| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Letters |
Dept. Internal Med. I, Univ. Hosp. Dykzigt, Dr Molewaterplein 40, 3015 GD Rotterdam, The Netherlands
a Author for correspondence.
To the Editor:
In an editorial comment (1) on our paper (2), Sealey and Laragh erroneously conclude that the IRMA is a step backward from the traditional enzyme-kinetic plasma renin activity (PRA) assay, and that the new IRMA is not suitable for measuring renin because it comeasures 1–2% of prorenin.
They regard the traditional PRA assay as an accurate measure of the in vivo production of angiotensin (Ang) II, the physiologically relevant end product of the renin–angiotensin cascade. In fact, the PRA assay measures the concentration of Ang I that is generated in plasma in vitro after a long incubation period (up to 18 h) under artificial conditions. The PRA assay is an indirect assay of renin and is complicated by the fact that Ang I is usually converted to Ang II and degraded into smaller inactive peptides. Ang I-to-II conversion and Ang I degradation are prevented by lowering the pH of plasma and by adding peptidase inhibitors before the incubation step. Whether this is 100% successful, particularly with such long incubation times, has not been formally tested. These difficulties are not encountered in the direct assay of renin by IRMA.
The prorenin concentration in plasma is higher than the renin concentration, and comeasurement of even a small percentage of prorenin may therefore lead to a sizable overestimation of renin. The important question is: What is the true magnitude of this problem? Our study was designed to address precisely this question, and the answer is simple: The problem is not important enough to render the IRMA unsuitable for clinical use. In contrast to what is stated in the editorial comment, our study demonstrates that the the new assay can readily distinguish low-, normal-, and high-renin hypertension. There was good agreement with the enzyme-kinetic assay not only in the normal- and high-renin ranges but also in the low-renin range, where overestimation would be the most troublesome. Our study also specifically addressed the possibility, suggested in the editorial comment, that comeasurement of prorenin may lead to unacceptably high variability of IRMA results. Between-patient variability in the low-, normal-, and high-renin subjects was similar for IRMA and the enzyme-kinetic assay.
Figure 1
(left) compares the results of the PRA assay according to the
method of Sealey (3), obtained after 3 h of
incubation, with the results of the same assay after 18 h of
incubation in plasma from 16 hypertensive patients with PRA
concentrations within the low to low-normal range of Sealey's method.
The 18-h results were 39% (20–59%) [mean (range)], lower than the
3-h results. To determine the recovery of Ang I in these plasma
samples, Ang I was added in a final concentration of 10 pmol/mL, and
renin activity was blocked by the specific renin inhibitor Remikiren,
in a final concentration of 10 nmol/mL (Ki =
3 x 10-4 nmol/mL) (2). The Ang I
recovery was 97% (90–102%) in the 3-h assay and 87% (84–90%) in
the 18-h assay. Thus the lower PRA result after 18 h of incubation
is, at least partly, caused by loss of Ang I.
|
Sealey (3) recommends prolonging the incubation to 18 h in samples with a PRA <1.0 ng Ang I/mL per hour (~0.8 pmol Ang I/mL per hour). In her opinion, the increase of Ang I is linear during this long incubation time, because comparison with the 3-h assay in samples with such low PRA showed no difference (4). We, however, are not convinced, because the value of 1 ng Ang I/mL per hour was at the lower limit of detection in the 3-h assay, which, it must be emphasized, is the rationale behind Sealey's advice to use the 18-h assay in these low PRA samples.
Figure 1
(right) compares the results of the IRMA of renin with those
of the enzyme-kinetic assay of renin in the same plasma samples as
presented on the left-hand side. It shows that the new IRMA is
perfectly capable of measuring the renin concentrations in these
samples with low to low-normal PRA, which confirms our earlier report
(2). The IRMA results were indeed higher than the results
of the enzyme-kinetic renin assay, probably because of comeasurement of
prorenin, but the error was certainly not greater than the error made
in the 18-h PRA assay as a consequence of the nonlinear increase of Ang
I.
In conclusion, we find no compelling theoretical arguments and no experimental data to support the statement that the new IRMA of renin is a decline in methodology. On the contrary, not only is the IRMA accurate, it also has the important advantage that results obtained in one clinical laboratory can be readily compared with those of other laboratories, because the IRMA results are expressed in units of the internationally accepted renin standard. Results of PRA assays cannot be expressed in this way, because the Ang I generation is determined not only by the plasma concentration of renin but also by the plasma concentration of angiotensinogen. We agree with the complaints of Robertson and Nicholls (5) that this has restricted the exchange of quantitative information among workers and has seriously stunted scientific progress in the field.
|
Footnotes
References
Cardiovascular Center, New York Hosp.–Cornell Univ. Med. College, 525 E. 68th St., New York, NY 10021
a Author for correspondence.
Drs. Derkx and Schalekamp must excuse us for not putting out the welcome mat for their new renin IRMA (1)(2). Renin methodology currently available is simple and precise(3). There is no reason to take a step backward to an assay that at the same time that it measures renin also detects prorenin, a fluctuating component that is always much higher than renin in plasma.
In their enthusiasm for their new tool, Derkx and Schalekamp have lost sight of the fact that determining a subnormal renin concentration in the individual patient is as important as identifying an increased one, especially for hypertensive patients and now also for normotensive patients prone to myocardial infarction or heart failure. Hypertensive patients with low renin concentrations are unlikely to have a myocardial infarction, whereas the amount of renin among those with medium or high renin values is directly related to the subsequent heart attack rate (4). Studies of >100 000 coronary unit patients (5)(6) show that the heart attack rate is promptly and persistently reduced by specific antirenin system drug therapies. This drug-related benefit involves subjects with medium or high renin concentrations and is not apparent in those with low renin (7). This relation between plasma renin concentration and cardiovascular pathology is also supported by previous demonstrations in hypertensive animal models of striking protection from cardiac, cerebral, and renal vascular injury after administration of a specific Ang II antagonist(8)(9).
As we previously pointed out (1), it is inopportune to introduce and proselytize a renin assay (2) that cannot with any certainty identify and discriminate the individual patient with subnormal renin concentrations. The many hypertensive patients with low renin concentrations are at considerable risk of being falsely classified by the proposed new assay as either medium (normal) or high renin. For example, three patients with the same low plasma renin (0.5) but with different plasma prorenins (5, 20, or 200) would be reported as having plasma renins of 0.6, 0.9, and 4.5, respectively. Thus two of the three would be erroneously classified as medium and high renin although all three have low renin. These wrong results are likely to mislead the physician in assessing cardiovascular risk and deciding on the urgency and mode of treatment. This risk of failure to identify the low-renin patient is unnecessary.
On another point, given Derkx and Schalekamp's expertise in the field,
it is most surprising that they are unable to repeat our work, first
published in 1975 and repeated many times since, showing that the 18-h
Ang I generation step of the traditional PRA assay is accurate when
carried out as described by us (Fig. 1
)(10). However, it is noteworthy that their
apparent underestimation of PRA with the 18-h incubation was consistent
across PRA concentrations (left side of their figure), whereas the IRMA
overestimation was much greater in the lower renin range (right side of
their figure). Derkx et al. still do not understand that the reason for
the 18-h incubation is not because the Ang I RIA is not
sensitive enough (it is sensitive to at least 0.3 ng/mL per hour for a
3-h incubation). Incubation for the sample with low renin is prolonged
because the greater amount of Ang I generated overwhelms any
nonspecific interference of plasma in the Ang I RIA and eliminates the
inaccuracy of measuring and subtracting a "blank" (0 °C Ang I
generation). Unfortunately, with their new IRMA, Derkx et al. have no
way of assessing or overcoming nonspecific interference, which becomes
a larger factor in the low range where interference by prorenin is
proportionally greater. Moreover, their claim that interlaboratory
comparison is more precise with a direct assay is wrong. The
"internationally accepted renin standard" to which they refer was
prepared long ago before modern purification techniques were available.
This renin "standard" is not pure, nor was it ever claimed to be
pure. When a pure standard of human renin becomes available that does
not contain potentially cross-reacting substances, restandardization of
direct renin assays will probably be necessary. Meanwhile a pure Ang I
standard for the enzymatic renin assay is readily available to all from
the National Bureau of Standards.
In any case, these issues are not nearly as germane as is the variable but consistently falsely high renin concentrations detected by their IRMA because of the large and variable prorenin intrusion, which we find wholly unacceptable for characterization of plasma renin concentrations not only of hypertensive patients but also of those with diabetes, hypertension, myocardial infarction, or heart failure. Altogether, therefore, this test does not improve on the presently available approach. If the proposed assay was entirely without cross-reactivity for prorenin, it could perhaps be used successfully to assay renin directly and to explore those rare esoteric situations in which the concentration of renin per se needs to be measured. However, to study disease states in which angiotensin formation is the critical issue, the PRA assay remains the gold standard because it incorporates the concurrent contribution of renin substrate to angiotensin formation and therefore is the most meaningful physiological and pathophysiological indicator of the activity of the renin–angiotensin system.
Acknowledgments
Supported by NIH grant HL 18323-SCR, HL 56400, and by grants from the Starr and Maxwell Foundations.
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  R. A. de Bruin, A. Bouhuizen, S. Diederich, F. H. Perschel, F. Boomsma, and J. Deinum Validation of a New Automated Renin Assay Clin. Chem., November 1, 2004; 50(11): 2111 - 2116. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Deinum, F. H.M. Derkx, and M. A.D.H. Schalekamp Improved Immunoradiometric Assay for Plasma Renin Clin. Chem., June 1, 1999; 45(6): 847 - 854. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
