Response

Extract

We appreciate Lamy et al.’s enquiry about the role of serine protease urokinase-type plasminogen activator (uPA) and plasminogen activator inhibitor type 1 (PAI-1) in our review (1). Indeed, uPA and PAI-1 have been studied as biomarkers in node-negative, early-stage breast cancer. American Society of Clinical Oncology guidelines state that immunohistochemistry for these markers is not accurate and the prognostic value of enzyme-linked immunosorbent assay has been validated with a minimum of 300mg of fresh or frozen breast cancer tissue (2). Readers should become familiar with the myriad and practical types of evidence that influence the overall assessment of an assay’s value and that extend beyond the Level-of-Evidence (LOE) determinations of clinical validity. These include evidence on analytical validity, decision impact, benefit in clinical outcomes, benefit outside research settings, and affordability (3). First, practice guidelines vary in recommendations on validity and use of uPA and PAI-1 as an option. The National Academy of Clinical Biochemistry Laboratory Medicine Panel stated that uPA and PAI-1 are the first biological factors to have been validated for prognosis in breast cancer (4). By contrast, the latest 2012 update of the National Comprehensive Cancer Network guidelines do not mention the test, with the panel stating that assays other than 21-gene recurrence score or the 70-gene assay have not been sufficiently validated (5). The 2011 St. Gallen’s Consensus Conference found that 23% of respondents voted in favor and 50% against the option of uPA and PAI-1 as a potential help in decision making (6). Moreover, the need for fresh frozen tissue was cited as a practical factor in the voting, even though there was awareness of the evidence on clinical validity. Second, criteria for including an assay in our review were that an assay “ha[s] achieved regulatory approval or [is] performed in a laboratory that is Clinical Laboratory Improvement Amendment certified,” targeting the US market (1). The tumor-associated antigen immunological test system for uPA and PAI-1 sold in the United States does not have clearance by the US Food and Drug Administration, nor do manufacturers offer to perform the test in a Clinical Laboratory Improvement Amendment–certified laboratory. The analytical validity and clinical validity of the enzyme-linked immunosorbent assay test for prognostic use in breast cancer then must be established by individual laboratories using the test system. Third, the final LOE determination framework proposed by Simon et al. makes no distinction for LOE 1 based on a prospective, controlled trial (Category A) vs a prospective trial using archived tissues (Category B); the distinction concerns the number of validation studies with consistent results that are needed, with none for Category A studies and one or more for Category B studies (7). Last, no published study provides evidence on the effect of the test on clinical decisions, benefit outside research settings, or on its affordability. In summary, although the clinical validity of the uPA and PAI-1 tumor markers is supported with Level I evidence, practice guideline panels, such as the National Comprehensive Cancer Network and St. Gallen’s Consensus Conference, also consider evidence on additional factors (ie, analytical validity, clinical utility, and affordability) as relevant when deciding whether to recommend an assay as an option for routine clinical use.