Novel Quality Metrics for Assessing the Reproducibility and Translational Relevance of Mouse Transverse Aortic Constriction Experiments: A Systematic Review

Abstract

Background

Transverse aortic constriction (TAC) is widely used in mice to study pressure overload–induced cardiac hypertrophy and heart failure. However, methodological variability limits reproducibility and complicates translation to human disease. We sought to develop and apply a set of quality metrics that evaluate reproducibility and translational relevance in contemporary TAC studies.

Methods

Following Preferred Reporting Items for Systematic Reviews and Meta‐Analyses guidelines, we reviewed articles with new mouse TAC experiments published between January 1, 2023, and December 31, 2024, across 6 leading cardiovascular journals. Studies were assessed for predefined quality metrics: bias‐reduction methods, reporting of TAC pressure gradients, exclusion of outlying gradients, early measurement of TAC severity and cardiac function, inclusion of female mice, and initiation of interventions ≥7 days post‐TAC. Reported average peak instantaneous pressure gradients were summarized.

Results

Eighty‐nine articles met inclusion criteria. Most (69.7%) reported at least 1 measure to reduce bias. Only 24.7% reported TAC pressure gradients, and 7.9% excluded mice with outlying TAC severity. Early cardiac function was evaluated in 20.2% of articles, and 31.5% initiated at least 1 intervention ≥7 days post TAC. Female mice were included in 28.1% of studies, with greater inclusion in National Institutes of Health‐funded articles. Reported average peak instantaneous pressure gradients varied widely (36.8–131 mm Hg).

Conclusions

Most contemporary TAC studies do not report or perform key parameters important for experimental reproducibility and translational validity. Adoption of standardized quality metrics, including routine pressure gradient measurement, early cardiac function assessment, delayed intervention timing, and inclusion of female mice, could improve the reproducibility and translational relevance of TAC research.

Nonstandard Abbreviations and Acronyms

NHLBI

National Heart, Lung, and Blood Institute

NIH

National Institutes of Health

PIPG

peak instantaneous pressure gradient

TAC

transverse aortic constriction

Research Perspective.

What Is New?

• This study introduces a novel set of quality metrics to evaluate reproducibility and translational relevance in mouse transverse aortic constriction (TAC) experiments, including reporting of TAC pressure gradients, exclusion of outlying gradients, early assessment of TAC severity and cardiac function, inclusion of female mice, and timing of interventions in a clinically relevant manner.

• Our review of TAC studies published in 2023 to 2024 in leading cardiovascular journals revealed persistent gaps in reporting across these key quality metrics.

What Question Should Be Addressed Next?

• Would the development and adoption of consensus guidelines for TAC experimental design and reporting meaningfully reduce methodological variability, improve reproducibility, and strengthen the translational relevance of future TAC studies?

Transverse aortic constriction (TAC) in mice is widely used to study pressure overload‐induced cardiac hypertrophy and heart failure. ¹ , ² By creating a fixed narrowing of the transverse aorta, TAC generates a pressure gradient that imposes left ventricular (LV) pressure overload and drives subsequent hypertrophic remodeling. Although TAC induces an abrupt increase in afterload, human conditions such as aortic stenosis typically develop gradually, limiting direct translational equivalence. Originally developed by Rockman et al. to study hypertrophic signaling, TAC is now commonly used to study heart failure mechanisms. ³ The Rockman technique involves tying a suture around the transverse aorta using a blunted needle as a spacer to create a standard narrowing, and subsequent modifications have introduced alternative approaches to control the constriction diameter. Across methods, the severity of pressure overload is determined principally by the degree of aortic narrowing. ²

Experimental outcomes in TAC experiments are highly sensitive to procedural variables, particularly the pressure gradient across the constriction site. This gradient can be measured noninvasively by echocardiographic Doppler‐derived peak instantaneous pressure gradient (PIPG, reported as millimeters mercury or mm Hg) or by the ratio of right‐to‐left carotid artery Doppler flow velocities (carotid flow ratios). Graded increases in PIPGs from TAC (60.2 ± 5.6 mm Hg, 85.5 ± 4.4 mm Hg, or 106.6 ± 5.2 mm Hg) produce distinct levels of myocardial fibrosis, LV systolic dysfunction, and heart failure features. ⁴ Other studies have reported substantial variability in PIPGs within experimental groups, including mice with gradients near sham control levels, raising concerns about reliability and reproducibility. ⁵ , ⁶ Collectively, this literature highlights the importance of measuring and reporting TAC pressure gradients, and excluding mice with outlying gradients. Despite this, a 2021 meta‐analysis by Bosch et al. found that 66% of studies using TAC either failed to measure or did not report the severity of pressure overload. ⁷ Among studies that reported measurements, the prior analysis did not evaluate average pressure gradients or exclusion of outliers, both of which are important parameters for reproducibility and cross‐study comparisons.

In humans, pressure‐overload disease follows a predictable trajectory from compensated concentric hypertrophy to eccentric remodeling and overt heart failure. ¹ , ⁸ Most patients demonstrate hypertrophy at the time of diagnosis, which is the earliest point at which treatment is clinically feasible. To enhance translational and clinical relevance, mouse TAC studies should therefore characterize whether their models follow this expected remodeling trajectory and should initiate interventions after hypertrophy is present. However, TAC experiments commonly induce molecular, genetic, or pharmacologic perturbations before or at the time of TAC surgery, limiting their applicability to real‐world therapeutic timing. The prior meta‐analysis did not examine whether TAC studies characterize early LV remodeling or time interventions in a clinically relevant manner, even though these parameters are measurable. ⁷ Additionally, pressure overload cardiac disease affects both sexes, yet the 2021 meta‐analysis showed that only 13% of TAC experiments included female mice, ⁷ underscoring the need for equitable sex representation as an objective marker of translational relevance.

Based upon these considerations, we developed quality metrics to assess reproducibility and translational relevance for mouse TAC studies. Whereas the systematic review by Bosch et al. evaluated factors influencing TAC‐induced remodeling in studies published before March 2019, ⁷ our analysis places greater emphasis on methodological rigor and translational validity. Our reproducibility metrics include pressure gradient reporting, minimizing bias, and excluding mice with outlying pressure gradients. Our translational metrics include assessment of early cardiac function, delaying interventions until established pressure overload, and inclusion of female mice. To provide an updated perspective beyond the previous systematic review, ⁷ we focus on recent publications in leading cardiovascular journals affiliated with major societies, offering insight into current practices among prominent research groups.

METHODS

Our review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta‐Analyses 2020 statement. ⁹ The review was not registered with PROSPERO (International Prospective Register of Systematic Reviews) or other registries because these platforms do not provide templates for methodological reviews of experimental techniques. A formal review protocol was not prepared a priori. Consequently, no amendments were made. As a systematic review of published literature, this study did not require ethics approval.

Data Availability Statement

The data, analytic methods, and study materials are available within the article and its online supplemental file. The full search strategy for PubMed and each journal website is provided in Data S1. The data‐collection template is provided in Data S2. Extracted study‐level data underlying all tables and analyses are provided in Table S1. Author‐derived estimates from articles that reported PIPG data exclusively in figures are provided in Table S2. Standard spreadsheet calculations were used for all descriptive analysis and no new analytic code was generated. Requests for clarification regarding the provided materials may be directed to the corresponding author.

General Search Approach

Two authors (A.E.F. and A.K.O.) performed screening independently at all stages of this review, including initial record identification, article selection process, and data extraction for all quality metrics. The only variation was that M.A.P. and A.K.O. independently performed the extractions for the assessment of the PIPG averages and experimental ranges. Any discrepancies were resolved by discussion between the 2 screening authors; if a consensus was not achieved, the nonscreening author adjudicated. Results were recorded in an Excel spreadsheet (Microsoft Corporation, Redmond, WA, USA).

Databases and Search Terms

Our review focused on journals from leading cardiovascular research societies using the rationale that these journals reflect the best practices to date for cardiovascular research methods. Specifically, we reviewed Circulation, Circulation Research, and Journal of the American Heart Association (JAHA) from the American Heart Association; the American Journal of Physiology–Heart and Circulatory Physiology from the American Physiological Society; Cardiovascular Research from the European Society of Cardiology; and the Journal of Molecular and Cellular Cardiology from the International Society for Heart Research. To capture contemporary research practice, we included articles published between January 1, 2023, and December 31, 2024.

We performed the initial search in PUBMED and on each journal's website. Our record search strategies are shown in Data S1. We used www.ahajournals.org for Circulation, Circulation Research, and Journal of the American Heart Association; journals.physiology.org/journal/ajpheart for the American Journal of Physiology Heart and Circulatory Physiology; www.jmcc‐online.com for Journal of Molecular and Cellular Cardiology; and academic.oup.com/cardiovascres for Cardiovascular Research. Advanced search functions were used to query the terms “mouse aortic constriction” and “mouse TAC,” entered separately. Publication dates were restricted to January 1, 2023, through December 31, 2024. All identified articles were recorded in an Excel file that included the title, journal, and article link when available. Review articles and editorials were excluded. The most recent search was completed on July 31, 2025.

Article Selection Process

From the records identified in the database searches, we first identified and removed all duplicate files. We then searched the remaining articles (both main text and supplemental materials) using the PDF (preferred) or Word Find/Search tool (magnifying glass symbol) for the terms “aortic constriction” and, separately, “TAC.” Articles without either of these terms were excluded. In the remaining articles, we then searched the methods section (both main article text and supplemental materials) for TAC experiments performed specifically using mice. We included articles that reported new TAC experiments in mice, including research articles or research letters. We excluded articles that (1) performed only abdominal aortic constriction, (2) used TAC in species other than mice, or (3) did not perform new TAC studies (for example, analyses of publicly available TAC transcriptomic data). The included articles were then subjected to the quality metric data collection process outlined next.

Quality Metric Assessments: General Considerations

Each TAC experiment included in a article was independently evaluated using prespecified decision rules for each predefined quality metric (described next). For articles reporting multiple TAC experiments, a quality metric was considered present if it was reported in at least 1 experiment. Metrics that were missing or inadequately described were classified as absent. No attempts were made to contact authors for clarification of unclear reporting. Specific decision rules, including our approach to graph only data, unclear experimental timing, unclear sex reporting, and other incomplete information, are detailed within the individual quality metric sections. Our prespecified decision rules were defined before full data extraction and applied consistently across all included articles.

Quality Metric Assessment: Minimizing Bias

We selected this metric because minimizing bias enhances research reliability and reproducibility. Articles were screened for the following established methods to minimize bias: implementing the Animal Research: Reporting of In Vivo Experiments guidelines, blinding investigators to the experimental groups during analysis, and randomly assigning mice to experimental groups. For each article (including supplemental files), we used the PDF or Word Find/Search tool to identify the following keywords: “ARRIVE” (for the Animal Research: Reporting of In Vivo Experiments guidelines), “blind,” and “random.” This search approach captured longer words containing these terms, which applies to all searches throughout this review. For example, the term “blind” would capture “blinding” or “blinded.”

We considered this metric present if authors reported any of the following: adherence to the Animal Research: Reporting of In Vivo Experiments guidelines, any form of blinding, or any randomization procedure. If none of these factors were stated in the article or supplemental materials, this metric was recorded as absent. The screening authors used a Word document template (Microsoft Corporation, Redmond, WA, USA) listing all keywords and screening methods for each article in the quality metric analyses (Data S2). The quality metric screening template was piloted on 5 initial articles and iteratively refined during this time to ensure accurate and consistent capture of the predefined metrics. Table 1 provides a summary of each quality metric and its corresponding rationale.

Table 1.

Novel Quality Metrics for Assessing Reproducibility and Clinical/Translational Relevance of Mouse Transverse Aortic Constriction Studies

Quality metric	Rationale
Minimizing bias	Adherence to established methods to minimize bias increases research reliability and reproducibility
Reporting TAC pressure gradients	TAC pressure gradients strongly influence cardiac phenotypes. Transparent reporting is essential for contextualizing results and ensuring experimental reproducibility
Excluding mice for outlying TAC pressure gradients	Pressure gradients can vary widely within experimental groups, leading to heterogeneous cardiac phenotypes. Excluding animals with outlying gradients improves internal consistency and enhances reproducibility
Measuring pressure gradients ≤7 d post TAC surgery	Significant systolic dysfunction can artifactually lower echocardiogram calculated pressure gradients. Measuring gradients within the first week post TAC, before substantial dysfunction develops, would provide more accurate and reproducible results
Assessing for early post‐TAC LV dysfunction (ejection fraction or fractional shortening ≤7 d post TAC surgery)	Human pressure‐overload disease follows a characteristic progression from compensated hypertrophy to dilation and systolic failure. Replicating this pattern in TAC experiments can enhance the clinical and translational relevance. Measuring early post‐TAC cardiac function is necessary to confirm whether experimental models follow this expected remodeling trajectory starting with compensated cardiac hypertrophy
Delaying interventions until established pressure‐overload	Patients with pressure‐overload disease often exhibit LV hypertrophy at diagnosis. Thus, delaying interventions better mirrors clinical practice and may enhance the translational relevance of TAC studies
Female mice inclusion	Incorporating female animals increases the generalizability of preclinical findings and supports equitable translation of cardiovascular therapies

LV indicates left ventricle or left ventricular; and TAC, transverse aortic constriction.

Quality Metric Analysis: Reporting TAC Pressure Gradients

We selected this metric because TAC pressure gradients strongly influence cardiac phenotypes, ⁴ making this information essential for contextualizing experimental findings and reproducing results. Articles were screened for the following established methods to measure TAC pressure gradients: echocardiogram PIPG or carotid doppler flows ratio. ¹⁰ We acknowledge that carotid Doppler flow ratios are a surrogate rather than a direct measurement of the pressure gradient. We also included direct LV pressure measurements because the difference in LV systolic pressure between TAC and sham mice is proportional to the TAC pressure gradient. No articles performed a direct pressure measurement across the TAC site.

To identify pressure gradient measurements, we searched each article (including supplemental methods) using the PDF or Word Find/Search tool for the following key terms: “echo,” “mm Hg,” “pressure gradient,” “carotid flow ratio,” “carotid,” “Doppler,” “velocity,” “catheter,” and “hemodynam.” We then reviewed the surrounding text or figures for the measurement values. Some articles included several separate TAC experiments. We considered this metric present if an article reported pressure gradient values for any TAC experiment, even if not all experiments included these data. We counted this metric as absent if pressure gradient values were not reported, even if the methods sections stated that the measurements were performed.

Quality Metric Analysis: Excluding Mice for Outlying TAC Pressure Gradients

We selected this metric because studies by Mohammed et al. and Merino et al. found a wide range in TAC PIPGs within their experimental groups, which could lead to inconsistent cardiac phenotypes. ⁵ , ⁶ Thus, excluding mice with outlying pressure gradients could improve reliability and reproducibility. For each article (including supplemental files), we used the PDF or Word Find/Search tool to identify the following keywords: “inclusion,” “inclu,” “exclu,” and “exclusion.” We counted this metric present if any exclusion or inclusion criteria were stated for pressure gradients regardless of the specific cutoff pressures, carotid flow ratios, or LV pressures. We counted this metric as absent if no criteria were included in the article or supplemental materials.

Quality Metric Analysis: Measuring Pressure Gradients ≤7 Days Post TAC Surgery

We selected this metric because many TAC mice have significantly reduced cardiac function at the study end point, which could cause underestimations of the echocardiogram derived pressure gradient if performed concurrently. ⁸ Measuring TAC pressure gradients early after surgery and without significant cardiac dysfunction provides a more accurate and reproducible measurement. We set ≤7 days post TAC as the cutoff for this metric based on the rationale that this time frame allows mice time to recover from surgery while also avoiding the expected nadir in cardiac function. We searched for this information in the methods sections (both the main article and supplemental materials) using the PDF or Word Find/Search tool for the terms: “echo” (echocardiogram for PIPG), “doppler” (for carotid flow ratios or PIPG), “catheter” (for direct LV pressure measurement), and “hemodynam” (for direct LV pressure measurement). We then reviewed the text surrounding these terms for the description of the timing post surgery. We counted this parameter as present if the publication stated that the pressure gradient was measured ≤7 days after TAC and absent if no time was provided, or the measurements were made >7 days post TAC.

Quality Metric Analysis: Assessing Early Post‐TAC LV Function

To enhance the clinical and translational relevance of TAC models, it is important for experimental remodeling to recapitulate the human trajectory of compensated hypertrophy preceding ventricular decompensation. In addition, severe TAC pressure gradients can trigger rapid LV failure, which likely reflects pathophysiologic processes distinct from those occurring in humans with pressure‐overload disease. ¹ , ² Accordingly, we selected early post‐TAC assessment of cardiac function as a quality metric to determine whether experimental hearts exhibit compensated hypertrophy within the early post‐TAC period.

We defined early post‐TAC assessment of LV dysfunction as the performance of echocardiography within ≤7 days post surgery. To determine whether studies met this quality metric, we searched the text and figures using the PDF or Word Find/Search tool for following key terms: “echo,” “ejection fraction,” and “fractional shortening.” We considered this quality metric positive if a publication showed values (in the text or figures) for either ejection fraction or fractional shortening ≤7 days post TAC and absent if no time frame was given for the echocardiograms or the first echocardiogram occurred >7 days post TAC.

Quality Metric Analysis: Delaying Interventions Until Established Pressure‐Overload

We selected this metric because most patients with pressure‐overload cardiac disease already exhibit LV hypertrophy at the time of diagnosis, which represents the earliest point at which therapeutic intervention is realistically feasible. Accordingly, delaying the initiation of experimental interventions in TAC models enhances clinical and translational relevance. For this review, we defined established pressure overload as ≥7 days post TAC surgery based on the rationale that mice have recovered from surgery by this time, and many have begun to exhibit measurable cardiac hypertrophy. Interventions were defined as (1) a genetic modulation, including inducible or constitutively active changes in gene expression; or (2) administration of medications or comparable biologically active substances. Some genetic modifications, such as those mediated by adeno‐associated viruses, require several weeks to manifest after administration. ¹¹ In such cases, we used the authors' reported time frame for the genetic change.

Because no single set of keywords reliably identified this parameter, we assessed it through manual review. Specifically, we examined (1) mouse descriptions for transgenic models or adeno‐associated virus strategies in the Methods section, (2) TAC procedural details, and (3) descriptions of each TAC experiment in the Results section, including any accompanying schematic depictions of experimental timelines. For articles that included multiple interventions, we counted this metric as positive if any intervention was initiated ≥7 days post TAC. This metric was considered negative if all interventions were performed <7 days post TAC or the article did not specify the timing of interventions relative to TAC surgery.

Quality Metric Analysis: Female Mice Inclusion

We selected this parameter because heart disease affects both sexes, and it is important to include female mice in preclinical research to enhance the generalizability of scientific findings and ensure equitable translation of cardiovascular therapies. We searched for this parameter within the articles (including supplemental methods) using the PDF or Word Find/Search tool for the key term “female.” We counted this parameter as positive if any TAC experiment within an article reported using female mice. We counted this parameter as negative if only males were used for TAC, or the sex was not clearly stated in the article or supplemental methods.

Because the National Institutes of Health (NIH) promoted inclusion of both male and female sexes in basic and clinical research, we further determined whether this policy increased use of females by comparing articles with NIH versus non‐NIH funding. To determine whether studies met this quality metric, we searched the articles for NIH funding using the PDF or Word Find/Search tool for the terms: “NIH,” “funding,” “R01,” and “NHLBI” for National Heart, Lung, and Blood Institute. We also reviewed the sources of funding section. We considered this parameter as positive if any author reported receiving NIH funding for the article. We considered this metric as negative if no authors had NIH funding or the funding agencies were not listed.

PIPG Averages and Experimental Range

To assess PIPGs within and across studies, we recorded the arithmetic mean reported for the PIPG and the corresponding range when available. We identified PIPG measurements by searching the main article and supplemental materials using the PDF or Word Find/Search tool for the key terms: “mm Hg,” “velocity,” and “pressure gradient.” We also manually reviewed all figures in the primary and supplemental files to identify echocardiographic PIPG values, as many studies presented PIPG data graphically without reporting numerical averages or ranges.

When PIPG values were available only in figure format, 2 authors (A.K.O. and M.A.P.) independently estimated the average and range by visual inspection. We report the unweighted arithmetic mean of these independent estimates. For articles containing multiple distinct TAC experiments, we extracted and reported PIPG values separately for each experiment, following the order in which they appeared in the article. When a single experiment included multiple TAC groups, we combined the groups to calculate the overall unweighted average and range. We did not attempt to estimate error bars (eg, SD or SEM) from graphical representations.

We also extracted key TAC surgical parameters potentially influencing pressure gradients, including age at surgery, sex, and constriction needle gauge. These data were obtained from the Methods sections of the main articles and supplemental materials, specifically from descriptions of the TAC procedure and mouse characteristics.

Reporting Bias, Risk of Bias, and Certainty of Evidence

Formal study‐level risk‐of‐bias assessments, reporting‐bias assessments, and certainty‐of‐evidence frameworks (eg, Grading of Recommendations Assessment, Development, and Evaluation) were not performed. These approaches are intended for syntheses of effect estimates and causal outcomes, whereas the present review focuses on descriptive assessment of experimental reporting practices and methodological features. Thus, results are reported as frequencies and ranges rather than effect estimates. Study quality was instead evaluated using our predefined quality metrics and no formal certainty ratings were assigned.

Statistical Analysis

All included articles contributed to the descriptive synthesis of the quality metrics. Values are reported as the percentage of articles that included a specific quality metric out of the total number of articles for which that metric was applicable. For each metric, we also present the lowest and highest reporting percentages observed across the individual journals; however, the specific journals corresponding to these values are not identified. When comparing TAC articles with and without NIH funding for their inclusion of female mice, a statistical comparison was performed using Fisher's exact test with statistical significance set at P<0.05. This comparison was exploratory and intended to identify differences in observed percentages rather than causal effects. No pooled effect measures were calculated. All analyses were conducted using GraphPad Prism version 10.2.3 (GraphPad Software, Boston, MA, USA).

RESULTS

Our search process is summarized as a Preferred Reporting Items for Systematic Reviews and Meta‐Analyses 2020 flow diagram ⁹ (FigureFigure). The initial database search identified 322 records of which 134 duplicate records were removed, leaving 188 articles for full‐text screening. Ninety‐nine articles were excluded because they did not contain new mouse TAC experiments (excluded articles shown in Data S3). The final systematic review included 89 articles that met inclusion criteria. ¹¹ , ¹² , ¹³ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ , ²⁹ , ³⁰ , ³¹ , ³² , ³³ , ³⁴ , ³⁵ , ³⁶ , ³⁷ , ³⁸ , ³⁹ , ⁴⁰ , ⁴¹ , ⁴² , ⁴³ , ⁴⁴ , ⁴⁵ , ⁴⁶ , ⁴⁷ , ⁴⁸ , ⁴⁹ , ⁵⁰ , ⁵¹ , ⁵² , ⁵³ , ⁵⁴ , ⁵⁵ , ⁵⁶ , ⁵⁷ , ⁵⁸ , ⁵⁹ , ⁶⁰ , ⁶¹ , ⁶² , ⁶³ , ⁶⁴ , ⁶⁵ , ⁶⁶ , ⁶⁷ , ⁶⁸ , ⁶⁹ , ⁷⁰ , ⁷¹ , ⁷² , ⁷³ , ⁷⁴ , ⁷⁵ , ⁷⁶ , ⁷⁷ , ⁷⁸ , ⁷⁹ , ⁸⁰ , ⁸¹ , ⁸² , ⁸³ , ⁸⁴ , ⁸⁵ , ⁸⁶ , ⁸⁷ , ⁸⁸ , ⁸⁹ , ⁹⁰ , ⁹¹ , ⁹² , ⁹³ , ⁹⁴ , ⁹⁵ , ⁹⁶ , ⁹⁷ , ⁹⁸ , ⁹⁹

Figure 1. Preferred Reporting Items for Systematic Reviews and Meta‐Analyses 2020 flow diagram ⁹ of the literature screening process for mouse transverse aortic constriction articles in journals from leading cardiovascular research societies with publication dates between January 1, 2023, through December 31, 2024.

Quality Metric: Minimizing Bias

Table 2 summarizes the overall quality metric results, with individual article details provided in Table S1. Overall, 69.7% of articles reported using at least 1 measure to reduce the research risk of bias.

Table 2.

Mouse Transverse Aortic Constriction Quality Metrics for Articles Published Between January 1, 2023 to December 31, 2024 by Leading Cardiovascular Research Societies^*

	Articles reporting parameter (n)	Articles with parameter applicable (n)	Percentage reporting applicable parameter (%)	Lower and upper range among the journals (%)
Addressed risk of bias reported	62	89	69.7	35.7–100
TAC PIPG reported	13	89	14.6	0–28.6
Right‐to‐left carotid flow ratio reported	3	89	3.4	0–14.3
Left ventricular direct systolic pressure reported	9	89	10.1	0–42.9
Any measurement of TAC pressure gradient reported†	22	89	24.7	5–71.4
Exclusion for low TAC pressure gradient	7	89	7.9	0–21.8
Exclusion for high TAC pressure gradient	1	89	1.1	0–4.3
Exclusion based upon TAC knot size	1	89	1.1	0–4.3
Reported PIPG or carotid flow ratio measured ≤7 d post TAC	6	18	33.3	0–100
EF or FS shown ≤7 d post TAC	18	89	20.2	0–30.4
Intervention performed ≥7 d post TAC	23	73	31.5	16.7–50
Female included for TAC	25	89	28.1	14.3–83.3
Sex for TAC not reported	15	89	16.9	0–35
Female included for TAC in articles with NIH funding	19	40	47.5‡	20–75
Female included for TAC in articles without NIH funding	6	49	12.2	0–100

The definition of an intervention is stated in the Methods section. LV indicates left ventricular; NIH, National Institutes of Health; PIPG, peak instantaneous pressure gradient; and TAC, transverse aortic constriction.

^* Circulation, Circulation Research, and Journal of the American Heart Association from the American Heart Association; the American Journal of Physiology–Heart and Circulatory Physiology from the American Physiological Society; Cardiovascular Research from the European Society of Cardiology; and the Journal of Molecular and Cellular Cardiology from the International Society for Heart Research.

^† “Any measurement of TAC pressure gradient reported” indicates that an article reported either PIPGs, carotid flow ratios, or LV systolic pressures. Three articles included invasive LV systolic pressures along with either PIPGs or carotid doppler flow ratios.

^‡ P < 0.015 compared with females included for TAC in articles without NIH funding.

Quality Metric: Reporting TAC Pressure Gradients

Only 24.7% of articles reported any measurement of TAC pressure gradients. Reporting varied substantially across journals, ranging from 5% to 71.4%. The most common method was echocardiographic PIPG followed closely by invasive LV pressure measurements. Three articles reported pressure gradient measurements using multiple methods, ²⁴ , ³⁷ , ⁸⁰ so the total number of articles with any TAC pressure gradient measurement is less than the sum of the individual methods.

Quality Metric: Excluding Mice for Outlying TAC Pressure Gradients

Few articles (7.9%) excluded mice for low TAC pressure gradients. The minimum PIPGs for exclusion ranged from 25 mm Hg ⁴⁵ to 60 mm Hg ⁴⁶ with 2 articles using 40 mm Hg ³⁷ , ⁵⁰ and another 2 using 50 mm Hg. ²⁹ , ⁸⁸ One article excluded mice for a high TAC severity, which they defined as carotid flow ratio >25. ⁴¹ Another study excluded mice based upon the TAC knot size measured after tissue collection, removing those beyond ±2 SDs from the mean. ⁴⁷ No studies reported using invasive LV systolic pressures to exclude mice.

Quality Metric: Reporting Cardiac Function ≤7 Days Post TAC

Eighteen articles (20.2%) reported cardiac functional measurements (ejection fraction or fractional shortening) ≤7 days after TAC surgery.

Quality Metric: Interventions Initiated ≥7 Days Post TAC

Among the 73 articles that performed an intervention with TAC, approximately one ‐third (31.5%) included at least 1 experiment initiating an intervention ≥1 week after TAC surgery.

Quality Metric: Female Mice Inclusion

Slightly more than a quarter of articles (28.1%) included female mice, either alone or with males in their TAC experiments. Fifteen articles (16.9%) did not clearly state the sex used. Inclusion of female mice was significantly higher in NIH‐funded studies (47.5%) compared with those without NIH funding (12.2%, P=0.0003).

Summary of PIPGs From Reporting Articles

Thirteen articles reported PIPGs, summarized in Table 3. ¹⁴ , ¹⁸ , ²⁴ , ²⁸ , ²⁹ , ³⁰ , ³⁷ , ⁴⁵ , ⁵⁰ , ⁵⁶ , ⁸¹ , ⁸³ , ⁹⁶ The PIPG estimates from articles showing these values only in figures are provided in Table S2. Average PIPGs ranged widely, from 36.8 mm Hg up to 131 mm Hg. ⁵⁶ , ⁸¹ Counterintuitively, studies using larger diameter needles for constriction (26‐gauge versus 27‐gauge) reported the highest average PIPGs. ⁸¹ , ⁸³ Seven articles presented individual mouse PIPGs, allowing estimation of ranges within experimental groups. ¹⁸ , ²⁸ , ³⁷ , ⁴⁵ , ⁵⁰ , ⁸³ , ⁹⁶ The largest estimated PIPG range was approximately 122 mm Hg whereas the smallest was ∼7 mm Hg.

Table 3.

Summary of Articles Reporting Echocardiogram Transverse Aortic Constriction Peak Instantaneous Pressure Gradients

Article	TAC needle gauge	Age at surgery (wk)	Sex	Weight at surgery (grams)	Average PIPG (mm Hg)	Range PIPG (mm Hg)
Li et al. 81	26‐gauge	9–11	Male	25.5–27	1. 127 2. 131	Not shown
Qian et al. 83	26‐gauge	9–11	Male	25.5–27	117	65–145
Bai et al. 37	26‐gauge	16	Both	Not reported	83	50–112
Ye et al. 56	27‐gauge	6–8	Male	Not reported	1. 36.9 2. 36.8	Not shown
Paulke et al. 50	27‐gauge	9–15	Both	Not reported	62	37–100
Liu et al. 45	27‐gauge	Not reported	Both	Not reported	1. 57 2. 59 3. 53 4. 57	1.49–67 2. 40–79 3. 38–70 4. 42–69
Strom et al. 96	27‐gauge	16	Male	Not reported	Not shown	23–145
Li et al. 14	27‐gauge	8	Both	Not reported	72	Not shown
Funk et al. 24	27‐gauge	8	Not reported	Not reported	71	Not shown
Pavlaki et al. 29	27‐gauge	9–12	Female	Not reported	60	Not shown
Liu et al. 28	25‐gauge	8–16	Not reported	Not reported	1. 75 2. 60	1. 64–81 2. 55–71
Ragni et al. 30	27‐gauge	8	Male	Not reported	74–78	Not shown
Apaydin et al. 18	Not reported	Not reported	Not reported	Not reported	91	87–94

For articles with multiple separate TAC experiments, values are numbered in order of appearance in the article. If >1 TAC group was included in a single experiment, we estimated the combined average for the groups. We did not estimate PIPG error bars which were typically used to show SD or SEM from figures. PIPG indicates peak instantaneous pressure gradient; and TAC, transverse aortic constriction.

DISCUSSION

The TAC model remains a cornerstone for investigating pressure overload‐induced cardiac remodeling and identifying potential therapeutic targets for human heart disease. To maximize the translational potential and prioritize the most promising interventions for clinical trials, our field must optimize this model's experimental reproducibility and relevance to human disease. In this review, we propose a set of TAC quality metrics based on literature review, scientific statements from the American Heart Association, and our own clinical and experimental experience. ¹ , ² , ⁴ , ⁶ , ⁷ , ¹⁰ , ¹⁰⁰ , ¹⁰¹ , ¹⁰² , ¹⁰³ , ¹⁰⁴ , ¹⁰⁵ Although we recognize that additional important experimental considerations may not have been captured by our quality metrics, we hope that our work stimulates broader discussion on how to improve the design and execution of TAC studies.

Reporting Pressure Gradients

In clinical cardiology, pressure gradients from aortic stenosis or other pressure‐overload lesions are essential for determining prognosis and guiding treatment decision. ⁸ Analogously, in mouse TAC models, the magnitude of the pressure gradient created by aortic constriction is a critical determinant of the subsequent cardiac response. Variations in TAC pressure gradients directly influence the severity of systolic and diastolic dysfunction, myocardial fibrosis, and other heart failure findings. ⁴ Thus, accurate quantification of TAC pressure gradients is necessary to enable meaningful comparison across studies and to ensure reproducibility, and we therefore considered this a key quality metric.

Despite its importance, only 24.7% of articles reported TAC pressure gradients using PIPG, carotid flow ratios, or invasive LV pressure measurements. This finding aligns with the 2021 meta‐analysis by Bosch et al., which reported that only 33% of TAC studies included pressure gradient data. ⁷ Accordingly, our analysis of more recent publications demonstrate that reporting of this critical information has not improved, even in high‐impact cardiovascular journals.

A common assumption among investigators is that using a specific needle gauge yields consistent pressure gradients across studies. However, our findings contradict this assumption. We observed substantial variability in the average PIPGs across studies, with higher gradients sometimes resulting from larger‐diameter needles (eg, 26‐gauge versus 27‐gauge). Considerable variability in PIPGs was also present within individual experiments. In our own experience, we achieved comparable average PIPGs (∼70–96 mm Hg) using needle sizes as large as 22‐ or 23‐gauge, albeit in mostly older mice. ¹⁰² , ¹⁰³ These discrepancies likely reflect the influence of additional biological and technical variables such as mouse age, sex, strain, surgical technique, and operator experience. Accordingly, our findings show the need to measure and report TAC pressure gradients for all experimental mice.

Fortunately, PIPGs or carotid flow ratios can be measured in most laboratories, as echocardiography is already widely used to assess cardiac function. Among these approaches, we recommend laboratories obtain PIPGs because they were the most used method in our review, and standardizing measurement techniques across laboratories would facilitate more reliable comparisons and improve reproducibility. Measuring the high Doppler velocities associated with TAC requires only minor imaging adjustments: using a lower‐frequency transducer, adjusting image depth and velocity scale, and shifting the pulse‐wave Doppler baseline. For example, we routinely measure PIPGs up to ∼90 mm Hg using a VisualSonics Vevo F2 (Fujifilm VisualSonics, Inc., Toronto, ON, Canada) with a UHF22x transducer, whereas we use a UHF46x to measure cardiac function. Although invasive LV pressure measurements can provide an alternative method for assessing gradients, they are more technically demanding and less practical for routine use.

Another important consideration is the timing of the pressure gradient measurements. In humans, impaired LV function and reduced cardiac output can lead to an underestimation of pressure gradients and similar concerns could apply to TAC mice. ⁸ For this reason, we recommend measuring pressure gradients soon after TAC rather than waiting until the conclusion of the study, when cardiac function has potentially worsened. However, only one third of the articles that reported PIPGs or carotid flow ratios performed these measurements soon after TAC (≤7 days post TAC), underscoring the need for improvement in this practice.

Excluding Mice for Outlying TAC Pressure Gradients

Prior studies demonstrated a large range in TAC PIPGs within an experimental group including mice with values only slightly above those of sham‐operated controls. ⁵ , ⁶ Mice with low PIPGs are unlikely to develop significant hypertrophy, cardiac dysfunction, or pathological remodeling, thereby diluting treatment effects and obscuring true biological differences. Conversely, mice with excessively high PIPGs would have greater degrees of cardiac dysfunction, myocardial fibrosis, and other heart failure features. ⁴ These extremes create internal variability that undermines interpretability and reproducibility.

In our review, only 8% of articles excluded mice with low PIPGs, and just 1 study (1.1%) excluded mice with high pressure gradients (measured by carotid flow ratio). This finding raises concerns about reproducibility and reliability, particularly given that distinct cardiac phenotypes from graded increases in TAC pressure gradients. ⁴ Some articles in our review reported PIPG ranges as high as 122 mmHg ⁹⁶ suggesting substantial heterogeneity in cardiac findings within a single experimental group. It is likely that similar variability exists in studies that did not report PIPGs.

To enhance consistency and reproducibility, investigators should exclude mice with outlying low or high TAC pressure gradients. Although the minimum PIPG necessary to generate reliable pressure‐overload stress remains uncertain, available data suggest that a lower PIPG cutoff of ∼40 to 50 mm Hg may be reasonable. ⁴ No studies provided an empirical upper limit, but excluding mice whose PIPGs fall >~20 mm Hg above or below the group mean may help standardize cardiac phenotypes and reduce experimental noise. ⁴ We acknowledge that implementing such exclusions would require including additional mice to account for anticipated losses; however, this approach would enhance model reliability and potentially strengthen the translational relevance of TAC studies.

Minimizing Bias

The earlier review by Bosch et al. reported low rates of randomization (12%) and blinding (23%). ⁷ In contrast, our combined metric indicated that ∼70% of recent articles included at least 1 measure to minimize bias. Although this represents an overall improvement, substantial opportunity for further progress remains.

Quality Metrics for Clinical and Translational Relevance

Translation of findings from mouse TAC models to human disease remains inherently complex. Multiple conceptual frameworks could be applied to define translational quality metrics for TAC studies, and the field would benefit from the development of expert consensus guidelines. In this review, we defined our quality metrics based on the premise that TAC models should reproduce key features of human pressure‐overload heart disease, including the characteristic patterns of LV remodeling and the clinically relevant timing of therapeutic interventions. ¹

However, fundamental biological differences between species complicate the direct translation of findings from mouse TAC experiments to human therapies. In general, mouse TAC is considered useful for elucidating mechanisms of cellular signaling or structural remodeling, as well as for interrogating sex differences or gene‐specific effects that may influence the hypertrophic response. However, transcriptomic comparisons indicate that TAC and human heart failure share limited overlap in differentially expressed genes, although upstream regulatory networks are partially conserved. ¹⁰⁶ , ¹⁰⁷ , ¹⁰⁸ Transcriptional similarity is further influenced by mouse strain, and the duration of pressure overload after TAC may also modulate overlap with human heart failure, with early‐to‐intermediate time points showing greater correspondence than prolonged TAC. ¹⁰⁷ , ¹⁰⁸ Importantly, most available human cardiac transcriptome data sets come from explanted hearts at the time of transplantation, representing end‐stage disease. ¹⁰⁷ , ¹⁰⁸ In contrast, the mouse samples could capture earlier phases of compensated hypertrophy or mild dysfunction, making direct comparison challenging.

To our knowledge, no studies have systematically evaluated how the different remodeling stages (eg, compensated hypertrophy, early dysfunction, or severe cardiomyopathy) affect transcriptional alignment with human disease. Such analyses could refine the translational relevance of mouse TAC and better determine the experimental conditions most predictive of human therapeutic relevance. Given the practical challenges of obtaining human myocardial tissue outside of end‐stage heart failure, incorporating large animal pressure‐overload models could be helpful for capturing the molecular continuum from compensated hypertrophy to overt heart failure and to contextualize the translational potential from mouse TAC studies.

In our review, only 20.2% of articles performed an early assessment of cardiac function to determine whether their TAC models recapitulate the typical human trajectory of compensated hypertrophy to ventricular decompensation. Given the wide variation in average PIPGs among the few studies providing these data, it is likely that a meaningful proportion of TAC experiments may develop early cardiac dysfunction. Incorporating echocardiographic measurements of cardiac function and PIPGs within the first 7 days after TAC would allow investigators to better characterize their models and confirm alignment with anticipated remodeling trajectories. If early dysfunction is present, investigators could reduce the degree of aortic constriction in subsequent surgeries to better reproduce the characteristic pattern of human LV remodeling.

We also found that most studies (68.5%) initiated interventions before or concurrent with TAC surgery. This design limits the ability to distinguish effects on the initial response to acute pressure‐overload from those on established hypertrophy or the transition to heart failure. Because patients usually present with existing hypertrophy or mild cardiac dysfunction, targeting mechanisms active during these later stages is more likely to yield clinically meaningful insights.

Inclusion of Females

Inclusion of both male and female animals in preclinical research is increasingly emphasized to enhance the generalizability of scientific findings. However, meaningful integration of both sexes in TAC studies presents practical and biological challenges. Female mice often develop less hypertrophy and milder systolic dysfunction after TAC, which can complicate direct comparisons with males. ¹⁰⁹ , ¹¹⁰ Additionally, investigators frequently select male mice for mechanistic studies involving genetic modifications to avoid variability related to estrous cycling. Despite these challenges, incorporating females is essential to ensure equitable translation of cardiovascular therapies.

The prior meta‐analysis by Bosch et al. reported that only 13% of TAC studies included female mice. ⁷ Our contemporary analysis shows modest improvement, with 28.1% of articles incorporating females. There was substantial variability among the journals with range of female inclusion from 14.3% to 83.3%. NIH‐funded studies demonstrated significantly greater inclusion compared with non‐NIH funded studies, suggesting that federal policy has positively influenced sex inclusion. Nevertheless, the persistently low overall rate underscores the need for broader adoption of sex‐inclusive study designs.

Investigators may struggle to incorporate both sexes due to increased cost, time, and uncertainty regarding optimal experimental design. Broader systemic efforts may be necessary to make meaningful progress in this area. For example, journals could consider having more dedicated issues on sex differences in cardiovascular research or editors might provide incentives for studies that include females.

For investigators working with both sexes, we recommend careful attention to achieving comparable TAC pressure gradients. In our experience, female mice often require constrictions with slightly smaller diameter needles to achieve equivalent PIPGs to their male counterparts. ¹⁰² No articles in our review reported using sex‐specific needle gauges or assessed for PIPG differences between sexes. It would be valuable for other investigators to report whether they observe a similar sex‐related phenomenon in their TAC models.

Mouse Substrains

Mouse substrains respond to TAC with varying degrees of cardiac hypertrophy and heart failure. Of the common substrains used in TAC, C57BL/6N mice tend to develop more severe cardiac dysfunction compared with C57BL/6J. ¹¹¹ , ¹¹² Thus, mouse substrains are another key factor affecting interpretation of experimental results, reproducibility, and translational relevance. Although we did not assess this factor in our systematic review, it is critical for investigators to report the specific mouse substrain used in their experiments.

Limitations

Our analysis has several limitations. First, we classified a quality metric as negative if it was not explicitly reported, which may underestimate actual compliance. Second, our keyword‐based search strategy may have missed relevant data. For example, our search for interventions ≥7 days post TAC was challenging because there were no specific keywords to identify individual experiments and intervention timing. Third, we considered a metric present if it was met in at least 1 experiment within a multi‐experiment article, which may overstate overall adherence. Fourth, we did not reevaluate all the quality parameters proposed by Bosch et al. such as sample size calculations and adequate presentation of outcome data because journals often screen for these parameters. ⁷ Fifth, although we provide a rationale to support our time frame post TAC for assessing early cardiac function (≤7 days) or intervention timing (≥7 days), we recognize that different time frames could potentially improve the validity of these quality metrics. Finally, as noted, we likely overlooked other important experimental parameters affecting reproducibility and translational potential. Nevertheless, we hope that our quality metrics and analysis spur a discussion focused on improving TAC experiments to enhance translation to human therapies.

CONCLUSIONS

Despite widespread use of the TAC model, most studies do not report or perform key parameters that influence reproducibility and translational potential. Pressure gradient measurements, early cardiac function assessments, experiments with delayed intervention timing, and equitable inclusion of the sexes are often lacking. Adoption of quality metrics could improve the reproducibility of TAC studies and promote the translation of preclinical discoveries into effective therapies for human heart failure.

Sources of Funding

Research reported in this publication was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under award number NIH R01HL122546 to Aaron K. Olson.

Disclosures

None.

Supporting information

Data S1–S3

Tables S1–S2