Genetics of 21-OH Deficiency and Genotype–Phenotype Correlation: Experience of the Hellenic National Referral Center

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The research investigation focusses on 21-hydroxylase deficiency (21-OHD), the most frequent kind of congenital adrenal hyperplasia (CAH), which is caused by pathogenic variants in the CYP21A2 gene. The study was to identify CYP21A2 variants in 500 Greeks suspected of having 21-OHD and to discover a biomarker that may discriminate between heterozygotes and negative genotyped patients. Genetic diagnosis confirmed 21-OHD in 27.4% of cases, with most exhibiting the non-classic form. Heterozygotes constituted 42.6%, and no pathogenic variant was found in 27% of cases. Five specific variants were discovered, and statistical analysis indicated that the difference in 17-hydroxyprogesterone levels following ACTH stimulation might be a biomarker, albeit no conclusive cut-off value was determined. The study emphasises the significance of extensive CYP21A2 genotyping for accurate diagnosis and genetic counselling. Although the article is extensive, a few discrepancies and sections might be clarified or improved.

The article redundantly repeats certain details in different sections. For example, the introduction discusses the role of 21-hydroxylase and the classification of CAH forms, which is reiterated in the results. Streamlining these sections would enhance clarity.

The introduction states an incidence of "1:200 to 1:1000" for NC 21-OHD and "1:14,000-1:18,000" for the classic type, however these ranges are huge. Clarifying the source or cause for such wide ranges would be helpful.

Furthermore, the introduction examines both the classic and NC types of CAH. However, it goes from addressing these forms to the diagnostic criteria without making a clear distinction.

The statistical relevance of Δ17-OHP60-0 as a biomarker is addressed, but not completely confirmed with defined cutoff values, limiting its usefulness. The report should either offer these numbers or clearly describe the biomarker's limitations.

The article lists 539 participants at first, then says that 39 were eliminated, leaving 500. However, there should be a clear connection made in the text between this last number and the participant distribution across age groups and the genetic findings breakdown (heterozygotes, negative instances, etc.). Additionally, "Five hundred subjects with clinical suspicion of 21-OHD" and "Genetic diagnosis was achieved in 27.4% of the probands" were included in the abstract. It's unclear, though, if the 500 individuals as a whole or only a portion of them are referred to as the "proband" here. This needs to be made clear. The elimination of 39 people on the grounds of "insufficient clinical/hormonal data" is also stated, however it is not explained. A succinct explanation of what "insufficient" means would give more context.

It would be more beneficial to emphasize the significance of the finding of five new variants and its consequences with a more thorough explanation of their possible effects. Furthermore, a more thorough examination could provide light on how these overlaps impede diagnosis and therapy.

Minor revision

 

The main conclusions are dispersed across the article's parts, giving it a somewhat fragmented organization. The results would be presented more clearly, and reading and understanding would be enhanced by a thorough discussion that came after.

 

The abstract mentions "overlapping clinical manifestations among all subjects tested" twice. This repetition could be removed to make the abstract more concise.

The text repeats the same content in two different sections. The paragraph starting with "Congenital adrenal hyperplasia (CAH) is a group of autosomal recessive disorders caused..." is duplicated. There's no need to keep this redundant.

 

Although the phrase "negative genotyped cases" is used, its definition is ambiguous. Think about substituting "cases with no pathogenic variants identified."

The sentence "using the existing hormonal assessment and genotypes to identify a biomarker that could differentiate heterozygotes from negatively genotyped cases" is quite ambiguous. It may be clarified by rephrasing it.

The phrase "0.4% (2/500) of the entire cohort tested harbored one pathogenic variant (resulting in the NC or SW form) in trans with a VUS" is unclear. It implies a relationship that could need more explanation between the prevalence of VUS and clinical manifestation.

 

The CYP21A2 gene is noted to be duplicated, however the effects of this duplication on clinical outcomes are not thoroughly investigated. In particular, the sentence "cases with duplication of the CYP21A2 gene with the p.Q319* on the same allele" lacks clarity and needs more explanation.

Author Response

Comments 1.The article redundantly repeats certain details in different sections. For example, the introduction discusses the role of 21-hydroxylase and the classification of CAH forms, which is reiterated in the results. Streamlining these sections would enhance clarity.

Response 1:We would like to thank you for your comment.

We have tried to avoid repeating among the different sections of the manuscript.

However, we should mention the different forms of 21OHD in the results section as we have to clarify the incidence of the variants according to their enzymatic activity among the different Age Groups.

Comments 2. The introduction states an incidence of "1:200 to 1:1000" for NC 21-OHD and "1:14,000-1:18,000" for the classic type, however these ranges are huge. Clarifying the source or cause for such wide ranges would be helpful.

Response 2:  We would like to thank you for your comment.

We absolutely agree that the range of the incidence is wide and we tried to explain and clarify the source.

Lines 44-46 are changed to: “The classic form, with an incidence of 1:14,000-1:18,000 live births, based on neonatal screening, is further divided into the classic salt wasting (SW) and the classic simple virilizing (SV) forms. “

Lines 53-56 are changed to: “The NC or late onset 21-OHD, is considered to be the most frequent endocrine disorder with a wide range of incidence of 1:200 to 1:1000, in various ethnic groups (Caucasians, Askenaki Jews, Yugoslavs, Mediterranean, Hispanics and Anglo-Saxons)”

Comments 3. Furthermore, the introduction examines both the classic and NC types of CAH. However, it goes from addressing these forms to the diagnostic criteria without making a clear distinction.

Response 3:  Thank you for your valuable comment.

We have now rearranged the classification paragraph and it now reads:

Lines 42-61: “Depending on the degree of enzymatic activity and consequently on the clinical manifestations, CAH is classified in two forms, the non-classic (NC) and the classic (C).

The classic form, with an incidence of 1:14,000-1:18,000 live births, based on neonatal screening, is further divided into the classic salt wasting (SW) and the classic simple virilizing (SV) forms [4]. In classic 21-OHD (SW and SV) females present with atypical/ambiguous genitalia at birth due to the intrauterine severe virilization of the external genitalia because of the prenatal excess androgen production and action, while males present with enlarged penis, that quite often escape proper medical attention. The SW form is further characterized after birth with severe salt loss, failure to thrive, potentially fatal hypovolemia and shock due to both cortisol and aldosterone insufficiency. Therefore, in both sexes, the SV form is mainly characterized by the signs of excess androgen production without salt loss since patients exhibit sufficient aldosterone production. The NC or late onset 21-OHD, is considered to be the most frequent endocrine disorder with a wide range of incidence of 1:200 to 1:1000, in various ethnic groups (Caucasians, Askenaki Jews, Yugoslavs, Mediterranean, Hispanics and Anglo-Saxons) [5,6]. NC form is manifested through a variety of clinical signs ranging from absence of clinical signs to premature adrenarche, hirsutism, acne, advanced bone age and menstrual irregularities [7].

However, the boundaries between the different forms of CAH are not strictly defined and it is therefore advisable to consider CAH as a continuum of phenotypes from asymptomatic to mild and severe [8].”

Comments 4:The statistical relevance of Δ17-OHP60-0 as a biomarker is addressed, but not completely confirmed with defined cutoff values, limiting its usefulness. The report should either offer these numbers or clearly describe the biomarker's limitations.

Response 4: We would like to thank you for your valuable comment, with which we agree.

The inability to provide cut-off values for this biomarker is a limitation of the study.

As observed in Figure 1, the median values of the cases with no pathogenic variant identified and the different groups of heterozygotes have a statistically significant difference. E.g., the median value for the normal group is 3.2, while the corresponding value for the SV heterozygotes is 9.7.  However, as shown in Figure 1, there is an overlap in the distributions of Δ17-OHP60-0  in cases with no pathogenic variant identified and each of the groups of heterozygotes; thus, Δ17-OHP60-0 cannot be used in clinical diagnosis, to differentiate heterozygotes from subjects with no pathogenic variant.

We have added a more detailed explanation of the Box Plots under Figure 1 Lines 183-186: “The horizontal black lines in every group represent the corresponding median values. As shown, although there is a statistically significant differences between the median values among the groups, there is an overlap in the range of the Δ17OHP60-0 values in the cases with no pathogenic variants with each group of the heterozygotes.”

We have also added in the Discussion Section lines 288-292: “This overlap in the values of Δ17-OHP60-0 of the Group of cases with no pathogenic variants and the different Groups of heterozygotes shows that this biomarker is not suitable to differentiate heterozygotes from the cases with no pathogenic variants identified thus it cannot be used in clinical practice.”

 

Comments 5. The article lists 539 participants at first, then says that 39 were eliminated, leaving 500. However, there should be a clear connection made in the text between this last number and the participant distribution across age groups and the genetic findings breakdown (heterozygotes, negative instances, etc.). Additionally, "Five hundred subjects with clinical suspicion of 21-OHD" and "Genetic diagnosis was achieved in 27.4% of the probands" were included in the abstract. It's unclear, though, if the 500 individuals as a whole or only a portion of them are referred to as the "proband" here. This needs to be made clear. The elimination of 39 people on the grounds of "insufficient clinical/hormonal data" is also stated, however it is not explained. A succinct explanation of what "insufficient" means would give more context.

Response 5: Thank you for your valuable comments.

It is now specified in the statistical analysis section (line 155), the results section (line 171) and the discussion (line 277) that the analysis was carried out for the 500 cases fulfilling the criteria for CYP21A2 genotyping.

The total number of cases has also been added in each age group participant number (lines 115-117)

The word proband is replaced by “subjects tested” throughout the manuscript.

The exclusion criteria for the 39 cases are now added in Section 2.1 Lines:110-113.

 

Comments 6. It would be more beneficial to emphasize the significance of the finding of five new variants and its consequences with a more thorough explanation of their possible effects. Furthermore, a more thorough examination could provide light on how these overlaps impede diagnosis and therapy.

Response 6: Thank you for your comment

We have now changed the interpretation and we state the clinical profile of the cases carrying these novel variants.

Discussion section, Lines 339-356: “Genotyping the CYP21A2 gene led to the identification of 5 novel variants. Three of them located in the promoter region, which is crucial for the transcriptional activity of the gene, the c.-127G>A and c.-82C>T in heterozygosity and the c.-115G>T in compound heterozygosity with the pathogenic variant p.P31L, were classified as VUS according to the ACMG criteria due to insufficient or conflicting evidence. The other two novel variants were located in exon 7, the p.R255K in heterozygosity being classified as VUS and the p.V282M in compound heterozygosity with p.L308Ffs*6, that was classified as Likely Pathogenic. Considering the hormonal data of the patients carrying these novel variants (presented in Table 2) only the p.V282M variant certainly contributes to the patient’s phenotype  and this patient was clinically considered and treated as non-classic form. It should be noted that the V282 residue is located in a very restricted space at the I-helix and we can assume that the substitution of Valine by Methionine could have the same effect as the substitution of Valine by Leucine due to the increased chain length of both Methionine and Leucine when compared to Valine [47]. The patient carrying the c.-115G>T in compound heterozygosity with the p.P31L, presented increased 17OHP concentrations with normal cortisol values, and was therefore considered as heterozygote, since there was no evidence for the effect of the VUS variant. The 3 remaining VUS variants were only present in heterozygosity, therefore irrespective of their potential impact on the enzymatic activity, their clinical significance is negligible.”

 

Minor Revision

 

 Comments 7. The main conclusions are dispersed across the article's parts, giving it a somewhat fragmented organization. The results would be presented more clearly, and reading and understanding would be enhanced by a thorough discussion that came after.

Response 7: We thank you for this constructive remark. Indeed, in this revised version we tried to organize our results more clearly by presenting first what is already known, for example the correlation of the severity of the enzymatic defect with the age of clinical presentation and then discussing the inability to clearly differentiate homozygotes from heterozygotes and unaffected individuals merely based on hormonal determinations. We hope that by restructuring our discussion section the reader can now follow the scope of the manuscript while still acknowledging the limitations of our current clinical and laboratory practice. We consider that presenting these still existing limitations despite the huge literature on this topic we can contribute to the discussion whether biochemical or hormonal assessments could be sufficient for providing the diagnosis to the affected individuals and their families.

 

Comments 8. The abstract mentions "overlapping clinical manifestations among all subjects tested" twice. This repetition could be removed to make the abstract more concise.

Response 8: Thank you for your comment.

The repetition is removed from the abstract and it now reads:

Lines 22-26: “Further analysis revealed overlapping clinical manifestations among all subjects tested. The presented phenotypic traits of the subjects tested and the inability to identify a discriminative biochemical marker highlight the importance of comprehensive CYP21A2 genotyping to ascertain the correct genetic diagnosis and proper genetic counselling.”

 

Comments 9. The text repeats the same content in two different sections. The paragraph starting with "Congenital adrenal hyperplasia (CAH) is a group of autosomal recessive disorders caused..." is duplicated. There's no need to keep this redundant.

Response 9: Thank you for your comment.

In case this point refers to line 72 the text is changed now to: “Most pathogenic variants (75%) in the CYP21A2 gene derive……”

Comments 10. Although the phrase "negative genotyped cases" is used, its definition is ambiguous. Think about substituting "cases with no pathogenic variants identified."

Response 10: Thank you for your comment.

We edited the manuscript and "cases with no pathogenic variants identified” substituted the “negative genotyped cases" throughout the text.

Comments 11. The sentence "using the existing hormonal assessment and genotypes to identify a biomarker that could differentiate heterozygotes from negatively genotyped cases" is quite ambiguous. It may be clarified by rephrasing it.

Response 11: Thank you for your comment.

Lines 12-15 has been changed to: “Τhe aim of this study is the identification of CYP21A2 variants in 500 subjects of Greek origin with suspicion of 21-OHD and, by using the existing hormonal assessment and genotypes of the 500 subjects tested, to identify a biomarker that could differentiate heterozygotes and cases with no pathogenic variants identified”.

 

Comments 12. The phrase "0.4% (2/500) of the entire cohort tested harbored one pathogenic variant (resulting in the NC or SW form) in trans with a VUS" is unclear. It implies a relationship that could need more explanation between the prevalence of VUS and clinical manifestation.

Response 12: Thank you for your comment.

We agree with you and we understand that this sentence is rather confusing, so we removed it from the result section.

We elaborate on these findings (novel VUS) in the discussion, lines 339-356.

Comments 13. The CYP21A2 gene is noted to be duplicated, however the effects of this duplication on clinical outcomes are not thoroughly investigated. In particular, the sentence "cases with duplication of the CYP21A2 gene with the p.Q319* on the same allele" lacks clarity and needs more explanation.

Response 13: Thank you for your comment.

We introduced a small paragraph (lines 81-87) in order to clarify the duplication of the CYP21A2 gene along with the p.Q319* and its effect on clinical outcomes. The corresponding references have also been added.

Reviewer 2 Report

Comments and Suggestions for Authors

I have a minor suggestion of correction in Table A1, in which it is described "Group Null variant in trans with a Group A variant (2) cases" => This should be corrected to "Group Null variant in trans with a Group A variant (2 cases)". 

Author Response

Comments 1. I have a minor suggestion of correction in Table A1, in which it is described "Group Null variant in trans with a Group A variant (2) cases" => This should be corrected to "Group Null variant in trans with a Group A variant (2 cases)".

Response 1: Thank you for your comment. We edited the text as suggested in Table A1.

 

Reviewer 3 Report

Comments and Suggestions for Authors

The article titled “Genetics of 21-OH deficiency and genotype-phenotype correlation: Experience of the Hellenic national referral center” by Fylaktou et al. studies CYP21A2 genotyping in 500 Greek subjects with suspected 21-hydroxylase deficiency (21-OHD). The authors implemented a multi-faceted approach, including clinical assessments, hormonal profiling, DNA extraction, PCR-based gene sequencing, and MLPA (Multiplex Ligation-dependent Probe Amplification). The study aims to identify pathogenic variants associated with 21-OHD and an exploration of genotype-phenotype correlations. Detailed comments are as follows:

1.        The study's failure to establish a clear cut-off value for the Δ17-OHP60-0 biomarker raises concerns about the validity of the statistical conclusions. The overlap in the Δ17-OHP60-0 values between heterozygotes and negative genotyped subjects suggests this marker is not ideal. This limitation undermines the strength of the study’s findings regarding the potential use of Δ17-OHP60-0 as a biomarker.

2. Excluding cases with novel variants from the statistical analysis could lead to a loss of valuable insights. These novel variants might have provided important information on the variability of the CYP21A2 gene and its association with clinical outcomes.

3.        The exclusion of cases with novel variants from the statistical analysis could lead to a loss of valuable insights. These novel variants might have provided important information on the variability of the CYP21A2 gene and its association with clinical outcomes. Can the authors elaborate more on this?

4.        The data presentation could be improved by providing more context for the statistical analyses. For example, the box plots in Figure 1 could benefit from additional information on the clinical significance of the findings in addition to just the statistical significance.

5.        The discussion could be expanded to explore alternative explanations for the findings, such as the potential influence of other genetic or environmental factors not considered in the study.

Addressing these concerns would significantly enhance the manuscript and make it more suitable for publication.

 

Comments on the Quality of English Language

The quality of the English language used is generally acceptable, with scientific concepts presented clearly and unambiguously. 

Author Response

Comments 1.The study's failure to establish a clear cut-off value for the Δ17-OHP60-0 biomarker raises concerns about the validity of the statistical conclusions. The overlap in the Δ17-OHP60-0 values between heterozygotes and negative genotyped subjects suggests this marker is not ideal. This limitation undermines the strength of the study’s findings regarding the potential use of Δ17-OHP60-0 as a biomarker.

Response 1: We thank you for your valuable comment with which we agree.

We do understand that this biomarker is not ideal (and this is clearly stated in the Discussion Section, Lines 292-302, as one of the limitations of this biomarker). Nevertheless, it was the only biomarker that had a statistically significant difference between the groups of interest. Hence, we believe it is useful for the clinicians to have this information that we present in the manuscript even with its limitations.

Comments 2. Excluding cases with novel variants from the statistical analysis could lead to a loss of valuable insights. These novel variants might have provided important information on the variability of the CYP21A2 gene and its association with clinical outcomes.

The exclusion of cases with novel variants from the statistical analysis could lead to a loss of valuable insights. These novel variants might have provided important information on the variability of the CYP21A2 gene and its association with clinical outcomes. Can the authors elaborate more on this?

Response 2: We thank you for your comment with which we absolutely agree.

The reason why novel variants identified in this study were excluded from our statistical analysis was their unknown effect on the enzymatic activity of 21-OHD.

However, we have now changed the interpretation and we state the clinical profile of the cases carrying these novel variants.

Discussion, Lines 339-356: “Genotyping the CYP21A2 gene led to the identification of 5 novel variants. Three of them located in the promoter region, which is crucial for the transcriptional activity of the gene, the c.-127G>A and c.-82C>T in heterozygosity and the c.-115G>T in compound heterozygosity with the pathogenic variant p.P31L, were classified as VUS according to the ACMG criteria due to insufficient or conflicting evidence. The other two novel variants were located in exon 7, the p.R255K in heterozygosity being classified as VUS and the p.V282M in compound heterozygosity with p.L308Ffs*6, that was classified as Likely Pathogenic. Considering the hormonal data of the patients carrying these novel variants (presented in Table 2) only the p.V282M variant certainly contributes to the patient’s phenotype and this patient was clinically considered and treated as non-classic form. It should be noted that the V282 residue is located in a very restricted space at the I-helix and we can assume that the substitution of Valine by Methionine could have the same effect as the substitution of Valine by Leucine due to the increased chain length of both Methionine and Leucine when compared to Valine [47]. The patient carrying the c.-115G>T in compound heterozygosity with the p.P31L, presented increased 17OHP concentrations with normal cortisol values, and was therefore considered as heterozygote, since there was no evidence for the effect of the VUS variant. The 3 remaining VUS variants were only present in heterozygosity, therefore irrespective of their potential impact on the enzymatic activity, their clinical significance is negligible.”

 

Comments  3. The data presentation could be improved by providing more context for the statistical analyses. For example, the box plots in Figure 1 could benefit from additional information on the clinical significance of the findings in addition to just the statistical significance.

Response 3:  We thank you for your valuable comment.

We have now added a more detailed explanation of the Box Plots under Figure 1 Lines 183-186: “The horizontal black lines in every group represent the corresponding median values. As shown, although there is a statistically significant differences between the median values among the groups, there is an overlap in the range of the Δ17OHP^60-0 values in the cases with no pathogenic variants with each group of the heterozygotes.”

We have also added in the Discussion Section lines 288-291: “This overlap in the values of Δ17-OHP60-0 of the Group of cases with no pathogenic variants and the different Groups of heterozygotes shows that this biomarker is not suitable to differentiate heterozygotes from the cases with no pathogenic variants identified thus it cannot be used in clinical practice.”

Comments 4. The discussion could be expanded to explore alternative explanations for the findings, such as the potential influence of other genetic or environmental factors not considered in the study.

Response 4:  We thank you for your valuable comment

We have added the following:

Discussion section, Lines 316-324: “The number of cases without any identifiable pathogenic variants reported in this study is not negligible and this could be explained by the fact that CAH-related clinical features may not exclusively be indicative of 21-OHD. One of the main clinical manifestations observed in Age Group II and III was premature adrenarche that can be also the result of several other factors such as childhood obesity or being born IUGR (Intrauterine Growth restricted).

Moreover, pathogenic variants in other CAH genes involved in the steroidogenic pathway, the presence of other clinical entities (such as polycystic ovarian syndrome), epigenetic influences as well as environmental factors might also account for  phenotypic characteristics indicative of CAH.”

 

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The article has been extensively revised following the suggestions and eliminating any possible inconsistencies. In my opinion, there are no further modifications to be made.