William Slade (Presenter)
LabCorp
Authorship: William O. Slade (1), Grace Van Der Gugten (2), Christopher M. Shuford (1) , Patricia Holland (1), Daniel T. Holmes (2), Russell P. Grant (1)
(1) LabCorp, Burlington, NC (2) St. Paul
| Short Abstract Plasma renin activity (PRA) has been traditionally measured using immunoassays for angiotensin I (AngI) with little attention paid to inter-method harmonization. These methods require generation steps of 3-18 h and include a portion of the specimen evaluated as a “cold sample” to account for AngI generation prior to analysis. Both of these requirements decrease throughput. We present manual and automated methods using a single internal standard to quantify PRA by LC-MS/MS. Our method development employed a double internal standard approach to evaluate the effect of generation conditions and pre-analytical processing on AngI recovery. Automation will be discussed. Finally, we demonstrate harmony with the LC-MS/MS method of St. Paul’s Hospital, Vancouver, despite numerous methodological differences. |
Long Abstract
Introduction
Plasma Renin Activity (PRA) measures the capacity peripheral blood to generate the peptide angiotensin I (AngI) and is critical in the diagnosis of disorders of the renin angiotensin aldosterone system (RAAS). In the RAAS, angiotensinogen, constitutively manufactured by the liver, is hydrolyzed by renin, an enzyme manufactured by the renal juxtaglomerular apparatus, to form AngI. AngI can then be cleaved by angiotensin converting enzyme (ACE) to yield Angiotensin II (AngII), a potent vasoconstrictor and mediator of a number of downstream RAAS effects. Current PRA assays measure the formation of intact AngI after a buffered generation period of 3hr+ and require blank-correction of a so-called “cold specimen” to account for AngI generated prior to analysis. Further, the intermethod comparability between current assays is largely unexplored which makes application of suggested aldosterone:PRA ratio thresholds a challenge. As a reference laboratory receiving thousands of PRA samples per week, we sought to develop a rapid, automatable PRA assay using a single internal standard and evaluate its performance against production PRA assays at other centers.
Results and Discussion
PRA assays have historically split specimens into two samples for the generation reaction: warm (37 °C) and cold (4 °C). The final PRA measurement was determined by subtracting the AngI measurement of the cold sample from the warm sample and dividing by the duration of the incubation. The rationale is that AngI results for the warm sample needed correction for any AngI present or generated prior to analysis.
To evaluate the effect of using a cold sample on PRA measurements, 105 patients were drawn in-house and processed according to our method. Comparing the warm AngI measurement to the blank-corrected measurement using Passing-Bablok Regression yielded a slope of 1.000, an intercept of 0 and a correlation coefficient of 0.9984. Further, we evaluated the effect of including the cold sample on the Aldosterone-to-Renin ratio of these patients. In only one case did a high level of endogenous Ang1 alter the ARR ratio from negative (< 30 ng/mL : ng/mL/hr) to positive (>30 ng/mL : ng/mL/hr). These results support the notion that blank correction is probably not required for the clinical utility of PRA.
PRA measurement obviously requires AngI to be sufficiently stable to allow reliable and repeatable results. To evaluate the recovery of AngI at all stages of the PRA assay, we employed a double internal standard approach. We evaluated the accuracy of AngI by specimen type, generation buffer, and preanalytical processing. Our results indicate that EDTA plasma is required for PRA, and we have established a simple, automated screen to distinguish non-EDTA plasma samples. We also observed striking differences in the stability of AngI during generation using maleic acid or tris-based buffers. For example, we observed a median recovery (N=336) of 62.5% (±58.7%) using a tris-based buffer compared to a median recovery (N=185) of 85% (±20.6%) using a maleic-acid based buffer. Further, we found a significant impact of preanalytical processing, i.e., thawing method, post-collection processing, and time-to-freeze, that increased the median recovery (N=36) to 97% (±14.4% ) using a maleic-acid based buffer compared to 85% without control of preanalytical variables.
Our validation parameters are compliant with FDA, GLP, C62 and CAP guidance. The calibration range was 0.167 to 66.7 ng/mL/hr, with intra- and inter-assay inaccuracy (to the nominal LLOQ concentration) and precision at the LLOQ of ≤± 5% and ≤ 7% CV, respectively. Intra- and inter-assay precision for quality control material was ≤ 8%.
To accommodate the required throughput, we developed an automated workflow. Across 125 patient specimens, 3 days, and 3 batches, we compared our manual procedure to an automated, double-plate batch procedure using a randomized design. Passing-Bablok Regression yields slopes of 0. 920 (CI: 0.898 to 0.940) and 0.949 (CI: 0.928 to 0.974), with intercepts of -0.0203 and -0.640 and correlation coefficients of 0.9943 and 0.9907, respectively for the two automated plates. The automated procedure includes a colorimetric confirmation of EDTA plasma.
Finally, inter-method comparison of LC-MS/MS based PRA assays is not reported. We compared our method with the LC-MS/MS method of St. Paul’s Hospital, which is substantially different from a methodological standpoint (buffer, generation duration, sample purification, calibrator). After assaying 107 patients at each site, Passing-Bablok regression yielded a slope of 1.039 (0.997 to 1.083 95% CI), intercept of -0.01191 and a correlation coefficient of 0.9950.
References & Acknowledgements:
| Description | Y/N | Source |
| Grants | no | |
| Salary | yes | LabCorp |
| Board Member | no | |
| Stock | no | |
| Expenses | no |
IP Royalty: no
| Planning to mention or discuss specific products or technology of the company(ies) listed above: | yes |