residual disease were significantly worse in women aged ≥40 years. Median follow-up was 66.7 months. At 5 years 65% of women below 40 years of age were alive compared with 20% of older women (95% confidence interval (CI) of the difference 27.1–63.0). The progression-free interval was 59% versus 16% (95% CI 24.3–60.8). No patient under 40 years of age relapsed after 18 months. Age ≥40 years was a poor prognostic variable, particularly for serous tumours, the commonest subtype in younger women (hazard ratio (HR): 3.33). Other prognostic factors were Eastern Cooperative Oncology Group (ECOG) performance status (HR: 1.25), presence of residual disease (HR: 1.43), histological grade (HR: 1.36) and International Federation of Gynaecology and Obstetrics (FIGO) stage (HR: 1.47). These results suggest that there are biological differences in the behaviour of serous carcinoma of the ovary in women of reproductive age compared with older women.

PURPOSE:To assess CA-125 as a measure of response in patients treated with paclitaxel.PATIENTS AND METHODS:One hundred forty-four patients treated with paclitaxel derived from four different trials and 625 patients treated with platinum from two trials were analyzed using precisely defined 50% and 75% reductions in CA-125. The standard and CA-125 response rates to paclitaxel and platinum were compared. In addition, we analyzed individual patient groups in which there was a difference in response according to the two response criteria.RESULTS:Patients with stable disease as determined by standard criteria who were treated with platinum and responded according to CA-125 criteria have an improved median progression-free survival compared with patients with stable disease who did not respond according to CA-125 criteria (10.6 v 4.8 months; P<.001). Standard and CA-125 response rates for patients treated with platinum (58.93% v 61.31%, respectively) and paclitaxel (30.65% v 31.67%, respectively) were very similar, as were rates of false-positive prediction of response by CA-125 (platinum 2.2% and paclitaxel 2.9%). Responders to paclitaxel had a significantly improved progression-free survival compared with non-responders by both standard criteria (median progression-free survival, 6.8 v 2.5 months; P<.001) and CA-125 criteria (median progression-free survival, 6.8 v 3.4 months; P<.001).CONCLUSION:Forassessing activity of therapy for ovarian cancer, these data show that precise 50% or 75% CA-125 response criteria are as sensitive as standard response criteria. We propose that they may be used as a measure of response in lieu of or in addition to standard response criteria in clinical trials involving epithelial ovarian cancer. Sensitivity is maintained whether patients are treated with platinum or paclitaxel.

BACKGROUND:There is a need to delineate how the best available serum marker for ovarian cancer, CA 125 should be used in monitoring the treatment of individual patients, and in clinical trials.PATIENTS AND METHODS:The situations where measurement of CA 125 could alter the management of individual patients during therapy and at relapse were critically analysed. Precise definitions for response according to CA 125 were first developed and tested on 989 patients receiving first line therapy and have since been tested on over 2000 patients in phase 2 trials. Precise definitions for progression were developed from 71 patients during and on 255 patients both during and after initial chemotherapy.RESULTS:In individual patient management, progression on initial chemotherapy can be demonstrated by serial rise in CA 125 suggesting the need for a change in therapy. A confirmed doubling of CA 125 from the upper limit of normal during follow up accurately predicts relapse. This endpoint needs prospective testing for use in clinical trials. Precise definitions of response based on a 50% or 75% fall in CA 125 levels accurately predicts whether a cytotoxic agent is active against ovarian cancer.CONCLUSION:Precise definitions of response based on CA 125 should be used in phase 2 trials of new cytotoxic drugs.

PURPOSE A phase II study was performed of oral altretamine in 71 patients with ovarian carcinoma who entered clinical complete remission with CA125 levels less than 35 U/mL after initial or second-line chemotherapy, and relapsed more than 6 months later. Response was compared between standard and CA125-based criteria. PATIENTS AND METHODS Altretamine 260 mg/m2 was given in divided doses daily for 14 days per month. Response was evaluated according to European Organization for Research and Treatment of Cancer (EORTC) criteria in 45 of 66 eligible patients. Response was assessed according to precise CA125 criteria in 51 patients based on either a confirmed > or = 50% or > or = 75% decrease in CA125 levels. RESULTS A combination of domperidone, dexamethasona, and chlorpromazine at night controlled toxicity in most patients, which was mainly nausea (National Cancer Institute criteria grade 2 or 3 in 27), vomiting (grade 2 or 3 in 19, grade 4 in one), and tiredness (grade 2 or 3 in 15). Responses (complete plus partial) were seen in 18 (40%; 95% confidence interval [CI], 25.4% to 54.6%) of those evaluated according to EORTC criteria and in 20 (39%; 95% CI, 25.5% to 52.9%) of those evaluated according to CA125 level. The overall response rate was 26 of 57 (45.6%) and was related to treatment-free interval: 6 to 12 months, 35%; 12 to 24 months, 52%; and greater than 24 months, 67%. The medium duration of response was 8 months. CONCLUSION Oral altretamine is a useful agent in patients-who relapse after previously responsive ovarian cancer. Response evaluation by a strict CA125 definition gave a similar estimate of the efficacy of altretamine as EORTC criteria.

BACKGROUND:There is no clear data on the optimum number of courses of platinum chemotherapy required for treating advanced ovarian cancer. A prospective randomised trial was performed to compare five with eight courses of either carboplatin 400 mg/m2 or cisplatin 75 mg/m2 given four-weekly to patients with stage IC-IV ovarian cancer. PATIENTS AND METHODS:225 patients were entered into the study and 233 were eligible for randomisation: 118 to five courses and 115 to eight courses. In each randomisation, half the patients received carboplatin and half received cisplatin. RESULTS:The mean number of courses received on the five arm was 4.74 and on the eight arm, 6.42, an increase of 35%. The median survival for all patients was 24 months with the median survival for the 156 patients with stage 3 disease being 21 months. No difference was detected in survival (P = 0.53) or time to progression from initial surgery (P = 0.29) between the two arms of the trial. False-negative calculations based on a multivariate analysis show that the trial currently has 95% power for excluding a difference of 10% in favour of the eight course arm at three years. CONCLUSION:There is insufficient evidence at the present time to justify more than five courses of first-line single agent platinum chemotherapy in the management of advanced ovarian cancer.

Background: Many studies have shown that CA 125 levels frequently rise prior to clinical evidence of progression of ovarian cancer. For clinical trials an accepted definition of progression according to CA 125 is required. We therefore determined what change in CA 125 level was the most accurate predictor of relapse in patients on follow up after therapy for ovarian cancer.Patients and methods: Serial CA 125 levels were studied from 255 patients entering the North Thames Ovary Trial of 5 versus 8 courses of chemotherapy. An initial analysis was made 2 months after closure of the trial, a more detailed analysis was made after 81 confirmed relapses among evaluable patients and a final analysis was made one year later with longer follow-up.Results: On the basis of the results from the interim analyses and the cut-off level of 22-35 U/ml used by different laboratories, 30 U/ml was chosen as the upper limit of normal. In the final analysis a doubling of CA 125 from the upper limit of normal was defined as progression. Using this method sensitivity was 85.9%, specificity 91.3%, positive predictive value 94.8%, and negative predictive value was 77.8%. Insisting on a confirmatory elevated CA 125 level reduced the false positive rate to <2% with a sensitivity of 83.9%. The median lead-time prior to clinical progression was 63 days.Conclusion: A confirmed rise of serum CA 125 level to more than twice the upper limit of normal during follow up after first line chemotherapy accurately predicts tumour relapse.

PURPOSE:To produce definitions based on serial CA 125 levels to measure response of ovarian carcinoma in patients receiving first-line chemotherapy.PATIENTS AND METHODS:Definitions were derived from analysis of 277 patients in North Thames Ovary Trial 3. Patient data were then incorporated into a computer program and tested against 254 patients in North Thames Ovary Trial 4 and 458 patients in Gynecologic Oncology Group (GOG) protocol 97. For optimum detection of response, three response definitions have been combined into a computer program. The precise definitions use mathematic logic and take account of factors such as intervening samples. Response to a specific treatment has occurred if after two samples there has been a 50% decrease, confirmed by a fourth sample (50% response), or a serial decrease over three samples of greater than 75% (75% response). The final sample has to be at least 28 days after the previous sample.RESULTS:Six hundred twenty of 989 patients were considered assessable for response according to CA 125 level. Only two patients (0.3%) had a CA 125 response at the time of clinical progression. The CA 125 response rate was 62% and 54% in the North Thames trials. In the GOG trial, it was 66% in all 317 patients assessable for CA 125 and 67% in 221 patients whose CA 125 level was not measurable according to GOG criteria, compared with a GOG-defined response rate of 62%. The sensitivity for detecting GOG-defined response was at least 68%.CONCLUSION:Definitions based on a 50% or 75% decrease of CA 125 levels have been shown reliably to define partial response of ovarian cancer in patients receiving first-line chemotherapy. These definitions should be used in addition to or instead of standard response criteria.

Twenty-two asymptomatic women with rising CA 125 levels after chemotherapy for ovarian cancer were entered into a trial of isotretinoin combined with calcitriol. Tumours were evaluated according to precise criteria based on serial CA 125 levels and by comparing regression slopes of CA 125 before and during therapy. There was no evidence based on CA 125 of any responses or significant change in tumour growth rate.

PURPOSE:To determine in a randomized trial of advanced ovarian carcinoma whether consolidation therapy with whole-abdominal radiotherapy (RT) after chemotherapy improves survival and disease-free survival compared with the continued chemotherapy.PATIENTS AND METHODS:Two hundred fifty-four patients with advanced epithelial ovarian cancer (stages IIB to IV) were entered onto a study of five monthly courses of 400 mg/m2 of carboplatin. One hundred seventeen patients with residual disease of 2 cm or less at second-look laparotomy or laparoscopy were then randomized to receive consolidation therapy, either five further courses of carboplatin at the same dosage or whole-abdominal RT (24 Gy). There was no control arm.RESULTS:Chemotherapy was well tolerated and was usually administered on an outpatient basis. Myelosuppression that was sufficient to delay chemotherapy occurred in only 3% of 1,418 courses analyzed. The main toxicity of carboplatin was nausea and vomiting, but this was easier to control than that with cisplatin. Although RT was well tolerated in the majority of the 58 patients, one patient who had been found to have multiple adhesions at second-look surgery developed fecal fistulae post-RT that resulted in the patient's death from peritonitis. Median survival for the whole group from date of surgery was 25 months. No statistical difference was found in either survival or disease-free survival between those patients who received consolidation chemotherapy and those who were treated with abdominal RT. Prognostic factors used to assess survival were stage, histology, amount of residual disease after primary surgery, and presence of tumor at second-look surgery.CONCLUSION:There seems to be no significant advantage for consolidation whole-abdominal RT compared with the continuation of the same chemotherapy in the management of advanced epithelial carcinoma of the ovary, even when no macroscopic residual disease is apparent at second-look surgery.

The prognostic value of serum CA-125 measurements was assessed in 54 patients with advanced ovarian adenocarcinoma. All patients received a minimum of two courses of carboplatin as part of the North Thames Cooperative Group trial. With a minimum follow-up of 6 months, 37 patients (69%) have clinical evidence of progressive disease and 28 have died. The absolute prechemotherapy level of CA-125 was of no value in predicting which patients would develop progressive disease. However, the change in CA-125 levels from before chemotherapy to 1 month later, after one course of carboplatin, could be used to divide patients into different prognostic groups. The best discrimination was found by dividing the patients into those who showed a greater than sevenfold decrease in CA-125 levels and those who showed a smaller change. Eight of 14 (58%) patients with a greater than sevenfold decrease in CA-125 levels remain disease-free compared with three of 36 (9%) patients with a lesser fall (P = .0005). The change in CA-125 levels during the first month of chemotherapy may indicate which patients should be offered alternative or symptomatic therapy and which should continue with the currently available toxic chemotherapy.