Table 2.

Summary of current data supporting changes in biomarkers as potential surrogates for clinical benefit in men with castration-resistant prostate cancer

Biomarker/outcome parameter	Evidence reference	Pros	Cons
PSA declines	Armstrong et al. [21]; Petrylak et al. [22]	Easily measurable Widely available Time <3 mo Evidence to support use with cytotoxic therapy	Not validated with novel agents (ie, PSA-independent benefits) PSA can rise after start therapy in minority Threshold of response unclear Does not allow for unique mechanism of novel agents (immunologic, differentiating, cytostatic) Subgroups of prostate cancer do not produce PSA
Progression-free survival	Halabi et al. [25]; Hussain et al. [26]; Scher et al. [28]	May capture clinical benefit as a delay in pain/tumor growth Improved measure of effect of cytostatic or antiangiogenic agents Flexible definitions	Exact definition is critical Composites likely necessary Lack of validation as surrogate for OS Censorship prevents current surrogate analyses
Pain improvements	Armstrong et al. [21]; Halabi et al. [82]	Direct patient measure	Qualitative thus requires validated scales Many men with CRPC are pain free Subjective, variable, and subject to change with narcotic analgesia alone Not validated Difficult to use as a marker by itself; many causes of pain independent of tumor progression
Bone turnover markers (urine N-telopeptide, bone alkaline phosphatase)	Coleman et al. [60]; Sonpavde et al. [66]	Reflects tumor-stromal interaction and prostate cancer microenvironment Linked to survival in multiple data sets	Normal in patients with visceral-only or node-only disease May be normal even in the face of bone metastases Unclear clinical implications if incompletely suppressed
Quality of life	Berthold et al. [34]	Direct patient measure	Qualitative; thus requires validated scales/measure Defining clinically significant changes Bias is inherent in non–placebo-controlled trials
Radiographic responses (including bone scan changes)	Scher et al. [15]; Sonpavde et al. [43]; Ryan et al. [74]; Scher et al. [83]	Well-defined criteria if measurable disease	No target lesions in patients with increasing PSA and localized disease or bone-only disease Not always measurable soft tissue disease in prostate cancer Modest correlation with overall survival Important treatment effects are missed Bone scan flare can be common, requiring confirmation scans
CTCs	de Bono et al. [42]; Scher et al. [44]	Early detection before PSA rise Allows tumor-specific biomarker assessment within CTCs Strongly prognostic and early signs of validity as surrogate	Only approximately 50% have detectable levels even with widespread metastases using FDA-approved CellSearch platform Not validated as surrogate yet Expensive; performed in specialized labs only; unable to bank/store Quick turnaround necessary due to expiration within 72 h

PSA = prostate-specific antigen; OS = overall survival; CRPC = castration-resistant prostate cancer; CTC = circulating tumor cells; FDA = US Food and Drug Administration.