AACC Annual Meeting - Chicago, IL - August 1, 2018
National Kidney Disease Education Program Laboratory Working Group, Joint Meeting with the International Federation of Clinical Chemistry and Laboratory Medicine Working Group for Standardization of Albumin in Urine Marriott Marquis Chicago Glessner A and B Rooms Chicago, IL August 1, 2018 12:00 p.m. – 1:30 p.m. final summary.
Welcome and Introductions
W. Greg Miller, Virginia Commonwealth University (VCU)
Dr. W. Greg Miller, Chair of the National Kidney Disease Education Program (NKDEP) Laboratory Working Group (LWG), welcomed the participants to the meeting and reviewed the day’s agenda. In March 2007, the LWG partnered with the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) Working Group for Standardization of Albumin in Urine (WG-SAU) to facilitate the development of reference measurement procedures and reference materials for urine albumin (UA) measurements. Upon completion of this project, the reference system components will be submitted to the Joint Committee for Traceability in Laboratory Medicine (JCTLM) for listing and then will be available for use by manufacturers for calibration traceability. The participants were invited to introduce themselves; a list of participants is provided as Attachment A.
National Institute of Standards and Technology (NIST) Reference Materials for UA
Lorin Bachmann, VCU
Ashley Beasley-Green, NIST
Karen Phinney, NIST
Standard Reference Material (SRM) 3666 in Frozen Human Urine
Dr. Lorin Bachmann, Chair, IFCC WG-SAU, provided an update on the development of the NIST SRM 3666 secondary reference materials for UA. She began by reviewing the joint NKDEP/IFCC effort to assess the status of harmonization and standardization of UA measurement methods, which culminated in a 2014 publication presenting the study’s findings (Bachmann et al., Clinical Chemistry, 2014;60:471–80). In this study, the agreement for UA measurements was assessed for 332 patient urine samples. The UA was measured using 16 in vitro diagnostic (IVD) immunoassay methods routinely used by clinical laboratories and a candidate comparative UA method, isotope dilution mass spectrometry (IDMS). Results showed that although some IVD manufacturers’ methods agreed with the IDMS method, other methods were biased low and high. On average, there was a 40 percent disagreement between the IVD manufacturers’ methods, suggesting a need for UA routine method standardization. The next steps are to (1) develop reference measurement procedures, with which NIST, Mayo Clinic’s Renal Reference Laboratory, and University of Minnesota’s Advanced Research and Diagnostics Laboratory are assisting, and (2) develop and evaluate commutable reference materials, which is in progress at NIST with collaboration from VCU to provide the urine materials.
Dr. Bachmann reminded LWG members that the approach for establishing traceability of a routine method to a reference system for purposes of standardization is based on the International Organization for Standardization 17511 requirements for the metrological traceability of values assigned to calibrators and control materials. To date, NIST has developed the purified primary reference material—SRM 2925—with a concentration of 0.8443 g/L (± 0.0320 g/L). The SRM 2925 is currently being verified by NIST, is expected to be released in late 2018, and will be used to prepare pure substance calibrators for the candidate reference measurement procedures.
Regarding the development of the NIST SRM 3666 albumin in frozen human urine standard, Dr. Bachmann detailed the procurement status for the urine samples. Four concentration levels of pooled native human urine samples are in the acquisition phase: Pool 1, 5–10 mg/L; Pool 2, 20–50 mg/L; Pool 3, 60–180 mg/L; and Pool 4, 275–375 mg/L. VCU is obtaining individual residual patient urine samples (leftover unadulterated individual urine samples) submitted for routine care within each of the target ranges; performing urinalysis for nitrites, leukocyte esterase, and blood on each sample; and centrifuging to remove sediment and storing approximately 50 mL aliquots of urine at -70 °C. Once the target procurement volume of 2.7 liters for each Pool has been reached, a NIST-contracted organization will conduct the pooling and bio-handling material processes. Next, the NIST reference measurement procedure will be used to value-assign the UA and creatinine concentrations. In the final step, the commutability study will be performed.
Dr. Bachmann reported that Pools 1, 2, and 3, consisting of 56–57 individual samples each, are completed and have exceeded the target volume by 13–20 percent. A total volume of 0.8 liters has been collected from 12 individual samples for Pool 4 (275–375 mg/L). The VCU group began collecting additional urine samples in concentration ranges, 200–275 mg/L and 375–600 mg/L. Currently, 25 individual samples with an average concentration of 387 mg/L and totaling 1.2 liters are available in this expanded concentration interval. Dr. Bachmann asked the LWG whether it would be acceptable to include samples from the expanded concentration interval to prepare Pool 4. Dr. Bachmann next described another issue impeding the Pool 4 collections: the presence of trace or small amounts of blood in the UA samples causing samples to be rejected. She asked the LWG whether these samples should be considered for the Pool 4 collections.
SRM 2925 Pure Albumin
Dr. Ashley Beasley-Green briefly noted that NIST is continuing to work on certifying candidate SRM 2925 as a suitable calibrator for the UA measurement procedures. Further updates will be provided at future meetings.
Including samples from the expanded concentration interval to accelerate the completion of Pool 4 was discussed. The following points were made:
- Dr. Joris Delanghe asked whether the impact of post-translational modifications (PTMs) to albumin, such as glycation or carbamylation, is being considered. Depending on the origin of the albuminuria—glomerular membrane defects or diabetic nephropathy—the type of albumin present in the Pools will be different. Dr. Anthony Killeen added that the higher albumin concentrations also would introduce higher concentrations of other proteins, which could be interferences.
- Dr. Bachmann wondered whether PTMs would be an ongoing concern if materials are analyzed using the IDMS method, which is specific for certain peptides, versus the immunoassays. Dr. Killeen suggested performing a serum protein electrophoresis analysis of the urine samples. Dr. Beasley-Green added that NIST will perform both quantitative and qualitative assessments of the reference materials.
- Dr. Miller asked whether the presence of carbamylated and/or glycated albumin would prohibit the use of the samples for developing the reference material. Dr. Delanghe noted that the charge of the albumin protein could affect the behavior of other interactions to influence an immunologic reaction and suggested that chemical comparative analysis would be helpful. Dr. Miller agreed that understanding the PTMs in the Pools is an important aspect, but he is not aware of any evidence regarding the modified albumin protein influencing immunoassays.
- Dr. Bachmann called attention to a larger issue regarding the measurements of proteins and peptides using mass spectroscopy that is being addressed by the Clinical and Laboratory Standards Institute (CLSI) C60 Document Development Committee that may helpful to the LWG. This CLSI Committee, of which Dr. Bachmann is a member, more than likely will recommend measurement based on the full-length protein. She also noted that it is unknown whether the reference material peptides being used in this project to value-assign patients’ samples are subject to PTMs or whether these PTMs correlate to the PTMs of endogenous albumin peptides. Dr. Karen Phinney commented on the challenge of identifying the degree of reference peptide PTMs early on. The goal is to develop fit-for-purpose reference material that best reflects the composition of the patient samples.
- Dr. Jesse Seegmiller questioned the use of high-UA concentration samples in the Pools, which might not reflect the critical range of patient care. Dr. Beasley-Green has encountered patient samples containing levels as high as 900 mg/L, which she obtained from NIDDK collaborators. Dr. Delanghe noted that the critical microalbuminuria range is 20–200 mg/L. Dr. Miller explained that the UA concentrations for each Pool were selected based on the measuring intervals of the IVD immunoassays to be standardized, which would be beneficial to manufacturers performing calibrations, rather than on clinical classifications.
- Given that achieving the Pool 4 levels using the current process will take a long time and that somewhat higher concentrations have the possibility to introduce more heterogeneity into the pool matrix, Dr. Miller asked whether it would be worth the risk to expand the concentration interval up to 600 mg/mL to accelerate the process.
- Dr. Seegmiller pointed out that the mass spectrum protein profiles may change across patient samples, but albumin has been observed as the predominant protein. Dr. Andrew Narva asked whether the excretion patterns were reflective of the protein profile changes. Dr. Delanghe explained that the albumin secreted is reflective of the glomerular membrane pore permeability, which increases with the size of the protein being excreted. Dr. Beasley-Green added that such types of characterizing information will be provided in addition to the certified values for the reference materials.
- Dr. David Seccombe asked the LWG to consider whether the ultimate goal of this project would be to optimize the value proposition for clinical decision making based on an albumin-to-creatinine ratio (ACR) or to establish a consistent ACR endpoint (i.e., single cutoff value). Dr. Seccombe suggested optimizing the Pool concentrations based on relevant proposed ranges. Dr. Narva commented that the guidelines and ACR limits are based on the data collected, which represent an arbitrary single cutoff value observed in the largest patient cohort often identified in the Chronic Kidney Disease (CKD) Prognosis Consortium that the NIDDK supports. These data may not necessarily represent the most useful clinically important cutoff values. Accurately understanding the UA levels, tracking how they are changing, and identifying interventions should remain priorities.
- Dr. Miller restated that the purpose of the Pools is to develop a four-level reference material to be used by IVD manufacturers to calibrate immunoassays. Although the concentration levels adequately cover the lower concentrations, higher level reference materials also are needed because some commercial assays have measuring intervals that extend to 400 mg/L. The role of the LWG is to provide the necessary tools to achieve UA measurement standardization. Problems regarding specificity from immunoassays are not anticipated based on data from the 332 individual assay assessment, which had small specimen-specific influences.
- Hearing no further concerns, Dr. Miller called for a recommendation from the LWG whether expanding the collection process to include samples 200–600 mg/L UA concentration is acceptable for the high-concentration pool. The consensus of the LWG was that it would be acceptable to expand the sample collection concentrations for Pool 4, barring no new concerns. Any issues on commutability will be identified when that assessment is performed as part of the pool’s characterization.
Including rejected samples that contain trace or small amounts of blood to accelerate the Pool 4 collections was discussed. The following points were made:
- Dr. Miller explained that expanding the concentration levels for Pool 4 would enable the target volume of 2.7 liters to be achieved in 2–3 months without use of samples with trace or small amounts of blood.
- Dr. Delanghe commented that excluding these samples could potentially omit patients with glomerular membrane defects or diabetic nephropathy from Pool 4. Carefully centrifuging the samples, he added, should remove the contaminated red blood cells. Dr. Bachmann explained that the rejected samples currently are being discarded but could be retained. Dr. Phinney noted that NIST knows of no measurement issues from traces of blood in the samples that are used for reference materials.
- Dr. Jian Dai explained that the preferred option is to exclude samples containing trace amounts of blood because it is not always known what influence blood might have on different commercial methods.
- Dr. Beasley-Green asked whether blood-traced urine samples were routine in a clinic laboratory and wondered how these samples are analyzed on the various platforms. Dr. Seegmiller replied that, clinically, how a patient sample presents cannot necessarily be predicted. Analytical handling of interferences would depend on the platform being used. Dr. Dai pointed out that the calibration should be precise, and Dr. Bachmann reiterated how the differences in selective versus nonselective assays might impact the commutability studies.
- Dr. Phinney recommended excluding the rejected samples for Pool 4 and noted the need to further explore this question on sample interference, adding that it would not necessarily be up to NIST to determine the sample quality. Dr. Miller added that the objective is to use samples without interfering contaminants and pointed out that UA variants will exist in the Pools.
- The consensus of the LWG was that it would not be acceptable to use samples that fail the quality screening criteria, including those samples containing trace levels of blood.
Reference Measurement Procedures (RMPs) for Urine Albumin
Ashley Beasley-Green, NIST
Jesse Seegmiller, University of Minnesota
Dr. Beasley-Green explained that the NIST Candidate RMP for UA is based the IDMS approach involving use of a nitrogen 15 (15N)-labeled full-length human serum albumin (HSA) internal standard that is added to the urine specimens. The reference materials are assessed quantitatively by the UA concentration and qualitatively based on their heterogeneity to the 10 peptides that span the amino acid sequence of HSA. The NIST–Mayo Clinic UA method comparison and validation studies analyzing reference materials using three peptides unique to human albumin have been completed. Dr. Beasley-Green reported that issues regarding updates to the Federal Policy for the Protection of Human Subjects (or the Common Rule) that temporarily suspended the NIST–Mayo Clinic collaborations have been resolved. Material Transfer Agreements between NIST and the Mayo Clinic have been established, and discussions are expected to resume in the upcoming weeks.
University of Minnesota and Mayo Clinic
Dr. Seegmiller provided an update on the UA standardization project and presented updates on behalf of Dr. John Lieske, Mayo Clinic, who was unable to attend the meeting. The goals of the project are threefold: (1) develop an RMP, (2) ensure that two separate laboratories are capable of performing the RMP, and (3) register the RMP with the JCTLM. Dr. Seegmiller reported that the bottleneck to progress has been in acquiring sufficient labeled HSA for the internal standard. Existing sources from commercial entities are too costly to support the large-scale validation and testing necessary for this project. In addition, peptide-based internal standards, which cost less to prepare, do not perform well. To address these issues, the Advanced Research and Diagnostic Laboratory at the University of Minnesota collaborated with the Mayo Clinic Renal Reference Laboratory to synthesize labeled HSA. The service contract award was finalized in April 2018, which coincided with the end of the award cycle, and an extension was granted to continue work past that deadline. To date, synthesis and purification of the 15N-labeled HSA has been completed. The stock solutions, which currently are being stored at -70 °C, have been divided between the Mayo Clinic and the University of Minnesota for further characterizations.
Dr. Seegmiller described the process of preparing for method validation that includes quantification of the internal standard, development of a calibration matrix, preparation of the sample, and utilization of a common liquid chromatography tandem mass spectrometry (LC-MS/MS) assay. Prior to validating a method, it would be necessary to understand the performance of the measurement procedure by performing preliminary analyses and fully assessing the lower limit of quantification and imprecision requirements. Because the median coefficient of variation (CV) of most LC-MS/MS measurement procedures ranges from 5–20 percent, achieving a CV of less than 6 percent for the RMP will be challenging. Efforts may need to focus on the use of nonstandard replicate sample averaging. In closing, Dr. Seegmiller emphasized that validation studies will commence as funding allows and that Mayo Clinic administrative issues regarding further comparisons with the NIST method reference materials, including calibrators and standards, recently was resolved.
- Dr. Miller remarked that as the RMP validations progress, the objective is to ensure that all studies and requirements for a JCTLM submission are completed. Monthly coordination meetings will be essential. Dr. Phinney commented that a list of the submission requirements would be helpful.
- Dr. Miller sought clarity on whether the NIST reference materials will be available for the Mayo Clinic–University of Minnesota validations. Dr. Beasley-Green clarified that the SRM 2925 and other calibrators would be made available, and discussions on validations at the three sites—NIST, the Mayo Clinic, and the University of Minnesota—are forthcoming.
Plan for Implementing Calibration Traceability to an RMP
W. Greg Miller, VCU
Dr. Miller described the general strategy for implementing the calibration and traceability to an RMP. He reminded the LWG members of the two-pronged approach for manufacturers to achieve UA measurement standardization that was discussed at the 2017 LWG meeting. One approach is the use of SRM 3666 as a commutable secondary reference material, and the second approach is the use of an RMP calibrated with SRM 2925 to value-assign UA in a panel of human urine samples. The primary reference material, SRM 2925, is soon to be released by NIST and is expected to be included in the next JCTLM cycle. By 2020, it is anticipated that a suitable JCTLM-listed primary reference material and reference measurement procedures will be available that will enable IVD manufacturers to use individual urine samples for calibrations. The secondary reference material, SRM 3666, should be released by 2022 and is expected to simplify the calibration traceability process and provide an additional tool to IVD manufacturers. Subsequent to implementing this general direction, a more structured planning process will be established. An LWG-IVD manufacturer’s collaborative submission to the U.S. Food and Drug Administration (FDA) more than likely will be considered. Dr. Miller recommended inviting an FDA representative to the next LWG meeting to advise on coordinating the FDA submission and discuss a streamlined approval process.
Recommendations for Reporting UA and ACR Values that Exceed the Measuring Interval
W. Greg Miller, VCU
Dr. Miller provided an update about the prior recommendation to generate an opinion paper addressing the importance of reporting UA and ACR values that exceed the measuring intervals because higher values are important for monitoring patients with kidney disease. A draft document developed with input from a subgroup of LWG members—including Drs. Bachmann, Lieske, Narva, James Fleming and Graham Jones—is being finalized. A final draft will be circulated to co-authors for comments, and additional input from other members is welcomed. The general recommendations are to dilute samples and report quantitative values rather than use the designation greater than a value.
Potential Collaboration with the National Kidney Foundation (NKF)
Andrew Narva, NIDDK
Dr. Narva described an NKF effort to improve adherence to evidence-based care in the United States by facilitating provider laboratory orders for urine albumin-to-creatinine ratio (UACR). He noted that the slides in the presentation on the NKF CKD initiative—CKDintercept—were provided by Dr. Joseph Vassalotti, Chief Medical Officer at NKF. He added that Dr. Elizabeth Montgomery, who attended this meeting, also is the program manager for this initiative and would be available to answer questions. The United States Renal Data System Annual Data Report indicates that although the routine reporting of estimated glomerular filtration rate (eGFR) is widely used for CKD across the clinical community, urine albumin tests are underutilized in at-risk groups for CKD.
Dr. Narva pointed out that the NKF convened a group of laboratory specialists currently known as the Laboratory Engagement Advisory Group (Advisory Group) that is similar in purpose to the NKDEP LWG, has members that are common to both groups, and shares two of the same recommendations: harmonizing the name of UA/ACR and harmonizing the reporting units for UA. Dr. Narva highlighted a third Advisory Group recommendation: creating a Kidney Profile laboratory order consisting of eGFR and ACR testing in one order. It is anticipated that the Kidney Profile will simplify the testing process for clinicians and significantly improve adherence to evidence-based care. Dr. Montgomery commented that the initial goal was to increase awareness in the clinical community and not on the science and standardization. The NKF proposed the Kidney Profile in response to feedback about the confusion about the term microalbumin, the lack of ACR reporting consistency from large health care payers using data analytics to track CKD patients, and the need for simple descriptions in patient records that UA is an assessment for CKD. The broader clinical community is collaborating with the NKF and has assisted with implementation.
- Dr. Delanghe called attention to the World Health Organization’s recently published list of essential diagnostic tests and the puzzling recommendation to use the UA analysis for assessing nutritional status.
- Dr. Killeen pointed out that Medicare rules may counter use of the Kidney Profile regarding reimbursements and claims. Dr. Montgomery explained that the Laboratory Engagement Advisory Group discussed the impact on Centers for Medicare & Medicaid Services (CMS) guidelines and concluded that no new billing code would be necessary for a combination of laboratory tests.
- Dr. Miller proposed that the LWG collaborate with the Laboratory Engagement Advisory Group to develop a communication strategy for laboratories to accelerate broad implementation of recommended changes for UA and ACR. Dr. Flavio Alcantara suggested also including strategies for international groups.
- Dr. Narva noted the opportunities to leverage the U.S. Department of Health and Human Services Healthy People 2020 Initiative and the CMS Quality Measures Program.
W. Greg Miller, VCU
Dr. Miller asked whether there were remaining issues to be discussed.
Dr. Johanna Camara informed the LWG members that NIST will develop and certify a replacement for the creatinine reference, SRM 967a, which is anticipated to be exhausted in 3 years. NIST is asking for input on the standard concentrations. The options are to continue the current two levels (Level 1, 0.85 mg/dL, and Level 2, 3.88 mg/dL) or develop a lower Level 1 concentration standard. Dr. Camara asked whether developing a lower Level 1 standard would be problematic for laboratories that are currently using the 0.85 mg/dL and 3.88 mg/dL standards.
Lowering creatinine concentration in NIST SRM 967a was discussed. The following points were made:
- Dr. Miller commented that the IVD manufacturers have long requested a low-level creatinine reference standard and does not anticipate any resistance to NIST taking this approach. Dr. Miller also remarked on the issues related to lack of standards for the pediatric population and suggested that 0.4 mg/dL be a target concentration. Dr. Camara explained that developing a 0.4 mg/dL standard based on the routine clinical samples in this range would require 4 liters of volume. The logistics of collecting and preserving samples until the target volume is achieved would need to be considered.
- Dr. Killeen noted a survey conducted by the College of American Pathologists’ Accuracy Based Testing Committee and the Instrumentation Resource Committee, which showed a 0.7 mg/dL creatinine standard as the lowest achievable limit from pooled donor samples.
- Dr. Miller asked the manufacturers for their perspective on a lower concentration standard. Dr. Dai pointed out that creatinine calibrations are generally linear across the Level 1 and 2 ranges, but a 0.4 mg/dL standard would be challenging. The current situation has more laboratories offering Jaffee-based assays than offering enzymatic assays. The Jaffe methods are challenged to have accurate measurement of low creatinine concentrations.
- The consensus of the LWG is that developing a low-level creatinine reference standard would be beneficial to the pediatric population and useful for the next generation of products. The technical limit that can be effectively achieved would need to be investigated. Implementing specific requirements for collections (e.g., meat restriction and donors with low muscle mass) might be helpful.
The meeting was adjourned at 1:37 a.m. CDT.
- The LWG will recommend that NIST modify the service contract to expand the albumin donor urine concentration range to 200–600 mg/L for the SRM 3666 4th-level material.
- Dr. Miller will organize a monthly call among laboratories developing the reference measurement procedures for UA to coordinate experiments needed to support JCTLM submission.
- Dr. Miller will prepare a list of the JCTLM submission requirements for review.
- The LWG will invite an FDA representative to the next LWG meeting to advise on the requirements for an FDA submission.
- Dr. Miller will circulate a final draft of the opinion paper to co-authors for review and comments.
- The LWG and the NKF Laboratory Engagement Advisory Group will collaborate on an implementation science strategy for broad adoption of the recommended changes for UA and ACR standardizations.
Flavio Alcantara, Ph.D.
Medical Chief of Section, Central Lab Division
Hospital das Clínicas da Faculdade de Medicina
University of São Paulo
Lorin Bachmann, Ph.D., DABCC
Chair, IFCC Working Group for Standardization of Albumin in Urine
Associate Professor, Department of Pathology
Co-Director, Clinical Chemistry Laboratory
Virginia Commonwealth University
Ashley Beasley-Green, Ph.D.
Staff Scientist, Biomolecular Measurement Division
National Institute of Standards and Technology
Johanna Camara, Ph.D.
Research Chemist, Chemical Sciences Division
National Institute of Standards and Technology
Jian Dai, Ph.D., FACB, FCACB
Siemens Healthcare Diagnostics
Joris Delanghe, M.D., Ph.D.
Professor, Department of Clinical Chemistry
University Hospital Ghent
Anthony Killeen, M.D., Ph.D.
Professor, Department of Laboratory Medicine and Pathology
Laboratory, Director of the Advanced Research and Diagnostics Laboratory
University of Minnesota Medical Center
W. Greg Miller, Ph.D.
Chair, Laboratory Working Group
Professor, Department of Pathology
Co-Director, Clinical Chemistry Laboratory
Director, Pathology Information Systems
Virginia Commonwealth University
Bogdan Milojkovic, Ph.D.,
European Director of Medical and Scientific Affairs
Elizabeth Montgomery, Ph.D.
Senior Project Director of the Primary Care Initiative, CKDintercept
National Kidney Foundation
Andrew Narva, M.D., FACP
Director, National Kidney Disease Education Program
National Institute of Diabetes and Digestive and Kidney Diseases
National Institutes of Health
Karen Phinney, Ph.D.
Research Chemist, Biomolecular Measurement Division
National Institute of Standards and Technology
David Seccombe, M.D., Ph.D., FRCPC
Canadian External Quality Assessment Laboratory
Jesse Seegmiller, Ph.D.
Assistant Professor, Department of Laboratory Medicine and Pathology
University of Minnesota