Endothelial nitric oxide synthase gene polymorphisms and renal survival.

To the Editor: Development of end stage renal disease (ESRD) and a blunted in vitro generation of nitric oxide (NO) have been recently associated with the 298Asp substitution in the endothelial NO synthase (eNOS) gene.1 Diabetic nephropathy was the culprit of ESRD in most patients (pts);1 however, the same association was recently reported in pts with ESRD due to polycystic kidney disease.2 Collectively these results would imply that a 298Asp allele–determined predisposition to generate less NO is mechanistically related with decreased renal survival. If proven, this appealing hypothesis could open new avenues to pharmacological prevention of ESRD. However, some caveats suggest caution in jumping to this conclusion. First, criteria for patient enrollment were not given, and therefore a selection bias cannot be excluded.1 Second, both studies were small since each entailed less than 200 ESRD pts. Sample size calculation showed that even in the larger study1 a two-group 2 test with a 0.05 two-sided significance level had only 60% power to detect a Glu298Asp genotype frequency difference between ESRD pts and controls. Thus, these were either lucky or serendipitous findings. Under both circumstances larger studies are mandatory. Furthermore, eNOS allele frequency differs markedly between Japanese and Caucasians; thus, these results must be replicated in populations with different ethnic backgrounds. Third, the restriction fragment length polymorphism analysis (RFLP) used for genotyping1 cannot be as accurate as it should be. We replaced this methodology with the melting curve analysis3 because of inconsistent amplicon cleavage with BanII resulting in misdiagnosis of GT and GG as TT and GT, respectively. According to the HardyWeinberg equilibrium, most pts would be heterozygous; therefore, the rate of misgenotyped pts could not be negligible. Sequencing of amplicons may circumvent this bias, but it was performed only in a minority, eg, on uncleaved fragments.1 Noiri et al stated that the Glu298Asp polymorphism is functionally relevant by quoting a study that claimed the 298Asp variant to be more vulnerable to intracellular cleavage.4 However, this contention was thereafter disproved by the demonstration that the intracellular cleavage found in cells harboring the 298Asp eNOS substitution was an in vitro artifact.5 According to both studies, the 298Asp replacement does not affect eNOS biological activity.4,5 Thus, it remains controversial whether the 298Asp variant implies a blunted eNOS activity; accordingly, the contention that the latter accounts for decreased renal survival cannot be taken for granted. Other functionally relevant eNOS polymorphisms, such as the T C variant in the eNOS promoter, exist. In the large cohort of Caucasian pts of the GENICA study6 and in another study,2 the Glu298Asp and T C genotypes were in linkage disequilibrium, albeit not strongly. We proposed that this genetic variant can be the most important for NO generation in essential hypertensive pts.3 This polymorphism was not investigated by Noiri et al,1 while it did not associate with ESRD in another study.2 However, the C allele was associated with multivessel coronary atherosclerosis,6 and since coronary events are the major cause of death in ESRD pts, premature mortality of C allele carriers could result in underrepresentation of this allele in the ESRD population. Recent findings by 2 independent groups support our contention.7,8