Public availability of trial-design and clinical-context information in contemporary lipoprotein(a)-targeted cardiovascular trials: A cross-sectional meta-research study

Abstract

Background

Direct lipoprotein(a) [Lp(a)]-lowering therapies are advancing toward cardiovascular outcomes testing, but the amount of publicly available trial-design and clinical-context information varies across development stages.

Objective

To characterize contemporary patient-enrolling direct Lp(a)-targeted pharmacologic trials and describe fixed-date public availability of predefined trial information.

Methods

We conducted a cross-sectional meta-research study anchored in ClinicalTrials.gov with linked public protocols, statistical analysis plans, design/baseline reports, supplements, and peer-reviewed manuscripts available through May 2, 2026. Contemporary trials started on or after January 1, 2017. Two reviewers independently abstracted 12 prespecified public-information elements. Elements were classified as publicly identified or not publicly identified; no composite score or trial ranking was used.

Results

Fourteen trials met inclusion criteria, representing 5 agents, 3 therapeutic platforms, and 56,983 planned or reported participants. Seven trials had Lp(a) biomarker primary endpoints, 5 had clinical cardiovascular outcomes, 1 assessed apheresis burden, and 1 assessed calcific aortic valve stenosis progression. Seven trials were published/protocol-supported and 7 registry-only. Cardiovascular/risk phenotype, prevention or special-population setting, and geographic scope were identified in all trials. Other elements were identifiable less consistently: Lp(a) unit (13/14), background lipid-lowering therapy (9/14), numeric Lp(a) threshold (8/14), PCSK9/inclisiran handling (8/14), demographic information (8/14), CKD/eGFR criteria (6/14), heart failure criteria (6/14), diabetes criteria (5/14), and LDL-C context (3/14).

Conclusions

Public availability varied across contemporary Lp(a)-targeted trials at the data lock, particularly for registry-only programs. These findings describe fixed-date public retrievability and should not be interpreted as protocol deficiency, registry noncompliance, trial quality, efficacy, safety, or regulatory readiness.

1. Introduction

Lipoprotein(a) [Lp(a)] is an apolipoprotein B-containing lipoprotein composed of an LDL-like particle covalently bound to apolipoprotein(a). Elevated Lp(a) is largely genetically determined, relatively stable across the life course, and associated with atherosclerotic cardiovascular disease, calcific aortic valve stenosis, and cardiovascular mortality [1], [2], [3], [4], [5]. Lp(a) concentrations are minimally modified by lifestyle intervention and only modestly reduced by currently available lipid-lowering therapies [3], [4], [5]. The use of both mg/dL and nmol/L further complicates interpretation because fixed conversion between mass and molar units is imprecise across apolipoprotein(a) isoform sizes [6].

Direct Lp(a)-targeted development now includes antisense oligonucleotides, small interfering RNA therapies, and oral small-molecule inhibitors of Lp(a) particle formation. Phase 2 and early clinical studies of pelacarsen, olpasiran, lepodisiran, zerlasiran, and muvalaplin have demonstrated substantial pharmacologic Lp(a) lowering, and several programs have advanced to event-driven cardiovascular outcomes trials [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18]. Before outcome results and full public documentation are available, registry records and linked documents define what external users can retrieve about who is being studied and in what treatment context.

Variation in public trial information is not unique to Lp(a). Meta-research using ClinicalTrials.gov has identified challenges related to record maturity, evolving fields, incomplete results, and discordance across public sources [19], [20], [21]. Drug-class audits have also examined safety reporting in PCSK9 inhibitor trials and demographic representation in glucagon-like peptide-1 trials [22], [23]. Lp(a) provides a timely case study because multiple therapeutic modalities are advancing simultaneously, thresholds and units vary, and several large outcomes programs remain in progress.

Prior Lp(a) reviews have focused mainly on biology, pharmacology, efficacy, and trial design [14], [15], [16], [17], [18]. The added value of the present analysis is a reproducible trial-level audit of what was publicly retrievable at a fixed date, an explicit separation of public reporting completeness from trial quality, and an expert-informed candidate checklist for future registrations and publications. We aimed to characterize the contemporary direct Lp(a)-targeted trial landscape, describe the public availability of these elements, distinguish public nonidentification from true absence in a sponsor protocol, and provide a reproducible baseline that can be reassessed as ongoing programs mature.

2. Methods

2.1. Study design, protocol, and ethical considerations

We conducted a cross-sectional, trial-level meta-research study of publicly available registry records and linked aggregate documents. The unit of analysis was the trial. The analytic protocol, eligibility criteria, source hierarchy, abstraction template, and coding dictionary were developed before final data abstraction, but the study was not prospectively registered or deposited in an external repository. The dated coding dictionary and search audit trail are provided in Supplementary Tables S1 and S2.

The study was designed to assess fixed-date public retrievability of predefined trial-design and clinical-context information. It did not assess therapeutic efficacy, safety, comparative value, risk of bias, trial conduct quality, cost-effectiveness, payer coverage, guideline adoption, or readiness for regulatory approval. Reporting was informed by STROBE, RECORD, PRISMA-S, CONSORT, SPIRIT, CONSORT-Outcomes, and SPIRIT-Outcomes principles [24], [25], [26], [27], [28], [29], [30]. These guidelines were used to improve transparency of the present report and were not used to validate the predefined information domains. Because the study used only public registry records and aggregate public documents and involved no identifiable private information or participant contact, institutional review board review and informed consent were not required. The coding dictionary, search audit trail, and trial-by-domain abstraction matrix are provided in Supplementary Tables S1, S2, and S9.

2.2. Data-source appropriateness and study identification

ClinicalTrials.gov served as the registry anchor because it provides an NCT-level public record that can be linked reproducibly to other public trial documents. The 12 predefined elements evaluated in this study were selected for descriptive external interpretation; they were not derived as a list of mandatory ClinicalTrials.gov registration fields and were not used as a compliance checklist. This distinction is particularly important for ongoing trials: a registry entry may precede public protocols, statistical analysis plans, baseline reports, and primary publications by years; records can change with protocol amendments; and information not identified in the registry may exist in nonpublic sponsor documents or become public later. We therefore used a fixed data lock, retained a source audit trail, classified the public source category, and used the term “not publicly identified” rather than inferring true absence.

The final search update and data lock were May 2, 2026. ClinicalTrials.gov searches covered condition, intervention, developmental-name, and acronym terms for lipoprotein(a), Lp(a), pelacarsen, TQJ230, AKCEA-APO(a)-LRx, IONIS-APO(a)-LRx, ISIS-APO(a)Rx, olpasiran, AMG 890, lepodisiran, LY3819469, zerlasiran, SLN360, muvalaplin, LY3473329, Lp(a)HORIZON, OCEAN(a), OCEAN(a)-DOSE, OCEAN(a)-Outcomes, OCEAN(a)-PreEvent, ALPACA, ALPACAR-360, KRAKEN, ACCLAIM-Lp(a), MOVE-Lp(a), Lp(a)FRONTIERS, and ADD-VANTAGE. No language, country, phase, recruitment-status, or results-status restriction was applied during candidate identification. The complete term set, fields searched, date, and restrictions are reported in Supplementary Table S2.

PubMed was searched using the agent names, developmental names, trial acronyms, and NCT identifiers listed in Supplementary Table S2 to identify primary reports, design or baseline reports, protocols, and supplements. ClinicalTrials.gov Study Documents were checked for protocols, statistical analysis plans, and appendices. Conference abstracts and sponsor pages were considered only when trial-specific and linked to an NCT identifier; they were used for discovery or corroboration and did not override a more detailed peer-reviewed report, posted protocol, or statistical analysis plan.

2.3. Source hierarchy and conflict resolution

Candidate records were deduplicated by NCT identifier. For relatively static design features, sources were prioritized in the following order: public protocol or statistical analysis plan; peer-reviewed primary report and supplementary material; peer-reviewed design or baseline report; ClinicalTrials.gov record; and trial-specific conference abstract or sponsor page. For dynamic fields such as recruitment status, planned enrollment, and site listings, the most recent dated ClinicalTrials.gov entry available at the data lock was used.

When sources conflicted, the value from the higher-priority source was retained and the discrepancy was recorded. If the conflict remained unresolved or the available text was too ambiguous to support a determination, the element was classified as not publicly identified. Public non-peer-reviewed materials were not used to infer efficacy or safety conclusions.

2.4. Eligibility criteria and definition of contemporary trials

Eligible studies were patient-enrolling interventional trials evaluating a direct pharmacologic therapy designed to lower Lp(a) or inhibit Lp(a) particle formation. “Contemporary” was operationally defined as a study start date on or after January 1, 2017, through the May 2, 2026 data lock. Eligible agents were pelacarsen or its direct precursor formulation, olpasiran, lepodisiran, zerlasiran, and muvalaplin.

We excluded studies enrolling only healthy volunteers, first-in-human or pharmacokinetic/pharmacodynamic-only studies without broader patient-enrolling design relevance, renal-impairment, QT, drug-drug interaction, or other ancillary pharmacology studies, open-label extensions, observational studies, apheresis-only studies without a direct pharmacologic Lp(a)-targeted intervention, and studies in which Lp(a) was only an exploratory biomarker or covariate for a therapy not designed to target Lp(a). Biomarker-primary trials of direct Lp(a)-targeted agents were retained because they characterize dose, threshold, treatment context, and special-population features within the direct therapeutic development program.

2.5. Data abstraction, public source category, and reproducibility

Two reviewers independently screened candidate records and abstracted trial-level variables using a structured template and coding dictionary (Supplementary Table S1). Disagreements were resolved by source review and consensus, with third-reviewer adjudication when required. Variables included trial identifiers, therapeutic platform, phase, status, planned or reported enrollment, prevention setting, qualifying phenotype, Lp(a) threshold and unit, background lipid-lowering therapy, LDL-C context, PCSK9 inhibitor or inclisiran handling, CKD/eGFR criteria, diabetes criteria, heart failure criteria, endpoint structure, geography, and demographic reporting.

Trials were classified as published or protocol-supported when at least one detailed public source beyond the registry was available, including a primary report, design or baseline report, public protocol, statistical analysis plan, or detailed supplement. Trials were classified as registry-only when ClinicalTrials.gov was the only detailed public design source. Source maturity was treated as a characteristic of documentation, not a characteristic of trial quality. Independent pre-consensus item-level ratings were not retained in an analyzable form; therefore, a valid inter-reviewer agreement statistic could not be calculated retrospectively. To maximize reproducibility despite this limitation, Supplementary Table S1 provides operational definitions and Supplementary Table S9 provides the complete trial-by-domain availability matrix.

2.6. Predefined public-information domains and interpretation

We prespecified 12 elements for descriptive abstraction. Six described core trial design: numeric Lp(a) threshold, Lp(a) unit, qualifying ASCVD/CVD or risk phenotype, classifiable prevention or special-population setting, background lipid-lowering therapy, and geographic scope. Six described clinical context: LDL-C criterion or optimization context, PCSK9 inhibitor or inclisiran handling, CKD/eGFR criteria, diabetes criteria, heart failure criteria, and demographic reporting or design-specific demographic eligibility [1], [2], [3], [4], [5], [6], [14], [15], [16], [19], [20], [21], [24], [25], [26], [27], [28], [29], [30], [31].

Each element was classified as publicly identified when the prespecified source set contained sufficient information for an unambiguous trial-level determination, or as not publicly identified when it did not. Because these elements differ in clinical meaning, expected timing of disclosure, and relationship to registry requirements, we did not combine them into a composite score, assign weights or cut points, rank trials, or assign qualitative completeness categories. The 12 elements are an investigator-developed abstraction framework rather than a validated reporting instrument. Supplementary Table S12 summarizes the stage-appropriate interpretation of these elements and emphasizes that they are descriptive abstraction fields rather than mandatory ClinicalTrials.gov fields.

Endpoint definitions and components were summarized descriptively in Supplementary Table S6. Endpoint adjudication methods, sample-size justification, patient and public involvement, the clinical appropriateness of eligibility exclusions, and actual population representativeness were not part of the 12-element abstraction framework. Table 3 presents stage-appropriate candidate elements for future public documents; it is explicitly not a validated or exhaustive reporting standard.

Table 3.

Expert-informed candidate elements for stage-appropriate public trial documents.

Domain	Relevance to external interpretation	Observed public availability at data lock	Candidate stage-appropriate public information
Lp(a) threshold and assay context	Defines the studied Lp(a) population and supports interpretation across mass and molar units.	Numeric threshold: 8/14; unit: 13/14.	When applicable to the document stage, report assay method, unit, numeric eligibility threshold, baseline distribution, and achieved Lp(a) values.
ASCVD/risk phenotype and prevention setting	Clarifies whether evidence applies to secondary prevention, high-risk primary prevention, or special populations.	Qualifying phenotype and prevention/special-population setting: 14/14.	Define qualifying ASCVD/CVD events, risk criteria, and prevention or special-population setting.
Background LDL-C therapy and residual-risk context	Helps readers understand background lipid management and treatment sequencing.	Background lipid-lowering therapy: 9/14; LDL-C criterion/context: 3/14; PCSK9/inclisiran handling: 8/14.	Describe LDL-C entry or optimization context and background statin, ezetimibe, PCSK9 inhibitor, inclisiran, or apheresis use when relevant.
Comorbidity eligibility	Supports later assessment of applicability to patients with common comorbidities.	CKD/eGFR: 6/14; diabetes: 5/14; heart failure: 6/14.	Report clinically relevant renal, diabetes, and heart-failure eligibility criteria in protocols and final reports when applicable.
Endpoint structure	Clarifies what benefit or biologic effect the trial is designed to measure.	7 biomarker; 5 clinical outcomes; 1 apheresis burden; 1 CAVS progression.	Provide full endpoint definitions and, at the appropriate protocol or results stage, adjudication procedures, component outcomes, follow-up, and absolute risks when applicable.
Participant characteristics and representation	Supports assessment of generalizability after enrollment data become available.	Demographic reporting or design-specific demographic eligibility: 8/14; geographic scope: 14/14.	In baseline or final reports, provide age, sex, race, ethnicity, geography, and representation goals or other relevant characteristics when available.

Note: These elements are expert-informed suggestions derived from this descriptive audit and existing reporting principles. They are not a validated or exhaustive reporting standard, are not a ClinicalTrials.gov compliance checklist, and do not imply that every element should be present in a registry record at trial initiation. The appropriate timing and document for disclosure depend on study stage and context.

2.7. Statistical analysis

Analyses were descriptive. Trial characteristics and element availability were summarized using counts and percentages, with exact 95% binomial confidence intervals for element-level proportions. No hypothesis tests were used to compare agents, source categories, or endpoint classes. Source category and endpoint class were described to contextualize which public documents were available at the data lock, not as measures of reporting performance. Analyses were performed at the trial level. Descriptive element-availability summaries stratified by public-source category and primary endpoint class are provided in Supplementary Tables S10 and S11. No between-group hypothesis testing was performed.

3. Results

3.1. Trial cohort and therapeutic platforms

Fourteen trials met inclusion criteria (Table 1; Supplementary Tables S3–S8). They evaluated 5 agents across 3 platforms: pelacarsen or its precursor in 6 trials, small interfering RNA therapies in 6 trials, and muvalaplin in 2 trials. Planned or reported enrollment totaled 56,983 participants and ranged from 51 to 17,300.

Table 1.

Analysis cohort of contemporary direct Lp(a)-targeted therapeutic trials.

NCT identifier	Trial/acronym	Agent	Therapeutic platform	Phase	Status	Enrollment, n	Population	Prevention setting	Comparator	Endpoint class	Public source basis
NCT03070782	ISIS 681257-CS6/AKCEA-APO(a)-LRx	Pelacarsen precursor/AKCEA-APO(a)-LRx	GalNAc-conjugated antisense oligonucleotide targeting LPA mRNA	Phase 2	Completed	286	Established CVD with elevated Lp(a)	Secondary prevention	Saline placebo	Lp(a) biomarker	Results publication
NCT04023552	Lp(a)HORIZON	Pelacarsen/TQJ230	GalNAc-conjugated antisense oligonucleotide targeting apolipoprotein(a) mRNA	Phase 3	Enrollment complete/active follow-up	8323	Established CVD with Lp(a) ≥70 mg/dL	Secondary prevention	Matching placebo	Clinical CV outcomes	Registry + design/baseline
NCT05305664	Lp(a)FRONTIERS APHERESIS	Pelacarsen/TQJ230	GalNAc-conjugated antisense oligonucleotide targeting apolipoprotein(a) mRNA	Phase 3	Completed	51	Established CVD, isolated elevated Lp(a), normal LDL-C, and weekly lipoprotein apheresis	Special population/secondary prevention/apheresis-dependent disease	Matching placebo	Apheresis-burden	Results publication + registry
NCT05646381	Lp(a)FRONTIERS CAVS	Pelacarsen/TQJ230	GalNAc-conjugated antisense oligonucleotide targeting apolipoprotein(a) mRNA	Phase 2	Recruiting	502 planned	Calcific aortic valve stenosis with elevated Lp(a)	Special population/valve disease	Matching placebo	Imaging surrogate	Registry-only
NCT06267560	Pelacarsen in US Black/African American and Hispanic Participants With Elevated Lp(a) and Established ASCVD	Pelacarsen/TQJ230	GalNAc-conjugated antisense oligonucleotide targeting apolipoprotein(a) mRNA	Phase 3/IIIb	Completed	422	US Black/African American and Hispanic participants with established ASCVD and elevated Lp(a)	Secondary prevention/equity-focused	Placebo	Lp(a) biomarker	Registry-only
NCT06813911	ADD-VANTAGE	Pelacarsen/TQJ230	GalNAc-conjugated antisense oligonucleotide targeting apolipoprotein(a) mRNA	Phase 3	Recruiting	340 planned	Established ASCVD, elevated Lp(a), elevated LDL-C, and background inclisiran	Secondary prevention/combination lipid therapy	Placebo	Lp(a) biomarker	Registry-only
NCT04270760	OCEAN(a)-DOSE	Olpasiran/AMG 890	GalNAc-conjugated siRNA targeting LPA/apo(a) synthesis	Phase 2	Completed	281	Established ASCVD with Lp(a) >150 nmol/L	Secondary prevention	Matching placebo	Lp(a) biomarker	Results publication
NCT05581303	OCEAN(a)-Outcomes	Olpasiran/AMG 890	GalNAc-conjugated siRNA targeting LPA/apo(a) synthesis	Phase 3	Active, not recruiting	7297	ASCVD with elevated Lp(a)	Secondary prevention	Placebo	Clinical CV outcomes	Registry-only
NCT07136012	OCEAN(a)-PreEvent	Olpasiran/AMG 890	GalNAc-conjugated siRNA targeting LPA/apo(a) synthesis	Phase 3	Recruiting	11,000 planned	Participants at risk for a first major cardiovascular event with elevated Lp(a)	High-risk primary prevention/pre-event	Placebo	Clinical CV outcomes	Registry-only
NCT05565742	ALPACA	Lepodisiran/LY3819469	Dicer-substrate siRNA targeting hepatic LPA mRNA/apo(a) synthesis	Phase 2	Completed	320	Adults ≥40 years with Lp(a) ≥175 nmol/L and high cardiovascular risk enrichment	Mixed high-risk primary prevention and secondary prevention	Placebo	Lp(a) biomarker	Results publication
NCT06292013	ACCLAIM-Lp(a)	Lepodisiran/LY3819469	Dicer-substrate siRNA targeting hepatic LPA mRNA/apo(a) synthesis	Phase 3	Active, not recruiting	17,300	Elevated Lp(a) with cardiovascular disease or risk of MI/stroke	Mixed secondary prevention + high-risk primary prevention	Placebo	Clinical CV outcomes	Registry-only
NCT05537571	ALPACAR-360	Zerlasiran/SLN360	GalNAc-conjugated siRNA targeting hepatic LPA mRNA/apo(a) synthesis	Phase 2	Completed	178 analyzed/180 randomized	Stable ASCVD with Lp(a) ≥125 nmol/L	Secondary prevention	Placebo	Lp(a) biomarker	Results publication
NCT05563246	KRAKEN	Muvalaplin/LY3473329	Oral small-molecule inhibitor of Lp(a) particle formation; disrupts apo(a)-apoB interaction	Phase 2	Completed	233	Lp(a) ≥175 nmol/L with ASCVD, diabetes, or familial hypercholesterolemia	Mixed high-risk primary prevention and secondary prevention	Placebo	Lp(a) biomarker	Results publication
NCT07157774	MOVE-Lp(a)	Muvalaplin/LY3473329	Oral small-molecule inhibitor of Lp(a) particle formation; disrupts apo(a)-apoB interaction	Phase 3	Recruiting	10,450 planned	Elevated Lp(a) with cardiovascular disease or risk of MI/stroke	Mixed secondary prevention + high-risk primary prevention	Placebo	Clinical CV outcomes	Registry-only

Note: Trial status and enrollment reflect publicly available information at the May 2, 2026 data lock. Enrollment values are planned, randomized, analyzed, or enrolled counts as reported in public sources. The public source basis describes the documentation available at the data lock and is not a measure of trial quality.

Abbreviations: ASCVD = atherosclerotic cardiovascular disease; CAVS = calcific aortic valve stenosis; CVD = cardiovascular disease; LDL-C = low-density lipoprotein cholesterol; Lp(a) = lipoprotein(a); MI = myocardial infarction; siRNA = small interfering RNA.

Six trials were phase 2 and 8 were phase 3 or phase 3b. Seven trials were completed or had publicly reported results, and 7 were recruiting, active but not recruiting, or in active follow-up at the data lock. Seven trials were published or protocol-supported and 7 were registry-only. Four of the 5 trials with clinical cardiovascular outcomes as the primary endpoint were registry-only at the data lock.

3.2. Eligibility, prevention settings, and endpoints

The cohort included secondary-prevention trials, mixed-risk or high-risk primary-prevention programs, apheresis-dependent disease, calcific aortic valve stenosis, an equity-focused enrollment design, and a combination-therapy setting with background inclisiran. A qualifying cardiovascular or risk phenotype and a classifiable prevention or special-population setting were publicly identified in all 14 trials.

Seven trials used Lp(a) biomarker reduction as the primary endpoint, 5 used clinical cardiovascular outcomes, 1 used apheresis-burden reduction, and 1 used imaging or physiologic progression of calcific aortic valve stenosis. Endpoint definitions and components are shown in Supplementary Table S6. Adjudication methods were not part of the 12-element framework and were not reduced to a single binary item across these heterogeneous endpoint classes.

3.3. Public availability of predefined elements

A numeric Lp(a) eligibility threshold was publicly identified in 8 of 14 trials (57.1%), and an Lp(a) unit in 13 of 14 (92.9%). Background lipid-lowering therapy was described in 9 of 14 (64.3%), whereas an LDL-C eligibility criterion or optimization context was identified in 3 of 14 (21.4%). PCSK9 inhibitor or inclisiran handling was identified in 8 of 14 (57.1%).

CKD/eGFR criteria were publicly identified in 6 of 14 trials (42.9%), diabetes criteria in 5 of 14 (35.7%), and heart failure criteria in 6 of 14 (42.9%). Demographic reporting or design-specific demographic eligibility was identified in 8 of 14 trials (57.1%). Geographic scope was identified in all 14 (Table 2, Supplementary Table S9).

Table 2.

Public availability of predefined trial-design and clinical-context elements.

Element	n/N	Percent	Exact 95% CI
Qualifying ASCVD/CVD or risk phenotype	14/14	100%	76.8%–100%
Prevention or special-population setting classifiable	14/14	100%	76.8%–100%
Geographic scope identified	14/14	100%	76.8%–100%
Lp(a) unit identified	13/14	92.9%	66.1%–99.8%
Background lipid-lowering therapy described	9/14	64.3%	35.1%–87.2%
Numeric Lp(a) threshold identified	8/14	57.1%	28.9%–82.3%
PCSK9 inhibitor/inclisiran handling identified	8/14	57.1%	28.9%–82.3%
Demographic reporting or design-specific demographic eligibility identified	8/14	57.1%	28.9%–82.3%
CKD/eGFR criteria identified	6/14	42.9%	17.7%–71.1%
Heart failure criteria identified	6/14	42.9%	17.7%–71.1%
Diabetes criteria identified	5/14	35.7%	12.8%–64.9%
LDL-C eligibility criterion or optimization context identified	3/14	21.4%	4.7%–50.8%

Note: An element was counted as publicly identified when the prespecified source set contained sufficient information for an unambiguous trial-level determination at the May 2, 2026 data lock. “Not publicly identified” does not establish absence from a sponsor protocol, noncompliance with ClinicalTrials.gov requirements, or trial-design deficiency.

These counts refer only to retrievability in the prespecified public-source set on May 2, 2026. A designation of not publicly identified does not establish that a criterion was absent from a full sponsor protocol, that a registry requirement was unmet, or that the underlying trial design was deficient. This qualification is particularly important for ongoing registry-only programs that had not yet produced a public protocol, baseline report, statistical analysis plan, or primary publication.

4. Discussion

In this fixed-date audit of 14 contemporary direct Lp(a)-targeted trials, broad trial identity and setting were usually retrievable from public sources: qualifying cardiovascular or risk phenotype, prevention or special-population setting, and geography were identifiable in all trials, and Lp(a) units were identifiable in 13 of 14. Other clinical-context elements were less often identifiable in the prespecified source set, including LDL-C eligibility or optimization context, CKD/eGFR criteria, diabetes criteria, heart failure criteria, and demographic information. These findings describe the public evidence layer available to an external user on May 2, 2026 rather than the content of complete sponsor protocols.

The most important interpretive constraint is source maturity. Several pivotal outcomes programs were still ongoing and registry-only at the data lock, and 4 of the 5 clinical-outcome trials fell into that category. Such trials may appropriately lack public baseline manuscripts, statistical analysis plans, or primary publications before completion, and their registry records may change through protocol amendments. It would therefore be inappropriate to treat nonidentification of one of our study elements as a reporting failure or to hold an ongoing registry-only program to the same public-document expectations as a completed, published trial. For this reason, the revised analysis does not use a composite completeness score, trial ranking, cut points, or source-category performance comparison.

The descriptive domains remain useful for a narrower purpose: they show which contextual details an external reader can and cannot currently retrieve. Numeric Lp(a) thresholds and units help define the studied population, especially because mass and molar units are not directly interchangeable [6]. Background LDL-C management and concomitant lipid-lowering therapy help clarify residual-risk context, while kidney, diabetes, heart failure, and demographic information can inform later assessment of applicability once a trial is reported. The present data do not establish when each element should first be disclosed or which document should contain it.

The issue is broader than Lp(a). ClinicalTrials.gov meta-research and drug-class studies have shown how record maturity, publication concordance, safety fields, and demographic reporting can shape what is publicly retrievable [19], [20], [21], [22], [23]. Reporting frameworks such as SPIRIT, CONSORT, SPIRIT-Outcomes, and CONSORT-Outcomes emphasize that different information becomes relevant at protocol and final-report stages, while oncology eligibility-modernization efforts provide a partial example of specialty-specific consensus development [27], [28], [29], [30], [31]. Table 3 therefore presents only an expert-informed set of candidate elements for stage-appropriate public documents as studies mature. It is not a ClinicalTrials.gov compliance checklist, is not validated or exhaustive, and requires stakeholder consensus before being treated as a reporting standard.

This study has several limitations. First, it used public trial-level information and cannot assess efficacy, safety, comparative value, subgroup effects, trial quality, actual clinical feasibility, payer decisions, guideline adoption, or regulatory readiness. Second, records and linked documents are dynamic, and public availability may change after the data lock. Third, not publicly identified may reflect delayed publication, a nonpublic protocol, a future statistical analysis plan or results report, or retrieval limitations rather than true absence. Fourth, the 12 domains were investigator-selected for descriptive abstraction and were not externally validated. Fifth, independent pre-consensus ratings were not retained in analyzable form, precluding a valid retrospective inter-reviewer agreement statistic. Sixth, we did not systematically assess every exclusion criterion, sample-size justification, patient and public involvement, or actual population representativeness. Finally, the cohort intentionally excluded early pharmacology and non-direct Lp(a)-targeted studies and is not a comprehensive review of every Lp(a)-related trial.

Within these limits, the study provides a transparent and updateable baseline of public retrievability. Repeating the same audit after public protocols, statistical analysis plans, baseline reports, and outcome publications become available would help distinguish temporary source immaturity from information that remains difficult to retrieve after study maturation.

5. Conclusion

Public availability of predefined trial-design and clinical-context information varied across contemporary direct Lp(a)-targeted trials at the May 2, 2026 data lock. The greatest interpretive caution is required for ongoing registry-only programs, for which nonidentification may simply reflect the stage and type of public documentation available. This analysis describes retrievability rather than registry compliance, protocol completeness, trial quality, efficacy, safety, comparative value, or regulatory readiness. Reassessment after major outcomes programs mature will be necessary to determine which apparent gaps were temporary and which persist in later public reports.

CRediT authorship contribution statement

Samuel Tan: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization, Investigation. Joyce Gu: Writing – review & editing, Writing – original draft, Methodology, Formal analysis, Data curation, Conceptualization, Investigation. Daniel Yow: Writing – review & editing, Writing – original draft, Data curation, Investigation. Lucy Zheng: Writing – review & editing, Writing – original draft, Data curation, Investigation.

Ethical statement

This study used only publicly available clinical trial registry records and linked publicly available aggregate sources. No patient-level data, protected health information, identifiable participant information, biospecimens, animal subjects, or direct participant contact were involved. Institutional review board approval and informed consent were therefore not required.

Disclosures

The authors report no relevant financial relationships.

Declaration of Generative AI and AI-assisted technologies in the writing process

During the preparation of this manuscript, ChatGPT (OpenAI) was used to supplement language editing and organization. Any suggestions made by this tool were first reviewed and modified as appropriate to preserve the originality of this manuscript. All authors full responsibility for the content of the publication.

Funding

No external funding was received for this analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary tables