Commentary to "chromogranin a: biological function and clinical utility in neuroendocrine tumor disease".

To the Editors:

We read with interest the study by Modlin et al., published online on March 9, 2010, and although overall we found it to be valid,1 we found an error. We refer to this assertion in the abstract: “Serum and plasma measurements are concordant.” This is not true because they differ markedly. Chromogranin A (CgA) measured in plasma appears to be far higher than in serum. This is supported by information provided on the Web site of CIS Bio International (Gif-sur-Yvette, France), the producer of immunoradiometric assay and enzyme immunoassay kits.

In the Discussion, Modlin et al. are right when they note that there is strong positive linear relationship between plasma and serum CgA levels, but this does not mean that the results are concordant and that the measurements undertaken in either sample type can be related to the same cutoff value.

According to the information presented on CIS Bio International’s Web site, the cutoff value for plasma CgA is approximately 150 ng/mL, which is approximately 50% higher than for the serum samples. The same observation appears in Fig. 1 in the article of Woltering et al.,2 which has been cited by Modlin et al. in relation to the discussed correlation between plasma and serum CgA levels. It is possible that Modlin et al. did not analyze this figure carefully enough; it shows that the serum CgA level of 100 ng/mL corresponds to 150 ng/mL plasma level.

A similar comparison of CgA measurements in serum and plasma samples obtained contemporaneously from the same people (volunteers and patients) was performed in our laboratory, and we can confirm that these differences are real.3