Association of Clinical Biochemists in Ireland Annual Conference

Aspergillus infections in Ireland and verification of A. fumigatus and A. flavus specific IgG on Phadia 250

K. Hutchinson (1), A. Ross (1), A. O’Machony (1), A. Zakharova (1), M. Louw (1), Faul JL (2), M.M. Hannan (1,3).

1. Eurofins Biomnis, Sandyford, Dublin 18, Ireland.

2. Asthma Research Centre, Connolly Hospital, Dublin 15, Ireland.

3. Mater Misericordiae University Hospital Dublin, Ireland.

Background:

Aspergillus species are the most frequent cause of fungal infections of the lungs with a broad spectrum of clinical presentations, including allergic, invasive and chronic pulmonary aspergillosis (PA). It has a high rate of misdiagnosis and has been reported to have an increasing rate of morbidity and mortality, specifically in immunocompromised and elderly populations. Detection of Aspergillus-specific IgG can be critical in diagnosis of PA.

Methods:

We have performed verification of Aspergillus fumigatus (m3) and Aspergillus flavus (m228) IgG assays on the Phadia 250 Thermo Fisher system. We also carried out an audit of m3 and m228 IgG in patients from GPs’ practices during 2023-2024 (total n=1326). All data were analysed in compliance with “Guidance on Anonymisation and Pseudonymisation”, June 2019. The software used for the statistical analysis were EP Evaluator and Microsoft Excel. The positivity rate (>27 mgA/L) of m3 and m228 tests was analysed overall (and in relation to sex and age groups <18; 18-40; 40-60; 60-75 and >75).

Results:

Within-run and between-run precision for m3 and m228 were <10%. Comparison with Eurofins France using the same methods showed acceptable >83.5% agreement for over 100 samples (Cohen’s Kappa >75%). McNemar test for symmetry passes for m3 with p=0.655; m228: p= 0.125.

The positivity rates for m3 and m228 were 44% and 40% respectively, more prevalent in females (m3: M=39%, F=49%; m228: M=34%, F=45%). The 60-75 age group had the highest number of positive results for males and in the 40-60 group for females, for both m3 and m228.

Conclusion:

Our study shows acceptable performance for both tests on the Phadia 250. The test repatriation to Ireland will improve our turnaround time for the screening for Aspergillus infection. More study is necessary to evaluate prevalence of Aspergillus infection in the Irish population.

Verification of faecal calprotectin measurement on the BioSystems A15s analyser at University Hospital Galway

Mariah Kelly, Verena Gounden, Damian Griffin

Background: This study aimed to verify the performance of the BioSystems A15s analyser for measuring faecal calprotectin as a diagnostic and monitoring tool for inflammatory bowel disease at University Hospital Galway.

Methods: The study evaluated the assay’s precision, linearity, functional sensitivity, accuracy, and compared it with the current external Diasorin Liaison XL method used by Viapath Analytics LLP in London. A faecal calprotectin stability study and a cost-benefit analysis for the introduction of faecal calprotectin testing onsite was conducted.

Results: Results demonstrated that the BioSystems faecal calprotectin assay met acceptable standards for within-assay and within-laboratory precision. The assay also demonstrated acceptable functional sensitivity and linearity up to 1632 µg/g. In terms of accuracy, a proportional bias was observed at higher faecal calprotectin concentrations. Implementing this onsite testing at University Hospital Galway is projected to reduce turnaround times from an average of 16 days to 2-3 days. It will significantly enhance patient management by reducing unnecessary colonoscopies and result in substantial cost savings estimated between €1.36 million and €4.09 million over five years.

Conclusion: While the study supports the integration of the BioSystems A15s analyser into routine clinical practice at University Hospital Galway, further research is needed to refine the faecal calprotectin cut-off values specific to the hospital’s clinical context. The findings underscore the analyser’s potential to improve diagnostic efficiency and patient outcomes in gastroenterology services at University Hospital Galway.

Two Cases of Galactokinase Deficiency diagnosed during Newborn screening (NBS) with elevated Galactose

Sabah Abdelfadil1, Rahidatul Fairuz Ibrahim1, Loretta O’Grady1, Ann O’Loughlin4, Jennifer Brady1, and Mohamed Elsammak1

1. National Newborn Screening Laboratory, Department of Laboratory Medicine, Temple Street Children Hospital, Dublin, Ireland

We report 2 babies born to a non-traveller Irish community families (Asian and Irish) with no family history of classical galactosaemia (CG) who were found to have a markedly elevated total Galactose (TGAL) level of > 3000umol/L on Newborn Screening (NBS) cards with normal Beutler; indicating normal galactose-1-phosphate uridyltransferase (GALT) enzyme activity. NBS for CG in Ireland uses 2 tier testing approach for non-high risk babies (no Family history and non-Irish travellers). Measuring Total galactose represents Tier-I testing. Babies with galactose level > the first cut off value COV (99th centile=574) had their T-Gal repeated in triplicates and if the T-Gal is > 2nd COV (99.9thcentile=874), second Tier-II testing (measuring enzyme activity) Beutler; is performed. Cases with no activity present Beutler (activity <950) repeated in triplicate are labelled as CG positive.

As those two babies’ TGAL level exceeded both COVs, they both had their GALT enzyme activity checked on their first NBS cards received and on the subsequent cards received by the lab. Presence of the enzyme activity was demonstrated on both cases excluding the diagnosis of CG. Molecular genetic testing was then performed to both cases revealed pathogenic variants in GALK1 that was consistent with Galactokinase (GALK) deficiency Galactosaemia. GALK is the first enzyme in the pathway of galactose metabolism, converting galactose to galactose-1-P. The only consequence of GALK deficiency is the development of cataracts that was identified on the second case. The pathogenic mechanisms underlying this localised defect is hypothesised to be secondary to galactitol accumulation. Patients with GALK deficiency; have elevated galactose level with low Gal-1-Phosphate and high urinary galactitol and glucose.

The first case was found to be homozygous for a variant in gene GALK1 and the second case was a heterozygous for two variants in gene GALK1.

While it is common to identify cases of CG during NBS, it is also important to rule out other galactosemias caused by different enzymes involved in the galactose metabolism in any unexplained elevation of total galactose during NBS. These 2 cases are an important example of GALK deficiency required further treatment (galactose restricted diet) and ophthalmology follow up for cataract.

A Case of Possible Cystic Fibrosis and Becker’s Muscular Dystrophy presented with persistently elevated Alanine Aminotransferase

Rahidatul Fairuz Ibrahim4, Sabah Abdelfadil4, Basil Elnazir1, 5, Nadeem Gadelseyed1, Sarah Murphy1, Denise McDonald2, Barry Linnane3, Loretta O’Grady4, Catherine Harvey4, Jennifer Brady4, Abbey Collins4, Paul Marsden4 and Mohamed Elsammak4

1. CHI at Tallaght Children University Hospital, Department of Pediatric, Dublin.

2. Tallaght Children University Hospital, Department of Neurology, Dublin

3. Limerick University Hospital Department of Pediatrics, Dublin.

4. National Newborn screening laboratory, Temple Street Children Hospital, Dublin

5. Trinity College Dublin

Cystic Fibrosis (CF) is one of the most common autosomal recessive monogenic diseases that may present with significant morbidity and mortality. The incidence of CF at birth is 1/14615 in Ireland with an approximately 1 in 19-carrier rate. The CF newborn screening (NBS) programme was introduced in Ireland in July 2011.It involves measuring the immunoreactive trypsinogen (IRT) level on a dried bloodspot (DBS) sample taken during the first 48-72 hours of life. A cut-off value of 99th percentile of IRT is used. For any baby with IRT value above the 99th percentile cut-off value, DBS is repeated in triplicate. If the mean of the triplicate is higher than the cut-off value, sample is then sent for molecular confirmation of DNA analysis. If the later reveals one or two pathogenic mutations, a sweat chloride testing is the gold standard for diagnosis. Herein, we report a case of CF that was undetected by CF NBS. The patient presented with myopathy at the age 9 months and persistently elevated alanine aminotransferase (ALT). A total creatine kinase (CK) was measured to assess for dystrophinopathies and was significantly elevated at 3000 U/L on repeat sample. Molecular testing confirmed the diagnosis of Duchene Muscle Dystrophy. During molecular testing, two cystic fibrosis mutations were incidentally detected, a F312DEL mutation and the classic cystic fibrosis causing mutation, Phe508del. Sweat chloride testing was repeatedly elevated in keeping with the diagnosis of cystic fibrosis. Despite of the significantly elevated sweat chloride test and molecular genetic profile showing heterozygosity for Phe508del and F312DEL mutations, patient did not show any clinical manifestation of cystic fibrosis. During newborn screening, IRT result was 26 ng/ml, which is way below the cutoff value used for screening (54 ng/ml).

In summary; patients’ heterozygous for F312DEL and Phe508del mutations may not be detected during newborn screening and may not have clinical manifestation of CF despite having unequivocally elevated sweat chloride test. An unexplained elevation of transaminases warrants CK testing to assess dystrophinopathies.

A case of Transient Neonatal Cystinuria (TNC) in a premature baby with negative Whole Exome Sequencing (WES)

Rahidatul Fairuz Ibrahim1, Clodagh Sweeny2, Robert.Kernan3, Sabah Abdelfadil1, Patricia Fitzsimons1 and Mohamed Elsammak1

1. Department of Paediatric Laboratory Medicine, CHI Temple Street Children Hospital, Dublin.

2. Department of Paediatric Nephrology, CHI Temple Street Children hospital, Dublin.

3. Department of Paediatric Neurology, CHI Temple Street Children hospital, Dublin.

Cystinuria is an autosomal recessive metabolic disorder characterized by defective reabsorption of cystine and dibasic amino acids (lysine, arginine, and ornithine) in the proximal renal tubules and small intestine. The condition involves mutations in two key genes: SLC3A1 and SLC7A9. The SLC3A1 gene encodes the rBAT protein, while the SLC7A9 gene encodes the b0,+AT protein. Cystinuria is classified into three types based on genetic mutations: Type A (SLC3A1 defects), Type B (SLC7A9 defects), and Type AB (mutations in both genes). Cystinuria is characterised by persistently elevated levels of cystine in the urine, with or without urinary cystine stones formation. In contrast, transient neonatal cystinuria (TNC) is observed in newborns with initially high cystine levels that decrease over the first two years of life due to immature renal amino acid transport. Confirmatory diagnosis involves measuring urinary amino acid levels, with cystine typically exceeding 1000 μmol/g creatinine. In this report, we describe a case of TNC that developed in a 15-month-old boy who was born extremely premature at 24 weeks and 3 days gestation. Immediately after birth, he required intubation and mechanical ventilation due to respiratory distress and poor Apgar score. At 40 days old, the boy was re-admitted with a non-functioning ileostomy, necessitating a stoma revision and adhesion lysis. During this admission, he experienced multi-system dysfunction including cardiac arrest, recurrent episodes of oliguria acute kidney injury, liver dysfunction, and septicemia (fungal and bacterial infections). Urine amino acid analysis revealed elevated levels of cystine and the dibasic amino acids suggestive of cystinuria. Ultrasound findings showed nephrocalcinosis. Whole exome sequencing (WES) was negative for cystinuria. Our findings suggest the diagnosis of TNC as WES was negative for any pathogenic variants in the Cystinuria gene. Furthermore, the intronic region of the clinically relevant gene was thoroughly examined with no variants detected. The gradual decline in the level of cystine and dibasic amino acids in urine makes TNC more likely. In this case, the TNC is likely due to the immaturity of the renal tubular transport system for dibasic amino acids.

Vitamin D Deficiency among Haemodialysis Patients in Ireland

A. Mc Carn1,2, A. O’Machony1, M. Louw1, JL Faul3, K. Hutchinson1.

1. Eurofins Biomnis, Sandyford, Dublin 18, Ireland.

2. School of Biological, Health & Sports Science, Technological University Dublin, Dublin, Ireland.

3. Asthma Research Centre, Connolly Hospital, Dublin 15, Ireland.

Background:

Vitamin D deficiency is common in patients with chronic kidney disease (CKD), specifically those on haemodialysis. It is also associated with a high rate of morbidity and mortality. The prognostic importance of vitamin D deficiency (<30nmol/L) in these patients is still unclear. The aim of our study was to analyse the levels of 25OHD in patients on haemodialysis, and to examine if there is an underlying correlation between 25OHD concentrations and a range of biomarkers.

Methods:

Serum samples of 413 patients on haemodialysis were obtained from several Renal Clinics in January-April 2024 (in compliance with “Guidance on Anonymisation and Pseudonymisation”, June 2019). We analysed 25OHD, renal and bone biomarkers on the Abbott Architect ci8200. Full Blood Count was measured on Sysmex CS-2500. 25OHD results for non-CKD patients were mined from the database in Eurofins during the same months and anonymised (n=34.174). The statistical analyses used were the GraphPad Prism 5 and Microsoft Excel.

Results:

The data showed that 59% of dialysis patients had vitamin D deficiency with significantly lower mean=10.9nmol/L, in comparison to 19.5% non-CKD 25OHD deficient patients with the mean=21.4nmol/L. There was significant difference in 25OHD concentration in non-CKD and patients on dialysis, with the median=57.6 and 26.3nmol/L respectively. All CKD patients had estimated glomerular filtration rate (eGFR)<15ml/min/1.73m2. We found a significant positive correlation between 25OHD concentration and e-GFR (p<0.001). A significant negative correlation was observed with Creatinine, White Blood Cells, Neutrophils (p < 0.001) and Neutrophil Lymphocyte Ratio (p=0.002).

Conclusion:

Our study shows that 85% of dialysis patients had low Vitamin D levels. The significant negative correlations with biomarkers of chronic inflammation may indicate the importance for 25OHD testing for monitoring and management CKD patients on dialysis, but further studies are still required to assess the usefulness of Vitamin D measurement and supplementation.”

Investigation of optimal LDL cholesterol formula for use in a paediatric population

Tara O’Brien (1), James O’Dwyer (2), Maggie Griffin (2) Jennifer J Brady (1,2)

1 School of Medicine, University College Dublin

2 Department of Biochemistry, Children’s Health Ireland, Temple Street

Dyslipidaemia, traditionally seen in adults, is rising in children due to inherited causes and increasing childhood obesity. The quantification of LDL-C is a cornerstone in the clinical management of hyperlipidaemia, typically done through direct measurement or equations like Friedewald, Martin Hopkins (MH), and Sampson. While these equations are validated in adults, there is limited evidence on the use of MH and Sampson equations in paediatrics.

This study aimed to determine the most appropriate equation for LDL measurement for use in a paediatric population, from 2,774 patients (retrospective gather over a 2-year period) and comparing Sampson and MH equations to the Friedewald equation. Additionally, data from 110 patients were prospectively collected and all three equations were compared to a direct LDL measurement (Siemens Atellica).

The mean differences between Friedewald and Martin Hopkins, Friedewald and Sampson, and Martin Hopkins and Sampson are -0.02 mmol/L (95% Confidence interval (CI) -0.31 to -0.27), -0.52 mmol/L (95%CI -0.36 to -0.68), and -0.50 mmol/L (95%CI - 0.34 to -0.65), respectively. Comparing Martin Hopkins and Friedewald equations, 2% of patients were discordant at 2.8mmol/L, much lower than the 18.2% discordance rate between Sampson and Friedewald equations. The Martin Hopkins and Sampson formulas aligned well for triglycerides >4.5 mmol/L (n=21). The Sampson equation showed the least agreement with direct measurement, having a mean positive bias of 0.37 mmol/L(95%CI -0.06 to -0.68).

Our study does not contradict other works that shows Martin Hopkins to be the preferred formula, however, it does not demonstrate its superiority over the Sampson equation. The Sampson formula is more suitable for use in the laboratory LIMS.

Groenendijk W1, Keane P1, Fields M1, Hunter A1, Crosbie B1, Tarrant H1, Fadalla K2, Twomey PJ1

1. Department of Clinical Chemistry, St. Vincent’s University Hospital

2. Department of Haematology, St. Vincent’s University Hospital

Title: A Case of Missing IgM

Background:

UK Myeloma Laboratory Best Practice Tool (2023) includes serum electrophoresis and total immunoglobulins as core tests for initial myeloma evaluation. We describe a case illustrating the importance of quantitative immunoglobulin and serum albumin for accurate interpretation of serum protein electrophoretograms.

Case:

78y male with chronic lymphocytic leukaemia presented with extremely elevated IgM (>58.50g/L, RI 0.4–2.4g/L), low IgG and IgA and serum albumin of 31g/L (RI 35–50g/L), measured by Roche Cobas c702. Subsequent capillary zone electrophoresis (CZE) showed a small monoclonal band (4g/L, Gamma region) and serum albumin of 55g/L with monoclonal IgM Kappa on immunofixation.

Significant discrepancy across Sebia and Roche platforms for albumin and monoclonal band quantification indicated assay interference. As total protein, albumin, immunoglobulins, and subsequent calculated globulins values were consistent, it was highly likely the interference lay with the CZE analysis.

After treatment with reducing agent (di-thiothreitol,DTT) to disrupt IgM polymerisation, CZE analysis confirmed the presence of a large monoclonal protein (approx.35g/L) in the Beta-2 region and albumin of 34g/L, concordant with the c702 estimation.

A retrospective analysis of all paired c702 and CZE albumin estimations was performed (Jan-Jun 2024, n=5327). Difference data was approximately normally distributed and described by: CZE Albumin = 4g/L + c702 Albumin (95% CI 4.25–4.35) on Bland Altman analysis. Application of the >99.5th centile (10g/L) and <0.5th centile (0g/L) of this difference data as the basis for an electronic rule on our LIMS is currently being trialled as an additional safeguard for highlighting interference.

Conclusion:

High concentrations of monoclonal immunoglobulin are universally recognised as a potential cause of interference in the different assays used in the evaluation of monoclonal gammopathies. We propose the use of an electronic rule flagging discrepant albumin results measured on the Roche c702 and Sebia CZE platforms as an additional method for highlighting possible assay interference.

Verification of the ARK Diagnostics methotrexate assay on the Roche Cobas c702 for clinical use

Kate Griffin, Sinead McDermott, Helen Doheny, Janice Reeve, Martina Doheny, Karen Heverin, Verena Gounden, Damian Griffin

Department of Clinical Biochemistry, Galway University Hospital

INTRODUCTION. Methotrexate (MTX) is a folate antimetabolite used to manage neoplasms, severe psoriasis and rheumatoid arthritis etc. MTX monitoring aims to avoid excessive toxicity and determine when to intervene with folinic acid rescue in patient’s requiring high dose MTX therapy. In the Department of Clinical Biochemistry, the soon-to-be decommissioned Abbott Architect quantifies serum MTX. To preserve the clinical service, the MTX assay required verification on the Roche Cobas c702.

METHODS. Before introduction all assays require verification; method comparison, precision, accuracy and linearity. Sera sent for routine MTX analysis were collected and frozen before testing on the Abbott Architect and Roche Cobas c702; n=56, range 0.04-0.85µmol/L. Precision was assessed using a 5 (replicate) x 5 (day) design with ARK Diagnostics Internal Quality Controls (IQC). The assay Limit of Quantitation (LoQ; 0.04µmol/L) was verified with repeat measures of patient serum with low MTX concentration (n=10). Accuracy and linearity were evaluated using RIQAS External Quality Control (EQA) material with known MTX concentrations. Data generated was assessed using Microsoft Excel.

RESULTS. Passing Bablok analysis produced equation of the line, Roche = 1.118(Abbott)+0.01; r 0.993. The mean absolute and relative differences between methods was 0.04µmol/L and 24.5%, respectively. IQC precision met manufacturers’ acceptance criteria; standard deviation (SD) ±0.1µmol/L when MTX <1.0µmol/L, coefficient of variation ±10% when MTX >1.0µmol/L. EQA analysed on the Roche (n=12) met with scheme provider criteria; SD Index <2, target score >50, %deviation <15%. Linearity was evident, y=1.02x+0.03, r 0.988.

CONCLUSION. At the clinically important low end, the Roche method, produced results higher than the Abbott Architect but this was considered clinically insignificant. The assay was considered fit for clinical use.

The Genetic Landscape of Porphyria in Ireland: Insights from a National Clinical Laboratory Service

Micheál Mac Aogáin1,2, Sarah Savage1, Brendan Lawlor3,4, Craig Maher1, Niamh Stein1, Aoife McConnon1, Nadia Brazil1, Robert O Byrne1, Vivion E.F. Crowley1,2.

1. Department of Biochemistry, St James’s Hospital, Dublin, Ireland.

2. School of Medicine, Trinity College Dublin, Ireland.

3. Department of Computer Science, Munster Technological University, Cork, Ireland.

4. Department of Biological Sciences, Munster Technological University, Cork, Ireland

Background: The Porphyrins Laboratory at the Biochemistry Department, St. James’s Hospital, Dublin, has served as a national reference service for porphyria in Ireland for over 40 years. This extensive experience has led to the establishment comprehensive data containing clinical, genetic, and biochemical information on more than 100 Irish families affected by porphyria.

Aim: To evaluate the genetic landscape of porphyria in Ireland by leveraging a comprehensive database that integrates clinical, genetic, and biochemical data.

Methods: We retrospectively analyzed data from 87 families with a biochemically confirmed porphyria phenotype (having excluded those with limited or incomplete clinical or biochemical data). Biochemical diagnosis was based on combined HPLC and fluorescence spectrophotometry across feacal, plasma, red blood cell, and urine samples. Genetic diagnoses were established through Sanger-based mutation scanning. The primary presenting phenotypes included Acute Intermittent Porphyria (AIP), Erythropoietic Protoporphyria (EPP), Variegate Porphyria (VP), familial Porphyria Cutanea Tarda (fPCT), and Hereditary Coproporphyria (HCP). Identified genetic variants were annotated and assessed for pathogenicity using a curated in-house genetic reference database, housed on a cloud-native, publicly accessible platform (PorphyriaDB.com).

Results: Among the 87 families analyzed, 74% had at least one member with a confirmed genetic diagnosis. The distribution of phenotypes was as follows: AIP (27%), EPP (27%), VP (20%), fPCT (11%), and HCP (10%) with several variants identified in the HMBS, PPOX, and CPOX genes, as well as FECH, and UROD highlighting a heterogeneous genetic landscape of porphyria in Ireland.

Conclusion: This study underscores the benefits of integrating molecular diagnostics and accurate variant annotation in the management of porphyria. The use of a comprehensive genetic reference database is imperative to ensuring genetic susceptibility is accurately identified and monitored in affected families, contributing to better patient care and understanding of porphyria in Ireland.

Compliance and Ethics: The study was approved by the SJH/TUH Joint Research Ethics Committee (REC: 2020-10 Chairman’s Action (40)).

Keywords: Porphyria, biochemical genetics, variant annotation, PorphyriaDB.

Introduction of Calculated Globulin – Practicalities and Considerations

Janice Reeve*, Verena Gounden*, Karen Heverin, Michael O’Meara, Damian Griffin

*Both authors contributed equally to this work

Department of Clinical Biochemistry, Galway University Hospital

Introduction. The HSE National Laboratory Handbook for Chemical Pathology Laboratory Profiles suggests that calculated globulin (CG) is determined when total protein (TP) and albumin (ALB) are measured. CG of ≥50g/L on General Practice (GP) adult patients are automatically sent to a clinical authorisation queue in the Laboratory Information System (LIS; APEX) occasionally, elevated CG from secondary care are manually referred for review. If a paraproteinaemia is considered likely, the sample is referred for immunoglobulin (Ig) and serum protein electrophoresis (SPEP) analysis. We aimed to determine how many monoclonal gammopathies were identified and derive a local population-based reference interval (RI) for CG, with commentary.

Methods. Specimens with CG (September 2023-August 2024) were extracted from the LIS using Cognos Impromptu software. Samples referred to Immunology were reviewed for a monoclonal gammopathy. GP TP and ALB results in patients >18years from March and April 2024 were retrieved. CG was calculated (TP–ALB; g/L). Analyse-IT software used the quantile method to derive 95th (2.5-97.5th) and 99th (1-99th) centile reference intervals with 90% confidence intervals (CI).

Results. From the 398,579 CGs, 878 specimens (227 patients) were ≥50g/L. 102 went to the authorisation queue for review; 73 GP results were sent automatically, 28 Secondary Care results were sent manually. Ten GP specimens (13.7%) were referred to Immunology; 7 (70%) had a monoclonal band, 3 (30%) did not. The 2/28 (7%) Secondary Care specimens with elevated CG referred had no evidence of a monoclonal band. From two months GP data, 29,222 TP and ALB results were available. The derived 95th (2.5-97.5th; 90% CI) and 99th centiles (1-99th; 90% CI) for CG were 20-35 g/L and 16-38 g/L, respectively. When implemented, CGs outside of 16-38 g/L would be highlighted as abnormal on the LIS and include interpretative comments indicating an abnormality with suggested follow-on testing.

Conclusions. Currently, all GP CGs ≥50 g/L are reflectively reviewed; review of Secondary Care CGs is ad hoc. Low CGs are not reviewed. Going forward, CG will be reported with a locally derived RI and interpretative comment to allow more appropriate decision making on follow-up investigations.

Audit of TFT requesting on two different platforms at University Hospital Galway Department of Clinical Biochemistry

Verena Gounden1, Karen Heverin1, Janice Reeve1, Michael O’ Meara1, Damian Griffin1

1Department of Clinical Biochemistry, University Hospital Galway

Background

Both Abbott Architect two-step FT4 and TSH assays and the Roche Cobas one-step FT4 and TSH assays have historically been available in the Clinical Biochemistry laboratory at UHG. The Abbott assay was utilised for cases where interference was suspected in the Roche assay. The aim of this study was to assess the pattern and magnitude of discrepancies between the two assays and determine if the basis for discrepancies in the platforms was methodological/clinical or both.

Methods

Data was extracted from the laboratory information system, APEX, using Cognos Impromptu software. Adult and paediatric TFTs performed on both platforms between 1/1/2023-31/12/2023 were reviewed against the in-use reference intervals (RI), as well as manufacturer-specific/published age-related RIs. Results were grouped into categories, A) both Abbott and Roche TSH within RIs; B) both Abbott and Roche TSH outside of RIs and C) discrepant TSH on the Abbott and Roche based on respective RIs. Categories were further subdivided by FT4 results.

Results

For the period of review, a total of 141 TFT sets were reviewed. The majority of samples (n=125) were concordant for TSH for both assays. 13 samples with concordant TSH within RI had FT4s greater than the Roche upper limit and Abbott FT4s within RI; when the 2.5 - 97.5th centile RI was applied to the Abbott FT4, 5 patients would then also be classified as above the RI (35.7%). A further 12 samples with concordant TSH within RI and Abbott FT4 within RI had FT4s below the Roche RI. When appropriate paediatric Abbott FT4 ranges were applied, 11 patients (91.7%) were also classified as having a low Abbott FT4.

Conclusion

The apparent differences in interpretation of the FT4 results were due to the RI used (Abbott FT4 RI) and the inherent methodological differences between the assays. Actual analytical interference was uncommon.

An audit of adjusted calcium practice in the HSE West and North West region

Karen Heverin1, Verena Gounden1, Janice Reeve1, Michael O’Meara1, Damian Griffin1

1. Department of Clinical Biochemistry, University Hospital Galway

Introduction.

The HSE Laboratory Services Reform Programme published ‘Reporting Units for Results of Laboratory Investigations’, which advises laboratories to harmonise reference intervals (RI) for the reporting of results, particularly within each health region. The HSE West and North West (W&NW) health region delivers healthcare services to counties Donegal, Galway, Leitrim, West Cavan, Mayo, Roscommon and Sligo. The aim of this audit was to obtain information on the adjusted calcium (ACA) practice, including RI, in W&NW in light of this recommendation.

Methods.

An audit questionnaire on ACA practice was drafted and emailed to a nominee in each of the following clinical biochemistry laboratories which covered the W&NW region; University Hospital Galway, Sligo University Hospital, Roscommon University Hospital, Portiuncula University Hospital, Mayo University Hospital, Letterkenny University Hospital. Survey questions captured information on; manufacturer, total calcium (CA) and albumin (ALB) methods, ACA equation where applicable, and RI for CA and ACA in use. Responses were compiled in Microsoft Excel to facilitate comparison.

Results.

Manufacturer breakdown: 5/6 Roche, 1/6 Beckman. 2/6 use an in house derived equation for ACA. 1/6 do not calculate or report an ACA. 3/6 use the standard Payne et al equation; ACA = CA mmol/L + 0.02 [40- ALB g/L]. 5/5 Roche users: NM-BAPTA CA and BCG ALB methods. 5/5 Roche users: CA age-related RI broadly based on manufacturer, with different laboratories employing varying age bins. 2/2 laboratories with derived ACA equations use derived RI based on 95th centile of normal population. 2/3 Roche users using standard equation for ACA, employ CA age-related RI (Roche) for ACA. 1/3 Roche user employing standard equation for ACA, use RI 2.10 – 2.60 mmol/L.

Conclusions.

Harmonisation of the ACA equation and associated ACA RI used across Roche laboratories employing the same methods for CA and ALB in the W&NW region is indicated.

Proof of concept study assessing first trimester urinary total and glycated CD59 to predict Gestational Diabetes (GDM) in second trimester

Karen Heverin1, Md Nahidul Islam2, Delia Bogdanet3, Fidelma P. Dunne4 Paula M O’Shea5

1. Department of Clinical Biochemistry, Galway University Hospital

2. Biochemistry Department, School of Biological and Chemical Sciences, University of Galway

3. Department of Endocrinology, Sligo University Hospital

4. College of Medicine Nursing & Health Sciences, University of Galway

4. Department of Clinical Biochemistry and Diagnostic Endocrinology, Mater Misericordiae University Hospital

Introduction.

A novel approach to diagnosis of GDM is required, replacing the traditional 1 step 75g oral glucose tolerance test (OGTT) with a biomarker that is easily obtained. CD59 is a complement regulatory protein present in all cells; it is anchored to the membrane by a lipid tail, exposing it to circulating glucose levels. In diabetes, CD59 is inactivated by non-enzymatic glycation forming glycated CD59 (gCD59). The aims of this study were; to examine whether urinary total CD59 (tCD59) or gCD59 was detectable in the urine of GDM patients, and whether it was superior to OGTT in diagnosing GDM in a small retrospective cohort.

Methods.

Ethical approval was granted by the research ethics committee at Galway University Hospital (Ref GUH: CA 2026). ELISA method for Human CD59 developed by AssayGenie was used for the analysis of tCD59 and gCD59. ThermoScientific Multiskan FC measured absorbance of 96well plates. 34 patients urine samples collected in the 1st trimester, stored at -20°C were tested; 26 normal glucose tolerance (NGT), 8 GDM. Results were compared with the OGTT performed in 2nd trimester using the WHO 2013 IADPSG criteria. Centrifugal Filter Devices were used to concentrate samples, 10K and 3K MWCO. Analyse-IT was used to generate receiver operator curve (ROC).

Results.

gCD59 in urine was undetectable on initial testing, 50 times concentration of primary sample was required for detectable gCD59. tCD59 was not a good predictor of GDM when compared with fasting glucose, 1hr glucose and 2hr glucose respectively; (AuROC 0.519 (95% CI 0.249 – 0.791), 0.788 (95% CI 0.548 – 1.029), 0.796 (95% CI 0.569 – 1.023), 0.784 (95% CI 0.557 - 1.010).

Conclusions.

gCD59 was not present at sufficient concentration to be assessed as an eligible biomarker for GDM. tCD59 showed no discrimination between groups rendering it ineligible as a biomarker for GDM screening.

Verification of Alanine Aminotransferase according to IFCC Standardised Method with Pyridoxal Phosphate Activation on the Roche Cobas c702 platform

Olivia Graham 1,2, Helen Doheny 1, Maria Prout 1, Martina Doheny 1, Damian Griffin 1, Verena Gounden 1.

1. Department of Clinical Biochemistry, University Hospital Galway, (UHG)

2. Medical Science, Atlantic Technical University

Introduction

Patients with liver disease commonly develop fibrosis. Fibroscans are used to determine if a patient has liver fibrosis. Many centres use a patient’s FIB4 score, based on ALT, AST, platelet count and age, to select which patients should be offered a fibroscan. To ensure the accuracy of the FIB4, accurate measurement of ALT levels is essential.

Aim

Our routine ALT method doesn’t incorporate pyridoxal phosphate (PLP) activation which is considered best practice to ensure accurate & precise ALT measurement. The aim of this project was to verify the performance of an ALT method with PLP activation in accordance with the International Federation of Clinical Chemistry (IFCC), on the Roche Cobas c702 platform, as a potential solution to this problem.

Methods

Precision, functional sensitivity, linearity, and accuracy for the measurement of ALT with PLP were performed. Precision study utilised the Clinical and Laboratory Standards Institute (CLSI) EP15-A3 guidelines. Accuracy was assessed by comparing UKNEQAS EQA specimen results achieved in UHG with their corresponding target means. Linearity was verified, using dilutions of a sample with an elevated ALT level and assessed using linear regression and mean recovery. Functional sensitivity was assessed by measuring a sample with low ALT activity using within-run and between-run replicates. A method comparison study as per CLSI-EP9 guidelines was conducted.

Results

Precision performance was considered acceptable (<5% for each QC). Assay was deemed linear up to 700U/L, in accordance with the manufacturers’ claims. Acceptable functional sensitivity with a between run CV<18.3% was achieved. Trueness demonstrated an acceptable performance within bias allowable limits of 5.3%. Bland Altman plots, Passing Bablok regression and Spearman’s coefficient of rank correlation, demonstrated an acceptable correlation r>0.95 with an expected positive bias observed of 19.4%.

Conclusion

Overall, new method was deemed fit for clinical use, with review/update of in-use reference intervals required.

Verification of Asparate Aminotransferase according to IFCC Standardised Method with Pyridoxal Phosphate Activation on the Roche Cobas c702 platform

Marcella Cahill 1,2, Helen Doheny 1, Sinead McDermott1, Martina Doheny 1, Damian Griffin 1, Verena Gounden 1.

1. Department of Clinical Biochemistry, University Hospital Galway (UHG),

2. Medical Science, Atlantic Technical University

Introduction

Patients with liver disease commonly develop fibrosis. Fibroscans are used to determine whether or not a patient has liver fibrosis. Many centres use a patient’s FIB4 score, based on ALT, AST, platelet count and age, to select which patients should be offered a fibroscan. To ensure the accuracy of the FIB4, it is essential that AST levels are measured as accurately as possible.

Aims

Our routine AST method doesn’t incorporate pyridoxal phosphate (PLP) activation which is considered best practice to ensure accurate & precise AST measurement. The aim of this project was to verify the performance of an AST method with PLP activation in accordance with the International Federation of Clinical Chemistry (IFCC), on the Roche Cobas c702 platform.

Method

Precision, functional sensitivity, linearity, and accuracy for the measurement of AST with PLP experiments were performed. Precision was assessed using the Clinical and Laboratory Standards Institute (CLSI) EP15-A3 recommendations. Linearity was verified using dilutions of a high sample, and assessed using linear regression and mean recovery. Functional sensitivity was assessed by measuring samples with a low AST activity using within/between run replicates. Accuracy was determined by comparing UKNEQAS target means with results obtained in UHG. A method comparison study as per CLSI-EP9 guidelines was conducted.

Results

Precision performance was considered acceptable (<5% for each QC). Assay proved linear up to 700U/L, in keeping with the manufacturer’s claims. Functional sensitivity was deemed acceptable with a between run CV of 18.1% achieved. Accuracy evaluation was within an allowable bias limit of 3%. Bland Altman plots, Passing Bablok regression analysis and Spearman rank correlation coefficient, demonstrated acceptable correlation r>0.95 with an expected positive bias of 22.7%. This was in keeping with UKNEQAS performance data.

Conclusion

Overall the method was deemed fit for clinical use, with review/update of in-use reference intervals required.

An Unusual case of Juvenile Nephropathic Cystinosis mimicking Bartter Syndrome

Dean Burrows1, Patricia E Fitzsimons1, Rahidatul Fairuz Ibrahim1 Sabah Abdelfadil1, Eirin Carolan2, Niamh Dolan3, Atif Awan3, Mohamed Elsammak1

1 Dept of Paediatric Laboratory Medicine, Children’s Health Ireland at Temple Street (CHI@TS)

2 Dept of Endocrinology, CHI@TS

3 Dept of Nephrology, CHI@TS

Background

We present a 5y old boy referred to Nephrology with a biochemical suspicion of Bartter syndrome based on a history of hyponatremia, hypochloremic, and persistent hypokalemic metabolic alkalosis after investigation for epistaxis. On examination he was on the 9th centile for weight/height, cardiovascular, BP, ENT and chest normal, no hepatosplenomegaly, non-dysmorphic, no deafness, good muscle tone, good appetite, dental caries and mild speech delay.

Results

Initial investigations confirmed above findings despite potassium supplementation; additionally low phosphate and TRP (on supplements); normal calcium, plasma osmolality, urea with slightly increased creatinine. Elevated renin, β-2-microglobulin, retinol-binding protein, proteinuria, aminoaciduria and confirmed renal tubular dysfunction. However, he was not polyuric, polydipsic, no history of lethargy, diarrhoea, vomiting or constipation. Aldosterone, abdominal US, bone age, calcium/creatinine ratio, were all normal and no evidence of nephrocalcinosis on renal US. Above negative findings were discordant with Bartter syndrome.

Additional investigations revealed primary hypothyroidism with increased thyroglobulin antibodies, but normal anti-thyroid-peroxide antibodies and thyroid binding globulin. PTH, cortisol, prolactin, IGF1/IGFBP3 and serum copper were also normal. Having considered a possible diagnosis of Cystinosis, opthalmic examination of the cornea revealed crystalline keratopathy. Genetics for Bartter Syndrome and differential panel including Cystinosis revealed patient was homozygous for a deletion on 17p13.2, which includes the first eight coding exons of CTNS gene previously reported, confirming nephropathic Cystinosis. Leucocyte cystine was increased 10.68 nmol ½ Cys/mg protein (ref <0.3).

Discussion

Cystinosis is a multisystem lysosomal transport disorder often presenting in first year of life with failure to thrive, polyuria, polydipsia, episodes of severe dehydration and electrolyte imbalance, vomiting, constipation and sometimes hypophosphatemic rickets. Cystine accumulation leads to cellular dysfunction with tissue and organ impairment. Laboratory findings may include hypokalemia, normal anion gap metabolic acidosis, hypophosphatemia, hypocalcemia, low carnitine and less frequently hyponatremia. Metabolic acidosis is the more common disturbance, though cases presenting with hypokalemic metabolic alkalosis mimicking Bartter syndrome have been reported. Our case highlights the importance of ruling out nephrogenic cystinosis in unexplained generalized tubular dysfunction.

Evaluation of the comparability of the IDS-iSYS and Roche Elecsys Growth Hormone assays

Khanum S1, Cullen MR2, Lee GR1,2

1School of Medicine, University College Dublin, Belfield, Dublin 4, Ireland

2Department of Clinical Chemistry and Diagnostic Endocrinology, Mater Misericordiae University Hospital, Dublin, Ireland.

Dynamic function tests (DFT’s) which rely on growth hormone (GH) analysis are used in the assessment and diagnosis of conditions of GH deficiency and excess. Such DFT’s include the insulin tolerance test, glucagon stimulation test for GH deficiency and the oral glucose tolerance test for GH excess. Associated diagnostic cut–offs as quoted in international best practice guidelines may not be universally applicable where assay standardisation and comparability is lacking.

Aim:

We aimed to evaluate the comparability of the IDS-iSYS and Roche Elecsys GH assays with a long-term retrospective evaluation of external quality assurance (EQA) data and method comparison. The study findings will inform if there is any bias between these assays and the implications for the transferability of these diagnostic cut-offs between these assays.

Methods:

Method laboratory trimmed means (MLTM) for the IDS- iSYS and Roche Elecsys GH was collated from the UKNEQAS peptide scheme. Results of analysis from both methods was undertaken following analysis of a clinically relevant range of redundant and anonymous patient samples (serum) also undertaken. Ethical approval was not required.

Results:

Passing-Bablok analysis from the most recent 12 distributions for GH EQA data indicated a positive bias of 22% for the Roche Elecsys MLTM compared to the IDS iSYS MLTM; (n=48; Roche Elecsys =0.0374+1.22 IDS iSYS). This was corroborated following analysis of patient samples |(n=11), with a positive bias of 25% (Roche Elecsys =0.008+1.25 IDS ISYS).

Conclusion:

We report a significant difference between the IDS-iSYS and Roche Elecsys GH assays which may have clinical consequence. We contend that method-specific cut-offs for dynamic function test cut-offs should be considered with further clinical evaluation of the transferability (or otherwise) and impact of using universal cut-offs across different methods.

Assessing HbA1c testing trends, prevalence of haemoglobin variants, and the impact of the Chronic Disease Management Programme on HbA1c monitoring

Luka Kostallas1, Marguerite MacMahon2, Graham Lee1,2

1School of Medicine, University College Dublin, Belfield, Dublin 4, Ireland

2 Department of Clinical Chemistry and Diagnostic Endocrinology, Mater Misericordiae University Hospital, Dublin

Background: Haemoglobin A1c (HbA1c) is a glycated form of haemoglobin used to monitor long-term glycaemic control in diabetes mellitus, reflecting blood glucose levels over the past two to three months. The accuracy of Hb1c measurements can be compromised by haemoglobin variants (HbS, HbC, HbE, HbD) and elevated foetal haemoglobin (HbF). Additionally, the Chronic Disease Management Programme (CDMP) advocates HbA1c testing in patients at high risk of diabetes. This study aimed to evaluate the prevalence of Hb variants and HbF and changes in HbA1c requesting trends to assess the impact of the CDMP.

Methods: HbA1c results from July-December 2019 and January-December 2023 at Mater University Hospital were analysed. Data included patient demographics, HbA1c levels, Hb variants and elevated HbF. Over- and under-requesting of HbA1c tests was assessed based on UK NICE and ADA guidelines. Data from January-June 2024 was included to evaluate under-requesting in patients with well-controlled diabetes.

Results: Hb variants increased >4 fold from 0.6% (2019) to 0.7% (2023). HbS was the most common (75%), followed by HbD (13.4%), HbC (7.7%) and HbE (4.3%). Patients with elevated HbF increased from 208 to 617, with a higher prevalence in women. GP requests accounted for 63.5% of activity in 2019 and 75.7% in 2023. Over-requesting was common in GP patients with well-controlled diabetes (34.4%) and under-requesting was prevalent in GP patients with poorly-controlled diabetes (45.2%).

Conclusion: This study provided insight into trends in HbA1c testing, Hb variant and HbF prevalence, and the effect of the CDMP programme on monitoring. The rise of Hb variants highlights the need to consider appropriate methodologies for HbA1c analysis methodologies in routine practice. The increased volume of requests may reflect uptake of the CDMP’s. However, 51% of repeat requests from were ordered inappropriately, suggesting a need for initiatives to control HbA1c retesting in line with published guidance.

Background/Aims: Alanine aminotransferase and aspartate aminotransferase are integral to the non-invasive Fibrosis Index 4 (FIB-4) score for assessing liver fibrosis. This study sought to determine whether different measurement methods for aminotransferases impacted the FIB-4 score and patient referral to liver clinics.

Methods: Patient samples (n=60) were analysed, and results were compared by method performance and principle. Combinations of assays (AAST/AALT, AAST/ALT, AAST/ALT2, AST/AALT, AST/ALT, AST/ALT2, AST2/AALT, AST2/ALT and AST2/ALT2) were input into the FIB-4 equation with the following platelet counts; 50 x109 /L, 150 x109 /L, 275 x109 /L, 400 x109 /L and 500 x109 /L, with age kept constant (age=40).

Results: No difference was observed between the recommended method of the International Federation for Clinical Chemistry (IFCC) and non-IFCC methods, both performed better than the claims made by the manufacturer. The difference between the reference interval in routine use (19-42 U/L) versus the recommended reference interval (5-34 U/L) showcased an extra 7.6% of patients that would be found above the upper limit of normal (34-42 U/L). Method combinations used for investigations on bias proved that those containing one or more IFCC method results produced a positive bias while non-IFCC method combinations produced a negative bias. The results of measurement using IFCC methods were consistently higher than non-IFCC methods.

Conclusions: Non-IFCC methods produce falsely low concentration values for liver aminotransferases. A larger prospective study is required to validate this among the patients referred to liver clinics.

Verification of Capillary plasma for commonly requested Biochemistry tests on the Siemens Atellica chemistry analyser

Names: Blessing Makwite1,2, Jennifer Brady1,3

1 Department of Clinical Biochemistry, Children’s Health Ireland at Crumlin, Dublin, IRELAND

2 School of Biomedical Sciences, University of Ulster, Coleraine, Northern Ireland

3 School of Medicine, University College Dublin, IRELAND

Abstract

Venepuncture can be painful, especially in children, and may be difficult to perform. It also carries risks such as needlestick injuries, nerve damage, scarring, haematomas, and iatrogenic anaemia from repeated attempts. In contrast, capillary blood collection is less painful, with modern lancets designed to reduce both pain and puncture depth.

In this study, 26 paired venous and capillary plasma samples were collected from sedated paediatric patients and analysed for biochemical parameters, including sodium, potassium, chloride, magnesium, calcium, creatinine, urea, bilirubin, total protein, albumin, AST, ALT, ALP, and C-reactive protein, to assess the interchangeability of these two matrices. The feasibility of direct sampling from the Sarstedt microtainer tube on the Siemens Atellica analyser was also examined.

The results showed excellent correlation (R² > 0.9) between venous and capillary samples for ALP, ALT, AST, creatinine, CRP, phosphate, magnesium, bilirubin, total protein, and urea, while albumin, calcium, chloride, direct bilirubin, and sodium showed good correlation (R² = 0.6–0.9). However, potassium and the haemolysis index (H index) had poor correlation (R² < 0.2). Haemolysis affected 59% of capillary samples (H index >2), preventing accurate reporting of potassium and AST levels due to interference. Linear regression confirmed capillary samples were more susceptible to haemolysis, significantly impacting potassium levels (P < 0.01 vs. 0.411 for venous plasma).

The study concluded that capillary plasma can be used interchangeably with venous plasma for most tests except potassium, which is affected by haemolysis. Additionally, the Atellica analyser can process 250 µL of plasma directly from a microtainer tube, with lower volumes flagged for sample integrity errors.

Reducing unnecessary testing of Vitamin D; A quality improvement initiative

M O’Meara, K Heverin, J Reeve, D Griffin, V Gounden

Department of Clinical Biochemistry, Galway University Hospital

Background

Vitamin D requesting is increasing exponentially. Data suggest approx. 40% of Europeans are Vitamin D deficient (<75 nmol/L), with 13% classified as severely deficient (<30 nmol/L). Vitamin D testing is recommended in specific patient groups only. The Department of Clinical Biochemistry, Galway University Hospital, introduced a quality improvement initiative in April 2024 to optimise testing for Vitamin D in both primary and secondary care.

Aims

Primary Aim:

To determine the effect of the initiative in reducing inappropriate (as per guidelines) vitamin D requests.

Secondary Aim:

To audit application of the protocol by lab staff (number of duplicates processed/correct clinical criteria applied)

Methods

Patient data, relating to Vitamin D test numbers pre- and post-policy introduction, was extracted from the Laboratory Information System using Cognos Impromptu software; April-June 2023 versus April-June 2024. Due to high duplicate requests and staff re-education on acceptance criteria, the quality process was re-reviewed and data examined for July – August 2024.

Results

Approximately 3600 samples were processed for Vitamin D April – June 2023 prior to the introduction of Vitamin D testing policy. Of these 15% were severely deficient (<30 nmol/L). Following introduction of Vitamin D requesting policy, 2800 samples were processed from April – June 2024, yielding overall severe deficiency rate of 10.5%. Of these 1973 (70%) had results >50 nmol/L, with more than one request processed in 6% of patients. The process was re-examined from July – August 2024, where 1369 samples had been processed, with severe deficiency rate of 9.6%. 74% of patients were >70 nmol/L. A significant number of patients (92, 6.7%) had >1 Vitamin D processed within this period.

Discussion

We have outlined a quality improvement policy relating to vitamin D testing for defined clinical indications which has resulted in a reduction in inappropriate requests. The majority of patients have adequate vitamin D levels.

Investigation of the impact of Icteric Interference in the Abbott Triglyceride 2 Allinity c Assay

1Rajesh, P., 2Byrne, B. & 2Lee, GR

1University College Dublin and 2Department of Clinical Biochemistry & Diagnostic Endocrinology, Mater Misericordiae University Hospital

Introduction: In 2021 Abbott Diagnostics released a new version of their Triglyceride (Tg) assay. Among the changes noted in the Instructions for Use (IFU) was a reduction in the Icteric indices threshold from 128 to 36. A forecasted increase in suppressed Tg results prompted a novel approach involving automated sample dilution of such icteric samples in attempt to reduce the Icteric index below 36 and permit Tg analysis.

Aim: The aim of this study was to determine the magnitude of any interference caused to the Abbott Diagnostics Triglyceride2 assay for samples with an Icteric index >36.

Methods: 353 plasma samples (lithium heparin) with an Icteric index > 36 were analysed using the TG2 assay prior to, and following dilution (1:4). Pre- and post-diluted Tg results were compared using Passing-Bablok regression analysis and against the critical difference (CD based on TG2 analytical variation (2.8 x 2.2 = 6.2%).

Results: Tg levels ranged from 0.25 to 5.01 mmol/L (median = 0.87 mmol/L). The icterus level ranged from 36 to 173 (median 40). There was no systematic, or proportional, differences identified between diluted and undiluted TG values (y = 0.02 + 1.00).

<5% of samples (13 of 353 samples) showed a difference of >10% and >0.1 mmol/L. However on further examination, these differences were not considered to be clinically significant. Therefore, there was no appreciable interference to the TG2 assay for icterus levels up to 95.

Conclusions: Our study supports using a higher icteric threshold, which may diminish need for sample dilution. Efforts to mitigate unnecessary suppression of test results are ongoing.

Methemoglobinemia: A Case of Discordant Pulse Oximetry and CO-oximetry Results in a Young Female Patient

O’Connor N1, Redahan L2,3, Lee GR.1,3

1Department of Clinical and Diagnostic Biochemistry, 2Nephrology Department, Mater Misericordiae University Hospital, 2University College Dublin

Background:

Pulse oximetry is a widely used, non-invasive method for monitoring patient oxygen saturation (SpO2). However, discrepancies in SpO2 analysis between pulse oxymetry and arterial blood gas (ABG) analyses can present significant diagnostic challenges.

Presentation:

During a nephrology clinical review, an unexpected low SpO2 (88-92%, reference range [RR: 95-99%) was obtained by pulse oximetry in a young female patient (18 years). ABG analysis did not corroborate this finding (P02: 12.5 [RR: 11-14.5 kPa]). Shortness of breath was denied and she was not cyanotic. Following laboratory consultation with the nephrologist, potential explanation for this discordance was considered. This prompted repeat ABG and co-oximetry. Accordingly, parameters of oxygen status remained normal but an elevated methaemoglobin (9.5%, RR: <1.5%) was reported. Methaemoglobinaemia secondary to recent dapsone treatment was considered a likely cause for the discrepancy, whereby MetHb can negatively interfere with pulse oximetry measurement.

Follow up:

Dapsone was promptly discontinued with a concomitant reduction in metHb of almost 2 fold (4.5%, RR: <1.5%) within just two days.

Conclusion:

Methemoglobinemia, although rare, should be considered in patients with low SpO2 by pulse oximetry but normal parameters of oxygen status by ABG analysis. This case further exemplifies the value of clinical liaison between the laboratory and clinical users to ensure appropriate management, particularly when results are clinically incongruent.

To Err is Good?

Byrne E1, Lee GR1

1Department of Clinical Biochemistry & Diagnostic Endocrinology, Mater Misericordiae University Hospital

Background

Paraprotein interference with clinical biochemistry assays is a well know phenomenon but it is often difficult to identify, particularly when there is no clinical suspicion. We report a case of a paraprotein interference resulting in a falsely decreased gamma-glutamyl transferase-2 (GGT-2) result in the Abbott Alinity-c assay.

Case

Routine bloods were received from a GP for a 79-year-old female patient with “review” stated as clinical details. Her bone profile and liver function tests appeared normal, but the GGT result flagged as <2 U/L raising suspicion by the Duty Biochemist. An abnormal pattern in the assay’s reaction curve was observed. Further testing showed an elevated creatinine (178 mmol/L, 46-86), total protein 97 g/L (65-83) and calculated globulin 61 g/L (26-39). She was also anaemic. These findings raised suspicion of a possible assay interference caused by a paraprotein. On discussion with the GP it emerged the patient was asymptomatic on presentation but urine dipstick analysis demonstrated haematuria and proteinuria prompting a nephrology referral. The Duty Biochemist suggested requesting Serum Protein Electrophoresis, Immunoglobulins and Serum Free Light Chains tests. which the GP requested promptly. Accordingly, results revealed an elevation in kappa free light chain, 10393.8 mg/L (3.3 – 19.4), with an abnormal kappa-lambda ratio, 124 (0.26 - 1.65) and the presence of a paraprotein (IgG-kappa monoclonal paraprotein)

Conclusion

This case has prompted implementation of a protocol involving a LIS rule that holds such suppressed GGT results for manual authorisation. This allows for inspection of analytical reaction curves and clinical reviewing to help identify assay interferences, in particular occult pathology caused by paraproteins. This case further exemplifies the added value of the duty biochemist services in promoting appropriate patient care and management.

Coming Out of the Dark: A Comprehensive Study of Vitamin A and E Stability

Conor Vaughan 1,2, David Green 1, Alison Bransfield 1, Karen Galvin 3, Janne Cadamurro 4, Seán J. Costelloe 1,2,5

1. Department of Clinical Biochemistry, Cork University Hospital, Cork, Republic of Ireland.

2. School of Medicine, University College Dublin, Dublin, Republic of Ireland.

3. School of Food and Nutritional Sciences, University College Cork (UCC), Cork.

4. Department of Laboratory Medicine, Paracelsus Medical University, Salzburg, Austria.

5. Department of Medicine, UCC.

Introduction. The preanalytical phase is the most significant source of error in laboratory medicine. High-quality studies of the effect of preanalytical variables are required, particularly where inadequate or conflicting published evidence exists.

Aim. Determine the effect of light exposure on the stability of fat-soluble vitamins A (VitA) and E (VitE) in several blood matrices over time.

Methods. EDTA plasma (EDTA-P), Lithium Heparin plasma (LH-P), and clot-activated serum gel tubes (Greiner Bio-one Vacuettes)), were taken from ten healthy adult volunteers (CUH staff). The effect of fluorescent light exposure (∼1000 lux) compared with light protection over 48 hours at 20-23OC was considered for VitA and VitE. Full ethical approval and informed written participant consent were obtained. Aliquots for each timepoint were stored and transported, light-protected, at -20OC. Vitamins were quantified by high-performance liquid chromatography. Reference change values incorporating analytical CV and intraindividual biological variabilities were 32.3% and 34.0% for VitA and VitE, respectively.

Results. VitA concentrations in EDTA-P and LH-P had a median change (∆M) <7% at 12 hours (hr), irrespective of light conditions. In these same matrices, ∆M for VitE up to 24 hr was <8%. In serum, the stability of VitA and VitE declined over time, and ∆M was <15% at 48 hr regardless of light exposure. Kruskall-Wallis analysis indicated no significant difference in stability between light conditions for either vitamin in any matrix examined.

Conclusion. VitA and VitE remain stable for up to 12 hr in all blood matrices examined, irrespective of light exposure, suggesting stringent light control in the preanalytical phase may not be necessary for these analytes. This finding, if reproducible, suggests fewer specimens might be rejected for VitA and VitE based on light stability, reducing the cost to the health system for repeat testing and potentially allowing for phlebotomy for these analytes in the community.

GPS for GP K

Alison Bransfield 1, Rory Howlin 1,2, Conor Vaughan 1,2, Seán J. Costelloe 1,2,3

1. Department of Clinical Biochemistry, Cork University Hospital, Cork.

2. School of Medicine, University College Dublin, Dublin.

3. Department of Medicine, University College Cork, Cork.

Introduction. The preanalytical phase is the area of laboratory medicine with the highest error rate. Laboratories accredited to ISO 15189:2022 must monitor the conditions of specimen transport periodically. However, the transport of diagnostic specimens from General Practice (GP) to Cork University Hospital (CUH) is not under laboratory control.

Aims. Consider the percentage of potassium (K+) requests fulfilled for each GP relative to their distance from CUH over one year.

Methods. All K+ results from GPs in 2023 were considered, and the distance from CUH was calculated from the Eircode for each GP using the ggmap package in R Studio. Only GP surgeries that sent >50 samples were included in the analysis. The percentage of K+ requests where numerical results were reported was calculated for each surgery and classified as high (>80%), medium (40-79%), or low (<40%) and coloured on a dot density geospatial map as green, blue and red, respectively, with dot size proportional to the number of requests.

Results. The data show no correlation between the distance from CUH and the percentage of K+ fulfilled. GPs located in Cork City, within 10 km of CUH, had fewer than 40% of their K+ results reported, while other GPs, as far as 120 km, had more than 80% of K+ requests successfully processed.

Conclusions. Delayed transport from GPs is a long-standing issue at CUH. This geospatial analysis shows that the assumption that proximity to CUH means timely processing of K+ (a standard proxy for poor preanalytical quality) is false. Sharing these data with GPs may facilitate better design of specimen transport services, ensuring their patients’ samples are received and processed as quickly as possible.

The Prolactin Puzzle: Macroprolactin and PEG Effects in Common Immunoassays

Rory Howlin 1,2, Emer Groarke 1, Graham Lee 3,4, Karen Heveren 5, Seán J. Costelloe 1,2,6

1. Department of Clinical Biochemistry, Cork University Hospital, Cork.

2. School of Medicine, University College Dublin, Dublin.

3. Department of Clinical Biochemistry and Diagnostic Endocrinology, Mater Misericordiae University Hospital, Dublin.

4. Department of Clinical Biochemistry, Midlands Hospital, Mullingar.

5. Department of Clinical Biochemistry, Galway University Hospitals, Galway.

6. Department of Medicine, University College Cork, Cork.

Introduction. Macroprolactinaemia is seen in conditions including drug-induced hypersecretion and prolactinoma. Commercial prolactin (PRL) immunoassays (PRLIAs) are differentially susceptible to interference from macro-complexes of PRL (macroprolactin (MPRL)), risking erroneous hyperprolactinaemia diagnoses. Commonly, identifying MPRL involves polyethylene glycol (PEG) precipitation, though PEG anecdotally interferes with specific PRLIAs.

Aims. Observe the interference from MPRL and PEG on four commercial PRLIAs.

Methods. Anonymised residual serum from adults with PRL in the 80-300 mlU/L range was gathered for base pools, which, along with macroprolactinaemic serum, 25% PEG solution, and phosphate-buffered saline, were used to construct specimens in nine concentration combinations such that MPRL and PEG interferences could be examined both independently and in concert. Following centrifugation, aliquoting, and storage (-20 °C), specimens were analysed on Beckman Coulter DxI9000, Abbott Architect, Roche Cobas, and Siemens Attelica PRLIAs. Relative Macroprolactin Estimates (RMEs) judged the percentage of spiked MRPL measured as PRL.

Results. When spiked with 0, 100, or 300 mlU/L MPRL alone, RMEs, in decreasing order, were: Abbott (0, 118, 117%); Roche (0, 78, 80%); Beckman (0, 59, 58%); and Siemens (0, 55, 54%). When spiked with PEG, RMEs in the presence of MRPL remained high (e.g. 60% (MRPL=300 mIU/L, 12.5% PEG) on Beckman, compared with <20% on all other platforms).

Conclusions. Interference in PRLIAs from MPRL and PEG differs significantly and consistently across platforms, with the Abbott assay most susceptible to MPRL interference. The new-to-market DxI9000 is subject to MPRL interference. However, it is unclear how MPRL might be managed in laboratories using this PRLIA, given that this study confirms a positive interference from PEG in the assay, as previously suspected at CUH. In the interests of patient safety, this issue requires careful operational, scientific, and clinical consideration.

Iatrogenic Interference in Macroprolactin Testing on the Beckman Coulter DxI9000

Rory Howlin 1,2, Emer Groarke 1, Alison Bransfield 1, Seán J. Costelloe 1,2,3 1. Department of Clinical Biochemistry, Cork University Hospital, Cork. 2. School of Medicine, University College Dublin, Dublin. 3. Department of Medicine, University College Cork, Cork.

Introduction. Significant prolactin (PRL) elevations may indicate the presence of a prolactinoma. It is essential, therefore, to exclude erroneous hyperprolactinaemia due to macro-complexes of PRL (macroprolactin (MPRL)). Clear guidance for MPRL detection on the new-to-market Beckman Coulter DxI9000 is lacking. By anecdote and observation, polyethylene glycol (PEG), routinely used to identify MPRL in immunoassays, causes positive interference in this assay. The manufacturer has, therefore, suggested that a post- PEG reference interval (RI) for PRL might help identify those patients with MPRL.

Aims. Develop a post-PEG RI for PRL on the DxI9000 platform.

Methods. Anonymised residual patient samples were chosen to develop the RI where the following criteria were met: patients were not pregnant, had no pituitary or thyroid testing in the preceding year, and were 18-75 years old. PEG precipitation was performed per local procedures involving 12.5% PEG for all included specimens. The bi-weight quartile methodology of Horne and Pesce was used to derive the RI for the post-PEG PRL using Analyse-IT statistical software. Outliers were identified and excluded in a two-step process using the Tukey and the Dixon and Reed methods. The need for partitioned by gender was assessed using Harris and Boyd’s test.

Results. One hundred and ten specimens were included in the analysis, and four were excluded as outliers. The post-PEG RI was 111.5-745.5 mlU/L, and partitioning by gender was not indicated.

Conclusions. The post-PEG RI for PRL on the Beckman DxI9000 is higher and much wider than the post-PEG range RIs quoted from CUH’s protocol (51-187 and 64-269 mlU/L for males and females respectively). This supports observations at CUH where the routine procedures for PEG precipitation have consistently and universally given overrecovery, making MPRL interpretation all but impossible.