W6 - HT CVD Biomarkers: study of CHD, Stroke and VTE - Phase I

Investigator Names and Contact Information

Introduction/Intent

W6 was conducted in several stages: Phase I: cases within 2 years of randomization; and Phase II: > 2 years after randomization.

See also

• W11-Assays on Strokes Identified Strokes After Feb 2001 and
• W14-Additional Assays on W6 CVD Biomarkers described below.

Initial recommendations came from the WHI Blood Biomarker Working Group for determining the potential impact of HT on blood biomarkers. Given that the working group is blinded to the WHI trial data, we did not have any explicit details about the specific scientific rationale for this effort. Thus, this document provides a more general overview of potential biomarkers for consideration in the context of the group’s experiences independent of WHI data. For the purpose of this activity the following assumptions were used:

1. Suggested biomarkers for a case-control study in HRT trial participants focused on identifying mediators of CVD events (MI, stroke, PE/DVT).
2. Criteria used to select markers included the following:

• The availability of established data linking the biomarker to CVD events;
• Evidence that suggest that estrogen may exert an effect on the biomarkers; and
• Prevalence/power for the biomarkers.

3. Potential biomarkers were evaluated in the following areas:

• Inflammation
• Thrombosis
• Lipids
• Other Blood Analytes
• Gene Polymorphisms

The group conducted a comprehensive discussion identifying the myriad of potential and informative biomarkers. For the purposes of this discussion it was noted that as of January 1999 there were 175 documented CHD cases, 120 documented stroke cases and 140 documented DVT/PE cases in participants in the HRT trial.

We discussed the potential for looking at change between baseline and the follow-up visits in the subset with blood drawn at follow-up visits. It was noted that given the longer follow-up time and larger number of participants with available blood at the baseline examination, the highest statistical power for these analyses are present when assessing the relationship between baseline markers and events. However, additional information on higher risk subgroups may be obtained by evaluating the impact of change in biomarkers on events. Therefore the final case-control study should consider a more global strategy when evaluating the utility of using baseline and follow-up blood markers.

Given the shared etiology of these outcomes and the need to have a reasonable number of events, the committee recommends that the case-control analyses should be done separately for two categories: CHD & Stroke cases; and DVT & PE cases. It is important to note that to enhance the cost-effectiveness of the study these groups may likely be able to share a common control group.

• W11-Assays on Strokes Identified Strokes After Feb 2001
• W14-Additional Assays on W6 CVD Biomarkers and

Initial and secondary tiers of analytes to explore the CVD outcomes from the E+P are complete or in progress. However, there have been scientific advances in the field of explanatory analytes that might be valuable addition or substitutions to the analytes already approved that would help WHI contribute to a greater understanding of this early CVD risk. Specific analytes are listed below.

1. NMR lipoprofiles (www.liposcience.com) for CHD/stroke baseline & year 1 quantifies 10 lipoprotein subfractions, as well as LDL particle size and LDL particle concentration, which predict coronary risk and are modulated by hormone therapy (JAMA 2003;290:2030).

2. Progesterone receptor polymorphisms – Very few previously reported. David Herrington has identified 75 SNPs, but does not know which (if any) are clinically relevant. He will have identified 10 or so most promising in next few months, and can run on same DNA samples (CHD/stroke) he is getting for estrogen receptor SNPs analysis.

3. Matrix metalloproteinases (MMPs) induce thinning of the fibrous cap in atherosclerotic plaques, thereby promoting plaque rupture (Am J Cardiol 2003;92:1461) – The current CVD study (W6)has done MMP-9; E+P-induced changes predicted CHD (p<0.01), although no interaction with drug was identified. Extensive literature on activation of MMPs by estrogen in breast cancer (J Steroid Biochem Mol Biol 2003;87:65); Medline found 0 references for MMP + estrogen + CHD. Dr. Francis Spinale (Medical Univ of South Carolina), an MMP expert, recommends running MMP-7, 8, 9 and TMP-1 (tissue inhibitor of MMP) on CHD/stroke cases/controls at baseline and year 1.

4. Panel of cytokines. IL-6 was associated with stroke risk in our study (OR ~10, p for interaction with E+P = 0.02). In unpublished paper from EPIC-Norfolk study, adjusted OR 1.98 (1.09, 3.61) for CHD by IL-8 quartile. Drs. Pieter Reitsma/Frits Rosendaal suggest assaying 10 interleukins.

5. APC – ETP (endogenous thrombin potential), currently performed in Dr. Rosendaal’s lab. The new APC-ETP has been shown in recent studies to be associated with hormone-induced changes that would make it very interesting to examine in the E+P and E-alone CT for all CVD outcomes. Dr. Rosendaal provided unpublished paper showing OR 4.7 (95% CI 1.4, 15.6) for venous thrombosis cases in postmenopausal women.

6. Free and total TFPI (tissue factor pathway inhibitor) antigen + TFPI activity in CHD/stroke/VTE cases/controls. TFPI falls 20% with hormones (J Thromb Haemost 2003;1:1208); OR 1.6 for VTE (Blood 2003;101:4387). Some suggestion of role in CHD (Circ 2003;108:2864).

• 1-4: W14 - New tests on W6 cases
• 5-6: W11 - New Strokes

a - tests done only on corresponding controls, not all controls

Results/Findings

See Publications: 204, 210, 222, 273, 345, 347, 350, 429, 445, 462, 526, 854, 866, 972, 1114. WHI publications by study lists published WHI papers that have been generated by ancillary studies. A complete list of WHI papers is available in the Papers of this website.