Neonatal Screening for CAH in Sweden—Results of Implementing Second-Tier Testing
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsSummary
This manuscript describes the adoption in Sweden of a second tier in newborn screening (NBS) for 21-hydroxylase deficient congenital adrenal hyperplasia (CAH). The authors begin by describing the major flaw in CAH NBS of having too many false positive screens when using only a single tier of testing 17-hydroxyprogesterone (17-OHP). The second tier they set up was using liquid chromatography with tandem mass spectrometry (LC-MS/MS) to measure other adrenal gland products.
In a retrospective analysis of stored dried blood spots (DBS) they found that the combination of two parameters, “steroid hormone ratio (androstenedione+17-hydroxyprogesterone)/cortisol and the concentration of 21-deoxycortisol” provide the best positive predictive value (PPV) without any apparent loss of sensitivity. They report increasing the PPV from 14% to 84% in full-term infants and decreasing the false positive results by 88%. CYP21A2 genotyping was done to classify babies into the categories of CAH, salt-wasting (SW), simple virilizing (SV) and late-onset.
Strengths
The LC-MS/MS technique developed in the author’s lab is described in sufficient detail. The approach using a large sample size over several years makes the results more valuable. The reduced false positive rate and increased PPV achieved are commendable.
Areas for Improvement
The manuscript does not make the target of their CAH NBS clear. Is it limited to SW CAH, all classic CAH, or classic and non-classic cases. For example, it is not clear in the first paragraph of 3.2 whether the 49 “confirmed CAH-cases” all had SW CAH. That affects the interpretation of the rest of the section. My opinion is that infants with non-classic CAH should be explicitly excluded from all of the results, with a statement to that effect in the Methods. It is mentioned in the Discussion, lines 308-309.
The manuscript is not consistent enough regarding preterm infants (less than 37 weeks gestational age). In line 100, the approach for preterm infants needs to be stated. In the first paragraph of 3.2, were any of the 49 + 155 + 46 newborns preterm?
The manuscript is not clear about case ascertainment. For example, lines 280-281 say, “based on the severity of the disease, based on their CYP21A2 genotype”. Is the classification based only on genotype or on some combination of genotype and phenotype. At best they are 90% in agreement. (My reference 1 below.) In section 3.6 and Figure 5, it is clear that only the genotype is used to classify severity.
It is not clear to me where the pilot and retrospective study end and the application of the findings start. Line 293 states that CAH NBS “was clearly improved” using the new algorithm, but I think the results only show that it can be improved. I did not see any prospective data showing a lower recall rate and higher positive predictive value without any missed cases. The extrapolation of the retrospective data is particularly concerning in the absence of true positive cases among the preterm infants as shown in Figure 4A. How can you know that no cases would be missed when there were no positives? The Discussion states, “In the three years follow up with the new algorithm we have not identified any missed cases.” Are you using the same case ascertainment methods described in reference 15? Are those data included in the Results or are the Results just from the retrospective analysis done by adding second-tier testing? In line 65, is the PPV for all three forms of CAH or just the target disease, SW CAH?
In Figure 2, if the numbers under each diagnostic category are the range of observed results, they do not match the figure. For example the 17OHP FP graph shows a point at about 250, but the range stated is “2.0-62”. Likewise for 21-deoxycortisol and TP and the ratio for FP. The Figure 2 caption should describe what the numbers represent.
If you have data on steroid stability in stored DBS, it should be stated in the results. If not, then others’ information about stability should be stated in the last paragraph of the Methods 2.2.
Minor Points
The Figures need to be renumbered so that they are mentioned sequentially in the text.
I think that “salt crisis” should be “SW crisis” in several places. SW is appropriately defined early in the manuscript.
“Gestational week” (GW) should be changed to “Gestational age” (GA) throughout to conform with common usage. That would also require adding “weeks” after the numbers and changing “in gestational week …” to “at gestational age…”
lines 46-48 should also include “short adult stature”. It is a large focus of CAH treatment.
lines 114-115 reporting screening of older children are irrelevant to this report and should be removed. Hopefully, the results from any older children are not included.
line 208 – “why another parameter” should be “which is why another parameter”.
line 250 and Figure 4 caption and other places – “rerun limit” should probably be changed to “cutoff” or “first-tier cutoff”.
Parts of section 3.2 are Methods and should be moved there.
3.4.1 I was confused by “Some of the remaining false-positive results had a ratio above the cut-off despite a low concentration of 17-OHP due to a low cortisol level.” After a few readings I understood that you mean, “Despite a low concentration of 17-OHP, some of the remaining false-positive results had a ratio above the cut-off due to a low cortisol level.” The low 17-OHP was not due to the low cortisol.
3.4.3 It is surprising that you retain DBS specimens from 1986 and remarkable that you detected steroids in those specimens despite storage at 4 degrees C.
Figure 2 “(17-OHP+A4/F) should be ((17-OHP+A4)/F). It is correctly shown that way everywhere else in the manuscript.
Figure 5 typo: “21-deoxykortisol” should be “21-deoxycortisol”
line 330 – “21-deoxycortison” should be “21-deoxycortisone”
supplementary figure S1 – The caption should include that the samples were only from those with a gestational age of 35 or 36 weeks and “weeks” should be added to “35” and “36” in the key. Conventionally descriptions of x-y plots describe them as variable Y plotted against variable X. The caption (and the text in the manuscript) describe it the other way, which may be confusing. “Correlates to” is better expressed as “represents”.
genotype/phenotype
- Atas N, Karaoglan M, Nacarkahya G. Genotype-Phenotype Correlation in Children With Congenital Adrenal Hyperplasia due to 21-Hydroxylase Deficiency Using Next Generation Sequencing. Mol Genet Genomic Med. 2025 Jun;13(6):e70110. doi: 10.1002/mgg3.70110. PMID: 40552539; PMCID: PMC12186092.
The overall genotype–phenotype concordance was found to be 73.1% (68/93). The expected concordance with the SW form of the null (n = 12) and A (n = 51) groups is 91.6% and 88.2%, respectively. While the percentage of the expected clinical form of SV in patients in group B (n = 5) was 80%, the concordance for the expected clinical form of NC for group C (n = 25) was not strong enough (32%).
Author Response
Response to reviewer 1
We thank the reviewers for constructive and helpful comments. We think that the manuscript has improved with the changes.
Summary
This manuscript describes the adoption in Sweden of a second tier in newborn screening (NBS) for 21-hydroxylase deficient congenital adrenal hyperplasia (CAH). The authors begin by describing the major flaw in CAH NBS of having too many false positive screens when using only a single tier of testing 17-hydroxyprogesterone (17-OHP). The second tier they set up was using liquid chromatography with tandem mass spectrometry (LC-MS/MS) to measure other adrenal gland products.
In a retrospective analysis of stored dried blood spots (DBS) they found that the combination of two parameters, “steroid hormone ratio (androstenedione+17-hydroxyprogesterone)/cortisol and the concentration of 21-deoxycortisol” provide the best positive predictive value (PPV) without any apparent loss of sensitivity. They report increasing the PPV from 14% to 84% in full-term infants and decreasing the false positive results by 88%. CYP21A2 genotyping was done to classify babies into the categories of CAH, salt-wasting (SW), simple virilizing (SV) and late-onset.
Strengths
The LC-MS/MS technique developed in the author’s lab is described in sufficient detail. The approach using a large sample size over several years makes the results more valuable. The reduced false positive rate and increased PPV achieved are commendable.
Areas for Improvement
The manuscript does not make the target of their CAH NBS clear. Is it limited to SW CAH, all classic CAH, or classic and non-classic cases. For example, it is not clear in the first paragraph of 3.2 whether the 49 “confirmed CAH-cases” all had SW CAH. That affects the interpretation of the rest of the section. My opinion is that infants with non-classic CAH should be explicitly excluded from all of the results, with a statement to that effect in the Methods. It is mentioned in the Discussion, lines 308-309.
Response
The primary aim of the screening program is to prevent neonatal salt crises and to identify classic CAH. We do not consider a not identified individual with non-classic CAH as a missed case. However, if a child with non-classic CAH is identified in the screening, we do think it may benefit to the child to have this knowledge. There are large variations in symptoms, from early androgen excess to no symptoms at all but among those that are identified in the screening there may be other factors affecting the hormone metabolism making them more prone to develop symptoms relatively early in life. We give instructions to the pediatric unit in charge about the possibility to avoid/postpone treatment if there are no symptoms. It is therefore not correct in a strict sense to call them false positive. Hence, we report the patients and consider this to be a secondary benefit of the screening. We have changed the wording to make this more clear.
The manuscript is not consistent enough regarding preterm infants (less than 37 weeks gestational age). In line 100, the approach for preterm infants needs to be stated. In the first paragraph of 3.2, were any of the 49 + 155 + 46 newborns preterm?
Response
Thank you for your comments. We have specified that separate approaches were used for those born GA < 34 or GA 35-36. We have now clarified this in the text.
The 49 (TP) + 155 (FP) + 46 (normal) newborn infants represent both full-term and preterm born infants. Among the 49 true positives 2 were born preterm. 78 of the 155 false positives were born preterm and 3 of the 46 controls (randomly sampled non-CAH healthy newborn) were preterm.
The manuscript is not clear about case ascertainment. For example, lines 280-281 say, “based on the severity of the disease, based on their CYP21A2 genotype”. Is the classification based only on genotype or on some combination of genotype and phenotype. At best they are 90% in agreement. (My reference 1 below.) In section 3.6 and Figure 5, it is clear that only the genotype is used to classify severity.
Response
Thank you for this comment. In our hands the CYP21A2 genotype and the use of genotype groups for dividing the patients into severity groups is consistent, with very few exceptions. In addition, this does not involve the judgement by different physicians and can be viewed as non biased. This approach has worked very well in all our previous studies, also on national population-based studies. In case of less common mutations the enzyme activity was used for the classification. This has now been clarified in the methods section and lines 280-281.
It is not clear to me where the pilot and retrospective study end and the application of the findings start. Line 293 states that CAH NBS “was clearly improved” using the new algorithm, but I think the results only show that it can be improved. I did not see any prospective data showing a lower recall rate and higher positive predictive value without any missed cases.
Response
For the full term infants, the PPV was calculated using the period with the algorithms B and C, which were the same during this whole period for full term infants. The recall rate was considerably lower resulting in the higher PPV but with a similar number of identified cases over this 3-year period. We did not identify any missed cases but as we mention in the discussion this cannot be ruled out until decades in the future.
The extrapolation of the retrospective data is particularly concerning in the absence of true positive cases among the preterm infants as shown in Figure 4A. How can you know that no cases would be missed when there were no positives? The Discussion states, “In the three years follow up with the new algorithm we have not identified any missed cases.” Are you using the same case ascertainment methods described in reference 15? Are those data included in the Results or are the Results just from the retrospective analysis done by adding second-tier testing? In line 65, is the PPV for all three forms of CAH or just the target disease, SW CAH?
Response
We are using the same case ascertainment as we used in ref 15, Gidlöf et al. The retrospective analyses of the previously identified CAH patients were used to set the referral levels. That data was not included in the calculations of recall or PPV. The data presented in Figure 4 are only the data from the 3 years of screening with the new algorithms (B and C).
In Figure 2, if the numbers under each diagnostic category are the range of observed results, they do not match the figure. For example the 17OHP FP graph shows a point at about 250, but the range stated is “2.0-62”. Likewise for 21-deoxycortisol and TP and the ratio for FP. The Figure 2 caption should describe what the numbers represent.
Response
Thank you for identifying this error. We have now corrected the numbers. We have also described what the numbers represent in the caption.
If you have data on steroid stability in stored DBS, it should be stated in the results. If not, then others’ information about stability should be stated in the last paragraph of the Methods 2.2.
Response
Thank you for this comment, we have now included this information.
Minor Points
The Figures need to be renumbered so that they are mentioned sequentially in the text.
We apologise for this mistake and are now not referring to the algorithms Figure 3 in the methods section.
I think that “salt crisis” should be “SW crisis” in several places. SW is appropriately defined early in the manuscript.
Salt-crises is a general description and can occur also in SV CAH why we think it is more correct to use SW only to describe the SW form of CAH. We have therefore not made changes.
“Gestational week” (GW) should be changed to “Gestational age” (GA) throughout to conform with common usage. That would also require adding “weeks” after the numbers and changing “in gestational week …” to “at gestational age…”
Thank you for this comment, we have made the appropriate changes though out the manuscript.
lines 46-48 should also include “short adult stature”. It is a large focus of CAH treatment.
This has now been included.
lines 114-115 reporting screening of older children are irrelevant to this report and should be removed. Hopefully, the results from any older children are not included.
We have removed the sentence about older children. Results from older children are not included in the manuscript.
line 208 – “why another parameter” should be “which is why another parameter”.
Thank you, this has been amended
line 250 and Figure 4 caption and other places – “rerun limit” should probably be changed to “cutoff” or “first-tier cutoff”.
Thank you, this has been amended
Parts of section 3.2 are Methods and should be moved there.
Thank you for this comment. We have now made it more clear in the methods section that the DBS cards of previously identified patients with CAH were retrospectively analyzed with the second-tier method. We think that it improves the readability of the procedure in 3.2 to state this specifically again,
3.4.1 I was confused by “Some of the remaining false-positive results had a ratio above the cut-off despite a low concentration of 17-OHP due to a low cortisol level.” After a few readings I understood that you mean, “Despite a low concentration of 17-OHP, some of the remaining false-positive results had a ratio above the cut-off due to a low cortisol level.” The low 17-OHP was not due to the low cortisol.
Thank you for this comment, the wording has now been changed.
3.4.3 It is surprising that you retain DBS specimens from 1986 and remarkable that you detected steroids in those specimens despite storage at 4 degrees C.
The Swedish PKU-biobank contains the DBS screening cards of all screened newborns since 1975. The stability of many of the analytes is surprisingly good.
Figure 2 “(17-OHP+A4/F) should be ((17-OHP+A4)/F). It is correctly shown that way everywhere else in the manuscript.
Thank you, this has been corrected
Figure 5 typo: “21-deoxykortisol” should be “21-deoxycortisol”
Thank you, this has been corrected
line 330 – “21-deoxycortison” should be “21-deoxycortisone”
Thank you, this has been corrected
supplementary figure S1 – The caption should include that the samples were only from those with a gestational age of 35 or 36 weeks and “weeks” should be added to “35” and “36” in the key. Conventionally descriptions of x-y plots describe them as variable Y plotted against variable X. The caption (and the text in the manuscript) describe it the other way, which may be confusing. “Correlates to” is better expressed as “represents”.
Thank you for this comment. The caption as well as the text in the manuscript has been updated accordingly.
genotype/phenotype
- Atas N, Karaoglan M, Nacarkahya G. Genotype-Phenotype Correlation in Children With Congenital Adrenal Hyperplasia due to 21-Hydroxylase Deficiency Using Next Generation Sequencing. Mol Genet Genomic Med. 2025 Jun;13(6):e70110. doi: 10.1002/mgg3.70110. PMID: 40552539; PMCID: PMC12186092.
The overall genotype–phenotype concordance was found to be 73.1% (68/93). The expected concordance with the SW form of the null (n = 12) and A (n = 51) groups is 91.6% and 88.2%, respectively. While the percentage of the expected clinical form of SV in patients in group B (n = 5) was 80%, the concordance for the expected clinical form of NC for group C (n = 25) was not strong enough (32%).
Response
See response above
Response to reviewer 2
We thank the reviewers for constructive and helpful comments. We think that the manuscript has improved with the changes.
Reviewer 2
The reviewer commends the authors for their valuable contribution regarding the implementation of second tier testing within Sweden’s neonatal screening program for CAH. The following points are provided to further enhance the quality and clarity of the manuscript:
- Title Revision: It is recommended that the title be amended to: Neonatal screening for CAH in Sweden. Results of implementing second tier testing.
The title has been changed.
- Abbreviation Usage: According to the IJNS author guidelines, abbreviations should be defined at their initial appearance in three sections: the abstract, the main text, and the first figure or table. The reviewer suggests a thorough review of abbreviation usage throughout the manuscript. For instance, certain abbreviations are defined repeatedly in the main text—particularly within the Introduction and Materials and Methods sections (e.g., DBS, cortisol (F), androstenedione (A4)).
Thank you for pointing this out. We have now gone through and made the corrections of the abbreviations.
- Figure Numbering: It is advisable to modify the figure numbering to reflect the order in which figures are first mentioned. Specifically, Figure 3 is the initial figure referenced in the Materials and Methods section and therefore should be renumbered as Figure 1.
We apologise for this mistake and have now corrected the numbering by omitting a reference to Figure 3 in the methods section.
- Additional Minor Edits:
- Abstract, first line: I suggest replacing “Newborn screening for Congenital” by “Newborn screening for congenital.”
- Thank you, this has been changed
- Materials and Methods, Subsection 2.1: I suggest replacing “salt-wasting congenital adrenal hyperplasia” with “salt-wasting CAH.”
- Thank you, this has been changed
- Materials and Methods: The supplier Perkin Elmer should be updated to Revvity Inc. in subsections 2.2 and 2.4.
- Thank you, this has been changed
- Results, Subsection 3.1, first line: I suggest replacing “5 steroid hormones” by “five steroid hormones.”
- Thank you, this has been changed
Author Response File:
Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsThe reviewer commends the authors for their valuable contribution regarding the implementation of second tier testing within Sweden’s neonatal screening program for CAH. The following points are provided to further enhance the quality and clarity of the manuscript:
- Title Revision: It is recommended that the title be amended to: Neonatal screening for CAH in Sweden. Results of implementing second tier testing.
- Abbreviation Usage: According to the IJNS author guidelines, abbreviations should be defined at their initial appearance in three sections: the abstract, the main text, and the first figure or table. The reviewer suggests a thorough review of abbreviation usage throughout the manuscript. For instance, certain abbreviations are defined repeatedly in the main text—particularly within the Introduction and Materials and Methods sections (e.g., DBS, cortisol (F), androstenedione (A4)).
- Figure Numbering: It is advisable to modify the figure numbering to reflect the order in which figures are first mentioned. Specifically, Figure 3 is the initial figure referenced in the Materials and Methods section and therefore should be renumbered as Figure 1.
- Additional Minor Edits:
- Abstract, first line: I suggest replacing “Newborn screening for Congenital” by “Newborn screening for congenital.”
- Materials and Methods, Subsection 2.1: I suggest replacing “salt-wasting congenital adrenal hyperplasia” with “salt-wasting CAH.”
- Materials and Methods: The supplier Perkin Elmer should be updated to Revvity Inc. in subsections 2.2 and 2.4.
- Results, Subsection 3.1, first line: I suggest replacing “5 steroid hormones” by “five steroid hormones.”
Author Response
Response to reviewer 2
We thank the reviewers for constructive and helpful comments. We think that the manuscript has improved with the changes.
Reviewer 2
The reviewer commends the authors for their valuable contribution regarding the implementation of second tier testing within Sweden’s neonatal screening program for CAH. The following points are provided to further enhance the quality and clarity of the manuscript:
- Title Revision: It is recommended that the title be amended to: Neonatal screening for CAH in Sweden. Results of implementing second tier testing.
The title has been changed.
- Abbreviation Usage: According to the IJNS author guidelines, abbreviations should be defined at their initial appearance in three sections: the abstract, the main text, and the first figure or table. The reviewer suggests a thorough review of abbreviation usage throughout the manuscript. For instance, certain abbreviations are defined repeatedly in the main text—particularly within the Introduction and Materials and Methods sections (e.g., DBS, cortisol (F), androstenedione (A4)).
Thank you for pointing this out. We have now gone through and made the corrections of the abbreviations.
- Figure Numbering: It is advisable to modify the figure numbering to reflect the order in which figures are first mentioned. Specifically, Figure 3 is the initial figure referenced in the Materials and Methods section and therefore should be renumbered as Figure 1.
We apologise for this mistake and have now corrected the numbering by omitting a reference to Figure 3 in the methods section.
- Additional Minor Edits:
- Abstract, first line: I suggest replacing “Newborn screening for Congenital” by “Newborn screening for congenital.”
- Thank you, this has been changed
- Materials and Methods, Subsection 2.1: I suggest replacing “salt-wasting congenital adrenal hyperplasia” with “salt-wasting CAH.”
- Thank you, this has been changed
- Materials and Methods: The supplier Perkin Elmer should be updated to Revvity Inc. in subsections 2.2 and 2.4.
- Thank you, this has been changed
- Results, Subsection 3.1, first line: I suggest replacing “5 steroid hormones” by “five steroid hormones.”
- Thank you, this has been changed
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsSummary
This manuscript describes the adoption in Sweden of a second tier in newborn screening (NBS) for 21-hydroxylase deficient congenital adrenal hyperplasia (CAH). The second tier uses liquid chromatography with tandem mass spectrometry (LC-MS/MS) to measure other adrenal gland products.
In a retrospective analysis of stored dried blood spots (DBS) they found that the combination of two parameters, provide the best positive predictive value (PPV) without any apparent loss of sensitivity. They report increasing the PPV from 14% to 84% in full-term infants and decreasing the false positive results by 88%.
Strengths
The LC-MS/MS technique developed in the author’s lab is described in sufficient detail. The approach using a large sample size over several years makes the results more valuable. The reduced false positive rate and increased PPV achieved are commendable.
Areas for Improvement
It is not clear to me whether after initiation of the 2nd tier assay the first tier assay was re-run on specimens whose 1st tier was abnormal. (Lines 137-138) If so, how were the re-run results used? It is not shown in Figure 3.
I do not understand why a criterion of “17-OHP >= 10 nM/L blood” was added into Algorithm C when the first tier test include a criterion of “17-OHP >= 50 nM/L blood”. Was there that large a discrepancy between the two assay techniques?
After several re-readings, the last paragraph of 3.4.2 does not make sense to me. Why was this comparison only done in those 35-36 weeks gestational age? I do not understand the comparison being made between the second and third sentences implied by the word “instead”. Why is the new cut-off of 80 nmol/L not the same as that shown in Figure 3C? What assurance is there that increasing the first tier cut-off did not reduce sensitivity?
Minor Points
In 3.4.2 “gestational age” is abbreviated as GA, but in at least two other places “GW” is used. I believe “GA” is the preferred abbreviation.
The Discussion should include the fact that specimens from 2015 were tested without confirmation of analyte stability for that long.
Presumably a multi-year population with 49 true positive specimens would have had more than 155 false positives. How were the false positive specimens that underwent retrospective second tier testing chosen? The “random healthy newborns” included in the analysis need to be mentioned in the Methods.
The y-axis label for the ratios (6th graph) of Figure 2 seems to be incorrect. A ratio is not expressed as “conc(nmol/L)”. Hopefully, the data shown are ratios. Similarly on that graph, the lower limits are unlikely to be zero, but should be stated as <0.x, the lower limit of detection of the numerator products divided by the measured denominator.
Line 245. Think the word “even” changes the meaning of the sentence and should be removed.
In the Discussion, line 308 does not make sense to me. Even two-tier CAH NBS does not “detect” NC CAH. It detects a baby at high risk for having CAH whose specific diagnosis is made by referring to a specialist.
In the Discussion, line 339 should not restrict the possibility of missed cases to those with SV CAH. SW cases might also be missed. The last two words of the sentence should be omitted.
The caption for Supplemental Figure S1 includes “The horizontal line” but there is no horizontal line.
The abbreviation for nanomoles per liter is not the same in Figure 3 as in the text.
The group labels for cortisol in figure 2 use commas instead of periods. Please be consistent.
Author Response
Response to Reviewers
Response to Reviewer 1
Areas for Improvement
It is not clear to me whether after initiation of the 2nd tier assay the first tier assay was re-run on specimens whose 1st tier was abnormal. (Lines 137-138) If so, how were the re-run results used? It is not shown in Figure 3.
Response
Thank you for this comment. Yes, the first-tier assay was re-run even after implementation of the 2nd-tier assay. The re-run results are used to validate that the initial screening result was correct, and with no mix-up of samples. We have added this information in section 2.4
I do not understand why a criterion of “17-OHP >= 10 nM/L blood” was added into Algorithm C when the first tier test include a criterion of “17-OHP >= 50 nM/L blood”. Was there that large a discrepancy between the two assay techniques?
Response
Thank you for your question. The first-tier criterion is 17-OHP >= 50 nM/L serum”, not blood. The difference in units accounts for half the difference, and most samples correlate well with a factor of about 2 in relation between serum and blood. The first-tier immunoassay is non-specific due to cross-reactivity with other steroids present as a result of metabolic stress. This is the reason that some samples can have a concentration of 17-OHP >= 50 nM/L serum in the first-tier, but later a concentration of 17-OHP <= 10 nM/L blood.
After several re-readings, the last paragraph of 3.4.2 does not make sense to me. Why was this comparison only done in those 35-36 weeks gestational age?
Response
The infants born at gestational age 35-36 weeks accounted for an unproportionally high share of the samples analyzed with 2nd tier. The difference between the previous re-run limit and the cut-off that resulted in a referral was large. This caught our attention, and was the reason for this comparison for newborns of GA 35-36 weeks, and not for the other infants
I do not understand the comparison being made between the second and third sentences implied by the word “instead”.
Response
The ”instead” has been changed to “on the contrary”.
Why is the new cut-off of 80 nmol/L not the same as that shown in Figure 3C?
Response
Thank you for finding this error. Figure 3C has been updated (something happened during the correction of that figure in the previous round of revisions).
What assurance is there that increasing the first-tier cut-off did not reduce sensitivity?
Response
With the first-tier algorithm (algorithm A), the cut-off was 100 nmol/L serum. The newborns with an initial screening result of 50-80 nmol/L serum, were extremely unlikely to result in a referral. Adjusting the first-tier cutoff of algorithm C to 80 nmol/L serum did not exclude any samples that would have resulted in a referral with algorithm A and therefore did not reduce sensitivity compared to algorithm A.
Previous first-tier data showed that all newborns with classic CAH born at GA 35-36 weeks had a first-tier 17-OHP above 100 nmol/L serum.
Minor Points
In 3.4.2 “gestational age” is abbreviated as GA, but in at least two other places “GW” is used. I believe “GA” is the preferred abbreviation.
Response
Thank you for your comment. This has now been corrected.
The Discussion should include the fact that specimens from 2015 were tested without confirmation of analyte stability for that long.
Response
Thank you for your comment. We have added this to the discussion.
Presumably a multi-year population with 49 true positive specimens would have had more than 155 false positives. How were the false positive specimens that underwent retrospective second tier testing chosen? The “random healthy newborns” included in the analysis need to be mentioned in the Methods.
Response
Thank you for the comment. We have now added this information in the methods section.
The y-axis label for the ratios (6th graph) of Figure 2 seems to be incorrect. A ratio is not expressed as “conc(nmol/L)”. Hopefully, the data shown are ratios. Similarly on that graph, the lower limits are unlikely to be zero, but should be stated as <0.x, the lower limit of detection of the numerator products divided by the measured denominator.
Response
Thank you for the comment. The data shows ratios. The label on the y-axis and the lower limits has been corrected.
Line 245. Think the word “even” changes the meaning of the sentence and should be removed.
Response
Thank you for your comment. The word “even” has been removed.
In the Discussion, line 308 does not make sense to me. Even two-tier CAH NBS does not “detect” NC CAH. It detects a baby at high risk for having CAH whose specific diagnosis is made by referring to a specialist.
Response
Thank you for your comment. This paragraph has been rewritten to clarify our standpoint that NC CAH individuals are not considered false positive.
In the Discussion, line 339 should not restrict the possibility of missed cases to those with SV CAH. SW cases might also be missed. The last two words of the sentence should be omitted.
Response
Thank you for your comment. We have omitted SV CAH.
The caption for Supplemental Figure S1 includes “The horizontal line” but there is no horizontal line.
Response
Thank you for your comment. Figure S1 has been updated.
The abbreviation for nanomoles per liter is not the same in Figure 3 as in the text.
Response
Thank you for your comment. The abbreviations in the figures have been changed to match the rest of the manuscript.
The group labels for cortisol in figure 2 use commas instead of periods. Please be consistent.
Response
Thank you for the comment. This has now been changed.
Response too Reviewer 2
Thank you for your valuable comments.
Reviewer 2 Report
Comments and Suggestions for AuthorsThis reviewer only wants to congratulate the authors for the edits made to the manuscript that have improved the quality of the document, and to suggest its publication in the IJNS.
Author Response
Response to Reviewers
Response to Reviewer 1
Areas for Improvement
It is not clear to me whether after initiation of the 2nd tier assay the first tier assay was re-run on specimens whose 1st tier was abnormal. (Lines 137-138) If so, how were the re-run results used? It is not shown in Figure 3.
Response
Thank you for this comment. Yes, the first-tier assay was re-run even after implementation of the 2nd-tier assay. The re-run results are used to validate that the initial screening result was correct, and with no mix-up of samples. We have added this information in section 2.4
I do not understand why a criterion of “17-OHP >= 10 nM/L blood” was added into Algorithm C when the first tier test include a criterion of “17-OHP >= 50 nM/L blood”. Was there that large a discrepancy between the two assay techniques?
Response
Thank you for your question. The first-tier criterion is 17-OHP >= 50 nM/L serum”, not blood. The difference in units accounts for half the difference, and most samples correlate well with a factor of about 2 in relation between serum and blood. The first-tier immunoassay is non-specific due to cross-reactivity with other steroids present as a result of metabolic stress. This is the reason that some samples can have a concentration of 17-OHP >= 50 nM/L serum in the first-tier, but later a concentration of 17-OHP <= 10 nM/L blood.
After several re-readings, the last paragraph of 3.4.2 does not make sense to me. Why was this comparison only done in those 35-36 weeks gestational age?
Response
The infants born at gestational age 35-36 weeks accounted for an unproportionally high share of the samples analyzed with 2nd tier. The difference between the previous re-run limit and the cut-off that resulted in a referral was large. This caught our attention, and was the reason for this comparison for newborns of GA 35-36 weeks, and not for the other infants
I do not understand the comparison being made between the second and third sentences implied by the word “instead”.
Response
The ”instead” has been changed to “on the contrary”.
Why is the new cut-off of 80 nmol/L not the same as that shown in Figure 3C?
Response
Thank you for finding this error. Figure 3C has been updated (something happened during the correction of that figure in the previous round of revisions).
What assurance is there that increasing the first-tier cut-off did not reduce sensitivity?
Response
With the first-tier algorithm (algorithm A), the cut-off was 100 nmol/L serum. The newborns with an initial screening result of 50-80 nmol/L serum, were extremely unlikely to result in a referral. Adjusting the first-tier cutoff of algorithm C to 80 nmol/L serum did not exclude any samples that would have resulted in a referral with algorithm A and therefore did not reduce sensitivity compared to algorithm A.
Previous first-tier data showed that all newborns with classic CAH born at GA 35-36 weeks had a first-tier 17-OHP above 100 nmol/L serum.
Minor Points
In 3.4.2 “gestational age” is abbreviated as GA, but in at least two other places “GW” is used. I believe “GA” is the preferred abbreviation.
Response
Thank you for your comment. This has now been corrected.
The Discussion should include the fact that specimens from 2015 were tested without confirmation of analyte stability for that long.
Response
Thank you for your comment. We have added this to the discussion.
Presumably a multi-year population with 49 true positive specimens would have had more than 155 false positives. How were the false positive specimens that underwent retrospective second tier testing chosen? The “random healthy newborns” included in the analysis need to be mentioned in the Methods.
Response
Thank you for the comment. We have now added this information in the methods section.
The y-axis label for the ratios (6th graph) of Figure 2 seems to be incorrect. A ratio is not expressed as “conc(nmol/L)”. Hopefully, the data shown are ratios. Similarly on that graph, the lower limits are unlikely to be zero, but should be stated as <0.x, the lower limit of detection of the numerator products divided by the measured denominator.
Response
Thank you for the comment. The data shows ratios. The label on the y-axis and the lower limits has been corrected.
Line 245. Think the word “even” changes the meaning of the sentence and should be removed.
Response
Thank you for your comment. The word “even” has been removed.
In the Discussion, line 308 does not make sense to me. Even two-tier CAH NBS does not “detect” NC CAH. It detects a baby at high risk for having CAH whose specific diagnosis is made by referring to a specialist.
Response
Thank you for your comment. This paragraph has been rewritten to clarify our standpoint that NC CAH individuals are not considered false positive.
In the Discussion, line 339 should not restrict the possibility of missed cases to those with SV CAH. SW cases might also be missed. The last two words of the sentence should be omitted.
Response
Thank you for your comment. We have omitted SV CAH.
The caption for Supplemental Figure S1 includes “The horizontal line” but there is no horizontal line.
Response
Thank you for your comment. Figure S1 has been updated.
The abbreviation for nanomoles per liter is not the same in Figure 3 as in the text.
Response
Thank you for your comment. The abbreviations in the figures have been changed to match the rest of the manuscript.
The group labels for cortisol in figure 2 use commas instead of periods. Please be consistent.
Response
Thank you for the comment. This has now been changed.
Response too Reviewer 2
Thank you for your valuable comments.
Round 3
Reviewer 1 Report
Comments and Suggestions for AuthorsNicely done.

