Circulating tumor cells (CTCs) have been recognized as a potential source of prostate cancer seeding to distant metastatic sites (typically bone) for over a century.
Circulating tumor cells (CTCs) have been recognized as a potential source of prostate cancer seeding to distant metastatic sites (typically bone) for over a century, and with ongoing improvements in isolation and characterization methodology, CTCs are now being more rigorously investigated as potential predictive biomarkers in men with metastatic, castration-resistant prostate cancer (mCRPC). The range of treatment options for men with mCRPC has expanded considerably in recent years to include hormonal therapies (eg, abiraterone, enzalutamide), chemotherapies (eg, cabazitaxel), immunotherapies (eg, sipuleucel-T), and those that target the bone microenvironment (eg, radium-223), and as such, predictive biomarkers to guide treatment decisions are urgently needed.1,2
CTCs: Definition and Use as a Biomarker
Biomarkers may be used either prognostically (to predict the outcome of a disease state independent of any given therapy) or predictively, to estimate overall outcome or response to a given therapy.1,3,4The benefit of using CTCs in the latter fashion is that this usage could potentially impact treatment decisions and allow for early discontinuation of ineffective therapies, reduce costs of treatment, decrease treatment-associated morbidities, and allow for the earlier implementation of alternative treatments.5As a prognostic indicator, CTCs would also be invaluable to determine which patients are at the highest risk of death from their disease, and therefore should be treated with the most aggressive therapies.5,6
The FDA has defined CTCs as a nucleated cell of over 4 µm in diameter isolated from blood using a ferromagnetic antibody directed against the epithelial cell surface marker (EpCAM), positive for the cytokeratins, and lacking expression of CD45.1,3,4,7Assessment of CTCs using this methodology (CellSearch System) has been FDA approved as a prognostic indicator for prostate and other cancers.4,5
Prognostic and Predictive Uses in mCRPC
In 2008, de Bono et al reported, in a multicenter prospective study, that CTCs were an accurate independent predictor of overall survival (OS) in patients with mCRPC.5In this study, patients (N = 276) with progressive mCRPC who were initiating a new line of chemotherapy were evaluated for the presence of CTCs by blood draw before treatment, and at monthly intervals thereafter. Prognostic groups were defined as being either favorable (<5 CTCs/7.5 mL) or unfavorable (>5 CTCs/7.5 mL). Those with favorable pretreatment CTCs were found to have significantly better OS than those with unfavorable pretreatment CTCs (21.7 months versus 11.5 months;P<.0001). Interestingly, the CTC counts were also significantly better (P= .0218) than prostate specific antigen (PSA) decrement algorithms (ie, 30% PSA reduction) in predicting OS, and those patients whose baseline unfavorable CTCs turned favorable improved, whereas those with favorable baseline CTCs who turned unfavorable, worsened.5The results confirmed the prognostic value of baseline CTCs, and also showed for the first time that post-treatment CTCs could predict survival after treatment in mCRPC.5
In 2011, a meta-analysis evaluating the detection of CTCs in peripheral blood as a prognostic factor in prostate cancer examined a total of four studies involving some 486 patients; all of the studies used the CellSearch method to isolate CTCs and used the <5-cell versus ≥5-in-7.5 mL peripheral blood criteria to define the favorable and unfavorable prognostic groups, respectively.6This meta-analysis showed a lower OS in the unfavorable versus the favorable group, with a relative risk of 2.51 (95% CI, 1.96-3.21). The mean survival time for the combined unfavorable groups was 12.5 months versus 29.5 months for the favorable groups, and the corresponding mean of 2-year survival rate was 23.4% and 62.9%.6
Docetaxel is the standard first-line treatment for mCRPC, and in 2014, Goldkorn et al examined the use of CTC enumeration as a prognostic and predictive marker in the Southwest Oncology Group (SWOG) trial S0421, a phase III study examining the use of docetaxel three times per week (with or without atrasentan) in men with mCRPC and bone involvement.8In the study, over 200 patients from the original SWOG S0421 had evaluable CTC counts, and Kaplan-Meier analysis for 2-year survival showed significantly better survival in those with <5 versus those with ≥5 CTC/7.5 mL (26 vs.13 months, HR, 2.74;P<.001) for the Cox-regression analysis of the association between CTC count and OS, after correcting for other clinical variables. A baseline CTC of ≥5 was also associated with higher tumor burden and unfavorable disease outcome indices, such as increased bone pain, increased PSA, lower hemoglobin, and higher alkaline phosphatase.
In terms of its use as a prognostic indicator, CTC increases observed as early as Day 21 were also associated with significantly worse OS (P= .041), although a drop in CTCs between Day 0 and 21 only tended (nonsignificantly) to improve OS (P= .324). The findings, although certainly requiring further validation, thus suggest that an early (cycle 1) rise in CTCs while on docetaxel might portend a poor outcome, and could influence management strategies, so as to potentially forego additional docetaxel in favor of another treatment option in such a patient.8Findings from the phase III trial of abiraterone, reported in 2011, also confirmed that CTC conversion from unfavorable to favorable was predictive of OS as early as 4 weeks after initiating therapy, and the observed treatment effect (ie, favoring abiraterone over placebo) could be explained by inclusion of this conversion at all post-treatment time points.3
Use With Radium Therapy
Among the recently approved therapies for mCRPC is radium-223 (Ra). In the phase III ALSYMPCA trial, men with mCRPC with two or more bone metastases, no visceral metastases, and who were unsuitable for, or had received prior docetaxel (N = 921) were randomized to Ra or placebo for 6 intravenous injections.9,10The results of the study showed a survival benefit of 3.6 months in the final analysis (14.9 months versus 11.3 months; HR, 0.70;P<.001) for patients on Ra versus placebo.10
At this year’s ASCO Annual Meeting, a report by Suárez et al examined the viability of using CTCs as a predictive biomarker of response to Ra treatment (abstract not presented at the meeting but published in conjunction with the meeting).11Blood samples were collected from 46 patients with mCRPC treated with Ra at baseline, at cycle 3, and at progression, and were centrally analyzed using the CellSearch methodology. Response was defined in the study as >50% decrease in CTCs at cycle 3 relative to baseline or ≤5 CTCs; progression was defined as >50% increase in CTCs or >5 CTCs, and stabilization was defined as neither response nor progression.11
Results of this study showed that a CTC response was observed in 29% (n = 8 patients), disease stabilization in 42% (n = 12 patients), and progression in 29% (n = 8 patients). Notably, those with 5 or fewer CTCs at cycle 3 had a greater likelihood of remaining on treatment, with 88% (14/16) remaining on treatment after cycle 3, versus 50% (6/12) with >5 CTCs at cycle 3 (P=.04).11CTCs at baseline also tended to identify patients likely to continue past cycle 3, with 87% (13/15) of those with ≤5 CTCs at baseline remaining on treatment after cycle 3 versus 54% (7/13) of those with >5 CTCs (P= .09). Although requiring further validation, the authors suggest the response to Ra and its use beyond cycle 3 can be predicted using CTCs as a biomarker.
Further advances in the characterization of CTCs, which move beyond their simple enumeration also have the potential to advance the use of CTCs in prostate and other cancers.12New technologies that aim to characterize genetic changes in CTCs (such as mutations that may confer resistance to selected therapies) will likely provide further information about drug resistance and enable clinicians to distinguish among available treatment choices in the increasingly complex setting of mCRPC.12