According to the Phoenix Criteria for biochemical failure (BCR), a PSA 2 ng/ml above nadir is defined as failure and the threshold where next line therapy should be initiated. (1)
I find the rationale for this criteria lacking because nadir after ADT and chemotherapy is correlated with time to recurrence and cancer specific survival (2). A nadir less or equal to 0.2 ng/ml typically indicates a good response to therapy and the likelihood of longer progression free and cancer specific survival. Adding 2 ng/ml to a nadir > 1.0 ng/ml, for example, as a criterion for progression is quantitatively different with respect to progression free survival and overall survival, compared to achieving a nadir of 0.2 ng/ml or less. Some will argue that the PSA test is not accurate and therefore 2 ng/ml threshold is necessary because of the scatter. In contrast, I have found that PSA results from two different laboratories in Canada have agreed fairly will and that the amount of scatter is low. In my opinion, a linear or polynomial correlation coefficient of 0.9 or better for 4 to 6 consecutive PSA results (1 – 2 months apart) indicates a reliable trend.
Furthermore, a consistent upward trend in monthly PSA results from any nadir, including a nadir <= 0.2 ng/ml, is an indication of failure and the doubling time an indication of the aggressiveness. Consistent increases and rates of increase from any nadir are indications of failure and aggressiveness and the need for immediate next line therapy. The exception to this rule of course would be the development of or increase in neuroendocrine-like character of the cancer at later stages, which does not express PSA at the same levels and is associated with metastatic progression according to bone and CT scans without significant increases in PSA. Additional measures such as ALP, LDH, neutrophil/lymphocyte ratio (inflammation marker), circulating tumor cells (CTCs), bone and CT scans to assess metastatic progression are necessary to determine the status of a patient.
(1) ncbi.nlm.nih.gov/pmc/articl...
(2) ncbi.nlm.nih.gov/pubmed/274...
Comments, opinions would be appreciated...
Cheers,
Phil