Laboratory Working Group

Meeting Minutes – Laboratory Working Group Meeting
July 30, 2008

National Institute Of Diabetes and Digestive and Kidney Diseases (NIDDK)
National Kidney Disease Education Program (NKDEP) Laboratory Working Group Meeting
AAAC Annual Meeting – Washington, DC

Participants: Greg Miller (Chair),George Schwartz, Leigh Ann Milburn, Mauro Panteghini, Chandra Jain, Neil Greenberg, David Lacher, John Lieske, Glen Hortin, Graham Jones, Joris Delanghe, Eileen Newman, David Seccombe, Karen Phinney, Matthew McQueen, David Bunk, Mary Robinson, John Eckfeldt, Yoshihisa Itoh, Andrew Narva, David Bruns, Gary Myers, Nancy Xu, Nancy Accetta, Christen Horn

Summary of Action Items:

• Glen Hortin and John Lieske will develop a proposal to evaluate the molecular forms of albumin.
• George Schwartz will communicate with Dave Bruns when the revised Schwartz equation is about to be published so he can submit a perspective article in Clinical Chemistry that will provide awareness to clinical chemists.
• Nancy Accetta will send an e-mail with information about the revised website.
• Proposals will be solicited to address information needed for urine albumin.

Meeting Minutes

1. Urine Albumin Group Activities, Greg Miller, Working Group Chair

Background: In 2007 there was a conference in collaboration with the IFCC on current issues and reporting of urinary albumin excretion. A manuscript on that conference has been submitted and favorably reviewed by Clinical Chemistry; once the manuscript has final acceptance, a draft will be circulated to those in the group that were not part of the albumin conference. Also in 2008, the IFCC’s and NKDEP’s working groups on albumin standardization merged with Greg Miller as the Chair of this joint committee. Membership in this joint working group will be comprised of those who take active roles on the urine albumin projects identified at this meeting.

Urine Working Group Conference Recommendations:

• “Urine albumin” should replace “microalbumin”
• First morning void provides less intra-individual biological variation than random collection
• Second morning void may be acceptable, but there is no evidence that it is superior to a first morning void
• Urine albumin should be measured in non-frozen urine within 7 days; refrigerated urine should be warmed to room temperature, examined for precipitate, and centrifuged if needed
• Long term storage should be at less than -70° C
• Albumin creatinine ratio (ACR) should be reported with all urine albumin measurements; ACR reporting units should be uniform within a country or geographic region; either mg/g or mg/mmol; harmonization of units may happen in the future, but is difficult to accomplish

Additional clarification needed in order to make final recommendations and build a reference system:

• Pre-analytical requirements: influence of container type, time of collection, degree of urine albumin degradation under various storage conditions
• Molecular forms of albumin and definition of the measurand
• Variation in urinary matrix composition over which measurement procedures must operate: influence of blood, seminal fluid, etc.
• Clinical requirements for total error in measurement of urine albumin, which is dependent on having a good estimate of the biological variability (see below)
• Reference system: measurement procedure, material for urine albumin, and material for urine creatinine

Subsequent steps that require the above clarifications and standardization components to be in place:

• Decision thresholds for ACR and albumin excretion rate (AER); ACR varies with age, gender, ethnicity
• Risk for CKD and CVD are continuous functions of urine albumin concentration
• Usefulness of age- and gender-specific equations to convert ACR to an estimated AER; possibly a single reference limit can be developed

2. Biologic variability, adsorption to containers, and frozen stability of urine albumin, Mary Robinson, Centers for Disease Control and Prevention

Evaluation of intra-individual biologic variation of urinary albumin is funded by the National Center for Chronic Disease Prevention and Health Promotion at CDC. Gary Myers and Mary Robinson are the lead investigators from CDC who wrote the proposal; additional project members are Greg Miller, James Richie, David Koch.

Part 1: Assess adsorption, storage, and freeze-thaw cycles:

• Surface adsorption of albumin onto containers:
• Polypropylene (PP),teflon (TF), polyethylene (PE), polysulfone (PF), polystyrene (PS), polycarbonate (PC), and glass (GL) will be evaluated; some products will be duplicated from different manufacturers
• Containers evaluated will be: Collection cups made of PP, PE, PS, GL, TF; centrifuge tubes made of PP, PE, PS, GL, TF, PF, PC; storage vials made of PP, PE, PS, GL; sample analysis cups made of PE, PS [if other materials are used for samples cups, please let Mary know the manufacturer(s)]
• Storage temperatures and freeze-thaw cycles (input to this plan from the group is welcome):
• Room temperature (25° C) for <1, 2, 4, 8, and 16 days
• Refrigerated (4° C) for 2, 4, 8, 16, and 32 days
• Frozen (-20° C) for 2 days, 1, 6, 12, and 24 months; refreeze and repeat at 3, 6, and 9 months
• Frozen (-80° C) for 2 days, 1, 6, 12, and 24 months; refreeze and repeat at 3, 6, and 9 months
• Material will be a pooled urine from CDC employee volunteers spiked with purified serum albumin at three levels: low, <15 mg/dL; medium, 30-50 mg/dL; and high, > 100 mg/dL; analyze identical samples after storage in containers made from a variety of materials at various temperatures and various freeze-thaw cycles; the Roche-Hitachi System is used for the analysis of both albumin and creatinine

Part 2: Assess biological variability in three ethnic groups:

• Short-term (3 months): recruit 20 subjects from Virginia Commonwealth University, Richmond, VA and Grady Health System, Atlanta, GA
• Long-term (12 months): recruit 10 subjects from Emory University, Atlanta, GA
• Each subject will collect four sample types (24-hour, 1st morning, 2nd morning, and random afternoon collections).
• Three populations will be targeted: 10 Caucasians, 10 African Americans, and 10 Hispanics will be recruited from the 3 sites; looking for stable diabetics with moderately elevated albumin levels; subjects with uncontrolled diabetes and other abnormalities will not be included
• Enrollment criteria: Hemoglobin A1c stable and well-controlled (<7%); blood pressure stable and well controlled (<140/90); medications stable, listed, and not likely to change in the next 3 months; serum creatinine stable (repeated results within 10%); ≥18 years old; subjects from “vulnerable” groups as defined by IRBs will not be considered
• Short-term biological variability study: each subject will collect three 24-hr samples with each void collected in a separate container (on wks 1, 6 and 12); and nine random, nine first-morning, and nine second-morning samples (on the intervening weeks)
• Long term biological variability study: each subject will collect six 24-hr samples with each void collected in a separate container (at months 4, 8, and 12); and 15 random, 15 first-morning, and 15 second-morning samples (at the intervening months)

Comments from LWG members:

• Details of the protocols are still under development and comments made at this meeting will be taken into consideration
• How the sample is mixed in the container may have an effect on the binding so this needs to be detailed
• A publication by Dr. Itoh suggests that hydrophilic materials adsorb less of the albumin, thus this material will be used wherever possible; however, nalgene containers for the biological variability study had to be purchased before adsorption study could be performed; may have to use the information determined by the adsorption study with regard to nalgene containers to correct biological variability data
• John Eckfeldt suggested using real urine rather than spiked urine for the adsorption study but pooled samples from volunteers is less expensive and more readily available than samples that require IRB approvals when collected from real patients; John Eckfeldt is uncomfortable with spiking the urine and would rather see samples of urine with low medium and high albumin levels pooled to make the pools that are used in the absorption study; Greg Miller suggested that the work group re-visit this issue
• David Bunk suggests storing the samples collected for the biological variability study at -80° C and re-evaluating them over several years to investigate stability
• Glen Hortin commented that each manufacturer’s process may affect the surface such that the properties of a given material may differ by manufacturer
• Dave Lacher commented that NHANES will do an ACR study in the next cycle beginning 2010; it may be possible to use this for reference intervals; participants will be shipping from their homes and this may result in temperatures above 25 ° C
• Dave Seccombe suggested using radio-labeled albumin for binding studies; also suggested including additional measurements such as pH and electrolytes to determine if the matrix changes between samples within an individual
• Mauro Panteghini suggested that subjects for the biological variability study should not have chronic disease and is uncomfortable with the plan to use subjects with stabilized diabetes; he suggests using a second group of subjects without chronic disease; however, only disease conditions cause albumin in urine
• Additional comments should be sent to Mary Robinson and Greg Miller; the finalized protocol will be circulated to the LWG for comments
• Mary Robinson commented that their statistician says that in order to have statistically meaningful data, 38 measurements per level, per container type, will be required for the stability study

3. Other Issues

Molecular forms of albumin, Glen Hortin NIH Clinical Center

• Fragmentation and dimerization of urine albumin: there are many publications that discuss this topic and many techniques are needed to evaluate this (e.g., SDS PAGE, Western blotting techniques, or LC-MS or CE-MS for small fragments)
• Side chain modifications are evaluated using cysteinylation, oxidation, and glycation with LC-TOF-MS
• Ligands of albumin contribute to mobility changes in non-denaturing electrophoresis in particular the binding of fatty acids; can guess what are ligands based on things that are present in urine and evaluate by spiking experiments
• Conformational variants/denatured forms of albumin are difficult to measure and we do not have a good handle on how to evaluate this
• Urine matrix effects: there are many components in urine that vary by individual; preliminary studies do not show anything that causes a huge effect
• LC-TOF-MS method shows promise as a technique to evaluate the albumin forms found in different samples from different types of patients
• Data from several sources shows that there is not a major problem with immunoassay methods; data suggest that minor variations will not have a major impact unless the method uses monoclonal antibodies or the variation causes a conformation change
• The goal is to be able to provide information to manufacturers regarding the albumin forms that their assay needs to measure; suggestions from manufacturers would be helpful
• Want to know if urine albumin is substantially different from serum albumin
• Different techniques are needed to evaluate different things (e.g., SDS-PAGE is a good method for looking at internal cleavages and dimerization, but side-chain modifications will not change mobility and thus not be detected by SDS-PAGE)
• Need suggestions on how to change this into a project; most important challenge is to know the measurand; Mary Robinson will try to find data from a study in which 2D-PAGE analysis of both urine and serum from the same individual was performed
• Dave Seccombe suggested that we need to know whether the small changes that can be detected by these different methods are clinically relevant
• Glen Hortin will develop a proposal to evaluate the molecular forms of albumin; finding appropriately stored samples is a challenge (e.g., effect of bacteria on the degrade albumin); may want to coordinate with Mary Robinson or NIDDK stored sample banks for a source of wellcategorized samples; large amounts of some compounds like ascorbic acid can degrade albumin
• There was discussion of confounders such as vitamins, supplements, statins, timing of ACE inhibitors in relation to urine samples collection

Understanding the matrix of urine:

• Using large data bases (e.g., commercial labs and large academic centers) to determine range of commonly measured molecules found in urine; these could be used to provide guidance to manufacturers about the ranges within which their assays must work
• Suggested approaches and development of a proposal is needed from volunteers who have the ability to conduct such data mining exercises
• Dave Seccombe recently did data mining with LabCorp; a statistical model is built; he will ask if they can send data related to urine testing
• Glen Hortin suggested another approach which is to identify urine samples with extreme values for specific characteristics (e.g., pH, specific gravity); then ultra-filter them and remove protein to generate a range of matrices that represents the urine matrix; also, after identifying what extremes of components are present, can perform spiking experiments with components that are deemed to be most likely to have an impact on assays; can develop some hypothesis about what are important components that manufacturers need to evaluate

4. Update: Status of Pediatric Equations, George Schwartz, University of Rochester

• Pilot study to evaluate equations to estimate GFR in children; used plasma iohexol disappearance as the gold standard; collected data at 0, 10, 30, 120, and 300 min; compared to the height/sCr, and included cystatin C, sCr, BUN
• Demographics of subjects is short and “chunky” which is characteristic of kids with CKD
• Using gender and height/sCR as was done in the original formula, 73% of estimates are within 30% of the measured iohexol GFR value
• Best formula adds cystatin C, BUN, gender and height by itself; with this improved formula, 83% of the estimates are within 30% of the measured value
• This improved formula did best when compared to other published formulas
• If height is not available, an estimate can be based on height percentiles (5%, 50%, 95%)
• The relatively low muscle mass and retarded Tanner status of the CKiD population may restrict the development of estimate formulas for adolescent males
• Mauro Panteghini suggests that the formula be re-evaluated after cystatin C standardization is established; George Schwartz commented that they need to work with Siemens Dade Behring to re-assay and introduce those results in the equation
• Dave Bruns requested that George Schwartz communicate with him prior to the publication of this data so he can submit a perspective article to Clinical Chemistry that will provide awareness to clinical chemists

5. Update: Status of Working Group to Address Pharmacy Issues, Andrew Narva, Director, NKDEP National Institute of Diabetes and Digestive and Kidney Diseases

• Cockcroft-Gault has traditionally been used by pharmacists and physicians for estimating drug doses for patients with CKD; with creatinine standardization, Cockcroft-Gault is not usable because of lower creatinine values; because of this some facilities are resisting change to standardized creatinine methods; NKDEP is developing an educational tool for physicians to help them understand and deal with this issue (these will not be formal guidelines for clinicians)
• The group working on this consists of Nancy Xu, Leigh Ann Milburn, John Eckfeldt, Neil Greenberg, Andy Narva, and Leslie Stevens (Chair); a draft document is in progress and it will be circulated more widely for comment once it is in a more final form
• NIDDK recognizes the importance of urine albumin standardization. Funds may be available to support some of these projects and, if available, will be allocated by 10/1/08.

6. Introduction of Revised Website (Lab Professionals Section), Nancy Accetta, NKDEP

No report was made due to lack of time; Nancy Accetta will send an e-mail with information

7. Next Steps, Greg Miller

• We need to turn project ideas into proposals to use funds that Andy Narva mentioned; the projects need to be focused in order to provide concrete information
• Submit proposals by early September; proposal drafts will be circulated through Nancy Accetta; proposals do not need to be R01 applications, just proposals that will allow the funding money to be encumbered and dispersed over the next year.

Meeting adjourned at 2:45 pm