| HOME | HELP | CONTACT US | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ASCO 1999: Critical Commentaries |
Division of Hematology and Oncology, Massachusetts General Hospital, Boston, Massachusetts, USA
Correspondence: Michael V. Seiden, M.D., Ph.D., Division of Hematology and Oncology, Massachusetts General Hospital, Boston, Massachusetts 02114, USA. Telephone: 617-724-3123; Fax: 617-724-3166; e-mail: Seiden.Michael{at}MGH.Harvard.edu
Key Words. Gynecologic oncology • ASCO • Review • Ovarian cancer • Cervical cancer
Introduction
The gynecologic oncology presentations at the 1999 ASCO meeting in Atlanta continue to focus on defining the optimal treatment strategies for women with advanced gynecologic malignancies. While there were no dramatic shifts in treatment paradigms at ASCO, a large number of well-controlled randomized trials addressed key areas and defined standard of care. Indeed, 10 of the 43 clinical trials presented were phase III trials that randomized a total of 5,664 patients. These studies add incrementally to both our knowledge base and improved outcomes. The most important data were from trials of patients with ovarian cancer. Two randomized studies provided evidence that optimal first-line chemotherapy is paclitaxel 175 mg/m2 over 3 h and carboplatin AUC 4 - 7.5. An increasing number of studies examined the incorporation of relatively new agents into first-line therapy. No studies addressed the potential role of intraperitoneal treatment. Lastly, although the role of high-dose chemotherapy in ovarian cancer is uncertain, for the most active agents the dose response curves in ovarian cancer and breast cancer are similar. High-dose chemotherapy in breast cancer was reviewed in ASCO's plenary session. Four studies evaluating high-dose therapy were presented and in general the results were not encouraging. A single study, presented by Dr. Bezwoda of South Africa demonstrated a disease-free and overall survival advantage for up-front tandem transplant, which raises the issue of whether immediate high-dose chemotherapy may overcome drug resistance and avoid the induction of resistance by moderate doses of the same agents [1]. While no specific randomized data were presented for ovarian cancer, it is notable that Gynecologic Oncology Group (GOG) 164, a study that investigates high-dose chemotherapy as consolidation for first-line treatment for ovarian cancer, has been closed because of poor accrual. Further information on the role of high-dose therapy in this disease awaits the completion of randomized trials ongoing in Europe.
Standard First-Line Chemotherapy for Ovarian Cancer
Dr. Robert Ozols of the Fox Chase Cancer Center in Philadelphia presented a phase III study comparing cisplatin and paclitaxel versus carboplatin and paclitaxel in patients with optimally debulked stage III epithelial ovarian cancer [2]. It was designed as an equivalency study in previously untreated patients with no residual tumor greater than 1 cm. The patient's surgeon also selected whether the patient would be evaluated by second look operation at completion of six cycles of treatment. The schedule of paclitaxel was 175 mg/m2 over 3 h in the carboplatin and paclitaxel arm while the cisplatin/paclitaxel regimen gave paclitaxel as a 135 mg/m2 dose over 24 h. The study was designed to detect a 30% difference in outcome between the two arms and enrolled 840 patients. The combination of carboplatin and paclitaxel was associated with more thrombocytopenia and pain, but cisplatin and paclitaxel caused a higher incidence of neutropenia, fever, and gastrointestinal toxicity, as well as metabolic toxicities. Although neurotoxicity was similar in both arms of this study, the follow-up is short (19.1 months). To date there have been 419 recurrences. Although it was a little early for survival analysis, progression-free survival was 22 months in both arms. Indeed, the relative risk of failure on carboplatin and paclitaxel was 0.90 with 95% confidence intervals of 0.75 to 1.1, suggesting that it may even be slightly superior to cisplatin and paclitaxel. The relative risk of death was 0.85 for carboplatin and paclitaxel and, though the confidence intervals were very broad, a significant survival advantage may develop with time for carboplatin and paclitaxel, which is clearly a more convenient outpatient based regimen.
A point of controversy was the selection of the carboplatin dose of AUC = 7.5, a dose defined in a previous phase I study done at Fox Chase Cancer Center. In the earlier trial, this dose was associated with a 75% response rate. This dose was higher than the apparently adequate doses investigated in two prior randomized controlled studies [3, 4]. Specifically, two randomized trials from Europe demonstrating that carboplatin dosed at an AUC of 6 was equivalent to carboplatin at an AUC of 12 (U.K. study) [3] and a study suggested equivalency of carboplatin at AUCs of 4 and 8 (Danish study) [4]. These two studies suggest that the AUC dose of 7.5, in GOG 158, is excessive.
A secondary objective of the study was to define the potential benefit of a second look procedure. The assignment was not randomized, but made pretherapy by the surgeon, presumably after discussion with the patient, and hence is subject to bias. Second look operation (SLO) in patients with optimally debulked disease had no impact on progression-free survival. Three hundred ninety-nine patients did not have SLO, while 395 were selected for second look procedure. Of 395 patients who had SLO, 136 were positive and of this group, 123 went on to have treatment that was not specified by the trial. Treatments for the subgroup of women with positive second look procedures included continuation of carboplatin and paclitaxel, intraperitoneal treatment, high-dose chemotherapy with bone marrow transplantation, and whole abdominal radiotherapy at the discretion of the patient and physician. There was no evidence that the information gained by the second look operation impacted positively on survival although this was not a randomized comparison. Hence it is possible that a subset of patients might benefit or that alternative salvage therapies may be worthwhile in this patient population.
Dr. Dubois presented a confirmatory phase III trial that compared paclitaxel 185 mg/m2 over 3 h with either carboplatin AUC 6 or cisplatin 75 mg/m2 as first-line treatment of epithelial ovarian cancer [5]. Nausea and neuropathy were more common in the group treated with cisplatin—19% versus 7% (nausea) and 19% versus 8% (neuropathy), respectively. During follow-up, patients receiving carboplatin appeared to have at least a 20% lower instance of any grade neuropathy. The hazard ratio for progression-free survival was 0.95 favoring carboplatin and the hazard ratio for overall survival was 1.17 (confidence intervals 0.92 to 1.19). The study supported the routine use of carboplatin because of the significantly improved quality of life reported using the EORTC questionnaire (p = 0.0048).
Clearly the equal progression-free and overall survival reported in GOG 158 and the AOG study, and the better toxicity profile and improved quality of life for patients receiving carboplatin and paclitaxel establish this regimen as the standard of care for first-line chemotherapy. This conclusion was directly challenged by the Third International Collaborative Oncology Neoplasm Study (ICON 3), the largest randomized control study ever conducted in ovarian cancer. Two thousand seventy-four patients were randomized in what were essentially two trials [6]. Paclitaxel 175 mg/m2 over 3 h with carboplatin (AUC = 6) was compared with a control arm of either carboplatin as a single agent (AUC of 6 based on the calculated GFR [glomerular filtration rate] or 5 if estimated) or CAP chemotherapy (500 mg/m2 cyclophosphamide, 50 mg/m2 adriamycin, and 50 mg/m2 cisplatin). The inclusion of two different control regimens reflected controversy regarding the effectiveness of CAP versus carboplatin. Prior to enrolling patients, each one of 130 centers in eight nations selected one of the control arms; the intent was to include such a large number of patients that there would be sufficient power to address the questions posed in subgroup analysis. Early analysis suggests progression-free survival and overall survival are equivalent for paclitaxel and carboplatin as compared to the outcomes of the two control arms.
The study has a number of methodological weaknesses including limited follow-up (18 months) and a design with two different control arms. Nevertheless, this very large study is difficult to reconcile with GOG 111 [7], a study that found a significant survival advantage favoring the regimen of paclitaxel and cisplatin over cytoxan and cisplatin. Dr. William McGuire, in discussion of ICON 3, evaluated subsets within ICON 3 and demonstrated that the study did support the hypothesis that patients with suboptimally debulked tumor (residual disease >2 cm) may have benefited from paclitaxel-based therapy. In the suboptimally debulked group, a 44% mortality gave greater statistical power, and in this group, there appeared to be a survival advantage for carboplatin and paclitaxel when compared to single-agent carboplatin (p = 0.05) or the CAP combination (p = 0.03). The fact that GOG 132 (cisplatin versus cisplatin/paclitaxel versus paclitaxel [8]) and ICON 3 suggest no benefit from the addition of paclitaxel, compared with GOG 111 and OV 10 (cisplatin/cyclophosphamide versus cisplatin/paclitaxel [9]) may be explained by other sources of bias. The availability of paclitaxel for women who relapse after non-Taxol based primary therapy poses a problem in all trials except for GOG 111. Indeed, in GOG 132, 48% of patients were immediately treated with a second-line agent, half of whom received paclitaxel. Salvage therapy essentially converted this randomized study into an investigation of sequential treatment that showed no clear advantage for concurrent versus sequential therapy. Though the addition of paclitaxel may reduce the dose intensity of platinum, the most active agent in ovarian cancer, 9 of 10 randomized controlled trials suggest no advantage of platinum dose intensity at the investigated doses. Finally, there may be a genuine population difference between European patients with ovarian cancer and patients from the United States.
A number of studies investigated novel up-front doublets (Abstract 1378), triplets (Abstracts 1379, 1397, and 1418), or quadruplets (Abstracts 1391 and 1417) and novel schedules (Abstracts 1416 and 1419). Although weekly paclitaxel appears particularly active in recurrent disease, Rosenberg et al. from Finland reported a phase III study of weekly paclitaxel 67 mg/m2 over 1 h versus 200 mg/m2 (equivalent dose intensity) given as a 3-h infusion on a three-week schedule [10]. The time to progression with the three-week schedule was significantly longer; 8.3 months, compared with 5.5 months for the weekly schedule. However, the weekly schedule was associated with a better toxicity profile. It has also been investigated in first-line treatment. The regimen of paclitaxel, at a relatively high dose of 100 mg/m2 over 1 h and carboplatin weekly at an AUC of 2 over 24 h was evaluated by Bremer et al. in 26 patients [11]. This was associated with tolerable toxicity and complete remissions (CRs) in four of seven suboptimally debulked patients.
Hansen et al. presented data on 28 patients in a phase II study of the triplet combination of gemcitabine 800 mg/m2 on days 1 and 8, carboplatin AUC-5 on day 1, and paclitaxel 175 mg/m2 over 3 h on day 1 [12]. For a first-line combination this had a remarkable response rate of 100% with 60% CRs. Eight of 28 patients had relapsed at a median follow-up of 12 months. Hematological toxicity was considerable with platelet transfusions required in three patients, one patient being admitted twice for fever and neutropenia. Gemcitabine dosing was delayed in eight patients. Other triplets included the addition of epirubicin (Abstract 1418) and the inclusion of ifosfamide or epirubicin (Abstract 1396), both with significant hematological toxicity and high response rates. A four-drug regimen was presented by Tung et al. with the CTEC protocol that combined cyclophosphamide, paclitaxel, etoposide, and carboplatin [13]. This regimen was myelosuppressive even with the required use of neupogen. Ten percent of cycles were associated with fever and neutropenia and 23% had grade IV thrombocytopenia, highlighting the issue of combining myelosuppressive agents.
The use of agents sequentially versus concurrently is attracting increasing attention. The Norton-Simon hypothesis predicts that sequential treatment is more likely to eliminate drug-resistant clones. However, although this approach is supported both by mathematical models and clinical experiments, the impact of new agents has not yet translated into a significant improvement for patients. Dr. Hoskins reported the results of a phase II study in 44 patients using sequential doublets [14]. Cisplatin 50 mg/m2 on day 1 and topotecan 0.75 mg/m2 on days 1 through 5 for four cycles and then followed by four cycles of paclitaxel 135 mg/m2 over 24 h and cisplatin 75 mg/m2 (day 2) as first-line treatment for advanced ovarian cancer. Although the doses of cisplatin and topotecan appear low, the regimen was based on a phase I study that showed synergistic myelosuppression. Fever and neutropenia occurred in approximately 10% of patients and was equally distributed between the two doublets of the study. Significant activity was demonstrated in a group of women with stage IV or suboptimally debulked ovarian carcinoma. Seven percent of patients progressed on cisplatin and topotecan, and 5% progressed on cisplatin and paclitaxel. In reviewing these novel combinations, discussants warned that the nonsurgical assessment of response, small clinical trial size, and lack of follow-up on durability of response made it difficult and perhaps dangerous to become enthusiastic about the high response rates.
Dr. Young presented GOG 95, a randomized clinical trial of adjuvant treatment of women with early stage ovarian cancer (FIGO I-IIA) [15]. This study evaluated 205 patients with high-risk ovarian cancer, defined as early stage disease with stage IC, IIC or grade II-III stage IA and B patients. Earlier studies in this patient population have compared intraperitoneal 32P with melphalan, both of which were associated with an 80% progression-free survival. Radioactive phosphorus was preferred because of the leukemogenic toxicity of melphalan [16]. Subsequently, Bolis et al. have reported a randomized trial of 32P and cisplatin, which again had similar activity but cisplatin was associated with less toxicity [17]. An ongoing study, ICON I has randomized early stage patients to immediate platinum-based therapy or treatment at relapse and should be analyzed soon. Young et al. reported a relatively high failure rate and poor compliance for 32P. There was no statistically significant difference between cisplatin and cyclophosphamide and 32P for progression-free interval. However, a trend favored systemic therapy and, in light of logistical issues complicating 32P administration, Dr. Young recommended that platinum-based treatment should be the treatment of choice for patients with high-risk early stage ovarian cancer.
Ovarian Cancer: After First-Line Therapy?
A burgeoning number of new agents have activity in ovarian cancer, a chemotherapy-sensitive disease marked by repeated recurrence. Although there has been an improvement in five-year survival in the United States for all stages of the disease from 32% in 1960 to 46% in 1992, it has essentially become a chronic disease associated with very significant morbidity [18]. In an educational session, Dr. Thigpen outlined treatment options for second line therapy using A) the same drugs, carboplatin and paclitaxel; B) alternative cytotoxics; C) high-dose chemotherapy using peripheral blood stem cell supported high-dose systemic therapy or intraperitoneal therapy, and D) alternative modalities such as radiation, biologics, and gene therapy.
One of the difficulties in comparing the relative worth of different treatments has been quantifying the degree of persisting chemosensitivity. Dr. Markman's definition of chemoresistant disease as progression or relapse within six months of completing platinum-based treatment has been most widely used [19]. The likelihood of responding to second-line treatment depends upon A) prior response; B) number of prior lines of therapy, and C) duration of progression-free or disease-free survival. Four drugs, etoposide, topotecan, doxil, and gemcitabine, have efficacy in both platinum- and paclitaxel-resistant disease. Other active agents include tamoxifen, vinorelbine, docetaxel, oxaliplatin, irinotecan, capecitabine, and ifosfamide ( Fig. 1).
|
Dr. Markman reviewed the rationale for dose intensity in the second-line treatment for ovarian cancer. At least nine randomized controlled studies have now shown that doubling of dose intensity of platinum fails to impact disease-free or overall survival, and studies of paclitaxel suggest that there is a greater effect from schedule than dose [21]. Whereas high-dose systemic treatment may increase the dose by 1/2 to 1 log, intraperitoneal treatment may increase local concentrations of cytotoxics from 1 to 3 logs. However, there is limited systemic exposure, and treatment is associated with local toxicity, increased cost, and a detrimental effect on the quality of life.
The panel was asked about two interesting and topical issues. The treatment of an asymptomatic patient with rising CA125 is common practice but is not supported by data that show any more then a lead time effect. Few patients accept watchful waiting. Secondly, the role of herceptin in ovarian cancer was discussed. Over 500 patients have being screened for the GOG study of single-agent herceptin, and only 25 patients with HER-2/neu positive tumors have been treated, suggesting a lower incidence of HER-2/neu positivity than in breast cancer, perhaps between 5% and 10%. There is a considerable interest in the agent, and despite a disappointing response rate, the study continues to accrue.
Additional topics of clinical interest included Dr. Markman's review of carboplatin-associated hypersensitivity reactions [22]. These reactions appeared to be mediated through an antigen-specific IgE mechanism in which carboplatin may act as a hapten rather than producing the complement-mediated acute hypersensitivity reaction that is characteristic of paclitaxel. Although carboplatin reactions had similar clinical characteristics to the reactions to paclitaxel, they occurred later, typically after six to eight cycles. Although paclitaxel rechallenge is reasonable with steroids and antihistamines, in two of three patients who were rechallenged with carboplatin, there were severe reactions and, in one case, a fatal reaction upon rechallenge with cisplatin, suggesting that there is cross-reactivity between the analogs.
Cervical Cancer Screening
In an educational session, Drs. Parhan and Seshadri reviewed the difficult issue of screening for cervical cancer in the developing world. Seventy percent of patients who are found to have cervical cancer in India are diagnosed with advanced stage disease and only 10% of the patients are operable. This is a clearly important finding with approximately 500,000 cases of cervical cancer worldwide. In some sub-Saharan and African nations, cervical cancer is the leading cause of death in women aged 33 to 44. Treatment of early stage disease in the third world tends to be by immediate cryotherapy. The application of low technology screening strategies was discussed. These include unaided visual inspection, use of 3% to 5% acetic acid, which whitens areas of nuclear proliferation and density, and magnification using speculoscopy. Increased diagnostic sensitivity is typically traded for poorer specificity. There are very significant issues of limited access and poor acceptance for screening programs and cervical screening still has a low priority in community clinics because of the social stigma, the belief that cancer cannot be cured and confusion between Pap smear testing and sterilization. The availability of an effective vaccine, and public health education programs to reduce the incidence of human papillomavirus (HPV) infection, are likely to have a greater impact on early stage disease than will attempts to decrease the false positive rate of refined screening tests.
Dr. Benedetti-Panici presented a study of neoadjuvant chemotherapy and radical surgery versus radiotherapy, in squamous cell carcinoma of the cervix [23]. This phase III trial commenced in 1990 and randomized 441 patients to receive cisplatin 240 - 320 mg/m2 in six to eight weeks followed by radical hysterectomy and pelvic lymphadenectomy or external beam radiotherapy to a total dose of 70 Gy at point A. The median follow-up was 47 months. There was a predicted 15% overall survival benefit for neoadjuvant chemotherapy and surgery. Dr. Sardi also presented at the educational session and updated his data, which suggest an overall survival advantage for neoadjuvant cisplatin, vincristine, and bleomycin in stage IB2 patients with a p value of <0.01 in subset analysis. However, in reviewing Abstract 1377, Dr. Eiffel reminded the audience that the control arm in this neoadjuvant study involved a suboptimal radiotherapy dose without concurrent cisplatin. Indeed, the recent reporting of five randomized controlled trials ( Table 1) that all show a clear survival advantage for the combination of chemotherapy and radiotherapy over radiotherapy alone now redefines standard therapy for women with locally advanced or local regional cervical cancer.
|
Acknowledgments
We gratefully acknowledge the cooperation of Kirin Pharmaceuticals and Excerpta Medica, Japan.
References
6) platinum courses: experience of the Cleveland Clinic Cancer Center Gynecologic Oncology Program. Proc Am Soc Clin Oncol 1999;18:1425a.
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | CONTACT US | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| THE ONCOLOGIST | STEM CELLS | CME | ALPHAMED PRESS JOURNALS |
