Marshall and Shaw in their editorial1 are correct to hypothesise that the normal range of serum angiotensin converting enzyme (SACE) would be narrower if ACE genotype (insertion/deletion; ID) is also taken into account. This would split its wide normal range into three narrower ranges defined by the genotypes II, ID and DD, with SACE concentrations rising co-dominantly with each copy of the D allele. However, they fail to realise that we have already shown that, not only is this the case, but also if SACE is measured in accordance with ACE genotype its diagnostic sensitivity in acute sarcoidosis is improved by 33.5%.2

In our paper we determined a genotype based normal range for SACE with 146 healthy white volunteers. Using this new normal range, SACE was measured in 29 patients with histologically proven sarcoidosis. The new SACE genotype based normal range identified 69% of these patients, an improvement of 33.5% on the previous normal range which identified only 51.7% of our patients with sarcoidosis.2Importantly, our work has also been independently confirmed.3

We suggest that the genotype based SACE normal range should be determined locally (especially for different racial populations4) to improve its sensitivity in the diagnosis of active sarcoidosis.