Population-Specific Pharmacogenomic Profiling of NAT2, CYP2E1, and SLCO1B1 in Tuberculosis Patients from Southern Peru: A Feasibility Pilot Study

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The brief report „Population-Specific Pharmacogenomic Profiling of NAT2, CYP2E1, and SLCO1B1 in Tuberculosis Patients from Southern Peru: Implications for Precision Dosing of Isoniazid and Rifampicin is a pilot study which investigates pharmacogenetic polymorphisms in NAT2, CYP2E1, and SLCO1B1 genes in 35 tuberculosis patients from southern Peru (Arequipa). It found a predominance of intermediate NAT2 acetylators (68.6%), rare CYP2E1 variants, and heterogeneity in SLCO1B1, with comparisons to Lima cohorts. The aim is to support personalized dosing of isoniazid and rifampicin in Andean populations, but clinical outcomes and statistical analyses are lacking.

The manuscript presents interesting findings but requires significant reorganization and refinement to enhance clarity and rigor.

Results Section Concerns: Specifically, the content in Section 3.1 appears misallocated and would be more appropriately classified under the Methods section. Additionally, Section 3.2 should be removed entirely, as it does not align thematically with the scope and objectives of the study. Moreover, the methods used in this section are not listed in methodology.

 

In Methodology section there is an absence of statistical tests that would correlate the presented results with the genotyping outcomes. This gap undermines the validity of the findings and should be addressed.

The pharmacogenetic component requires a complete changed. Instead of merely summarizing SNPs within genes, the authors should focus on reconstructing haplotypes, utilizing tools such as PHASE. Once haplotypes are established, they should be translated into 'star alleles,' which is essential for discussing allele frequencies. To convert SNPs to star alleles, resources such as PHarmVar.org should be utilized, ensuring that the methodology is transparent and accessible. Nomenclature Understanding: It is also advisable for the authors to consult ClinPGx Vocabulary for a better understanding of nomenclature, as this will enhance the clarity and professionalism of the paper.

Discussion and Conclusions: The discussion and conclusions must be grounded in the frequency and comparison of these star alleles, providing a more precise and relevant context for the findings.

Author Response

Reviewer 1

Comment 1.1

The brief report „Population-Specific Pharmacogenomic Profiling of NAT2, CYP2E1, and SLCO1B1 in Tuberculosis Patients from Southern Peru: Implications for Precision Dosing of Isoniazid and Rifampicin is a pilot study which investigates pharmacogenetic polymorphisms in NAT2, CYP2E1, and SLCO1B1 genes in 35 tuberculosis patients from southern Peru (Arequipa). It found a predominance of intermediate NAT2 acetylators (68.6%), rare CYP2E1 variants, and heterogeneity in SLCO1B1, with comparisons to Lima cohorts. The aim is to support personalized dosing of isoniazid and rifampicin in Andean populations, but clinical outcomes and statistical analyses are lacking.

The manuscript presents interesting findings but requires significant reorganization and refinement to enhance clarity and rigor.

Response 1.1

Thank you for this comment and appreciation of our study. We agree that the study is exploratory and descriptive in nature. According to our Methods section (Section 2.2), this work was designed as a pilot pharmacogenetic profiling study aimed at establishing baseline allele and phenotype frequencies in a previously uncharacterized regional population. The primary objective was not to evaluate clinical outcomes but rather to characterize the distribution of functional variants in NAT2, CYP2E1, and SLCO1B1 in Southern Peru. Regarding the absence of clinical outcomes, we have clarified in the revised Discussion and Limitations sections that hepatotoxicity markers and pharmacokinetic data were not collected as part of this pilot design. We now emphasize that genotype–phenotype associations and clinical correlations represent an essential next step and are currently being considered for future studies. Concerning statistical analyses, we have refined the manuscript to improve clarity and rigor. Specifically, Hardy–Weinberg equilibrium testing and quality control procedures are now more clearly presented.

 

Comment 1.2

Results Section Concerns: Specifically, the content in Section 3.1 appears misallocated and would be more appropriately classified under the Methods section. Additionally, Section 3.2 should be removed entirely, as it does not align thematically with the scope and objectives of the study. Moreover, the methods used in this section are not listed in methodology.

Response 1.2

Thank you for this comment. In the revised manuscript, the 3.1 section has been revised, and part of this content was reallocated to Section 2.1 to clearly describe participant recruitment, inclusion/exclusion criteria, and final sample size before the presentation of results. Regarding Section 3.2 (Clinical and Epidemiological Findings), we acknowledge that this content was not directly aligned with the primary pharmacogenetic objective of the study. In order to improve thematic coherence, this section has been removed from the main manuscript. The Results section has been reorganized to focus exclusively on pharmacogenetic analyses (NAT2, CYP2E1, and SLCO1B1).

 

 Comment 1.3
In Methodology section there is an absence of statistical tests that would correlate the presented results with the genotyping outcomes. This gap undermines the validity of the findings and should be addressed.

Response 1.3

Thank you for this thoughtful comment. We appreciate the opportunity to clarify the scope of the study and to improve the methodological description. Following the reviewer’s suggestion, we have expanded the analytical framework to strengthen the interpretation of the results. In the revised manuscript, we now include Hardy–Weinberg equilibrium (HWE) analyses and genotype-based clustering patterns, which provide statistical support for the internal consistency and structure of the genotyping data. These analyses offer an additional layer of validation for the observed genotype distributions, while remaining aligned with the exploratory and hypothesis-generating nature of the study. We believe that these additions improve the clarity and robustness of the manuscript, and we thank the reviewer for helping us refine the presentation and interpretation of our results.

 

Comment 1.4

The pharmacogenetic component requires a complete changed. Instead of merely summarizing SNPs within genes, the authors should focus on reconstructing haplotypes, utilizing tools such as PHASE. Once haplotypes are established, they should be translated into 'star alleles,' which is essential for discussing allele frequencies. To convert SNPs to star alleles, resources such as PHarmVar.org should be utilized, ensuring that the methodology is transparent and accessible. Nomenclature Understanding: It is also advisable for the authors to consult ClinPGx Vocabulary for a better understanding of nomenclature, as this will enhance the clarity and professionalism of the paper.

Response 1.4

We thank the reviewer for this important and constructive comment. In response, we substantially revised the pharmacogenetic component of the manuscript to move beyond a simple SNP-level description toward a haplotype-oriented framework. In the revised version, we now present genotype-based clustering analyses that capture haplotype-like structures, which are subsequently interpreted in the context of established STAR allele definitions. Allele nomenclature and functional interpretation were guided by publicly available resources, including PharmVar and the ClinPGx Vocabulary, to ensure consistency, transparency, and adherence to current pharmacogenetic standards. Corresponding methodological details and expanded results sections have been added to clearly document this workflow.

Regarding haplotype reconstruction, we carefully evaluated PHASE alongside other contemporary phasing tools. While PHASE is a well-established method, it is computationally intensive and less suited for datasets with moderate sample sizes and multiple loci across several genes. We therefore opted to use BEAGLE, which provides robust and widely validated statistical phasing, efficiently handles missing genotypes, and is extensively used in modern pharmacogenomic and population-genetic studies. BEAGLE’s performance and scalability make it particularly appropriate for reconstructing haplotypes in multi-gene panels and for downstream translation into STAR alleles. Importantly, BEAGLE-based phasing has been shown to yield results comparable to PHASE in similar contexts, while offering improved computational efficiency and reproducibility.

 

Comment 1.5
Discussion and Conclusions: The discussion and conclusions must be grounded in the frequency and comparison of these star alleles, providing a more precise and relevant context for the findings.

Response 1.5

We thank the reviewer for this valuable comment. The Discussion and Conclusions have been revised to more explicitly anchor our interpretation in the observed NAT2, CYP2E1, and SLCO1B1 allele and star-allele frequency distributions. We now provide quantitative comparisons with the previously published Lima cohort, highlighting the predominance of intermediate NAT2 acetylators (68.6%) and the lower proportion of slow acetylators (11.4%) in Southern Peru.

Reviewer 2 Report

Comments and Suggestions for Authors

This manuscript presents a descriptive pharmacogenomics study of NAT2, CYP2E1 and SLCO1B1 polymorphisms in a cohort of 35 tuberculosis patients from Southern Peru. The manuscript addresses an important topic and provides novel baseline pharmacogenomics data for Southern Peru. The methodology is generally sound and clearly described. However, revisions are needed before considering publication:

• The final analyzed cohort consists of 35 patients. While acceptable for a pilot frequency study, no confidence intervals are provided for allele or phenotype frequencies. This limits interpretability and comparability. Furthermore, no Hardy–Weinberg equilibrium testing is reported.
• The large discrepancy in slow acetylator frequency between Southern Peru and Lima is potentially important. However, the manuscript does not explore possible explanations (ancestral composition, sampling bias, demographic differences). The authors should either expand this discussion or avoid implying strong regional differentiation without statistical confirmation.
• The limitations section is present but could be strengthened. In particular, the following should be emphasized, especially: the small sample size, the lack of hepatotoxicity data, the absence of pharmacokinetic metrics, etc.
• Finally, in the “Results” section, the “Clinical and Epidemiological findings” sub-section should be shortened.

Author Response

Reviewer 2

Comment 2.1
This manuscript presents a descriptive pharmacogenomics study of NAT2, CYP2E1 and SLCO1B1 polymorphisms in a cohort of 35 tuberculosis patients from Southern Peru. The manuscript addresses an important topic and provides novel baseline pharmacogenomics data for Southern Peru. The methodology is generally sound and clearly described. However, revisions are needed before considering publication:

Response 2.1

We sincerely thank the reviewer for the positive evaluation of our manuscript and for recognizing the relevance of generating baseline pharmacogenomic data in an underrepresented population from Southern Peru. In this revised version, we have carefully addressed all of your comments and incorporated the suggested changes throughout the manuscript. We sincerely appreciate your time, and we hope that the revised version now aligns with your assessment criteria.

 

Comment 2.2
The final analyzed cohort consists of 35 patients. While acceptable for a pilot frequency study, no confidence intervals are provided for allele or phenotype frequencies. This limits interpretability and comparability. Furthermore, no Hardy–Weinberg equilibrium testing is reported.

Response 2.2

Thank you for this observation. In this revised version, we have expanded the description of allele distributions and genotype-based clustering patterns. We have also added a new Figure 1, which presents a Hardy–Weinberg equilibrium (HWE) Q–Q plot for the analyzed variants.

Figure 1 shows that most variants follow the expected HWE distribution, with deviations confined to the upper tail of the p-value distribution. As this study is based on a clinically selected patient cohort, departures from HWE are not unexpected and may reflect population structure, non-random sampling, or selection related to disease status or pharmacogenetically relevant traits, in addition to technical factors.

In addition, we added the New Figure 3 to show the genotype-based clustering patterns, which are subsequently interpreted in the context of established STAR allele definitions. Allele nomenclature and functional interpretation were guided by publicly available resources, including PharmVar and the ClinPGx Vocabulary, to ensure consistency, transparency, and adherence to current pharmacogenetic standards. Corresponding methodological details and expanded results sections have been added to clearly document this workflow.

 

Comment 2.3
The large discrepancy in slow acetylator frequency between Southern Peru and Lima is potentially important. However, the manuscript does not explore possible explanations (ancestral composition, sampling bias, demographic differences). The authors should either expand this discussion or avoid implying strong regional differentiation without statistical confirmation.

Response 2.3

We thank the reviewer for this important observation. We agree that the apparent difference in slow acetylator frequency between Southern Peru and Lima requires cautious interpretation. The Discussion has been revised to moderate language, suggesting strong regional differentiation, and to acknowledge that the study was not designed to formally assess population genetic structure.

We have also incorporated a paragraph discussing possible contributing factors, including differences in ancestral composition, sampling variability due to limited cohort size, and demographic heterogeneity.

 

Comment 2.4
The limitations section is present but could be strengthened. In particular, the following should be emphasized, especially: the small sample size, the lack of hepatotoxicity data, the absence of pharmacokinetic metrics, etc.

Response 2.4

We thank the reviewer for this constructive suggestion. The Limitations section has been expanded to explicitly emphasize the relatively small sample size (n = 35), which limits the precision of frequency estimates and precludes robust inferential analyses. We have also clarified that hepatotoxicity data and other clinical toxicity endpoints were not systematically collected, and that pharmacokinetic parameters such as plasma drug concentrations or clearance metrics were not assessed.

 

Comment 2.5
Finally, in the “Results” section, the “Clinical and Epidemiological findings” sub-section should be shortened.

Response 2.5

We thank the reviewer for this suggestion. In order to improve clarity and maintain alignment with the primary pharmacogenetic focus of the manuscript, the “Clinical and Epidemiological Findings” subsection has been removed from the Results section. The revised version now concentrates exclusively on pharmacogenetic analyses.

Reviewer 3 Report

Comments and Suggestions for Authors

This study aim to characterize clinically relevant polymorphisms in NAT2, CYP2E1, and SLCO1B1 in tuberculosis patients from Southern Peru. Thirty-five adults receiving first-line therapy (isoniazid and rifampicin) underwent targeted Sanger sequencing of key functional variants. NAT2 acetylator phenotypes were predominantly intermediate (68.6%), 23 followed by rapid (25.7%) and slow (5.7%) profiles. CYP2E1 functional promoter variants were infrequent, whereas 25 SLCO1B1 exhibited notable allelic heterogeneity, suggesting potential variability in rifampicin transport. 

Data from this population were compared with previously reported data for Peruvian subjects from the Lima region. The authors claim that regional differences in acetylator distribution were detected. 

Although the study is interesting, there is a major drawback: the small sample size of only 35 subjects. The authors justify this small sample size arguing that it is a pilot study. If such is the case, the title must be changed to reflect this issue. Moreover, a pilot study is performed to provide data that can be then used for a definitive study. Data from a pilot study cannot be used to perform comparisons between populations. Moreover, data from a pilot study cannot be used to implement national or local treatment strategies. The manuscript should hence be rewritten to reflect these limitations, avoiding  overconclusions.

The role of sample size in pharmacogenomic studies begs to be discussed. References such as Rao CD. Sample size considerations in genetic polymorphism studies. Hum Hered 2001; 52(4):191-200 and Hong & Park. Sample size and statistical power calculations in genetic association studies. Genomic Infor 2012 Jun 30;10(2):117–122 should be considered.

Author Response

Reviewer 3

Comment 3.1
This study aim to characterize clinically relevant polymorphisms in NAT2, CYP2E1, and SLCO1B1 in tuberculosis patients from Southern Peru. Thirty-five adults receiving first-line therapy (isoniazid and rifampicin) underwent targeted Sanger sequencing of key functional variants. NAT2 acetylator phenotypes were predominantly intermediate (68.6%), 23 followed by rapid (25.7%) and slow (5.7%) profiles. CYP2E1 functional promoter variants were infrequent, whereas 25 SLCO1B1 exhibited notable allelic heterogeneity, suggesting potential variability in rifampicin transport. Data from this population were compared with previously reported data for Peruvian subjects from the Lima region. The authors claim that regional differences in acetylator distribution were detected. 

Response 3.1

 Regarding the comparison with the Lima cohort, we have revised the Discussion to moderate the wording suggesting strong regional differentiation. We now frame the observed differences in acetylator distribution as preliminary evidence of possible regional variation based on descriptive comparison, rather than definitive proof of intra-country genetic stratification. In this revised version, we have carefully addressed all of your comments and incorporated the suggested changes throughout the manuscript. We sincerely appreciate your time, and we hope that the revised version now aligns with your assessment criteria.

 

Comment 3.2
Although the study is interesting, there is a major drawback: the small sample size of only 35 subjects. The authors justify this small sample size arguing that it is a pilot study. If such is the case, the title must be changed to reflect this issue. Moreover, a pilot study is performed to provide data that can be then used for a definitive study. Data from a pilot study cannot be used to perform comparisons between populations. Moreover, data from a pilot study cannot be used to implement national or local treatment strategies. The manuscript should hence be rewritten to reflect these limitations, avoiding  overconclusions.
Response 3.2

Thank you for this relevant observation. We fully acknowledge that the relatively small sample size (n = 35) represents an inherent limitation of this exploratory study. In response to the reviewer’s suggestion, we have revised the manuscript to more clearly frame it as a pilot descriptive investigation. The title has been modified accordingly to explicitly reflect its pilot nature. Furthermore, we have carefully revised the Discussion and Conclusions to avoid overinterpretation. 

 

Comment 3.3
The role of sample size in pharmacogenomic studies begs to be discussed. References such as Rao CD. Sample size considerations in genetic polymorphism studies. Hum Hered 2001; 52(4):191-200 and Hong & Park. Sample size and statistical power calculations in genetic association studies. Genomic Infor 2012 Jun 30;10(2):117–122 should be considered.

Response 3.3

We thank the reviewer for this point. In this new version, we addressed the study’s limitations in our discussion and included the recommended literature as follows:

‘As previously discussed in the context of genetic polymorphism studies (Rao, 2001; Hong & Park, 2012), adequately powered cohorts are essential for detecting modest genotype–phenotype effects and for conducting reliable inter-population comparisons. Nevertheless, given the exploratory nature of the present study, the cohort size was not intended to support formal association testing or definitive conclusions regarding regional genetic differentiation. Rather, the primary objective was to establish preliminary allele and phenotype frequency data in a previously underrepresented population. Confirmation of these patterns and assessment of their clinical implications will require larger, multicenter studies incorporating pharmacokinetic and outcome-based endpoints.’

Round 2

Reviewer 3 Report

Comments and Suggestions for Authors

The authors have replied to this reviewer's comments in a satisfactorily manner.