Literature review

The evidence on your question found, appraised and set in one table, then argued in prose to the gap your aims fill. Every claim cited to its source

Customize this Agent

DOCXXLSX15 steps

The problem

Pelacarsen lowered lipoprotein(a) and missed its primary endpoint. Your Significance says lowering cuts events

In Lp(a)HORIZON, a trial of 8,323 patients, pelacarsen lowered Lp(a) and missed its primary endpoint. Mendelian randomization puts the benefit in the absolute fall; two studies put the fall that matches 1 mmol/L lower LDL cholesterol at about 66 and 100 mg/dL. Quote percent lowering and the trial cannot be read against that; convert nmol/L to mg/dL at one factor and it is wrong, because the factor depends on isoform

The request

Lp(a)HORIZON missed its primary endpoint last month, and the Significance in our R01 resubmission argues Lp(a) is causal and lowering it will cut events. Rewrite the literature review behind it: Mendelian randomization, cohorts and every Lp(a) lowering trial, published or registered. Read each trial on absolute lowering against what the genetics say is needed, in the unit it reported; never convert nmol/L to mg/dL at a flat factor. Treat the Novartis topline as a release until full results are out. Give me the strengths and weaknesses of each line of evidence, which claims in our Research Strategy still stand, and the gap our aims fill now. Evidence table in Excel, every claim cited. Research Strategy and summary statement attached.

Research Strategy.docx
Summary statement.pdf

How Axe works

ApprovedChanges asked

Your reviewApprove, or say what to change

Approved or changes

Literature review and evidence tableReview, table and citation check

Final reviewApprove the revised files

Revised literature review and evidence tableRevised files and citation check

Step

Reconcile genetics and trials

Prompt

Sent to the model at this step

Read the trials against the genetics on absolute lowering, not percent. Two Mendelian randomization studies put the fall needed to match lowering LDL cholesterol by 38.67 mg/dL, 1 mmol/L, at about 100 mg/dL (Burgess and colleagues, JAMA Cardiology 2018, a 5.8 percent lower risk of coronary heart disease per 10 mg/dL) and 65.7 mg/dL (Lamina and Kronenberg, JAMA Cardiology 2019). For each trial reported in mg/dL, set its baseline and absolute lowering against both, convert the lowering into its LDL cholesterol equivalent on each estimate, and predict the trial's effect from the Cholesterol Treatment Trialists' 22 percent lower risk of major vascular events per 1 mmol/L over about five years, scaled to the trial's follow up. Show the formula and report the range. The 5.8 percent per 10 mg/dL is lifelong genetic exposure, not a trial's expected effect. For trials reported in nmol/L, follow the European Atherosclerosis Society's 2022 consensus: the conversion factor depends on isoform, so never convert at a flat factor. Use mg/dL only where the trial itself reports it, and otherwise state that the comparison cannot be made on the genetic scale. Then weigh each explanation for Lp(a)HORIZON missing its primary endpoint only as far as the evidence carries it: absolute lowering achieved, baseline Lp(a) of those enrolled, follow up set against the genetic estimates, the composite endpoint, background LDL lowering and events accrued. Name each registered outcome trial still running that could settle it, with the unit and threshold it uses and its expected completion. Keep the topline open until the full results are presented or published.

Uses

Evidence table
Appraisal
Search log

Produces

Reconciliation

What you get back

DOCXLiterature reviewState of the evidence, genetics and cohorts, trials, open questions, the gap, revised claims

XLSXEvidence tableSearch log, studies and trials in their own units, reconciliation, claims, sources

Citation checkOpen items first, then each claim and figure marked checked or corrected

Make it yours

Point it at your question and your grant or thesis, add the databases your library holds, or set your citation style. Every step and every prompt can be edited

Customize this Agent

Opens a private copy of this Agent in your workspace, ready to adapt to your own sources

More agent examples

ResearchersSystematic reviewDOCXXLSX20 steps

Capital markets analystsBroker opinion of valuePPTXXLSX16 steps

Financial planning analystsBudget variance analysisXLSXDOCX13 steps

Credit research analystsCapital structure analysisXLSX12 steps

Life sciences analystsClinical trial landscapeXLSXDOCX14 steps