Comparison of Referral Rates and Costs Using Fibrosis-4 and Enhanced Liver Fibrosis (ELF) Testing Strategies for Initial Evaluation of Metabolic Dysfunction-Associated Steatotic Liver Disease (MASLD) in a Veteran Population

Abstract

Background

Global metabolic dysfunction-associated steatotic liver disease (MASLD) prevalence is estimated at 30% and projected to reach 55.7% by 2040. In the Veterans Affairs (VA) healthcare system, an estimated 1.8 million veterans have metabolic dysfunction-associated steatohepatitis (MASH).

Methods

Adult patients at risk for MASLD in a VA healthcare system underwent Fibrosis-4 (FIB-4) and Enhanced Liver Fibrosis (ELF^®) testing. Referral rates and cost savings were compared among 6 noninvasive testing (NIT) strategies using these 2 tests independently or sequentially at various cutoffs.

Results

Enrolled patients (N = 254) had a mean age of 65.3 ± 9.3 years and mean body mass index (BMI) of 31.7 ± 6, 87.4% male: 78.3% were non-Hispanic/Latino, and 96.5% had type 2 diabetes mellitus (T2DM). Among the 6 evaluated strategies, using FIB-4 followed by ELF at a 9.8 cutoff yielded the highest proportion of patients retained in primary care without need of referral to hepatology clinic (165/227; 72.7%), and was associated with the lowest costs ($407.62). Compared to the FIB-4 only strategy, FIB-4/ELF with a 9.8 cutoff strategy resulted in 26% fewer referrals and 8.47% lower costs. In the subgroup of patients with BMI >32, there were 25.17% fewer referrals and costs were 8.31% lower.

Conclusions

Our study suggests that sequential use of ELF with a 9.8 cutoff following indeterminate FIB-4 tests results in lower referral rates and lower care costs in a veteran population at risk of MASLD. Adding ELF as a sequential test after indeterminate FIB-4 might help reduce the number of referrals and overall cost of care.

INTRODUCTION

The worldwide prevalence of metabolic dysfunction-associated steatotic liver disease (MASLD) is projected to increase to 55.7% by 2040, representing a significant global health issue (1). Defined by excessive (>5%) fat accumulation within the liver unrelated to alcohol consumption in conjunction with at least one of 5 cardiometabolic risk factors, MASLD is recognized as a leading progenitor of severe liver-related and cardiovascular complications. Considered to be the hepatic manifestation of metabolic syndrome, MASLD has intricate connections to obesity and metabolic disorders such as type 2 diabetes mellitus (T2DM): increasing prevalence is closely linked to increasing obesity and T2DM prevalence (2–9). This strong, bidirectional relationship has prompted healthcare providers and researchers to seek innovative methods for early identification, prognostication, and effective MASLD management in patients with metabolic-dysregulation indicators. Current management of early-stage MASLD mostly entails lifestyle modifications, although a therapy (Rezdiffra^TM) was recently FDA-approved for the treatment of patients with stage 2 or 3 fibrosis. Several other promising pharmacological treatments are in development, making early, cost-efficient diagnosis more valuable (10, 11).

Noninvasive tests (NITs) have emerged as promising liver fibrosis staging tools. NITs in conjunction with evidence-based care pathways can help establish patient risk profiles for individuals with or at risk of developing MASLD. In contrast to liver biopsy, NITs can provide earlier, safer, more accessible, and potentially more cost-effective ways to assess liver fibrosis and track disease progression (12–15).

NITs commonly used in clinical practice for fibrosis staging and monitoring include both blood-based tests and imaging. The most frequently used blood-based NITs are the fibrosis index (FIB-4) based on liver function tests [aspartate aminotransferase (AST), alanine aminotransferase (ALT)], platelet count, and patient age; and the Enhanced Liver Fibrosis (ELF^®) test that assesses 3 assays measuring direct markers of fibrogenesis (hyaluronic acid, N-terminal protein of procollagen type 3) and fibrinolysis (tissue inhibitor of metalloproteinase 1), and is currently the only FDA-authorized commercial assay. Other blood-based NITs include NAFLD fibrosis score (NFS), NIS4, NIS2+, and Fibrosure, however, FIB-4 and ELF are the only tests included in the European Association for the Study of the Liver (EASL), American Association for the Study of Liver Diseases (AASLD), and American Gastroenterology Association (AGA) guidelines (2, 12–19). Frequently used diagnostic imaging methods include vibration-controlled transient elastography (VCTE) and magnetic resonance elastography (MRE) (20). Despite method-specific limitations (accuracy, availability, accessibility, cost, lack of standardization, and limited long-term data on predictive ability), NITs offer valuable information about the extent of liver fibrosis, enabling informed decision-making for patient management and treatment. The EASL, the AASLD, and the AGA recommend a combination of NITs as an alternative to liver biopsy for primary fibrosis staging, patient follow-up, and management (16–19). However, there is currently no accepted consensus on which, if any, NIT-based strategies yield the best performance for early detection of advanced liver fibrosis (stages F3/F4). Lack of consensus results in unnecessary hepatology referrals of patients with a low likelihood of progression to advanced disease, and delayed referrals or long waiting times for patients with greater need.

Patients in primary care and endocrinology practices often do not have easy, cost-effective access to VCTE or MRE as these technologies are typically limited to specialty locations and tertiary care centers in predominantly major urban centers.

In this study, we focused on comparing blood-based NIT strategies easily accessible to primary and endocrine practices, regardless of location. As MASLD prevalence continues to rise and NITs become increasingly incorporated into clinical practice, it is important to compare the diagnostic performance and cost impact of employing NIT strategies for triaging hepatology referrals from primary and endocrinology-care. Our primary goal was to compare referral rates and associated costs of 6 blood-based NIT-based strategies using data collected in a real-world study.

MATERIALS AND METHODS

This cross-sectional study prospectively enrolled patients with T2DM and/or obesity recruited from primary care clinics in the Veteran Affairs Palo Alto Healthcare System (VAPAHCS, Palo Alto, CA). At-risk individuals with a body mass index (BMI) ≥30 and/or receiving medication for T2DM were identified using electronic medical records. Individuals were initially contacted via US mail, followed by telephone contact. All study participants underwent a preliminary standardized research visit performed by a trained investigator, which included history, physical exam, laboratory testing with FIB-4 and ELF, and completion of a detailed patient questionnaire. BMI was defined as the body weight (in kg) divided by height (in m) squared. Alcohol consumption was assessed using the Alcohol Use Disorders Identification Test (AUDIT-C). Other types of liver disease were systematically ruled out based on history and laboratory tests. Exclusion criteria included: cirrhosis or liver transplantation, hepatitis B or C, excessive alcohol use as determined by AUDIT-C, a life-limiting malignancy or life expectancy <10 years, type 1 diabetes, current pregnancy, or inability to consent for research (e.g., dementia). The study was approved by the VAPAHCS and Stanford Institutional Review Board. Written informed consent was obtained from all participants.

A subset of consecutive patients willing to undergo MRE were assessed based on a grant-limited prospective protocol. Our weighted sampling of 58 patients included 5 patients with high-risk FIB-4 (≥2.67), 30 patients with indeterminate FIB-4 (>1.30–2.67) and ELF ≥ 9.8, 9 patients with indeterminate FIB-4 and ELF <9.8, and 14 patients with low-risk FIB-4 (<1.3).

Data for the study were collected between May 31, 2022, and October 25, 2023 at VAPAHCS. This study and the primary clinical study were funded by Siemens Healthineers imaging and laboratory diagnostics divisions. Siemens Healthineers employees were involved in the study design and analysis.

NIT Strategies

Enrolled patients underwent baseline FIB-4 and ELF testing. Six NIT strategies were evaluated and grouped into 2 categories: 3 single-test strategies, and 3 two-test sequential strategies (Fig. 1). FIB-4 was calculated on site as previously described (21). ELF testing was completed on serum samples and performed in batches and automatically calculated using a single Siemens Atellica IM Analyzer at the Siemens Healthineers laboratory in Tarrytown, NY. We analyzed one single-test FIB-4 strategy with a cutoff of 1.3 (16) and 2 single-test ELF strategies at cutoffs of 9.8 (13) and 9.0 (22). In each the single-test strategies (strategies 1–3), patients below the respective cutoffs were considered at low risk for advanced fibrosis (F3/F4) (22). Conversely, patients with FIB-4 ≥ 1.3, or ELF ≥ 9.0 or ELF ≥ 9.8 were considered high risk.

In the 2-test strategies (strategies 4–6), patients were initially screened using FIB-4. A second test using ELF was applied sequentially if the initial FIB-4 test gave an indeterminate result (>1.30–2.67). We evaluated 3 different cutoffs for ELF following indeterminate FIB-4 scores: 7.7 per AASLD (19) and AACE (23) low-risk guidance; 9.8 per AACE high-risk guidance, per current VAPAHCS clinical practice and recommended by the manufacturer (24); and 9.0 as a balanced cutoff between these 2 previous cutoffs. For each of the 6 strategies, patients were categorized as either low or high risk for significant fibrosis (≥F2) based on scores below or above each respective cutoff.

In all scenarios, patients at low risk of advanced fibrosis (F3/F4) were advised to follow up with their primary care physician or endocrinologist for lifestyle modification counseling and monitoring as per current recommendations (19, 23). Patients considered at high risk of advanced fibrosis were referred to a hepatologist for additional testing and fibrosis staging. Figure 1 illustrates the referral pathways for the 1- and 2-test strategies.

Cost Components

Cost values were based on the Centers for Medicare and Medicaid Reimbursement Fee Schedule, corresponding to the Current Procedural Terminology (CPT) codes. Analyses were performed from a US healthcare perspective using Medicare reimbursement rates. We used a 1-year time horizon and captured costs related to resources used in the primary care/endocrinology triage and hepatology staging work-up, as described in Table 1, which lists cost components, associated costs, resource utilization units, and cost source using 2024 reimbursement rates. No discount rate or cost adjustment was used due to the short-term focus of the analysis. The annual cost of FIB-4 was set at $0 because its component tests were assumed to be available as part of routine blood work ordered for patients with suspected MASLD.

For patients who continued to receive care within the primary care clinic, expenses comprised the visit and test costs. By contrast, patients who were referred to hepatology incurred the initial primary care visit and test costs, in addition to hepatology and further staging test costs. The costs of complications, hospitalizations, or treatment related to diabetes, obesity, or other possible disease states were not considered because of the short-term focus of the analysis. Table 2 shows the cost calculation algorithms for patients remaining in primary care and those referred to hepatology.

Subgroup Analyses

We performed additional subgroup analyses of patients with (a) BMI >32, and (b) patients ≥65 years remaining in primary care or referred to hepatology using 2.0 as a lower FIB-4 threshold (25).

Statistical Analysis

Baseline demographic and outcome characteristics are presented as mean ± standard deviation (SD) or as frequencies and percentages. We calculated the percentage of patients according to each blood-based NIT strategy who continued to be managed by primary care providers following testing, and those referred to the hepatology clinic.

The subset of patients (N = 58) described previously underwent a noncontrast MRI exam with liver fat quantification and liver stiffness assessment using MRE at VAPAHCS based on FIB-4 and ELF scores (described above). Imaging was performed using a 3T research scanner (Siemens MAGNETOM Prisma 3T; Siemens Healthineers), and liver stiffness data were obtained using 2D MRE at 60 Hz. Acquired magnetic resonance images were interpreted by an experienced radiologist to quantify liver fat and stiffness, as described (26). MRE was performed to serve as a gold standard for diagnostic purposes. The diagnostic performance of each blood-based NIT strategy is reported in terms of sensitivity, specificity, accuracy, positive predictive value, and negative predictive value using MRE as an adjudicator. All statistical analyses were performed using R Statistical Software (v.4.3.1; R Core Team 2021).

RESULTS

Of 254 enrolled patients, 227 had both FIB-4 and ELF test results available by the time of data draw on October 31, 2023. The mean ± SD age of the cohort was 65.3 ± 9.3 years (range, 34–76 years); 87.4% were male, 78.3% were non-Hispanic White, the mean BMI was 31.7 ± 6.1 (range, 19.2–53.0), and 96.5% had a diagnosis of T2DM. FIB-4 data were available for 99.2% (252/254) of patients, and ELF data were available for 89.4% (227/254) of patients. MRE data were available for 22.8% (58/254) of patients because of the limited access to MRE in this study. Only a subset of patients was scheduled to undergo MRE examination. Two patients had missing FIB-4 and ELF data because they had just been enrolled and had not received a blood draw at the time of analysis. Twenty-seven patients had missing ELF test results because they were not yet available at the time of the analysis. The mean ± SD scores were: FIB-4 1.2 ± 0.7 (range, 0.26–4.6), ELF 9.9 ± 0.8 (range, 7.7–12.2), and MRE 2.6 ± 0.8 kPa (range, 1.6–7.0 kPa). Among all patients, FIB-4 was <1.3 for 63.1% (159/252), 1.3–<2.67 for 32.9% (83/252), and ≥2.67 for 4% (10/252). For ELF, 0.5% (1/227) of patients had a score <7.7, 15% (34/227) were between 7.7 and 9.0, 33% (75/227) were between 9.0 and 9.8, and 51.5% (117/227) had a score ≥9.8.

Referral Pathways and Associated Costs

Table 3 presents the data on the distribution of patients remaining in primary care vs those referred to hepatology, along with the corresponding associated costs for each NIT strategy.

The per-patient cost for each blood-based NIT strategy varied significantly, ranging from $407.62 using FIB-4 followed by ELF (9.8 cutoff), to $1089.70 using ELF as the sole test at a 9.0 cutoff. Similarly, primary care retention rates for blood-based NIT strategies varied, ranging from 72.7% (165/227) for FIB-4/ELF (9.8 cutoff) to 15.4% (35/227) using only ELF (9.0 cutoff). By contrast, FIB-4 alone (the previous standard of care at VAPAHCS before ELF test adoption) resulted in a 63.1% (93/252) primary care retention rate at a per-patient cost of $445.35 (16).

We gained additional insight by examining the per-patient costs for referral and nonreferral using each blood-based NIT strategy. The per-patient cost of referral was similar across all strategies because it was dominated by the high cost of the hepatology staging work-up. The most expensive per-patient cost of referral (ELF only) was 17% higher [($1241.10/$1064.91) – 1] than the least expensive strategy (FIB-4 only). There was greater variation in nonreferral per-patient costs: ELF only at any cutoff value was by far the most expensive and was 212% higher [($259.16/$82.97) – 1] than the least expensive strategy (FIB-4 only). FIB-4 only was the least expensive method as the FIB-4 test was considered cost free.

For both ELF only and FIB-4/ELF, increasing the ELF cutoff led to a decrease in the referral rate and per-patient costs because fewer patients had an ELF value greater than the cutoff threshold. Due to the strong NPV of FIB-4 under 1.3 (27, 28), the FIB-4/ELF strategy referred fewer patients to hepatology and at a lower per-patient cost than using ELF alone for referral, regardless of the ELF cutoff applied.

Subgroup Analyses

We performed an additional subgroup analysis of the patient population with BMI >32. In this subgroup, the FIB-4/ELF strategy at the 9.8 cutoff continued to be the least costly ($344.44) and retained more patients in primary care than any other strategy (73/94; 77.7%). By contrast, using FIB-4 alone resulted in retention of 70.2% (73/104) of patients in primary care at a per-patient cost of $375.66 (Table 4).

Comparing the entire and BMI >32 populations across the blood-based NIT strategies, the per-patient cost of all ELF-only strategies increased, whereas the per-patient cost for all other strategies decreased. This could possibly be because abdominal adiposity contributing to liver steatosis is more likely in patients with a higher BMI, increasing the probability of finding ELF sufficiently elevated to trigger referral to hepatology. By contrast, the FIB-4 test seemed to be less sensitive to BMI.

In addition, we performed a subgroup analysis of patients ≥65 years old using 2.0 as a lower threshold of FIB-4 (25). The results of the analysis are shown in the lower part of Tables 3 and 5. In this subgroup FIB-4 followed by ELF (9.8 cutoff) still had the highest primary care retention rate 90.7% (137/151) and the lowest cost ($182.64). This strategy also achieved the highest diagnostic accuracy (83.7%).

Diagnostic Performance

The diagnostic performance, including the positive predictive value and NPV for the single- and 2-test blood-based NIT strategies for detecting advanced fibrosis (Table 5), was assessed using MRE ≥3.6 kPa as the adjudicator. The sensitivity for diagnosing ≥F3 fibrosis using only FIB-4 was 100.0%, with 25.5% specificity, and with 29.3% diagnostic accuracy. The ELF-only strategy demonstrated 100.0% sensitivity, 15.5% specificity, and 19.0% diagnostic accuracy using 9.0 as the cutoff, whereas at the 9.8 cutoff, sensitivity decreased to 66.7%, while specificity increased to 32.7% and diagnostic accuracy increased to 34.5%.

When using FIB-4 and ELF sequentially, an ELF cutoff of 7.7 resulted in 100.0% sensitivity, 25.5% specificity, and 29.3% diagnostic accuracy, while ELF at the 9.0 cutoff resulted in 100.0% sensitivity, 29.1% specificity, and 32.8% diagnostic accuracy. Again, using the 9.8 ELF cutoff improved specificity (40.0%) and diagnostic accuracy (41.4%) at the expense of sensitivity (66.7%). Thus, the highest specificity and diagnostic accuracy were achieved when indeterminate FIB-4 was followed by ELF ≥9.8.

DISCUSSION

Patient access to VCTE or MRE before specialty referral can be hampered by location and costs. To substantiate a referral strategy enabling more efficient and equitable access to health care, we prospectively compared referral rates and associated costs of blood-based NIT strategies for determining MASLD risk in a real-world VA patient population using FIB-4 and ELF testing strategies as proposed in the recent AASLD (19) and AACE (23) practice guidance documents. The results of our study indicate that using FIB-4 followed by ELF at a 9.8 cutoff is the least costly and most accurate method among the strategies tested, and generated the lowest hepatology referral rate. This blood-based NIT strategy demonstrates an ability to appropriately risk-stratify patients with T2DM in this well-phenotyped prospective cohort. This finding was also true in the patient subgroup with either BMI >32, or ≥age 65 years using the >2.0 FIB-4 cutoff.

It was demonstrated previously that using ELF at the 9.8 cutoff as an alternative to initial transient elastography evaluation has acceptable performance and may be more feasible in health systems where access or availability is limited due to expense, population, or patient travel distance (29). As in previous studies, we found that utilizing FIB-4 as the first test in a 2-test strategy can help avoid follow-up testing by leveraging the NPV of FIB-4 <1.3 (22). Although using a 9.8 ELF cutoff is not included in the sequential algorithms proposed in the AASLD document (19), it is supported by prior studies (29–32) and the manufacturer’s guidance.

In cost-comparison studies conducted in the United Kingdom, Srivastava et al. and Crossan et al. found that serial FIB-4/ELF testing resulted in the lowest costs and the least referrals to hepatology among the strategies they evaluated (15, 33). Our results are consistent with these findings. Contrary to these studies, we found that using a single-test strategy, such as ELF only, is inefficient in reducing costs and referrals compared to the FIB-4 only strategy. Nevertheless, decreased referral rates should be weighed against the human and economic repercussions of missing advanced-stage MASLD, resulting in delayed hepatology referral and care. All diagnostic tests have the potential for generating false positive and/or false negative results. In liver fibrosis assessment, a false positive result conveys limited harm to the patient, including patient concern and stress, some additional diagnostic expenses, and potential risks and pain associated with additional procedures, such as liver biopsy. There is greater potential harm to patients associated with false negative results, mainly that advanced fibrosis is missed and mis-classified as no or mild fibrosis, delaying treatment and lifestyle counseling. This is especially consequential as the therapeutic landscape is rapidly changing with approval of the first drug for treating MASLD (Rezdiffra, Madrigal Pharmaceutical), and the promise of others currently in clinical trials. Even more critically, patients with undiagnosed cirrhosis are at risk for complications associated with decompensation and hepatocellular carcinoma. Early identification of these patients can aid appropriate management and potentially mitigate or delay these risks. Regardless, all patients with MASLD are advised to lose weight and adopt other lifestyle modifications regardless of disease severity.

Among the total patient cohort, when FIB-4 was used followed by ELF, the per-patient cost was lower and the percentage of patients remaining in primary care was higher than when using only FIB-4. This was also true in the subgroup of patients with a BMI >32, and those ≥65 years. This is an important finding because it shows that, counterintuitively, adding follow-up tests can actually decrease downstream costs by generating fewer referrals to hepatology with enhanced accuracy. Our findings agree with the current guidelines endorsing FIB-4 followed by either ELF or VCTE as preferred NIT strategies for risk stratification of individuals at risk for MASLD (19, 23, 34).

The primary strength of this study is its use of real-world patient data drawn from the VA patient population. Using real-world patient data helps demonstrate the value of NITs for evaluating patients at risk of MASLD in primary care and endocrinology settings, and reduction of downstream costs. This study had several notable limitations, however. First, we lacked biopsy data as a reference standard for staging fibrosis and diagnosing advanced fibrosis. Performing liver biopsies routinely in such a low-prevalence setting is unethical. Having MRE data on the subset of patients with high-, indeterminate-, and low-risk FIB-4 scores in combination with higher-risk and lower-risk ELF scores partially mitigates this limitation. Ideally, we would have MRE data on all patients in the study, however, this was precluded due to financial constraints imposed by the study grant. Second, our study utilized real-world data from a single center with a predominantly older male veteran population with T2DM. Thus, our findings might not be generalizable to other populations. We also did not test using other blood-based NITs, such as NIS4 or Fibrosure, as we wanted to limit the scope of this analysis only to blood-based NITs recommended by major hepatology guidelines (EASL, AASLD, and AGA). Finally, we only captured the direct costs associated with blood-based NITs associated with physician visits to primary care practitioners and hepatology referrals. We did not include other potential resource inputs, such as hospitalizations or emergency department visits following the initial evaluation, costs of missed diagnoses, long-term care costs, or impact on the quality of life.

Future studies comparing different NIT strategies across diverse data resources from multiple centers are needed to enable more generalizable findings applicable to a broader patient population at risk of MASLD for both short- and long-term assessments.

CONCLUSION

Our study suggests that, among the strategies we tested, using ELF at a 9.8 cutoff following indiscriminate FIB-4 results (>1.3–2.67) results in the lowest rate of referrals and per-patient costs. In the total population using 2-step testing compared to FIB-4 alone yielded direct cost savings of 8.47% and 26% fewer referrals.

Nonstandard Abbreviations: MASLD, metabolic dysfunction-associated steatotic liver disease; VA, Veterans Affairs; MASH, metabolic dysfunction-associated steatohepatitis; FIB-4, fibrosis-4; ELF, enhanced liver fibrosis; NIT, noninvasive test; BMI, body mass index; T2DM, type 2 diabetes mellitus; AST, aspartate aminotransferase; ALT, alanine aminotransferase; EASL, The European Association for the Study of the Liver; AASLD, The American Association for the Study of Liver Diseases; AGA, The American Gastroenterology Association; VCTE, vibration-controlled transient elastography; MRE, magnetic resonance elastography; VAPAHCS, Veteran Affairs Palo Alto Healthcare System; AUDIT-C, alcohol use disorders identification test, concise; CPT, current procedural terminology.

Author Contributions:The corresponding author takes full responsibility that all authors on this publication have met the following required criteria of eligibility for authorship: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; (c) final approval of the published article; and (d) agreement to be accountable for all aspects of the article thus ensuring that questions related to the accuracy or integrity of any part of the article are appropriately investigated and resolved. Nobody who qualifies for authorship has been omitted from the list.

Samrat Yeramaneni (Conceptualization-Equal, Data curation-Equal, Investigation-Equal, Methodology-Equal, Project administration-Equal, Resources-Equal, Software-Equal, Validation-Equal, Writing—original draft-Equal, Writing—review & editing-Equal), Stephanie Chang (Conceptualization-Equal, Data curation-Equal, Investigation-Equal, Methodology-Equal, Resources-Equal, Writing—review & editing-Equal), Ramsey Cheung (Conceptualization-Equal, Data curation-Equal, Formal analysis-Equal, Investigation-Equal, Methodology-Equal, Resources-Equal, Supervision-Equal, Writing—original draft-Equal, Writing—review & editing-Equal), Donald Chalfin (Conceptualization-Equal, Data curation-Equal, Formal analysis-Equal, Investigation-Equal, Methodology-Equal, Resources-Equal, Validation-Equal, Writing—review & editing-Equal), Kinpritma Sangha (Conceptualization-Equal, Data curation-Equal, Formal analysis-Equal, Investigation-Equal, Methodology-Equal, Project administration-Equal, Resources-Equal, Software-Equal, Visualization-Equal, Writing—review & editing-Equal), H. Roma Levy (Conceptualization-Equal, Data curation-Equal, Formal analysis-Equal, Investigation-Equal, Methodology-Equal, Project administration-Equal, Resources-Equal, Supervision-Equal, Validation-Equal, Writing—original draft-Equal, Writing—review & editing-Equal), Artem Boltyenkov (Conceptualization-Equal, Data curation-Equal, Formal analysis-Equal, Investigation-Equal, Methodology-Equal, Project administration-Equal, Resources-Equal, Software-Equal, Supervision-Equal, Validation-Equal, Writing—original draft-Equal, Writing—review), and & editing-Equal)

Authors’ Disclosures or Potential Conflicts of Interest:Upon manuscript submission, all authors completed the author disclosure form.

Research Funding: This study was supported by Siemens Healthineers.

Disclosures: D.B. Chalfin and A.T. Boltyenkov are employees of Siemens Healthcare Diagnostics Inc. K. Sangha is an employee of Siemens Medical Solutions USA Inc. A.T. Boltyenkov, H.R. Levy, and K. Sangha are shareholders of Siemens Healthineers. S.T. Chang and R.C. Cheung receive research support from Siemens Healthineers. H.R. Levy is a previous employee of Siemens Healthcare Diagnostics Inc.

Role of Sponsor: The funding organization played a direct role in the design of study and preparation of manuscript. The funding organizations played no role in the choice of enrolled patients, review and interpretation of data, or final approval of manuscript.

References