Clinical Cancer Research Vol. 11, 5195-5198, July 15, 2005
© 2005 American Association for Cancer Research
Imaging, Diagnosis, Prognosis |
Prognostic Significance of Baseline Reverse Transcriptase-PCR for Prostate-Specific Antigen in Men with Hormone-Refractory Prostate Cancer Treated with Chemotherapy
Robert W. Ross1,
Judith Manola1,
Kristen Hennessy1,
Matthew Galsky2,
Howard Scher2,
Eric Small3,
W. Kevin Kelly2 and
Philip W. Kantoff1
Authors' Affiliations: 1 Dana-Farber Cancer Institute, Harvard Medical School Boston, Massachusetts; 2 Memorial Sloan Kettering Cancer Center, New York, New York; and 3 University of California, San Francisco, California
Requests for reprints: Robert W. Ross, Dana-Farber Cancer Institute, Smith 353, 44 Binney Street, Boston, MA 02115. Phone: 617-632-6648; Fax: 617-643-2272; E-mail: rwross{at}partners.org.
 |
Abstract
|
|---|
Purpose: Methods accurately categorizing the diverse biology of prostate cancer are needed. A positive baseline reverse transcriptase-PCR for prostate-specific antigen (RT-PCR PSA) in the androgen-independent setting is an independent prognostic marker of survival. The objectives of the current study were to examine the prognostic implication of baseline RT-PCR PSA positivity during treatment with an active chemotherapeutic agent and explore whether an RT-PCR PSA "response" provides prognostic information.
Materials and Methods: In a combined analysis of a phase I and a randomized phase II trial of BMS-247550 (an epothilone B analogue), 104 patients with hormone-refractory prostate cancer had whole blood samples collected at baseline, then with each cycle of therapy. RT-PCR PSA was assessed and related to time to progression (TTP).
Results: From 100 evaluable patients, 368 samples were received, of which 90.8% were evaluable for RT-PCR PSA status. Baseline RT-PCR PSA status was significantly associated with TTP (hazard ratio, 2.22; 95% confidence interval, 1.40-3.52). Twenty-six of 38 patients positive at first assessment had at least one follow-up RT-PCR PSA that was negative ("response"). In univariate analysis, RT-PCR PSA response was not significantly associated with TTP, but in multivariate analysis, RT-PCR PSA response was of borderline statistical significance in predicting TTP (hazard ratio, 0.41; 95% confidence interval, 0.16-1.01).
Conclusion: These results provide further confirmation that baseline RT-PCR PSA is a statistically significant predictor of TTP in hormone-refractory prostate cancer. Moreover, this is the first report to suggest that RT-PCR PSA response during chemotherapy treatment may predict TTP.
Prostate cancer is a remarkably heterogeneous disease, explaining why accurate pretreatment predictors of outcome are of utmost importance in clinical decision making. Moreover, as treatments for metastatic disease improve, predictors of response during treatment may allow for appropriate treatment modifications. In the setting of hormone-refractory prostate cancer, Halabi et al. have developed and validated a model that divides men into risk groups that predict median survival (1). Stratification into these risk groups depends upon factors that consider overall tumor burden and the condition of the patient. Only Gleason grade reflects the inherent biological aggressiveness of the tumor.
Over the last decade, the significance of the detection of circulating prostate cancer cells as determined by the presence of prostate-specific antigen (PSA) transcripts in the blood has been evaluated. In the localized disease setting, a positive reverse transcriptase-PCR for PSA (RT-PCR PSA) is of unclear significance (2–7); however, in hormone-refractory prostate cancer, several reports have confirmed that RT-PCR PSA is a significant and independent prognostic factor for survival (8–10). To date though, no study has considered the prognostic value of a positive RT-PCR PSA before treatment with effective chemotherapy; moreover, no study has considered the prognostic significance of an RT-PCR PSA "response" (converting a positive result to a negative one) while receiving therapy. The objectives of the current study were to both examine the prognostic implication of baseline RT-PCR PSA positivity in the setting of treatment with an active chemotherapeutic agent and to explore whether an RT-PCR PSA response provides prognostic information.
|
Patients and Methods
|
|---|
Patient population. Blood samples from two sequential studies of BMS-247550 (a semisynthetic analogue of epothilone B) were studied. The first was a phase I study of BMS-247750 with estramustine and the second a randomized phase II study of BMS-247550 with or without estramustine. Both studies included only patients with chemotherapy naive progressive hormone-refractory prostate cancer. For the phase I study, 13 patients were treated with 280 mg estramustine orally thrice daily on days 1 to 5, with BMS-247550 at doses of 35 to 40 mg/m2 infused over 3 hours on day 2 of a 21-day cycle. Additional information on this study can be found in Smaletz et al. (11). In the phase II study, 92 patients with chemotherapy naive progressive hormone-refractory prostate cancer were randomized equally to 35 mg/m2 BMS-247550 every 3 weeks alone or with 250 mg estramustine orally thrice daily on days 1 to 5. Additional information on this study can be found in Galsky et al. (12). Both studies were Institutional Review Board approved at all sites and patients consented to participate.
RNA isolation and reverse transcriptase-PCR assay. Details regarding RNA isolation and the RT-PCR assay have been published elsewhere (8). Briefly, samples of whole blood were collected from patients at baseline and then serially with each cycle of therapy. Samples were sent by overnight courier to a central laboratory at the Dana-Farber Cancer Institute for RNA extraction. RNA was isolated from the mononuclear cell fraction by single-step guanadinium thiocyanate extraction. The RNA was reverse transcribed using a primer previously described and Superscript Rnase H-Reverse Transcriptase (Invitrogen, Carlsbad, CA) was used to synthesize cDNA for PSA and actin separately. The PCR products were electrophoresed through a 3.0% gel.
Statistical methods. Descriptive statistics such as medians, ranges, and frequencies were used to characterize patients at baseline. RT-PCR PSA response was defined as a transition from positive at first assessment to negative at any subsequent assessment. Fisher's exact test was used to test for associations among categorical variables. The log-rank test and Cox proportional hazards models were used to characterize associations among RT-PCR, baseline covariates, PSA response, and time to progression. Landmark analysis was considered for the analysis of PSA response but was not used because results were identical to those of a standard proportional hazards model. Exact binomial confidence intervals were computed for response rates. All statistical tests were two sided. P = 0.05 was considered statistically significant.
|
Results
|
|---|
Patient characteristics and samples evaluated. Figure 1 presents the specimen flow for this trial. A total of 100 patients had at least one evaluable RT-PCR PSA for analysis and are included. Table 1 presents baseline characteristics of these 100 patients. They represented a typical, if advanced, patient cohort in this population, with a median age of 70, a baseline good performance status, and a median PSA of 73.8. Most had received three or more prior hormonal treatments.
Reverse transcriptase-PCR for prostate-specific antigen results. A total of 368 samples were received, of which 34 (9.1%) were not evaluable because of the absence of RNA as demoted by a negative actin control. Of the 334 evaluable samples, 110 (32.9%) were positive at baseline. Table 2 shows the patterns of RT-PCR PSA results.
Association of baseline reverse transcriptase-PCR for prostate-specific antigen with efficacy variables. Of the 100 patients included in the analysis, 93 had a baseline RT-PCR PSA measurement, defined as a measurement done within 8 days after registration. The proportion of positive measurements was 38 of 93 or 40.9%. The overall response rate as measured by a drop in PSA of >50% was 65.5% among patients who were RT-PCR PSA negative at baseline (95% confidence interval, 51.4-77.8) and 50.0% among patients who were RT-PCR PSA positive at baseline (95% confidence interval, 33.4-66.6). Using a two-sided Fisher's exact test, P = 0.062 for the association between PSA response and baseline RT-PCR status.
Time to PSA progression was defined according to the criteria outlined in Bubley et al. (13). Of the 93 patients with a baseline RT-PCR PSA measurement, information on time to progression was available on 90. Figure 2 shows time to progression by baseline RT-PCR PSA status. Patients who were negative at baseline had longer time to progression than patients who were positive at baseline (5.3 versus 3.4 months, P = 0.001; see Table 3).
Association of reverse transcriptase-PCR for prostate-specific antigen response with efficacy variables. As described above, RT-PCR PSA response was defined as a transition from positive at first assessment to negative at any subsequent assessment. Of the 100 evaluable patients, 93 had their first assessment at baseline. For the other seven patients, the first evaluable assessment occurred 3 weeks to 4 months after registration. Of these 100 patients, there were 34 who were RT-PCR PSA positive at baseline and had at least one follow-up RT-PCR PSA assessment. Eight patients remained positive (nonresponders) and 26 had at least one follow-up assessment that was negative (responders). There was no association between RT-PCR PSA response and PSA response (two-sided Fisher's exact P = 1.00).
Figure 3 shows time to progression by RT-PCR PSA response category. Patients who responded seemed to have longer time to progression; however, in univariate analysis, this was not statistically significant (two-sided log-rank P = 0.22). The median number of assessments among patients who remained positive was 2 (range, 2-5) compared with 3 (range, 2-7) among patients who went from positive to negative. The association between number of assessments and RT-PCR PSA response was not statistically significant (two-sided Fisher's exact P = 0.67).
In univariate analysis, there were no significant associations between any of the baseline factors (age >70, performance status <90, measurable disease, prior prostatectomy, prior radiotherapy, baseline PSA greater than the median of 70 ng/mL, or >2 prior hormone regimens) and time to PSA progression. In multivariate analysis, when controlling for PSA response, RT-PCR PSA response approached statistical significance in predicting time to PSA progression (P = 0.053, Table 3).
|
Discussion
|
|---|
This study is the first attempt to prospectively evaluate the prognostic significance of baseline RT-PCR PSA results in hormone-refractory prostate cancer patients beginning effective chemotherapy. Consistent with the growing literature regarding the use of RT-PCR PSA in hormone-refractory prostate cancer, we found that RT-PCR PSA positivity is a statistically significant deleterious predictor of time to PSA progression in men beginning epothilone analogue-based therapy. In fact, patients who were RT-PCR PSA positive at baseline (40.9% of the population, consistent with earlier reports; refs. 8, 9) were 2.22 times more likely to progress than those who were baseline RT-PCR PSA negative.
As studies like this one continue to reinforce the value of RT-PCR PSA in identifying a poorer prognosis subset of patients, several questions should be considered. First, how might this information be incorporated into the Halabi stratification model? This assay is a sensitive measure of the disease dissemination, which is influenced by both the biology of the cancer and characteristics of the host. As the assay becomes more widely available, it should be incorporated into future models and nomograms.
Second, might knowledge of baseline RT-PCR PSA status be of value in drug development in advanced prostate cancer? For example, a cytostatic drug that requires some period of time to have an effect (i.e., vaccine therapy, or an antiangiogenesis drug) might be best considered in patients with a negative baseline RT-PCR PSA. On the other hand, a therapy focused on prevention of the development of metastasis might be best tested in a nonmetastatic rising PSA population with a positive RT-PCR PSA.
Third, in the setting of chemotherapy with a proven survival benefit, trials should consider whether, in patients with a rising PSA despite androgen suppression but no radiologic evidence of disseminated disease, RT-PCR PSA status may define a population that would benefit from earlier chemotherapy. Currently, the standard of care in the United States is to treat with secondary hormonal manipulations or investigational therapies until clear radiographic evidence of distant disease, although there is no definitive evidence that any secondary hormonal manipulation extends survival. Using a positive baseline RT-PCR PSA as an entry criteria, this standard could be challenged with a well-designed trial.
Our second goal was to explore the relationship between RT-PCR PSA response and time to progression. This relationship is of scientific interest, as if the elimination of circulating prostate cancer cells is of prognostic value, then it further implicates those cells in the progression of disease. Our trial was limited by the relatively small number of patients who did not respond but still suggests that an RT-PCR PSA response may contribute independent information from PSA response in the determination of time to progression. Larger studies are needed to confirm this result.
Only eight patients in this study had detectable circulating prostate cancer cells at baseline and did not clear these cells with treatment. These patients did particularly poorly, with a median time to progression of 2.1 months. Future studies should isolate these cells and characterize the gene expression of this treatment resistant subset. Such a study would be of particular interest in comparison with circulating prostate cancer cells in responders (both at baseline and again after recurrence).
|
Footnotes
|
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Note: This study has been presented in part at the American Society of Clinical Oncology 2005 Prostate Symposium in Orlando, Florida and at the American Society of Clinical Oncology 2005 Annual Meeting in Orlando, Florida.
Received 2/25/05;
revised 4/14/05;
accepted 4/22/05.
|
References
|
|---|
- Halabi S, Small EJ, Kantoff PW, et al. Prognostic model for predicting survival in men with hormone-refractory metastatic prostate cancer. J Clin Oncol 2003;21:1232–7.[Abstract/Free Full Text]
- de la Taille A, Olsson CA, Buttyan R, et al. Blood-based reverse transcriptase polymerase chain reaction assays for prostatic specific antigen: long term follow-up confirms the potential utility of this assay in identifying patients more likely to have biochemical recurrence (rising PSA) following radical prostatectomy. Int J Cancer 1999;84:360–4.[CrossRef][Medline]
- Ennis RD, Katz AE, de Vries GM, et al. Detection of circulating prostate carcinoma cells via an enhanced reverse transcriptase-polymerase chain reaction assay in patients with early stage prostate carcinoma. Independence from other pretreatment characteristics. Cancer 1997;79:2402–8.[CrossRef][Medline]
- Katz AE, de Vries GM, Begg MD, et al. Enhanced reverse transcriptase-polymerase chain reaction for prostate specific antigen as an indicator of true pathologic stage in patients with prostate cancer. Cancer 1995;75:1642–8.[CrossRef][Medline]
- Mejean A, Vona G, Nalpas B, et al. Detection of circulating prostate derived cells in patients with prostate adenocarcinoma is an independent risk factor for tumor recurrence. J Urol 2000;163:2022–9.[CrossRef][Medline]
- Seiden MV, Kantoff PW, Krithivas K, et al. Detection of circulating tumor cells in men with localized prostate cancer. J Clin Oncol 1994;12:2634–9.[Abstract/Free Full Text]
- Thomas J, Gupta M, Grasso Y, et al. Preoperative combined nested reverse transcriptase polymerase chain reaction for prostate-specific antigen and prostate-specific membrane antigen does not correlate with pathologic stage or biochemical failure in patients with localized prostate cancer undergoing radical prostatectomy. J Clin Oncol 2002;20:3213–8.[Abstract/Free Full Text]
- Kantoff PW, Halabi S, Farmer DA, Hayes DF, Vogelzang NA, Small EJ. Prognostic significance of reverse transcriptase polymerase chain reaction for prostate-specific antigen in men with hormone-refractory prostate cancer. J Clin Oncol 2001;19:3025–8.[Abstract/Free Full Text]
- Halabi S, Small EJ, Hayes DF, Vogelzang NJ, Kantoff PW. Prognostic significance of reverse transcriptase polymerase chain reaction for prostate-specific antigen in metastatic prostate cancer: a nested study within CALGB 9583. J Clin Oncol 2003;21:490–5.[Abstract/Free Full Text]
- Ghossein RA, Rosai J, Scher HI, et al. Prognostic significance of detection of prostate-specific antigen transcripts in the peripheral blood of patients with metastatic androgen-independent prostatic carcinoma. Urology 1997;50:100–5.[Medline]
- Smaletz O, Galsky M, Scher HI, et al. Pilot study of epothilone B analog (BMS-247550) and estramustine phosphate in patients with progressive metastatic prostate cancer following castration. Ann Oncol 2003;14:1518–24.[Abstract/Free Full Text]
- Galsky M, Small EJ, Oh W, Kelly WK. Multi-institutional randomized phase II trial of the epothilone B analogue Ixabepilone (BMS-247550) with or without estramustine phosphate in patients with progressive castrate metastatic prostate cancer. J Clin Oncol 2005;23:439-46.
- Bubley GJ, Carducci M, Dahut W, et al. Eligibility and response guidelines for phase II clinical trials in androgen-independent prostate cancer: recommendations from the Prostate-Specific Antigen Working Group. J Clin Oncol 1999;17:3461–7.[Abstract/Free Full Text]
This article has been cited by other articles:
|
|
|
|
 
D. C. Danila, G. Heller, G. A. Gignac, R. Gonzalez-Espinoza, A. Anand, E. Tanaka, H. Lilja, L. Schwartz, S. Larson, M. Fleisher, et al.
Circulating Tumor Cell Number and Prognosis in Progressive Castration-Resistant Prostate Cancer
Clin. Cancer Res.,
December 1, 2007;
13(23):
7053 - 7058.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Patients and Methods
