Genotyping increases the yield of angiotensinconverting enzyme in sarcoidosis – a systematic review

Sarcoidosis is an inflammatory, granulomatous disease. Its pathogenesis is unknown, but probably involves genetic predisposition as well as external factors [1]. Approximately 500 new cases are diagnosed in Denmark annually. Diagnosing sarcoidosis is challenging and includes radiological changes, clinical manifestations and paraclinical findings, including measurement of serum angiotensin-converting enzyme activity (s-ACE, peptidylpeptidase A). ACE has a number of metabolic effects; most notably it catalyses the modification of angiotensin I to angiotensin II, a potent vasoconstrictor [2] and inactivates bradykinin through the kallikrein-kininogen system [3]. It is also a potent pro-inflammatory modulator [4] secreted by activated cells of the monocyte-macrophage cell lineages, which are crucial in the process of granuloma formation. S-ACE is elevated in about 60% of sarcoidosis patients [5], but also in other granulomatous diseases like Gaucher’s disease and tuberculosis [6]. Though the level of s-ACE reflects the mass of granuloma in the body [7], the clinical use of s-ACE in monitoring disease activity is controversial, and recommendations differ between guidelines.

The activity of ACE can be measured by enzyme kinetic methods which most commonly utilise the polypeptide FAPGG (furyl-acryloyl-phenylalanyl-glycyl-glycine), which acts as a synthetic substrate for ACE. The degradation of FAPGG to FAP is visualised by a changed absorption spectrum by spectrophotometry [8]. Several commercial kits for automatic analysis are available.

The normal level of ACE depends on genetic variation in the ACE gene. In intron 16, a common insertion/deletion polymorphism, varying in a 287 base pair sequence, is of importance [9]. The genotypes are termed DD (homozygote for deletion), ID (heterozygote) and II (homozygote for insertion).

The I/D polymorphism is responsible for almost half of the biological variation in s-ACE among healthy individuals [10], s-ACE being highest in individuals carrying the genotype DD and lowest in genotype II.

Analysis of the genotype was previously performed by restriction fragment length polymorphism (RFLP) [9], which has now been replaced by PCR-based methods for identification of the I and D alleles [11] and most recently by high-resolution melting (HRM) technique [12].

Since the I/D polymorphism impacts the normal level of s-ACE, we aimed to summarise current evidence for genotype-based differences in mean values of s-ACE in different ethnic populations.

METHODS

We used PubMed to search the MEDLINE library for articles providing genotype-based reference intervals of s-ACE until June 2013. We applied the following search terms: ”sarcoidosis, pulmonary” (Mesh) AND “peptidyl-dipeptidase A” (Mesh), and “sarcoidosis” AND “ace” AND “genotype” (free text search). We restricted the search to articles in English, German and Danish. No limits were set regarding entry year

From the studies selected, genotype-based mean values of s-ACE and standard deviations were obtained, as well as size and ethnicity of the study populations. Between-laboratory and between-assay variation in measurement of s-ACE is substantial [8] and, furthermore, results were reported in different units between studies. Aa quantitative meta-analysis of the mean s-ACE levels would therefore not provide useful information for comparing the genotype groups. Instead, we used ln-transformed mean values of s-ACE from each study for the groups II, ID and DD to calculate ratios of s-ACE between the groups. These ratios (II/DD, II/ID and ID/DD) do not provide clinically meaningful information in themselves, but they serve to evaluate whether the differences between the groups are significant. This is the case if the confidence interval does not include the value one.

We summarised II/DD, II/ID and ID/DD ratios as weighted mean values as the sample sizes varied considerably. Data were expressed as means and two-sided 95% confidence intervals. Furthermore, we calculated ethnicity-stratified, weighted estimates of the II/DD ratio.

Finally, we summarised genotype frequencies from studies in which such data were reported, and ethnicity-grouped median values were obtained.

Where applicable, the review process complied with the Prisma guidelines [13].

Ethics approval

The review included only data from prior studies. Therefore, ethics approval was not needed. All included studies documented that appropriate ethics approvals had been obtained.

RESULTS

The study selection process is outlined in Figure 1. We identified 102 journal articles. By review of title and abstract, 12 articles were relevant for this analysis. Seven articles [14-20] presented new genotype-based reference intervals for s-ACE based on genotyping and s-ACE measurements in healthy individuals. Furthermore, one article [9] presenting genotype-based reference intervals was identified from reference lists and, in addition, data from a recent Danish study [12] were included. The nine studies represented 2,052 healthy individuals. Genotype-based mean values of s-ACE, and standard deviations for all nine studies are shown in Table 1. One study [14] measured ACE activity with two assays. These are provided individually in Table 1.

All studies found significantly different levels of s-ACE between genotype groups, with DD having the highest mean ACE value, II having the lowest mean ACE value and ID having intermediate values. The distribution of ratios of s-ACE (DD/II, DD/ID and ID/II) is shown in Figure 2. The weighted mean DD/II ratio was 1.85 (range: 1.79-1.92) for all studies, 2.01 (1.92-2.10) for Caucasians and 1.64 (1.55-1.73) for Asians. The mean DD/ID ratio and ID/II ratio were both significantly different from one. Therefore, the mean s-ACE was significantly higher for the DD genotype than for the ID genotype, which was again significantly higher than that for the II genotype.

Two studies [16, 20] provided genotyping and s-ACE data for 310 sarcoidosis patients. These are shown in Table 2. Weighted mean DD/II ratio was 1.45 (1.30-1.62) for these patients. Severity of sarcoidosis was indicated roentgenologically ad modem DeRemee, but genotype-based s-ACE levels were not stratified for roentgenologic disease stage.

All included papers documented that none of the participants were being treated with ACE inhibitors.

Frequency of I/D genotypes was reported in eight of nine study populations. These are shown in Table 3. When adjusting for ethnicity, the mean frequency of I/D genotypes among Caucasians was 27.6% II, 45.7% ID and 26.7% DD, while the mean frequencies among two Asian studies was 36.1% II, 49.9% ID and 14.0% DD.

DISCUSSION

This review revealed that significant differences of s-ACE between I/D genotype groups were seen in all included studies. On this basis, it seems rational to recommend genotyping of patients if the value of s-ACE is considered part of the diagnostic process for sarcoidosis or for monitoring disease activity in confirmed cases. It is important to note, though, that the benefit of I/D genotyping is primarily derived from the significant differences of s-ACE between the genotypes in healthy people. In this analysis, two studies supported that genotyping will also improve the yield of s-ACE in sarcoidosis patients, but more clinical studies are needed to clearly confirm this finding.

The mean s-ACE level for the genotype DD is almost twice that of genotype II. Variance analyses od previous data have shown that almost 50% of the normal variation in s-ACE in healthy people is attributable to the polymorphism [9, 17]. It is noteworthy that by evaluating s-ACE in 129 sarcoidosis patients and sarcoidosis suspect patients after genotyping Kruit et al [17] found that 8.5% of these were misclassified as having either normal or elevated s-ACE by application of standard (non-genotype based) reference intervals. Sharma et al [19] evaluated their genotype-based reference intervals on 47 sarcoidosis patients and found 33.5% more patients to have elevated s-ACE than by applying standard reference intervals. These findings indicate that routine genotyping would increase the yield of s-ACE in diagnosing sarcoidosis, though they do not show whether routine genotyping will have a similar impact on the clinical management of these patients. More prospective studies are needed to clarify this.

There is a considerable variation in the results obtained by photometric measurement of s-ACE between analytical methods. This has potential implications for the ability to compare values between laboratories for clinical as well as for scientific purposes. This variation is reduced by using commercial kits, traceable calibrators and by applying external quality control programmes for laboratories [5]. The introduction of routine genotyping increases the clinical significance of small variations in s-ACE, which makes efforts to reduce between-laboratory variation even more important. Whenever I/D genotyping is introduced, genotype-based reference values should be verified with the laboratory kits used for genotyping as well as for measurement of s-ACE activity, preferably at the laboratory performing the analyses, but at least with identical kits in a comparable population group.

Genotyping will certainly incur additional costs to the investigation of sarcoidosis. This has to be taken into account when considering the rationale for performing genotyping. In this context, one approach would be to restrict genotyping to the group of patients in which the impact would expectedly be greatest; at least theoretically, this would be in persons in whom s-ACE is between the upper 97.5 percentile for the genotype II and DD since values in this interval are at the greatest risk of being misclassified as normal or elevated if genotyping is not performed. Prospective studies of the clinical impact of genotyping would help clarify the cost-effectiveness of the analysis and would also help define whether a routine or a selective genotyping approach should be chosen. The use of s-ACE in diagnosing and monitoring of sarcoidosis is challenging itself since it is neither specific, nor sensitive [21], and the clinical justification of the test is a matter of ongoing debate. Since genotyping seems to improve the accuracy of the test, this may very well be cost-effective, though no present studies clearly address this question.

As shown in Table 2, a significant difference in s-ACE between genotypes was also present among a smaller number of verified sarcoidosis cases. It has been demonstrated that the genotype DD is associated with a higher increase in s-ACE than the genotype II [22] in sarcoidosis. The data included in this review did not confirm this trend as the DD/II ratio of s-ACE was lower in sarcoidosis populations than in healthy study populations. However, this assumption is based on two studies only, and it only represents Asian patients. In neither of the two studies were the genotype-based ACE levels stratified for severity of sarcoidosis; therefore, they did not reveal whether the impact of the I/D genotype on s-ACE reflects sarcoidosis severity.

We found the I-allele to be more frequent in Asian than in Caucasian populations, which is in concordance with prior findings [23]. As there is also a well-known geographical and ethnic variation in the incidence of sarcoidosis, it has been hypothesised that the I/D polymorphism may play a role in the pathogenesis of sarcoidosis [16, 24]. Data are conflicting, but most recent studies generally do not support such a correlation [25, 26]. The function of the I/D polymorphism, though, is not clearly understood, but its location in an intronic position suggests a linkage disequilibrium with other transcription-regulating genes [27].

For genotyping of the I/D polymorphism to be used routinely, the method has to be stable and reliable. Previous data have shown that prior PCR techniques misclassified 4-5% [28, 29] of heterozygote individuals because the shorter D-allele is amplified more efficiently that the I-allele. The vast majority of these can be detected by running a confirmatory genotyping on all DD patients. A recently introduced high-resolution meltingtechnique provides more robust genotyping results, with primary HRM results performing at par with primary and confirmatory RT-PCR results in combination [12].

Certain limitations apply to this review. Since the studies were performed over a span of 23 years, the methods for genotyping and ACE measurement have changed, which makes direct comparison of studies difficult. Also, this review only included Asian and Caucasian subjects. As the I/D prevalence varies between ethnic groups, the external validity in other ethnic groups might be limited. All studies showed a significant difference of s-ACE between genotype groups. We assume that this effect is due to the I/D genotype playing a major role for the normal level of s-ACE. This effect could potentially be enforced by publication bias if studies showing no significant difference between groups have not been published. At the single study level, results could be biased by misclassification of genotypes. In all the identified studies except one [18], PCR-based genotyping (RT-PCR or conventional PCR with agarose gel electrophoresis) was confirmed with a second genotyping.

CONCLUSION

This literature search unequivocally demonstrates that among Asian and Caucasian persons, the mean s-ACE activity is significantly higher in individuals with the DD genotype than in individuals with the II genotype, with the ID genotype having intermediate values. Few studies have evaluated the impact of genotyping on the management of sarcoidosis and though more studies are clearly needed, the present data suggest that a significant amount of sarcoidosis patients are misclassified because non-genotype-based reference values are applied. Genotyping will expectedly be of greatest impact if s-ACE is between the upper 97.5 percentile for the genotypes II and DD since values in this interval are at risk of being misclassified, but clinical validation studies are needed to clarify cost-effectiveness.

Whenever implementing I/D genotyping, genotype-specific reference levels should always be verified locally due to great assay variation.

Correspondence: Andreas Fløe, Lungemedicinsk Afdeling LUB, Hjertecentret, Aarhus Universitetshospital, Nørrebrogade 44, 8000 Aarhus C, Aarhus, Denmark. E-mail: andrniel@rm.dk

Accepted: 30 January 2014

Conflicts of interest:Disclosure forms provided by the authors are available with the full text of this article at www.danmedj.dk

Acknowledgements: The study was financed by the Department of Clinical Biochemistry, Aarhus University Hospital, and Department of Pulmonary Medicine, Aarhus University Hospital. It did not involve any external funding.

Funding: The study was financed by the Department of Clinical Biochemistry, Aarhus University Hospital, Denmark, and the Department of Pulmonary Medicine, Aarhus University Hospital, Denmark. No external funding was received.