Prognostic Importance of Preoperative CA-125 in International Federation of Gynecology and Obstetrics Stage I Epithelial Ovarian Cancer: An
http://www.100md.com
《临床肿瘤学》
the Queensland Centre for Gynaecological Cancer, Medical School University of Queensland, Herston
Royal Brisbane Hospital, Brisbane
Department of Gynaecological Oncology, Royal Hospital for Women, Melbourne
Western Australian Gynaecological Cancer Service, King Edward Memorial Hospital, Subiaco
Gynaecological Cancer Centre, Royal Hospital for Women, Randwick
Hunter Centre for Gynaecological Cancer, John Hunter Hospital, New Lambton
Department of Gynaecological Oncology, Royal Adelaide Hospital, Adelaide
Department of Gynaecological Oncology, King George V Hospital, Camperdown, Australia
ABSTRACT
PURPOSE: To evaluate the prognostic significance of preoperative CA-125 levels on overall survival of patients with International Federation of Gynecology and Obstetrics (FIGO) stage I epithelial ovarian cancer (EOC).
PATIENTS AND METHODS: Data from 518 patients with FIGO stage I EOC treated in seven gynecologic oncology centers throughout Australia between 1990 and 2002 were analyzed. Patients with borderline tumors and nonepithelial ovarian carcinomas were excluded, as were women in whom CA-125 had not been determined preoperatively. Preoperative CA-125 levels were studied in surgically staged and incompletely staged patients and compared with prognostic factors, such as substage, grade, and histologic type. Multivariate Cox models were calculated.
RESULTS: CA-125 levels more than 30 U/mL were associated with higher grade, substage 1B and 1C, nonmucinous histologic type, and older age. In univariate analysis, higher histologic grade, the absence of surgical staging, and preoperative CA-125 levels more than 30 U/mL were associated with impaired survival. Multivariate analysis identified histologic grade, preoperative CA-125, and surgical staging as independent predictors for survival. In the subgroup of completely surgically staged patients, the 5-year overall survival rate was 82% (95% CI, 76% to 88%) for patients with CA-125 levels more than 30 U/mL and 95% (95% CI, 90% to 99%) for patients with CA-125 levels of 30 U/mL or less (P = .028). In the group of incompletely staged patients, the 5-year survival rates were similar for patients with elevated and normal serum CA-125 levels.
CONCLUSION: Complete surgical staging, histologic grade, and preoperative serum CA-125 levels are independent prognostic factors and should be included in the decision making for chemotherapy.
INTRODUCTION
Although the prognosis for early-stage epithelial ovarian cancer (EOC) is good, only 15% to 25% of all patients with EOC are diagnosed with disease confined to the ovaries.[1,2] The standard treatment for stage I EOC is total abdominal hysterectomy, bilateral salpingo-oophorectomy, and extended surgical staging, which includes pelvic and aortic lymph node dissection, omentectomy, peritoneal washings, and biopsies. Whether adjuvant chemotherapy is superior to a policy of no immediate treatment for patients with stage I EOC remains a matter of debate.[3] However, patients with stage IB or IC disease, histologic grade 2 or 3 tumors, or densely adherent tumors commonly receive chemotherapy postoperatively at present.[4,5]
CA-125 is an antigenic determinant of a high molecular weight glycoprotein recognized by the murine monoclonal antibody OC-125 as performed by a routine blood test.[6] It has an established role in monitoring treatment and detecting recurrence of ovarian cancer and has been advocated as a prognostic marker for advanced ovarian cancer.[7-9] Serum levels of CA-125 are believed to correlate with the volume of disease,[10] and elevated serum CA-125 levels can be found in 50% of patients with disease confined to the ovaries.[11] Only one study claims that CA-125 has prognostic significance for stage I ovarian cancer.[12] The aim of the present study was to evaluate the prognostic significance of preoperative CA-125 on the survival of patients with International Federation of Gynecology and Obstetrics (FIGO) stage I EOC.
PATIENTS AND METHODS
Patients
Between 1990 and 2001, 518 patients underwent primary surgery for stage I EOC at seven gynecologic cancer centers in Australia (Queensland Centre for Gynaecological Cancer, n = 202; Royal Hospital for Women, Melbourne, n = 90; The King Edward Memorial Hospital, Perth, n = 73; Royal Women Gynaecological Cancer Centre, Sydney, n = 55; Hunter Centre for Gynaecological Cancer, n = 47; Royal Prince Alfred Hospital, n = 29; and Royal Adelaide Hospital, n = 22). All data for this retrospective analysis were obtained from the respective cancer registry database, and additional information was retrieved from chart reviews. Patients had to have histologically proven, FIGO stage I EOC, and the preoperative serum CA-125 level had to be available. We excluded patients with nonepithelial tumors, patients with higher than stage I disease, and patients in whom CA-125 was not recorded. Histologic grading was performed as outlined by Day et al,[13] with highly differentiated tumors classified as grade 1, moderately differentiated tumors classified as grade 2, and poorly or undifferentiated tumors classified as grade 3. The histologic type was classified as defined by WHO.[14] Histologic slides were reviewed by specialized gynecologic pathologists. The CA-125 levels were measured approximately 1 week before the operation by a commercially available radioimmunoassay.[15] Institutional review board approval was obtained from the Queensland Centre for Gynaecological Cancer.
Treatment
Standard surgical treatment consisted of total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and pelvic and para-aortic lymphadenectomy. Peritoneal washings, diaphragmatic scrapings, and thorough exploration of the peritoneal cavity, including biopsies, were performed as part of the surgical staging process. Standard staging operation was performed in 346 patients; 172 patients did not receive full staging. Reasons for incomplete surgical staging included refusal of patients, medical and anesthetic risk as a result of medical comorbidities, and patient choice of chemotherapy over further surgery. Postoperative chemotherapy commenced approximately 10 days postoperatively and consisted of a combination of cisplatin 75 mg/m2 and cyclophosphamide 700 mg/m2 until 1998; thereafter, carboplatin (area under the curve = 5) as a single agent or in combination with paclitaxel 175 mg/m2 over a 3-hour infusion was administered. Patients were observed every 3 months for the first 2 years, then every 6 months for up to 5 years, and annually thereafter. Patients were observed based on clinical, radiologic, biochemical, and imaging techniques.
Statistical Analysis
The 2 trend test was used to examine patient characteristics for discrete categoric variables or factors. Overall survival was the end point for the study, with a statistical event defined as death from any cause. Survival time was calculated using date of first diagnosis and date last seen or date of death. Preoperative CA-125 serum levels greater than 30 U/mL were considered elevated, and values 30 U/mL were considered normal.[16,17] Actuarial survival plots and time to event curves were constructed using the Kaplan-Meier product-limit method, with the log-rank test applied to detect differences between groups. A generalized linear interactive modeling package (GLIM-4; Numerical Algorithms Group, Oxford, United Kingdom) was used to explore the data.
Univariate and multivariate Cox proportional hazards models were fitted to the data to determine the importance of recognized explanatory variables. Factors that were prognostically relevant in univariate analysis were included. These factors were serum CA-125 ( 30 v > 30 U/mL), full surgical staging as part of primary surgery (no v yes), and histologic grade (grade 1 v 2 v 3). The stepwise addition of other covariates, such as FIGO substage, age at diagnosis, postoperative chemotherapy, or histologic type, to this model did not result in a reduction of deviance and, consequently, did not contribute to a significant improvement to the model. Therefore, the best model fit isolated the main effects of grade, surgical staging, and CA-125. Considering that CA-125 was prognostic in surgically staged patients but not in incompletely staged patients, an interaction term between CA-125 and surgical staging was created in the multivariate model. Inclusion of the interaction term did not improve the final model (deviance test). Finally, a new variable was created with the four possible combinations of surgical staging and CA-125 (complete surgical staging and CA-125 30 U/mL; incomplete surgical staging and CA-125 30 U/mL; complete surgical staging and CA-125 > 30 U/mL; and incomplete surgical staging and CA-125 > 30 U/mL). A main effects model, including histologic grade (grade 1 v 2 and 1 v 3) and this new variable, was fitted. Only the combination of complete surgical staging, CA-125 30 U/mL, and histologic grade remained statistically significant. All P values are the result of two-sided tests, with P < .05 considered statistically significant.
RESULTS
Five hundred eighteen patients were entered onto this study. From those patients, 176 patients (33.9%) had preoperative CA-125 levels 30 U/mL, and 342 patients (66.1%) had CA-125 levels more than 30 U/mL. Patients with higher grade and higher substage were more likely to have elevated CA-125 levels, and patients with mucinous tumors were less likely to express CA-125. Patients with elevated CA-125 were more likely to receive chemotherapy. Patient characteristics are listed in [Table 1].
Within a median follow-up of 46.5 months, 87 patients died. On univariate analysis, higher histologic grade, the absence of extended surgical staging, and preoperative CA-125 level more than 30 U/mL were significantly associated with impaired survival, whereas age, substage, histologic type, and chemotherapy failed to be prognostically significant ([Table 2]).
Multivariate analysis, including the covariates of extended surgical staging, preoperative serum CA-125 more than 30 U/mL, and histologic grade, revealed prognostic information ([Table 3]). Interaction analysis between surgical staging and CA-125 level revealed that only the combination of complete surgical staging and CA-125 30 U/mL remained statistically significant, whereas all other combinations (incomplete surgical staging and CA-125 30 U/mL; complete surgical staging and CA-125 > 30 U/mL; and incomplete surgical staging and CA-125 > 30 U/mL) did not reach statistical significance.
In the subgroup of completely surgically staged patients, the 5-year overall survival rate was 82% (95% CI, 76% to 88%) for patients with CA-125 levels more than 30 U/mL and 95% (95% CI, 90% to 99%) for patients with CA-125 levels of 30 U/mL (P = .028; [Fig 1]). For completely staged patients, the hazard or relative risk of death was 2.4 (95% CI, 1.26 to 4.6) for patients with elevated serum CA-125 levels. In the group of incompletely staged patients, the 5-year survival rates were 78% (95% CI, 66% to 88%) and 70% (95% CI, 59% to 80%) for patients with elevated and normal serum CA-125 levels, respectively (P = .237; [Fig 2]). Of the 86 deaths among the 518 patients in this study, 21 (12%) and 65 (19%) were in the groups of patients with CA-125 levels and more than 30 U/mL, respectively.
DISCUSSION
Patients with surgical stage I EOC and a serum CA-125 level 30 U/mL have an excellent prognosis, and this prognostic effect is independent of other prognostic factors. Hence, preoperative serum CA-125 should be included as part of the decision making for postoperative chemotherapy in stage I EOC. These patients may even represent a subgroup of patients in whom further postoperative chemotherapy may not further improve overall survival.
Chemotherapy for surgical stage I ovarian cancer is still a controversial issue. Although chemotherapy is known to be effective in patients with advanced ovarian cancer, to date, no controlled study has been able to demonstrate a survival benefit of postoperative chemotherapy for patients with stage I EOC.[18,19] Most commonly, postoperative therapy is advised on an ad hoc basis. Traditionally, patients with high-grade tumors or bilateral ovarian cancers; patients with tumors densely adherent to surrounding pelvic structures; patients with positive peritoneal cytology, tumor rupture, or spillage of tumor contents; and patients with high-risk histologic features are adminsitered postoperative chemotherapy.[4,5] Patients with unilateral and well- or moderately differentiated tumors are often spared adjuvant chemotherapy in the absence of adverse factors.
To date, only one study demonstrated an association of CA-125 with prognosis in EOC.[12] By using an uncommonly used cutoff level for serum CA-125 of 65 U/mL, Nagele et al[12] found 44% of patients with stage I disease to be positive for CA-125. Patients with a preoperative serum CA-125 level of less than 65 U/mL demonstrated a survival probability of 91.2% at 5 years. In accordance, our report confirms a strong prognostic effect of preoperative CA-125 in stage I EOC. For surgically staged patients, the 5-year survival rate was 95% for patients with normal CA-125. In contrast, for patients with preoperative serum CA-125 levels greater than 30 U/mL, the overall survival rate at 5 years was 13% less, and the relative risk of death was 2.4 (95% CI, 1.26 to 4.6) compared with patients with preoperative CA-125 levels 30 U/mL. Regardless of other prognostic factors, the overall survival rate was more than 95% at 5 years if patients had undergone a full surgical staging, had a preoperative serum CA-125 level 30 U/mL, and presented with grade 1 or 2 tumors or stage IA disease, were more than 50 years of age, or had endometrioid or clear cell histologic type. Patients who had preoperative serum CA-125 levels 30 U/mL but who did not undergo a full surgical staging procedure and patients who presented with CA-125 more than 30 U/mL had a significantly worse survival probability.
The results presented here are derived from an Australian, multi-institutional effort to collect data on FIGO stage I EOC. The usual limitations of a retrospective study also apply to this report. Fifty-seven percent of patients received chemotherapy postoperatively, and patients with preoperative serum CA-125 levels more than 30 U/mL were more likely to be at risk for recurrence and, hence, were more likely to be selected to receive chemotherapy postoperatively.
Up to 30% of patients with disease apparently confined to the ovaries will have occult metastases with poor prognosis.[2] In our series, 33% of patients did not receive an extended surgical staging, and CA-125 was prognostically irrelevant in those patients. Considering the fact that recent US patterns of care studies revealed that 39.3% of patients with apparent stage I ovarian cancer were also denied an extended surgical staging at initial surgery, this information is important and relevant.[20,21] The prognostic information of the preoperative serum CA-125 level is confined to surgical stage I EOC. Patients who received extended surgical staging and in whom FIGO stage I EOC was confirmed had an excellent prognosis if their serum CA-125 level was 30 U/mL or less preoperatively.
Although this study is retrospective and was not designed to evaluate the effectiveness of chemotherapy in patients with stage I EOC, the data support the hypothesis that the prognosis of patients with preoperative serum CA-125 30 U/mL is extremely good if they receive a full surgical staging. Therefore, the preoperative serum CA-125 value should be included into the decision making for postoperative chemotherapy. The overall survival probability of these patients might already be so good that it can hardly be improved by further postoperative measures.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Pettersson F, Coppleson M, Creasman W, et al: Annual Report on the Results of Treatment in Gynecologic Cancer, Vol. 20. Stockholm, Sweden, International Federation of Gynaecology and Obstetrics, 1988, pp 110-151
Young RC, Decker DG, Wharton JT, et al: Staging laparatomy in early ovarian cancer. JAMA 250:3072-3076, 1983
Young RC, Pecorelli S: Management of early ovarian cancer. Semin Oncol 25:335-339, 1998
Schilder RJ, Young RC: Management of early stage ovarian cancer. Hematol Oncol Clin North Am 6:867-877, 1992
Young RC, Walton LA, Ellenberg SS, et al: Adjuvant therapy in stage I epithelial Ovarian cancer. N Engl J Med 322:1021-1027, 1990
Bast RC, Klug T, John E, et al: A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 309:883-887, 1983
Rustin GJ, Nelstrop AE, Mclean P, et al: Defining response of ovarian carcinoma to initial chemotherapy according serum CA 125. J Clin Oncol 14:1545-1551, 1996
Rustin GJ, Nelstrop AE, Tuxen MK, et al: Defining progression of ovarian carcinoma during follow-up according to CA 125: A North Thames Ovary Group study. Ann Oncol 7:361-364, 1996
Ind Tej, Iles R, Shephard JH, et al: Serum concentration of cancer antigen 125, placental alkaline phosphatase, cancer associated serum antigen and free beta human chorionic gonadotrophin as prognostic markers for epithelial ovarian cancer. Br J Obstet Gynaecol 104:1024-1029, 1997
Makar A, Kristensen GB, Kaern J, et al: Prognostic value of pre- and postoperative serum CA 125 levels in ovarian cancer: New aspects and multivariate analysis. Obstet Gynecol 79:1002-1010, 1992
Jacobs I, Bast RC Jr: The CA 125 tumour-associated antigen: A review of the literature. Hum Reprod 4:1-12, 1989
Nagele F, Petru E, Medl M, et al: Preoperative CA 125: An independent prognostic factor in patients with stage I epithelial ovarian cancer. Obstet Gynecol 86:259-264, 1995
Day TG, Gallager HS, Rutledge FN: Epithelial carcinoma of the ovary: The prognostic importance of histologic grade. Natl Cancer Inst Monogr 42:15-21, 1975
Serov SF, Skully RE, Sorbin LH: Histological typing of ovarian tumours, in International Histological Classification of Tumours. Geneva, Switzerland, WHO, 1973, p 17
Einhorn N, Sjovall K, Knapp RC, et al: Prospective evaluation of serum CA 125 levels for early detection of ovarian cancer. Obstet Gynecol 80:14-18, 1992
Jacobs IJ, Skates S, Davies AP, et al: Risk of diagnosis of ovarian cancer after raised serum CA 125 concentration: A prospective cohort study. BMJ 313:1355-1358, 1996
Jacobs I, Davies AP, Bridjes J, et al: Prevalence screening for ovarian cancer in postmenopausal women by CA 125 measurement and ultrasonography. BMJ 306:1030-1034, 1993
Trope C, Kaern J, Hogberg T, et al: Randomized study on adjuvant chemotherapy in stage I high-risk ovarian cancer with evaluation of DNA-ploidy as prognostic instrument. Ann Oncol 11:259-261, 2000
Finn CB, Luesley DM, Buxton EJ, et al: Is stage I epithelial ovarian cancer overtreated both surgically and systemically? Results of a five-year cancer registry review. Br J Obstet Gynaecol 99:54-58, 1992
Carney ME, Lancaster JM, Ford C, et al: A population-based study of patterns of care for ovarian cancer: Who is seen by a gynecologic oncologist and who is not? Gynecol Oncol 84:36-42, 2002
Harlan L, Clegg LX, Trimble EL: Trends in surgery and chemotherapy for women diagnosed with ovarian cancer in the United States. J Clin Oncol 21:3488-3494, 2003(Sellva Paramasivam, Lee T)
Royal Brisbane Hospital, Brisbane
Department of Gynaecological Oncology, Royal Hospital for Women, Melbourne
Western Australian Gynaecological Cancer Service, King Edward Memorial Hospital, Subiaco
Gynaecological Cancer Centre, Royal Hospital for Women, Randwick
Hunter Centre for Gynaecological Cancer, John Hunter Hospital, New Lambton
Department of Gynaecological Oncology, Royal Adelaide Hospital, Adelaide
Department of Gynaecological Oncology, King George V Hospital, Camperdown, Australia
ABSTRACT
PURPOSE: To evaluate the prognostic significance of preoperative CA-125 levels on overall survival of patients with International Federation of Gynecology and Obstetrics (FIGO) stage I epithelial ovarian cancer (EOC).
PATIENTS AND METHODS: Data from 518 patients with FIGO stage I EOC treated in seven gynecologic oncology centers throughout Australia between 1990 and 2002 were analyzed. Patients with borderline tumors and nonepithelial ovarian carcinomas were excluded, as were women in whom CA-125 had not been determined preoperatively. Preoperative CA-125 levels were studied in surgically staged and incompletely staged patients and compared with prognostic factors, such as substage, grade, and histologic type. Multivariate Cox models were calculated.
RESULTS: CA-125 levels more than 30 U/mL were associated with higher grade, substage 1B and 1C, nonmucinous histologic type, and older age. In univariate analysis, higher histologic grade, the absence of surgical staging, and preoperative CA-125 levels more than 30 U/mL were associated with impaired survival. Multivariate analysis identified histologic grade, preoperative CA-125, and surgical staging as independent predictors for survival. In the subgroup of completely surgically staged patients, the 5-year overall survival rate was 82% (95% CI, 76% to 88%) for patients with CA-125 levels more than 30 U/mL and 95% (95% CI, 90% to 99%) for patients with CA-125 levels of 30 U/mL or less (P = .028). In the group of incompletely staged patients, the 5-year survival rates were similar for patients with elevated and normal serum CA-125 levels.
CONCLUSION: Complete surgical staging, histologic grade, and preoperative serum CA-125 levels are independent prognostic factors and should be included in the decision making for chemotherapy.
INTRODUCTION
Although the prognosis for early-stage epithelial ovarian cancer (EOC) is good, only 15% to 25% of all patients with EOC are diagnosed with disease confined to the ovaries.[1,2] The standard treatment for stage I EOC is total abdominal hysterectomy, bilateral salpingo-oophorectomy, and extended surgical staging, which includes pelvic and aortic lymph node dissection, omentectomy, peritoneal washings, and biopsies. Whether adjuvant chemotherapy is superior to a policy of no immediate treatment for patients with stage I EOC remains a matter of debate.[3] However, patients with stage IB or IC disease, histologic grade 2 or 3 tumors, or densely adherent tumors commonly receive chemotherapy postoperatively at present.[4,5]
CA-125 is an antigenic determinant of a high molecular weight glycoprotein recognized by the murine monoclonal antibody OC-125 as performed by a routine blood test.[6] It has an established role in monitoring treatment and detecting recurrence of ovarian cancer and has been advocated as a prognostic marker for advanced ovarian cancer.[7-9] Serum levels of CA-125 are believed to correlate with the volume of disease,[10] and elevated serum CA-125 levels can be found in 50% of patients with disease confined to the ovaries.[11] Only one study claims that CA-125 has prognostic significance for stage I ovarian cancer.[12] The aim of the present study was to evaluate the prognostic significance of preoperative CA-125 on the survival of patients with International Federation of Gynecology and Obstetrics (FIGO) stage I EOC.
PATIENTS AND METHODS
Patients
Between 1990 and 2001, 518 patients underwent primary surgery for stage I EOC at seven gynecologic cancer centers in Australia (Queensland Centre for Gynaecological Cancer, n = 202; Royal Hospital for Women, Melbourne, n = 90; The King Edward Memorial Hospital, Perth, n = 73; Royal Women Gynaecological Cancer Centre, Sydney, n = 55; Hunter Centre for Gynaecological Cancer, n = 47; Royal Prince Alfred Hospital, n = 29; and Royal Adelaide Hospital, n = 22). All data for this retrospective analysis were obtained from the respective cancer registry database, and additional information was retrieved from chart reviews. Patients had to have histologically proven, FIGO stage I EOC, and the preoperative serum CA-125 level had to be available. We excluded patients with nonepithelial tumors, patients with higher than stage I disease, and patients in whom CA-125 was not recorded. Histologic grading was performed as outlined by Day et al,[13] with highly differentiated tumors classified as grade 1, moderately differentiated tumors classified as grade 2, and poorly or undifferentiated tumors classified as grade 3. The histologic type was classified as defined by WHO.[14] Histologic slides were reviewed by specialized gynecologic pathologists. The CA-125 levels were measured approximately 1 week before the operation by a commercially available radioimmunoassay.[15] Institutional review board approval was obtained from the Queensland Centre for Gynaecological Cancer.
Treatment
Standard surgical treatment consisted of total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and pelvic and para-aortic lymphadenectomy. Peritoneal washings, diaphragmatic scrapings, and thorough exploration of the peritoneal cavity, including biopsies, were performed as part of the surgical staging process. Standard staging operation was performed in 346 patients; 172 patients did not receive full staging. Reasons for incomplete surgical staging included refusal of patients, medical and anesthetic risk as a result of medical comorbidities, and patient choice of chemotherapy over further surgery. Postoperative chemotherapy commenced approximately 10 days postoperatively and consisted of a combination of cisplatin 75 mg/m2 and cyclophosphamide 700 mg/m2 until 1998; thereafter, carboplatin (area under the curve = 5) as a single agent or in combination with paclitaxel 175 mg/m2 over a 3-hour infusion was administered. Patients were observed every 3 months for the first 2 years, then every 6 months for up to 5 years, and annually thereafter. Patients were observed based on clinical, radiologic, biochemical, and imaging techniques.
Statistical Analysis
The 2 trend test was used to examine patient characteristics for discrete categoric variables or factors. Overall survival was the end point for the study, with a statistical event defined as death from any cause. Survival time was calculated using date of first diagnosis and date last seen or date of death. Preoperative CA-125 serum levels greater than 30 U/mL were considered elevated, and values 30 U/mL were considered normal.[16,17] Actuarial survival plots and time to event curves were constructed using the Kaplan-Meier product-limit method, with the log-rank test applied to detect differences between groups. A generalized linear interactive modeling package (GLIM-4; Numerical Algorithms Group, Oxford, United Kingdom) was used to explore the data.
Univariate and multivariate Cox proportional hazards models were fitted to the data to determine the importance of recognized explanatory variables. Factors that were prognostically relevant in univariate analysis were included. These factors were serum CA-125 ( 30 v > 30 U/mL), full surgical staging as part of primary surgery (no v yes), and histologic grade (grade 1 v 2 v 3). The stepwise addition of other covariates, such as FIGO substage, age at diagnosis, postoperative chemotherapy, or histologic type, to this model did not result in a reduction of deviance and, consequently, did not contribute to a significant improvement to the model. Therefore, the best model fit isolated the main effects of grade, surgical staging, and CA-125. Considering that CA-125 was prognostic in surgically staged patients but not in incompletely staged patients, an interaction term between CA-125 and surgical staging was created in the multivariate model. Inclusion of the interaction term did not improve the final model (deviance test). Finally, a new variable was created with the four possible combinations of surgical staging and CA-125 (complete surgical staging and CA-125 30 U/mL; incomplete surgical staging and CA-125 30 U/mL; complete surgical staging and CA-125 > 30 U/mL; and incomplete surgical staging and CA-125 > 30 U/mL). A main effects model, including histologic grade (grade 1 v 2 and 1 v 3) and this new variable, was fitted. Only the combination of complete surgical staging, CA-125 30 U/mL, and histologic grade remained statistically significant. All P values are the result of two-sided tests, with P < .05 considered statistically significant.
RESULTS
Five hundred eighteen patients were entered onto this study. From those patients, 176 patients (33.9%) had preoperative CA-125 levels 30 U/mL, and 342 patients (66.1%) had CA-125 levels more than 30 U/mL. Patients with higher grade and higher substage were more likely to have elevated CA-125 levels, and patients with mucinous tumors were less likely to express CA-125. Patients with elevated CA-125 were more likely to receive chemotherapy. Patient characteristics are listed in [Table 1].
Within a median follow-up of 46.5 months, 87 patients died. On univariate analysis, higher histologic grade, the absence of extended surgical staging, and preoperative CA-125 level more than 30 U/mL were significantly associated with impaired survival, whereas age, substage, histologic type, and chemotherapy failed to be prognostically significant ([Table 2]).
Multivariate analysis, including the covariates of extended surgical staging, preoperative serum CA-125 more than 30 U/mL, and histologic grade, revealed prognostic information ([Table 3]). Interaction analysis between surgical staging and CA-125 level revealed that only the combination of complete surgical staging and CA-125 30 U/mL remained statistically significant, whereas all other combinations (incomplete surgical staging and CA-125 30 U/mL; complete surgical staging and CA-125 > 30 U/mL; and incomplete surgical staging and CA-125 > 30 U/mL) did not reach statistical significance.
In the subgroup of completely surgically staged patients, the 5-year overall survival rate was 82% (95% CI, 76% to 88%) for patients with CA-125 levels more than 30 U/mL and 95% (95% CI, 90% to 99%) for patients with CA-125 levels of 30 U/mL (P = .028; [Fig 1]). For completely staged patients, the hazard or relative risk of death was 2.4 (95% CI, 1.26 to 4.6) for patients with elevated serum CA-125 levels. In the group of incompletely staged patients, the 5-year survival rates were 78% (95% CI, 66% to 88%) and 70% (95% CI, 59% to 80%) for patients with elevated and normal serum CA-125 levels, respectively (P = .237; [Fig 2]). Of the 86 deaths among the 518 patients in this study, 21 (12%) and 65 (19%) were in the groups of patients with CA-125 levels and more than 30 U/mL, respectively.
DISCUSSION
Patients with surgical stage I EOC and a serum CA-125 level 30 U/mL have an excellent prognosis, and this prognostic effect is independent of other prognostic factors. Hence, preoperative serum CA-125 should be included as part of the decision making for postoperative chemotherapy in stage I EOC. These patients may even represent a subgroup of patients in whom further postoperative chemotherapy may not further improve overall survival.
Chemotherapy for surgical stage I ovarian cancer is still a controversial issue. Although chemotherapy is known to be effective in patients with advanced ovarian cancer, to date, no controlled study has been able to demonstrate a survival benefit of postoperative chemotherapy for patients with stage I EOC.[18,19] Most commonly, postoperative therapy is advised on an ad hoc basis. Traditionally, patients with high-grade tumors or bilateral ovarian cancers; patients with tumors densely adherent to surrounding pelvic structures; patients with positive peritoneal cytology, tumor rupture, or spillage of tumor contents; and patients with high-risk histologic features are adminsitered postoperative chemotherapy.[4,5] Patients with unilateral and well- or moderately differentiated tumors are often spared adjuvant chemotherapy in the absence of adverse factors.
To date, only one study demonstrated an association of CA-125 with prognosis in EOC.[12] By using an uncommonly used cutoff level for serum CA-125 of 65 U/mL, Nagele et al[12] found 44% of patients with stage I disease to be positive for CA-125. Patients with a preoperative serum CA-125 level of less than 65 U/mL demonstrated a survival probability of 91.2% at 5 years. In accordance, our report confirms a strong prognostic effect of preoperative CA-125 in stage I EOC. For surgically staged patients, the 5-year survival rate was 95% for patients with normal CA-125. In contrast, for patients with preoperative serum CA-125 levels greater than 30 U/mL, the overall survival rate at 5 years was 13% less, and the relative risk of death was 2.4 (95% CI, 1.26 to 4.6) compared with patients with preoperative CA-125 levels 30 U/mL. Regardless of other prognostic factors, the overall survival rate was more than 95% at 5 years if patients had undergone a full surgical staging, had a preoperative serum CA-125 level 30 U/mL, and presented with grade 1 or 2 tumors or stage IA disease, were more than 50 years of age, or had endometrioid or clear cell histologic type. Patients who had preoperative serum CA-125 levels 30 U/mL but who did not undergo a full surgical staging procedure and patients who presented with CA-125 more than 30 U/mL had a significantly worse survival probability.
The results presented here are derived from an Australian, multi-institutional effort to collect data on FIGO stage I EOC. The usual limitations of a retrospective study also apply to this report. Fifty-seven percent of patients received chemotherapy postoperatively, and patients with preoperative serum CA-125 levels more than 30 U/mL were more likely to be at risk for recurrence and, hence, were more likely to be selected to receive chemotherapy postoperatively.
Up to 30% of patients with disease apparently confined to the ovaries will have occult metastases with poor prognosis.[2] In our series, 33% of patients did not receive an extended surgical staging, and CA-125 was prognostically irrelevant in those patients. Considering the fact that recent US patterns of care studies revealed that 39.3% of patients with apparent stage I ovarian cancer were also denied an extended surgical staging at initial surgery, this information is important and relevant.[20,21] The prognostic information of the preoperative serum CA-125 level is confined to surgical stage I EOC. Patients who received extended surgical staging and in whom FIGO stage I EOC was confirmed had an excellent prognosis if their serum CA-125 level was 30 U/mL or less preoperatively.
Although this study is retrospective and was not designed to evaluate the effectiveness of chemotherapy in patients with stage I EOC, the data support the hypothesis that the prognosis of patients with preoperative serum CA-125 30 U/mL is extremely good if they receive a full surgical staging. Therefore, the preoperative serum CA-125 value should be included into the decision making for postoperative chemotherapy. The overall survival probability of these patients might already be so good that it can hardly be improved by further postoperative measures.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Pettersson F, Coppleson M, Creasman W, et al: Annual Report on the Results of Treatment in Gynecologic Cancer, Vol. 20. Stockholm, Sweden, International Federation of Gynaecology and Obstetrics, 1988, pp 110-151
Young RC, Decker DG, Wharton JT, et al: Staging laparatomy in early ovarian cancer. JAMA 250:3072-3076, 1983
Young RC, Pecorelli S: Management of early ovarian cancer. Semin Oncol 25:335-339, 1998
Schilder RJ, Young RC: Management of early stage ovarian cancer. Hematol Oncol Clin North Am 6:867-877, 1992
Young RC, Walton LA, Ellenberg SS, et al: Adjuvant therapy in stage I epithelial Ovarian cancer. N Engl J Med 322:1021-1027, 1990
Bast RC, Klug T, John E, et al: A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 309:883-887, 1983
Rustin GJ, Nelstrop AE, Mclean P, et al: Defining response of ovarian carcinoma to initial chemotherapy according serum CA 125. J Clin Oncol 14:1545-1551, 1996
Rustin GJ, Nelstrop AE, Tuxen MK, et al: Defining progression of ovarian carcinoma during follow-up according to CA 125: A North Thames Ovary Group study. Ann Oncol 7:361-364, 1996
Ind Tej, Iles R, Shephard JH, et al: Serum concentration of cancer antigen 125, placental alkaline phosphatase, cancer associated serum antigen and free beta human chorionic gonadotrophin as prognostic markers for epithelial ovarian cancer. Br J Obstet Gynaecol 104:1024-1029, 1997
Makar A, Kristensen GB, Kaern J, et al: Prognostic value of pre- and postoperative serum CA 125 levels in ovarian cancer: New aspects and multivariate analysis. Obstet Gynecol 79:1002-1010, 1992
Jacobs I, Bast RC Jr: The CA 125 tumour-associated antigen: A review of the literature. Hum Reprod 4:1-12, 1989
Nagele F, Petru E, Medl M, et al: Preoperative CA 125: An independent prognostic factor in patients with stage I epithelial ovarian cancer. Obstet Gynecol 86:259-264, 1995
Day TG, Gallager HS, Rutledge FN: Epithelial carcinoma of the ovary: The prognostic importance of histologic grade. Natl Cancer Inst Monogr 42:15-21, 1975
Serov SF, Skully RE, Sorbin LH: Histological typing of ovarian tumours, in International Histological Classification of Tumours. Geneva, Switzerland, WHO, 1973, p 17
Einhorn N, Sjovall K, Knapp RC, et al: Prospective evaluation of serum CA 125 levels for early detection of ovarian cancer. Obstet Gynecol 80:14-18, 1992
Jacobs IJ, Skates S, Davies AP, et al: Risk of diagnosis of ovarian cancer after raised serum CA 125 concentration: A prospective cohort study. BMJ 313:1355-1358, 1996
Jacobs I, Davies AP, Bridjes J, et al: Prevalence screening for ovarian cancer in postmenopausal women by CA 125 measurement and ultrasonography. BMJ 306:1030-1034, 1993
Trope C, Kaern J, Hogberg T, et al: Randomized study on adjuvant chemotherapy in stage I high-risk ovarian cancer with evaluation of DNA-ploidy as prognostic instrument. Ann Oncol 11:259-261, 2000
Finn CB, Luesley DM, Buxton EJ, et al: Is stage I epithelial ovarian cancer overtreated both surgically and systemically? Results of a five-year cancer registry review. Br J Obstet Gynaecol 99:54-58, 1992
Carney ME, Lancaster JM, Ford C, et al: A population-based study of patterns of care for ovarian cancer: Who is seen by a gynecologic oncologist and who is not? Gynecol Oncol 84:36-42, 2002
Harlan L, Clegg LX, Trimble EL: Trends in surgery and chemotherapy for women diagnosed with ovarian cancer in the United States. J Clin Oncol 21:3488-3494, 2003(Sellva Paramasivam, Lee T)