- Email: [email protected]
Effects of Paclitaxel on CA-125 Serum Levels in Ovarian Cancer Patients
Gynecologic Oncology 76, 326 –330 (2000) doi:10.1006/gyno.1999.5699, available online at http://www.idealibrary.com on
Effects of Paclitaxel on CA-125 Serum Levels in Ovarian Cancer Patients Torbjørn Paulsen, M.D.,* ,1 Christian Marth, Ph.D.,† Janne Kærn, Ph.D.,* Kjell Nustad, Ph.D.,‡ Gunnar B. Kristensen, Ph.D.,* and Claes Trope´, Ph.D.* *Department of Gynecologic Oncology, The Norwegian Radiumhospital, Oslo, Norway; †Department of Obstetrics and Gynecology, Innsbruck University Hospital, Austria; and ‡Department of Laboratory Medicine, The Norwegian Radiumhospital, Oslo, Norway Received June 14, 1999
Objective. As in vitro activation of ovarian carcinoma cells in terms of CA-125 secretion by taxanes has been demonstrated, we were interested in whether taxanes also modulate CA-125 expression in vivo. Methods. Serum CA-125 was determined immediately before and 24 h after paclitaxel-containing chemotherapy in 53 ovarian carcinoma patients. To test the quality of the analysis methods and the biological variation of untreated patients, serum CA-125 levels of two control groups were analyzed. Results. Median CA-125 concentration was 107 kU/liter 24 h after chemotherapy treatment compared with 99 kU/liter the day before paclitaxel treatment. Changes in CA-125 serum levels observed immediately after paclitaxel treatment were not correlated to treatment response. However, overall change in CA-125 serum concentration was a good predictor of response to paclitaxel containing treatment. Patients achieving a complete or partial response had a significant reduction of median CA-125 levels, whereas tumor progression was associated with increased CA-125 levels. Only for the group of patients obtaining a complete response was a decrease in the median relative CA-125 value observed. Conclusion. Paclitaxel-induced modulation of CA-125 expression could not be confirmed in vivo. © 2000 Academic Press Key Words: paclitaxel; CA-125; tumor marker; ovarian cancer.
INTRODUCTION The antigen CA-125, first described by Bast et al. [1], is elevated in serum in approximately 80% of women presenting with advanced nonmucinous epithelial ovarian cancer. This tumor marker has evolved into one of the most useful diagnostic tests in the management of individuals with this malignant disease. Rising CA-125 levels may predict disease recurrence many months before clinical evidence of tumor progression. Determination of CA-125 serum levels in ovarian cancer patients has recently gained major interest for prediction
of response to chemotherapy [2– 4]. Decreasing concentrations have been shown to be associated with tumor regression, while increasing CA-125 levels indicate resistance to the treatment and disease progression. Moreover, chemotherapy-induced rapid clearance of CA-125 from the serum is associated with good prognosis [5, 6]. The close correlation with clinical and surgical findings, e.g., at second-look surgery, has evoked a discussion about using CA-125 in addition to or instead of standard criteria for response [4, 7]. A prerequisite for this proposal is that the release of CA-125 by ovarian cancer cells is dependent only on the number of living tumor cells. However, it has been shown that the biosynthesis and shedding of CA-125 are not a constitutively stable process but may be modulated by several factors. It has been demonstrated that ovarian carcinoma cells reduce CA-125 production during the mitotic cycle and increase shedding in the G 0/1 phase [8]. Moreover, interferon ␥ is able to augment biosynthesis and release of CA-125 in cultured ovarian carcinoma cells [9]. This modulation of CA-125 production in vivo could impair the predictive value of tumor marker determination. Taxanes represent a new class of antineoplastic agents being evaluated in several malignant tumors. Paclitaxel in combination with a platinum compound has been shown to induce a high remission rate and to prolong survival in ovarian cancer patients and is, therefore, currently regarded as the standard therapy [10]. Assessment of response by measuring CA-125 has, however, been shown to have limited value in paclitaxeltreated refractory ovarian cancer patients [11]. One possible explanation for this finding was our in vitro experiments showing direct taxane-mediated induction of CA-125 release from ovarian carcinoma cells. We were therefore interested in studying whether paclitaxel increases the expression of CA-125 in ovarian cancer patients in vivo. We chose the endpoint of 24 h for in vivo measurements as in vitro measurements had shown maximum changes after 24 h. MATERIALS AND METHODS
1
To whom correspondence and reprint requests should be addressed at Department of Gynaecologic Oncology, The Norwegian Radiumhospital, Montebello, N-0310 Oslo, Norway. Fax: ⫹47 22 93 4469. E-mail: [email protected]. 0090-8258/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
326
Patients. Patients having biopsy-proven primary or recurrent epithelial ovarian cancer, a WHO performance status
327
PACLITAXEL AND CA-125 SERUM LEVELS
TABLE 1 Characteristics of 53 Patients with Ovarian Cancer Number of patients Age (in years) Median Range FIGO stage I II III IV Histology Serous Mucinous Endometrioid Clear cell Unclassified Other Grade 1 2 3 Not graded Operative treatment Extensive debulking Suboptimal debulking Biopsy only Residual disease Macroscopic negative ⬍2 cm ⬎2 cm Treatment T CT TEC Response CR PR SD PD Line of treatment First Second Third
53
60 39–76 3 (6%) 5 (9%) 36 (68%) 9 (17%) 39 (73%) 3 (6%) 3 (6%) 4 (7%) 3 (6%) 1 (2%) 5 (9%) 14 (27%) 28 (53%) 6 (11%) 19 (36%) 25 (47%) 3 (17%) 14 (26%) 11 (21%) 28 (53%) 13 (25%) 35 (66%) 5 (9%) 21 (39%) 11 (21%) 17 (32%) 4 (8%) 37 (70%) 9 (17%) 7 (13%)
lower than 3, and adequate bone marrow and renal function were enrolled in this prospective trial at The Norwegian Radiumhospital (Table 1). All histological slides were reviewed at our pathology department. Histological classification was based on criteria defined by WHO [12]. Clear cell carcinomas were not graded. Treatment. The patients were treated with either paclitaxel alone 175 mg/m 2 (T), or a combination of carboplatin [AUC (area under the curve) ⫽ 6] and paclitaxel (175 mg/m 2) (CT), or paclitaxel (175 mg/m 2), epirubicin (75 mg/m 2), and carboplatin (AUC ⫽ 4) (TEC). Conventional definitions of response were used according to WHO criteria [12]. A complete response (CR) required the disappearance of all detectable tumors. A partial response (PR) required a ⱖ50% reduction in
the cross-sectional area of the indicator lesion(s) without growth or appearance of any other lesion. Stable disease (SD) was defined as a less than 25% increase or a less than 50% reduction in the cross-sectional area of any lesion without appearance of any new lesion. Progression (PD) consisted of a ⱖ25% increase in the cross-sectional area of any lesion or the appearance of any new lesion. Either category of response required documentation by two observations at least 4 weeks apart. CA-125 determination. Blood samples were obtained the day before and the day after chemotherapy. Serum was collected and CA-125 was determined by means of a fluoroimmunometric assay using an AudoDelfia (Wallac, Finland) as outlined by Norum et al. 1998 [13]. The interassay and intraassay coefficients of variation were 5.3 and 4.2%, respectively. The lower limit of detection as defined by 2SD of the zero standards was 1.2 kU/liter serum. To test the quality of the analysis methods and the biological variation of untreated patients, two control groups were selected. In control group 1 we reanalyzed 78 serum samples after storage at ⫹4°C for 48 h. In control group 2, two separate blood samples were taken with a 48-h interval from 10 patients without any treatment during the last month. Both groups consisted of ovarian cancer patients. Statistical analysis. Data were analyzed with nonparametric procedures. A paired Wilcoxon test was used to analyze differences between measured values of CA-125 before and after paclitaxel treatment. A Spearman correlation coefficient was applied for estimation of correlations. Probability levels less than 0.05 were considered significant. The sample size needed to detect a chemotherapy-mediated induction of CA125 of 10% was estimated to be 44 patients, using a significance level of 0.05 (␣) and a power of 90%. An assumption of a standard deviation of 20% for differences in CA-125 serum levels was applied. RESULTS Fifty-three ovarian cancer patients were included in this trial (Table 1). The distribution of major prognostic factors was typical, with the majority of patients with advanced ovarian cancer. Seventy percent of the patients were tested for CA-125 modulation during first-line treatment. CA-125 serum levels were elevated above the cutoff value of 35 kU/liter in 80% of patients with a median value of 227 kU/liter (Table 2), after primary laparotomy. CA-125 was above the cutoff value in 66% of the patients the day before chemotherapy and this percentage was not increased the day following paclitaxel treatment (66%, n.s.). Median CA-125 concentration was 107 kU/liter 24 h after chemotherapy compared with 99 kU/liter the day before paclitaxel (n.s. tested by paired Wilcoxon test). CA-125 levels measured before and after paclitaxel treatment are illustrated in Fig. 1A. The results are given as percentages relative to the first of the two CA-125 values. Figure
328
PAULSEN ET AL.
TABLE 2 CA-125 Serum Levels in Ovarian Carcinoma Patients a CA-125 (kU/liter)
Postoperative Before paclitaxel After paclitaxel Change before/after (%) Next course Change before/next (%) Last control
Minimum
Quartile 1
Median
Quartile 2
Maximum
9 3 3 54 3 6 4
68 18 15 87 8 49 9
227 99 107 100 54 80 60
1021 571 466 111 260 99 287
13420 3990 3940 135 4300 374 25040
a Serum CA-125 levels were determined in 53 patients postoperatively, before the first course of chemotherapy, before and after paclitaxel-containing chemotherapy, one course thereafter, and after termination of chemotherapy.
1B shows the data from control group 1, which consists of samples from 78 patients with the same samples reanalyzed after 2 days, and control group 2, which consists of 10 patients with two samples taken 48 h apart without paclitaxel treatment. The vertical dotted line is drawn at the reference value for CA-125 (35 kU/liter). The two horizontal dotted lines indicate a difference of ⫾10%. The mean difference and standard deviation for (1) paclitaxel-treated patients, (2) reanalyzed samples, and (3) untreated patients were 0 (16.5), 2.5 (13.2), and ⫺3.2 (18.8) kU/liter, respectively. The largest relative differences existed in samples with low CA-125 values. There was no difference between the mean values. Analysis of subgroups—patients with CA-125 concentration above or below the cutoff value, patients with first-line che-
motherapy, and patients with any other line of treatment— did not change this pattern. Changes in CA-125 serum levels observed immediately after paclitaxel treatment were not correlated to treatment response. However, overall change in CA-125 serum concentration was a good predictor of response to paclitaxel-containing treatment (Table 3). Patients achieving a CR or PR had a significant reduction of median CA-125 levels, whereas tumor progression was associated with increased CA-125 levels (P ⬍ 0.05). Three weeks after treatment with paclitaxel the reduction of CA-125 levels was not detectable for patients obtaining PR, SD, or PD after induction chemotherapy. Only for the group of patients obtaining a CR was a decrease in the median relative CA-125 value observed. To test our hypothesis, we analyzed patients with an elevated CA-125 and a change in relative CA-125 value of more than 10%. We could not detect a significant difference between patients with and without a CA-125 increase with respect to response to chemotherapy, FIGO stage, grade of differentiation, residual disease after primary surgery, line or type of chemotherapy, age, or postoperative CA-125. However, an
TABLE 3 Changes in S-CA-125 Levels in Relation to Chemotherapy Response a Median CA-125 (%) Clinical response CR (n ⫽ PR (n ⫽ SD (n ⫽ PD (n ⫽ P value
FIG. 1.
Results of repeated CA-125 measurements.
21) 11) 17) 4)
Before/24 h after paclitaxel
Before/until next cycle
After induction chemotherapy
100 (67–135) 100 (54–112) 102 (81–125) 80 (65–92) n.s.
53 (6–120) 89 (26–158) 90 (21–372) 88 (37–127) n.s.
7 (1–188) 18 (1–228) 99 (2–786) 1600 (50–3300) P ⬍ 0.05
a Changes are expressed as percentages of initial value. The starting value is 100%. Relative changes in CA-125 serum levels are depicted for immediate paclitaxel-induced effects (before/24 h after), or until next cycle representing a 3-week interval (before/until next cycle), or after induction chemotherapy (six to nine courses).
329
PACLITAXEL AND CA-125 SERUM LEVELS
increase of more than 10% in CA-125 serum levels after paclitaxel was observed in 12 of 49 patients with non-clear cell tumors, compared with 3 of 4 patients with clear cell tumors ( 2 ⫽ 4.649, P ⬍ 0.03). DISCUSSION A combination of paclitaxel with a platinum compound is now considered the standard first-line chemotherapy in advanced epithelial ovarian cancer [10]. High response rates in cisplatin-refractory tumors in second-line treatment have been attributed to the unique mechanism of action. Taxanes promote the assembly of microtubules and also stabilize them, thus preventing their depolymerization [13]. In addition to the antimitotic process there is evidence that other activities may be important as well. Lee et al. observed paclitaxel-dependent transcriptional activation of the interleukin-8 gene in human ovarian carcinoma cells [14]. Expression of human chorionic gonadotropin, the leading tumor marker in gestational trophoblastic disease, was also induced by paclitaxel in choriocarcinoma cells [15]. More recently, we showed that taxanes, namely, paclitaxel and docetaxel, showed a similar antiproliferative activity in human ovarian carcinoma cell lines. In CA-125-positive cell lines both taxanes augmented the release of this tumor marker into the culture medium. We were unable to clarify the mechanism; however, we could affirm that taxanes effectively induce the secretion of CA-125 in ovarian carcinoma cells by their specific action and not simply by release from damaged cells. We were therefore interested in whether the observed taxane-mediated induction of CA-125 in vitro is of clinical importance. However, in our study we were unable to detect any short-term effects of paclitaxel-containing chemotherapy on CA-125 serum levels. This is in agreement with Van der Burg et al., who recently reported CA-125 to be an unreliable marker for response monitoring in patients with relapsed ovarian cancer, especially when the tumor marker is assessed at short intervals, i.e., weekly [11]. We did see a reduction in CA-125 in responding patients already 3 weeks after the first chemotherapy course. At this time the reduction was not statistically significant, but was clearly significant after commencement of induction chemotherapy. Other authors, however, found a good correlation between clinical tumor response and serological marker response in patients treated with either paclitaxel or docetaxel [16 –19]. Looking at overall response we could confirm a correlation to CA-125. It is interesting, however, to note that three of four patients with clear cell carcinoma but only one of four patients with non-clear cell carcinoma showed an increase in CA-125 of more than 10% after one course of chemotherapy. We are therefore currently performing in vitro experiments with ovarian carcinoma cells derived from clear cell tumors to study further this interesting observation. It has been shown that several endogenous as well as exogenous factors can modulate CA-125 expression. CA-125 shedding is not a constitutive and stable process but may be affected
by cell cycle and proliferation as well as by various growth factors and chemotherapeutic drugs. Not only interferons, interleukin-1, tumor necrosis factor ␣, and transforming growth factor ␣, but also taxanes may induce CA-125 expression [9, 20, 21]. Our data show that CA-125 is a reliable serological marker for prediction of treatment response to paclitaxel if it is used with long intervals (3– 4 weeks) as the long-term fluctuations are minor [16 –19]. One possible explanation is that the biological variation of CA-125 level in each patient is greater than the random variation of the analysis. Our control groups indicate that the biological variation in CA-125 serum levels is considerable over a 2-day period. This agrees with the data of van der Burg et al. [11]. Paclitaxel-induced modulation of CA-125 expression observed in vitro was not detectable in this study of ovarian cancer patients in vivo. REFERENCES 1. Bast RC Jr, Klug TL, St John E, Jenison E, Niloff JM, Lazarus H, Berkowitz RS, Leavitt T, Griffiths CT, Parker L, Zurawski VR Jr, Knapp RC: A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 309: 883– 887, 1983 2. Fish RG, Shelley MD, Maughan T, Rocker I: The clinical value of serum CA125 levels in ovarian cancer patients receiving platinum therapy. Eur J Cancer Clin Oncol 23:831– 835, 1987 3. Makar AP, Kristensen GB, Bormer OP, Trope CG: Serum CA 125 level allows early identification of nonresponders during induction chemotherapy. Gynecol Oncol 49:73–79, 1993 4. Rustin GJ, Nelstrop AE, McClean P, Brady MF, McGuire WP, Hoskins WJ, Mitchell H, Lambert HE: Defining response of ovarian carcinoma to initial chemotherapy according to serum CA 125. J Clin Oncol 14:1545– 1551, 1996 5. Buller, RE, Vasilev S, Disaia PJ: CA 125 kinetics: A cost-effective clinical tool to evaluate trial outcomes in the 1990s. Am J Obstet Gynecol 174:1241–1254, 1996 6. van der Burg ME, Lammes FB, van Putten WL, Stoter G: Ovarian cancer: The prognostic value of the serum half-life of CA125 during induction chemotherapy. Gynecol Oncol 30:307–312, 1988 7. Markman M: CA-125: An evolving role in the management of ovarian cancer. J Clin Oncol 14:1411–1412, 1996 8. Marth C, Zeimet AG, Bock G, Daxenbichler G: Modulation of tumor marker CA-125 expression in cultured ovarian carcinoma cells. Eur J Cancer 28A:2002–2006, 1992 9. Marth C, Brattia M, Muller-Holzner E, Mayer I, Zech J, Tabarelli M, Daxenbichler G: Modulation of secretion of human chorionic gonadotropin by biologic response modifiers on term placenta and choriocarcinoma cells. Mol Biother 1:140 –144, 1989 10. McGuire WP, Hoskins WJ, Brady MF, Kucera PR, Partridge EE, Look KY, Clarke-Pearson DL, Davidson M: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 334:1– 6, 1996 11. van der Burg ME, Myles JD, Hoskins PJ, Ten Bokkel Huinik WW and Eisenhauer E: CA 125 is an unreliable marker for monitoring response to taxol therapy in patients with relapsed ovarian cancer. Eur J Cancer 29A:133–133, 1993 12. WHO Handbook for Reporting Results of Cancer Treatment. WHO Offset Publication 48. Geneva, WHO, 1979
330
PAULSEN ET AL.
13. Schiff PB, Horwitz SB: Taxol stabilizes microtubules in mouse fibroblast cells. Proc Natl Acad Sci USA 77:1561–1565, 1980 14. Lee LF, Schuerer-Maly CC, Lofquist AK, van Haaften-Day C, Ting JP, White CM, Martin BK, Haskill JS: Taxol-dependent transcriptional activation of IL-8 expression in a subset of human ovarian cancer. Cancer Res 56:1303–1308, 1996 15. Marth C, Lang T, Widschwendter M, Muller-Holzner E, Daxenbichler G: Effects of taxol on choriocarcinoma cells. Am J Obstet Gynecol 173: 1835–1842, 1995 16. Francis P, Schneider J, Hann L, Balmaceda C, Barakat R, Phillips M, Hakes T: Phase II trial of docetaxel in patients with platinum-refractory advanced ovarian cancer. J Clin Oncol 12:2301–2308, 1994 17. Pearl ML, Yashar CM, Johnston CM, Reynolds RK, Roberts JA: Exponential regression of CA 125 during salvage treatment of ovarian cancer with taxol. Gynecol Oncol 53:339 –343, 1994 18. Pudjade Lauraine E, Piccart M, Rustin G, Gore M, Ten Bokkel Huining
W, Fumoleau P, Van Osteroom A, Kerweij YA, Kaye S, Belpomme D, Van Glabbeke M, Wanders R, Chevallier B, and Franklin H: Comparison of CA 125 kinetics and WHO criteria to evaluate chemotherapy response (R): An EORTC ovarian cancer study with taxotere. Proc Am Soc Clin Oncol 13:269, 1994 19. Bridgewater JA, Nelstorp AE, Rustin GJS, Gore ME, McGuire WP, Hoskins WJ: Comparison of standard and CA-125 response criteria in patients with epithelial ovarian cancer treated with platinium or paclitaxel. J Clin Oncol 17:501–508, 1999 20. Kurachi H, Adachi H, Morishige K, Adachi K, Takeda T, Homma H, Yamamoto T, Miyake A: Transforming growth factor-alpha promotes tumor markers secretion from human ovarian cancers in vitro. Cancer 78:1049 –1054, 1996 21. Zeimet AG, Marth C, Offner FA, Obrist P, Uhl-Steidl M, Feichtinger H, Stadlmann S, Daxenbichler G, Dapunt O: Human peritoneal mesothelial cells are more potent than ovarian cancer cells in producing tumor marker CA-125. Gynecol Oncol 62:384 –389, 1996
Effects of Paclitaxel on CA-125 Serum Levels in Ovarian Cancer Patients Torbjørn Paulsen, M.D.,* ,1 Christian Marth, Ph.D.,† Janne Kærn, Ph.D.,* Kjell Nustad, Ph.D.,‡ Gunnar B. Kristensen, Ph.D.,* and Claes Trope´, Ph.D.* *Department of Gynecologic Oncology, The Norwegian Radiumhospital, Oslo, Norway; †Department of Obstetrics and Gynecology, Innsbruck University Hospital, Austria; and ‡Department of Laboratory Medicine, The Norwegian Radiumhospital, Oslo, Norway Received June 14, 1999
Objective. As in vitro activation of ovarian carcinoma cells in terms of CA-125 secretion by taxanes has been demonstrated, we were interested in whether taxanes also modulate CA-125 expression in vivo. Methods. Serum CA-125 was determined immediately before and 24 h after paclitaxel-containing chemotherapy in 53 ovarian carcinoma patients. To test the quality of the analysis methods and the biological variation of untreated patients, serum CA-125 levels of two control groups were analyzed. Results. Median CA-125 concentration was 107 kU/liter 24 h after chemotherapy treatment compared with 99 kU/liter the day before paclitaxel treatment. Changes in CA-125 serum levels observed immediately after paclitaxel treatment were not correlated to treatment response. However, overall change in CA-125 serum concentration was a good predictor of response to paclitaxel containing treatment. Patients achieving a complete or partial response had a significant reduction of median CA-125 levels, whereas tumor progression was associated with increased CA-125 levels. Only for the group of patients obtaining a complete response was a decrease in the median relative CA-125 value observed. Conclusion. Paclitaxel-induced modulation of CA-125 expression could not be confirmed in vivo. © 2000 Academic Press Key Words: paclitaxel; CA-125; tumor marker; ovarian cancer.
INTRODUCTION The antigen CA-125, first described by Bast et al. [1], is elevated in serum in approximately 80% of women presenting with advanced nonmucinous epithelial ovarian cancer. This tumor marker has evolved into one of the most useful diagnostic tests in the management of individuals with this malignant disease. Rising CA-125 levels may predict disease recurrence many months before clinical evidence of tumor progression. Determination of CA-125 serum levels in ovarian cancer patients has recently gained major interest for prediction
of response to chemotherapy [2– 4]. Decreasing concentrations have been shown to be associated with tumor regression, while increasing CA-125 levels indicate resistance to the treatment and disease progression. Moreover, chemotherapy-induced rapid clearance of CA-125 from the serum is associated with good prognosis [5, 6]. The close correlation with clinical and surgical findings, e.g., at second-look surgery, has evoked a discussion about using CA-125 in addition to or instead of standard criteria for response [4, 7]. A prerequisite for this proposal is that the release of CA-125 by ovarian cancer cells is dependent only on the number of living tumor cells. However, it has been shown that the biosynthesis and shedding of CA-125 are not a constitutively stable process but may be modulated by several factors. It has been demonstrated that ovarian carcinoma cells reduce CA-125 production during the mitotic cycle and increase shedding in the G 0/1 phase [8]. Moreover, interferon ␥ is able to augment biosynthesis and release of CA-125 in cultured ovarian carcinoma cells [9]. This modulation of CA-125 production in vivo could impair the predictive value of tumor marker determination. Taxanes represent a new class of antineoplastic agents being evaluated in several malignant tumors. Paclitaxel in combination with a platinum compound has been shown to induce a high remission rate and to prolong survival in ovarian cancer patients and is, therefore, currently regarded as the standard therapy [10]. Assessment of response by measuring CA-125 has, however, been shown to have limited value in paclitaxeltreated refractory ovarian cancer patients [11]. One possible explanation for this finding was our in vitro experiments showing direct taxane-mediated induction of CA-125 release from ovarian carcinoma cells. We were therefore interested in studying whether paclitaxel increases the expression of CA-125 in ovarian cancer patients in vivo. We chose the endpoint of 24 h for in vivo measurements as in vitro measurements had shown maximum changes after 24 h. MATERIALS AND METHODS
1
To whom correspondence and reprint requests should be addressed at Department of Gynaecologic Oncology, The Norwegian Radiumhospital, Montebello, N-0310 Oslo, Norway. Fax: ⫹47 22 93 4469. E-mail: [email protected]. 0090-8258/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved.
326
Patients. Patients having biopsy-proven primary or recurrent epithelial ovarian cancer, a WHO performance status
327
PACLITAXEL AND CA-125 SERUM LEVELS
TABLE 1 Characteristics of 53 Patients with Ovarian Cancer Number of patients Age (in years) Median Range FIGO stage I II III IV Histology Serous Mucinous Endometrioid Clear cell Unclassified Other Grade 1 2 3 Not graded Operative treatment Extensive debulking Suboptimal debulking Biopsy only Residual disease Macroscopic negative ⬍2 cm ⬎2 cm Treatment T CT TEC Response CR PR SD PD Line of treatment First Second Third
53
60 39–76 3 (6%) 5 (9%) 36 (68%) 9 (17%) 39 (73%) 3 (6%) 3 (6%) 4 (7%) 3 (6%) 1 (2%) 5 (9%) 14 (27%) 28 (53%) 6 (11%) 19 (36%) 25 (47%) 3 (17%) 14 (26%) 11 (21%) 28 (53%) 13 (25%) 35 (66%) 5 (9%) 21 (39%) 11 (21%) 17 (32%) 4 (8%) 37 (70%) 9 (17%) 7 (13%)
lower than 3, and adequate bone marrow and renal function were enrolled in this prospective trial at The Norwegian Radiumhospital (Table 1). All histological slides were reviewed at our pathology department. Histological classification was based on criteria defined by WHO [12]. Clear cell carcinomas were not graded. Treatment. The patients were treated with either paclitaxel alone 175 mg/m 2 (T), or a combination of carboplatin [AUC (area under the curve) ⫽ 6] and paclitaxel (175 mg/m 2) (CT), or paclitaxel (175 mg/m 2), epirubicin (75 mg/m 2), and carboplatin (AUC ⫽ 4) (TEC). Conventional definitions of response were used according to WHO criteria [12]. A complete response (CR) required the disappearance of all detectable tumors. A partial response (PR) required a ⱖ50% reduction in
the cross-sectional area of the indicator lesion(s) without growth or appearance of any other lesion. Stable disease (SD) was defined as a less than 25% increase or a less than 50% reduction in the cross-sectional area of any lesion without appearance of any new lesion. Progression (PD) consisted of a ⱖ25% increase in the cross-sectional area of any lesion or the appearance of any new lesion. Either category of response required documentation by two observations at least 4 weeks apart. CA-125 determination. Blood samples were obtained the day before and the day after chemotherapy. Serum was collected and CA-125 was determined by means of a fluoroimmunometric assay using an AudoDelfia (Wallac, Finland) as outlined by Norum et al. 1998 [13]. The interassay and intraassay coefficients of variation were 5.3 and 4.2%, respectively. The lower limit of detection as defined by 2SD of the zero standards was 1.2 kU/liter serum. To test the quality of the analysis methods and the biological variation of untreated patients, two control groups were selected. In control group 1 we reanalyzed 78 serum samples after storage at ⫹4°C for 48 h. In control group 2, two separate blood samples were taken with a 48-h interval from 10 patients without any treatment during the last month. Both groups consisted of ovarian cancer patients. Statistical analysis. Data were analyzed with nonparametric procedures. A paired Wilcoxon test was used to analyze differences between measured values of CA-125 before and after paclitaxel treatment. A Spearman correlation coefficient was applied for estimation of correlations. Probability levels less than 0.05 were considered significant. The sample size needed to detect a chemotherapy-mediated induction of CA125 of 10% was estimated to be 44 patients, using a significance level of 0.05 (␣) and a power of 90%. An assumption of a standard deviation of 20% for differences in CA-125 serum levels was applied. RESULTS Fifty-three ovarian cancer patients were included in this trial (Table 1). The distribution of major prognostic factors was typical, with the majority of patients with advanced ovarian cancer. Seventy percent of the patients were tested for CA-125 modulation during first-line treatment. CA-125 serum levels were elevated above the cutoff value of 35 kU/liter in 80% of patients with a median value of 227 kU/liter (Table 2), after primary laparotomy. CA-125 was above the cutoff value in 66% of the patients the day before chemotherapy and this percentage was not increased the day following paclitaxel treatment (66%, n.s.). Median CA-125 concentration was 107 kU/liter 24 h after chemotherapy compared with 99 kU/liter the day before paclitaxel (n.s. tested by paired Wilcoxon test). CA-125 levels measured before and after paclitaxel treatment are illustrated in Fig. 1A. The results are given as percentages relative to the first of the two CA-125 values. Figure
328
PAULSEN ET AL.
TABLE 2 CA-125 Serum Levels in Ovarian Carcinoma Patients a CA-125 (kU/liter)
Postoperative Before paclitaxel After paclitaxel Change before/after (%) Next course Change before/next (%) Last control
Minimum
Quartile 1
Median
Quartile 2
Maximum
9 3 3 54 3 6 4
68 18 15 87 8 49 9
227 99 107 100 54 80 60
1021 571 466 111 260 99 287
13420 3990 3940 135 4300 374 25040
a Serum CA-125 levels were determined in 53 patients postoperatively, before the first course of chemotherapy, before and after paclitaxel-containing chemotherapy, one course thereafter, and after termination of chemotherapy.
1B shows the data from control group 1, which consists of samples from 78 patients with the same samples reanalyzed after 2 days, and control group 2, which consists of 10 patients with two samples taken 48 h apart without paclitaxel treatment. The vertical dotted line is drawn at the reference value for CA-125 (35 kU/liter). The two horizontal dotted lines indicate a difference of ⫾10%. The mean difference and standard deviation for (1) paclitaxel-treated patients, (2) reanalyzed samples, and (3) untreated patients were 0 (16.5), 2.5 (13.2), and ⫺3.2 (18.8) kU/liter, respectively. The largest relative differences existed in samples with low CA-125 values. There was no difference between the mean values. Analysis of subgroups—patients with CA-125 concentration above or below the cutoff value, patients with first-line che-
motherapy, and patients with any other line of treatment— did not change this pattern. Changes in CA-125 serum levels observed immediately after paclitaxel treatment were not correlated to treatment response. However, overall change in CA-125 serum concentration was a good predictor of response to paclitaxel-containing treatment (Table 3). Patients achieving a CR or PR had a significant reduction of median CA-125 levels, whereas tumor progression was associated with increased CA-125 levels (P ⬍ 0.05). Three weeks after treatment with paclitaxel the reduction of CA-125 levels was not detectable for patients obtaining PR, SD, or PD after induction chemotherapy. Only for the group of patients obtaining a CR was a decrease in the median relative CA-125 value observed. To test our hypothesis, we analyzed patients with an elevated CA-125 and a change in relative CA-125 value of more than 10%. We could not detect a significant difference between patients with and without a CA-125 increase with respect to response to chemotherapy, FIGO stage, grade of differentiation, residual disease after primary surgery, line or type of chemotherapy, age, or postoperative CA-125. However, an
TABLE 3 Changes in S-CA-125 Levels in Relation to Chemotherapy Response a Median CA-125 (%) Clinical response CR (n ⫽ PR (n ⫽ SD (n ⫽ PD (n ⫽ P value
FIG. 1.
Results of repeated CA-125 measurements.
21) 11) 17) 4)
Before/24 h after paclitaxel
Before/until next cycle
After induction chemotherapy
100 (67–135) 100 (54–112) 102 (81–125) 80 (65–92) n.s.
53 (6–120) 89 (26–158) 90 (21–372) 88 (37–127) n.s.
7 (1–188) 18 (1–228) 99 (2–786) 1600 (50–3300) P ⬍ 0.05
a Changes are expressed as percentages of initial value. The starting value is 100%. Relative changes in CA-125 serum levels are depicted for immediate paclitaxel-induced effects (before/24 h after), or until next cycle representing a 3-week interval (before/until next cycle), or after induction chemotherapy (six to nine courses).
329
PACLITAXEL AND CA-125 SERUM LEVELS
increase of more than 10% in CA-125 serum levels after paclitaxel was observed in 12 of 49 patients with non-clear cell tumors, compared with 3 of 4 patients with clear cell tumors ( 2 ⫽ 4.649, P ⬍ 0.03). DISCUSSION A combination of paclitaxel with a platinum compound is now considered the standard first-line chemotherapy in advanced epithelial ovarian cancer [10]. High response rates in cisplatin-refractory tumors in second-line treatment have been attributed to the unique mechanism of action. Taxanes promote the assembly of microtubules and also stabilize them, thus preventing their depolymerization [13]. In addition to the antimitotic process there is evidence that other activities may be important as well. Lee et al. observed paclitaxel-dependent transcriptional activation of the interleukin-8 gene in human ovarian carcinoma cells [14]. Expression of human chorionic gonadotropin, the leading tumor marker in gestational trophoblastic disease, was also induced by paclitaxel in choriocarcinoma cells [15]. More recently, we showed that taxanes, namely, paclitaxel and docetaxel, showed a similar antiproliferative activity in human ovarian carcinoma cell lines. In CA-125-positive cell lines both taxanes augmented the release of this tumor marker into the culture medium. We were unable to clarify the mechanism; however, we could affirm that taxanes effectively induce the secretion of CA-125 in ovarian carcinoma cells by their specific action and not simply by release from damaged cells. We were therefore interested in whether the observed taxane-mediated induction of CA-125 in vitro is of clinical importance. However, in our study we were unable to detect any short-term effects of paclitaxel-containing chemotherapy on CA-125 serum levels. This is in agreement with Van der Burg et al., who recently reported CA-125 to be an unreliable marker for response monitoring in patients with relapsed ovarian cancer, especially when the tumor marker is assessed at short intervals, i.e., weekly [11]. We did see a reduction in CA-125 in responding patients already 3 weeks after the first chemotherapy course. At this time the reduction was not statistically significant, but was clearly significant after commencement of induction chemotherapy. Other authors, however, found a good correlation between clinical tumor response and serological marker response in patients treated with either paclitaxel or docetaxel [16 –19]. Looking at overall response we could confirm a correlation to CA-125. It is interesting, however, to note that three of four patients with clear cell carcinoma but only one of four patients with non-clear cell carcinoma showed an increase in CA-125 of more than 10% after one course of chemotherapy. We are therefore currently performing in vitro experiments with ovarian carcinoma cells derived from clear cell tumors to study further this interesting observation. It has been shown that several endogenous as well as exogenous factors can modulate CA-125 expression. CA-125 shedding is not a constitutive and stable process but may be affected
by cell cycle and proliferation as well as by various growth factors and chemotherapeutic drugs. Not only interferons, interleukin-1, tumor necrosis factor ␣, and transforming growth factor ␣, but also taxanes may induce CA-125 expression [9, 20, 21]. Our data show that CA-125 is a reliable serological marker for prediction of treatment response to paclitaxel if it is used with long intervals (3– 4 weeks) as the long-term fluctuations are minor [16 –19]. One possible explanation is that the biological variation of CA-125 level in each patient is greater than the random variation of the analysis. Our control groups indicate that the biological variation in CA-125 serum levels is considerable over a 2-day period. This agrees with the data of van der Burg et al. [11]. Paclitaxel-induced modulation of CA-125 expression observed in vitro was not detectable in this study of ovarian cancer patients in vivo. REFERENCES 1. Bast RC Jr, Klug TL, St John E, Jenison E, Niloff JM, Lazarus H, Berkowitz RS, Leavitt T, Griffiths CT, Parker L, Zurawski VR Jr, Knapp RC: A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 309: 883– 887, 1983 2. Fish RG, Shelley MD, Maughan T, Rocker I: The clinical value of serum CA125 levels in ovarian cancer patients receiving platinum therapy. Eur J Cancer Clin Oncol 23:831– 835, 1987 3. Makar AP, Kristensen GB, Bormer OP, Trope CG: Serum CA 125 level allows early identification of nonresponders during induction chemotherapy. Gynecol Oncol 49:73–79, 1993 4. Rustin GJ, Nelstrop AE, McClean P, Brady MF, McGuire WP, Hoskins WJ, Mitchell H, Lambert HE: Defining response of ovarian carcinoma to initial chemotherapy according to serum CA 125. J Clin Oncol 14:1545– 1551, 1996 5. Buller, RE, Vasilev S, Disaia PJ: CA 125 kinetics: A cost-effective clinical tool to evaluate trial outcomes in the 1990s. Am J Obstet Gynecol 174:1241–1254, 1996 6. van der Burg ME, Lammes FB, van Putten WL, Stoter G: Ovarian cancer: The prognostic value of the serum half-life of CA125 during induction chemotherapy. Gynecol Oncol 30:307–312, 1988 7. Markman M: CA-125: An evolving role in the management of ovarian cancer. J Clin Oncol 14:1411–1412, 1996 8. Marth C, Zeimet AG, Bock G, Daxenbichler G: Modulation of tumor marker CA-125 expression in cultured ovarian carcinoma cells. Eur J Cancer 28A:2002–2006, 1992 9. Marth C, Brattia M, Muller-Holzner E, Mayer I, Zech J, Tabarelli M, Daxenbichler G: Modulation of secretion of human chorionic gonadotropin by biologic response modifiers on term placenta and choriocarcinoma cells. Mol Biother 1:140 –144, 1989 10. McGuire WP, Hoskins WJ, Brady MF, Kucera PR, Partridge EE, Look KY, Clarke-Pearson DL, Davidson M: Cyclophosphamide and cisplatin compared with paclitaxel and cisplatin in patients with stage III and stage IV ovarian cancer. N Engl J Med 334:1– 6, 1996 11. van der Burg ME, Myles JD, Hoskins PJ, Ten Bokkel Huinik WW and Eisenhauer E: CA 125 is an unreliable marker for monitoring response to taxol therapy in patients with relapsed ovarian cancer. Eur J Cancer 29A:133–133, 1993 12. WHO Handbook for Reporting Results of Cancer Treatment. WHO Offset Publication 48. Geneva, WHO, 1979
330
PAULSEN ET AL.
13. Schiff PB, Horwitz SB: Taxol stabilizes microtubules in mouse fibroblast cells. Proc Natl Acad Sci USA 77:1561–1565, 1980 14. Lee LF, Schuerer-Maly CC, Lofquist AK, van Haaften-Day C, Ting JP, White CM, Martin BK, Haskill JS: Taxol-dependent transcriptional activation of IL-8 expression in a subset of human ovarian cancer. Cancer Res 56:1303–1308, 1996 15. Marth C, Lang T, Widschwendter M, Muller-Holzner E, Daxenbichler G: Effects of taxol on choriocarcinoma cells. Am J Obstet Gynecol 173: 1835–1842, 1995 16. Francis P, Schneider J, Hann L, Balmaceda C, Barakat R, Phillips M, Hakes T: Phase II trial of docetaxel in patients with platinum-refractory advanced ovarian cancer. J Clin Oncol 12:2301–2308, 1994 17. Pearl ML, Yashar CM, Johnston CM, Reynolds RK, Roberts JA: Exponential regression of CA 125 during salvage treatment of ovarian cancer with taxol. Gynecol Oncol 53:339 –343, 1994 18. Pudjade Lauraine E, Piccart M, Rustin G, Gore M, Ten Bokkel Huining
W, Fumoleau P, Van Osteroom A, Kerweij YA, Kaye S, Belpomme D, Van Glabbeke M, Wanders R, Chevallier B, and Franklin H: Comparison of CA 125 kinetics and WHO criteria to evaluate chemotherapy response (R): An EORTC ovarian cancer study with taxotere. Proc Am Soc Clin Oncol 13:269, 1994 19. Bridgewater JA, Nelstorp AE, Rustin GJS, Gore ME, McGuire WP, Hoskins WJ: Comparison of standard and CA-125 response criteria in patients with epithelial ovarian cancer treated with platinium or paclitaxel. J Clin Oncol 17:501–508, 1999 20. Kurachi H, Adachi H, Morishige K, Adachi K, Takeda T, Homma H, Yamamoto T, Miyake A: Transforming growth factor-alpha promotes tumor markers secretion from human ovarian cancers in vitro. Cancer 78:1049 –1054, 1996 21. Zeimet AG, Marth C, Offner FA, Obrist P, Uhl-Steidl M, Feichtinger H, Stadlmann S, Daxenbichler G, Dapunt O: Human peritoneal mesothelial cells are more potent than ovarian cancer cells in producing tumor marker CA-125. Gynecol Oncol 62:384 –389, 1996