Supplemental Data Table 4: Summary of included studies on PTH stability in plasma and serum measured by second generation assays. Studies in which anticoagulated plasma and serum were compared directly against each other are highlighted in blue. * Clarification of published data were sought directly from corresponding author; ** Unless indicated otherwise, values are the range of concentrations observed on EDTA plasma reference samples; *** Table of results unclear on the specific time point at which PTH is considered significantly different compared to baseline in serum. BD: Becton Dickinson; CKD: chronic kidney disease; ECLIA: electrochemiluminescence immunoassay; EDTA: ethylenediaminetetraacetic acid; GST: gel separator tube; ICMA: immunochemiluminometric assay; IRMA: immunoradiometric assay; PTH: parathyroid hormone; PST: plasma separator tube; RT: room temperature; ULN: upper limit of normal. Reference Anderson et al, 2001, 32 Number/nature/ concentration range** of samples Sample type studied/tube manufacturer n=9, CKD patients, predialysis patients Serum (plain, GST) (Sarsted and plain glass tubes) Storage of samples Comparator Blood centrifuged after venepuncture and separated plain serum stored at -20oC before analysis Test GST serum separated after venepuncture and stored at 20oC before analysis Assay type (analyser, manufacturer) ICMA (Immulite, Siemens) Overall conclusion Statistical analysis ANOVA with Tukey-Kramer post-test comparison PTH concentration higher (13%) in plain than in GST serum, suggesting stability over 20 min at RT better in plain serum. Brinc et al, 2012, 44 Children of 0-18 years of age Serum (GST) Blood allowed to clot for 20 min, centrifuged and stored at 4oC until baseline measurement (no baseline measurement performed for “all ages” and “1-30 days” pools) Serum stored at 80oC until analysis. Sample analysed once monthly over 1013 month period ICMA (Immulite, Siemens) Stability of analytes at monthly intervals after storage was assessed by calculating the percentage change from baseline (tbaseline) as follows: ((tx−tbaseline)/tbas eline)*100; x=storage time in months. Linear regression of analyte concentration vs. time was also performed and examined for significant slope PTH unstable at -80 oC; up to 16% decrease after 2 months and 27.2% after 10 months. Serum (GST), plasma (EDTA) (BD) Blood centrifuged and serum/plasma assayed immediately Serum and plasma stored at -20oC for 24 and 120 h ECLIA (Elecsys, Roche) Applied the criterion of the Royal Australasian College of Pathologists Quality Assurance Program (<25% difference between the sample and the target) GST and EDTA values equivalent at baseline; PTH stable in serum and plasma for 24 h at -20oC but after 120 h greater loss in EDTA than serum (10.8% vs -7.2%) occurs Sample pooled into a single pool or into age-group specific pool (3-30 days, 1 month – 1 year, 15 years, 6-10 years, 11-14 years and 15-18 years) Cavalier et al, 2007, 16 n=16, hemodialysis patients, predialysis samples, 0800h, 298.5 pmol/L Evans et al, 2001, 33 minimum of n=6, non-fasting, 08.0010.00 h, 9-6 pmol/L Serum (plain), plasma (EDTA, lithium heparin, sodium fluoride/potassium oxalate) (Greiner; BD) EDTA blood centrifuged immediately and plasma stored at -20oC Separation immediately (plasma) or 10 min (serum) after collection. Storage of serum/plasma aliquots at 4oC for 24 h, 4oC for 120 h, 30oC for 24 h or 30oC for 120 h. IRMA (Nichols Institute) A difference of 10% was selected because changes of less than this were deemed unlikely to be clinically significant At 30oC PTH stable for 3 h (serum), 16 h (fluoride plasma), 33 h (heparin plasma) and 36 h (EDTA plasma). At 4oC PTH stable for 55 h (serum) and 5 days (heparin, fluoride and EDTA). Overall conclusion: PTH more stable in EDTA plasma than serum at both 30oC and 4oC Glendennin g et al, 2002, 34 n=36, patients, 0.927 pmol/L Plasma (EDTA ), serum (GST) (BD) EDTA plasma assayed within 3 h of collection Serum assayed within 3 h of collection and serum and EDTA plasma assayed after 72 h at RT ICMA (Immulite, Siemens) ANOVA and Dunnett’s post hoc test PTH concentration 35% lower in serum than plasma at 3 h and 60% lower at 72 h; PTH concentration EDTA plasma did not change significantly over 72 h. Overall conclusion: PTH more stable in EDTA plasma than serum Hermsen et al, 2002, 35 n=15, healthy individuals and CKD patients Serum, plasma (EDTA, heparin) Aliquots of serum and plasma frozen at -20oC at baseline Aliquots of serum and plasma stored at RT or 4oC for varying lengths of time up to 168 h, then frozen at -20oC ECLIA (Elecsys, Roche) Stability was defined as <10% difference between the initial PTH concentration and the measured PTH concentration PTH stable for at least 8 h (serum), 24 h (heparin plasma), 48 h (EDTA plasma) at RT; stable for at least 24 h (serum, heparin plasma) and 72 h (EDTA plasma) at 4oC. Overall conclusion: PTH more stable in EDTA plasma than serum Holmes et al, 2005, 36 n=31, CKD patients, predialysis samples, 08.0009.30 h, 29.1-57.8 pmol/L (95% range) Serum (GST), Plasma (EDTA) (BD) 2 tubes (serum, plasma) collected on ice and centrifuged immediately at 4oC, 2 tubes (serum, plasma) collected at RT and centrifuged at RT, 1 tube (serum) collected at RT and centrifuged at 4oC (manufacturer indications); Analysed within 3 h Analysis at 24 h, 48 h and 72 h for each of the five conditions/ sample types ICMA (Immulite, Siemens) Joly et al, 2008, 37 n=34, hemodialysis patients, predialysis samples, 1.6-136.1 pmol/L serum (plain), plasma (EDTA, citrate) (BD) All samples separated and frozen at -20oC in aliquots within 2.5-4 h of venepuncture. After storage, aliquots thawed and assayed immediately. Same aliquots reassayed after 18 h at RT. For each sample type and each assay, results were compared against their equivalent that had been assayed immediately after thawing. ECLIA (Elecsys, Roche), ICMA (ADVIA Centaur, Siemens), ICMA (Liaison, DiaSorin), ICMA (Immulite, Siemens), ICMA (Architect, Abbott), ICMA (Access, Beckman Coulter) ANOVA with Bonferroni corrections for multiple comparisons Not stated Higher PTH concentration in plasma at RT compared to 4oC (9.8% higher at RT at baseline); PTH stable over time in EDTA at 4oC (at least for 72 h), increases slightly in EDTA at RT (8.3% increase at 24 h and 8.8% at 48 h): concentration at 72 h not different basline); PTH stable over time in serum sample kept at 4oC (at least for 72 h), stable in serum at RT for 24 h (significant decrease by -17.3% at 48 h and 24.9% at 72 h from baseline) Across methods, mean change of PTH after 18 h +3.3% in thawed EDTA plasma, -9.5% in serum, +2.8% in citrate plasma . Concentrations in citrate plasma lower than EDTA plasma or serum. Concentrations in serum higher than EDTA plasma with some assays (Elecsys, Liaison, Architect, Beckman) and lower with others (Immulite, Centaur); PTH more stable in EDTA plasma after thawing than in serum. Levin GE et al, 1994***, 38 n=9, patients with primary hyperparathyroidis m (8 with CRF), n=8, healthy volunteers, n=3 patients with acute pancreatitis (mean concentration 13.5 pmol/L) Serum (plain), plasma (protease inhibitors, EDTA, lithium heparin) (Nichols Institute) Samples separated after collection (refrigerated centrifugation) and immediately frozen at -20oC Separated serum and plasma kept at RT for 6, 24, 48 or 72 h and then frozen at 20oC IRMA (Nichols Institute) Wilcoxon signedrank test PTH concentration decreased significantly in serum upon storage at RT (26% loss after 24 h) but not in any of the plasmas tested. PTH concentration higher in serum than plasma at 0 h. Morales Garcia et al, 2009, 19 n=49, CKD patients (249 blood samples), 08.0009.00 h, fasting, predialysis samples, 21.5 ± 21.2 pmol/L Plasma (EDTA) Blood separated after collection (refrigerated centrifugation) and plasma immediately frozen IRMA (total intact assay, Scantibodies) Paired t-test including the Bonferroni correction PTH stable in all storage conditions tested; PTH stable for 24 h in EDTA plasma at 4oC Newman et al, 1988, 20 n=4 normal volunteers, n=4 hyperparathyroid patients Serum 1 aliquot clotted for 30 min before centrifugation, then serum frozen (-20oC) Sample left 1 h at RT after collection, then separation and aliquoting. 1 tube frozen, 1 tube kept 8 h at 4oC then frozen, 1 tube kept > 24 h at 4oC then frozen; Sample left > 3 h at RT after collection, then separation and aliquoting; 1 tube frozen, 1 tube kept 8 h at 4oC then frozen Serum kept at 4oC or 23oC for 18 h before freezing IRMA (Nichols Institute) Not stated 10% loss in serum incubated at 4oC. Greater increases in serum stored at 23oC (14%). Overall conclusion: PTH unstable in serum Oddoze et al, 2012, 29 n=50 healthy volunteers Serum (plain and GST), plasma (EDTA) Sample centrifuged 30 min after blood clotting (serum) and immediately after collection (plasma) Samples were stored at 4oC or RT for 2 h, 4 h, 6 h, 24 h or 72 h. ICMA (Cobas, Roche) Combined analytical and intra-individual imprecisions by the estimation of the square root of the sum of the squared analytic and biologic imprecisions, defined as the total change limit (TCL). If the results for an analyte had a mean percentage difference which exceeded the TCL, then the difference was judged to be significant and not to meet the stability criteria. At RT PTH stable for 6 h in serum (plain and GST), and 72 h in EDTA plasma; At 4oC PTH stable for 72 h in EDTA plasma and serum (plain and GST) Parent et al, 2008, 31 Parent et al, 200, 22 (1) n=53, hemodialysis patients (1.6-89.4 pmol/L), n=46, renal calculi patients (1.5-11.1 pmol/L); (2) n=17 (2.7-93.4 pmol/L) + 21 (1.2-93.4 pmol/L) hemodialysis patients; predialysis samples n=31, hemodialysis patients, predialysis samples 6.6-129.4 pmol/L (1) Serum (plain), plasma (EDTA); (2) Serum (plain, GST) (BD) (1) plain serum separated at 4oC within 30 min and stored at -20oC; (2) plain serum separated within 30 min and stored at -20oC (1) EDTA plasma separated at 4oC within 30 min and stored at -20oC; (2) Plain serum and spun GST tubes left for 4 h at RT or 4oC before separation and freezing at 20oC ECLIA (Elecsys, Roche) Student t test (1) PTH concentration in plasma and serum equivalent; (2) Lower PTH concentration in plain serum kept at RT; No effect of temperature on PTH concentration in GST serum; PTH stable in plain serum for 4 h at 4oC but not at RT; stable for 4 h in GST serum at both 4oC and RT. Serum (plain, GST), plasma (EDTA, EDTA + antiprotease = aprotinin) (BD) Plain serum, separated and frozen at -20oC immediately GST serum spun and frozen immediately or kept at RT or 4oC for 18 h before freezing; EDTA and aprotinin plasma either centrifuged immediately and separated and frozen, or left as whole blood at RT or 4oC for 18 h and then separated and frozen ECLIA (Elecsys, Roche), ICMA (Immulite, Siemens), ICMA (Liaison, DiaSorin) Wilcoxon non parametric test No effect of assay used on PTH concentration (plain serum); 30-35% bias between Elecsys and Immulite with EDTA and aprotinin tubes; PTH stable in samples separated immediately for all tube types (Elecsys), lower PTH concentration with EDTA or aprotinin tubes (Liaison), higher PTH concentration with EDTA or aprotinin tubes (Immulite); At 4oC, PTH stable in serum for 18 h; at RT significant losses in serum after 18 h. Ratcliffe et al, 1989, 23 n=6, primary hyper parathyroidism patients, 11.1 pmol/L (mean) Polypropylene tube (whole blood, serum) (Sarstedt) Serum frozen immediately Whole blood or serum kept at 20oC for up to 30 h before storage at -20oC and analysis IRMA (N-tact, INCSTAR) Compared individual means by Student’s t-test, adjusted according to the Bonferroni method for multiple comparisons PTH stable in serum kept at 20oC for up to 6 h, after 24 h 64% of PTH concentration remains. Russell et al, 1995, 24 n=18, hyperparathyroidism or renal failure patients, 1.9-23.7 pmol/L Serum (GST) Serum allowed to clot for 2 h and then centrifuged, separated and stored frozen at 20oC. Two tubes centrifuged after 2 h at RT and then stored at RT or 4oC, and two tubes left uncentrifuged at RT or 4oC. Serum samples removed before analysis. All samples divided into 2 groups: samples analysed within 2-6 h or within 24-30 h ICMA (Magic Lite, Ciba Corning Diagnostics) Paired t-test Serum stored separated stable for 6 h at both RT and 4oC. Significant losses seen in samples after 24-30 h at RT but not at 4oC. Overall conclusion: PTH stable in serum for at least 6 h after collection and for longer at 4oC. Scharnhors t et al, 2004, 39 n=137, volunteers, 09.00-13.00 h, 2.08.3 pmol/L (95% reference interval) Serum (GST), plasma (EDTA) (BD) 1 EDTA tube kept on ice (centrifuged at 4oC), 1 EDTA tube and 1 GST kept at RT; centrifugation, separation and analysis within 2.5 h after collection Serum and plasma kept at 4oC, analysis 48 h after initial measurement ICMA (Immulite, Siemens) Wilcoxon signedranks test PTH more stable in plasma than serum; PTH concentration decreases over 48 h in serum (2.6 to 2.4 pmol/L), increases over 48 h in plasma (3.8 to 4.1 pmol/L), stable in EDTA plasma kept on ice (3.2 pmol/L) Stokes et al, 2011*, 25 n=18, metabolic bone disease patients (divided into 3 groups of 6) Serum, plasma (lithium heparin, EDTA) (Sarstedt) Immediate separation, storage of serum/plasma at -70oC Immediate separation and storage at RT for 2, 4, 8, 24, 48 h or 7 days; Immediate separation and storage at 2-8oC for 24, 48 h, 7 days, 14 days, 28 days; Samples then stored at 70oC until analysis ECLIA (Elecsys, Roche) 1-way ANOVA with Bonferronicorrected post hoc t-tests to compare analyte concentrations at each time point with the baseline concentration. PTH stability was defined by the amount of time a change of <10% occurred Teal et al, 2003, 40 n=13, hemodialysis patients, predialysis samples (2.1-115.8 pmol/L) Serum (plain, GST), plasma (EDTA) (BD) Aliquots of serum and plasma centrifuged, separated and frozen rapidly at 20oC after venepuncture Aliquots of serum and plasma kept at RT for 2, 4, 8, 12, 24 and 48 h and then frozen at -20oC ICMA (Nichols Institute) 1-way ANOVA and Wilcoxon’s matched pairs signed ranks test In separated plasma/serum PTH stable for 48 h (lithium heparin, EDTA) or 24 h (serum); In separated plasma/serum at 2-8oC PTH stable 28 days (lithium heparin, EDTA), 7 days (serum); [PTH] not different from baseline at 1, 2, 4 h at RT and 2-8oC (serum, plasma). Overall conclusion: PTH more stable in EDTA plasma than serum. PTH stable in clotted whole blood at RT for only 4 h. Hence measure PTH on plasma (EDTA, lithium heparin), on serum if sample centrifuged immediately and frozen until analysis. PTH concentration at 0 h similar in serum and plasma; median PTH concentration 4% lower in plasma; PTH unstable after 2 h (plain serum), at 4 h (GST), stable in plasma for 48 h. PTH concentration decreases after 48 h by 47% (plain serum), 26% (GST) Wood, 1992, 27 n=6, normal subjects and primary hyper parathyroidism patients, 3.9-56.6 pmol/L serum Whole blood samples centrifuged and serum frozen at 20oC as soon as possible Serum stored at RT for 2 and 4 h, 1, 2 or 4 days before freezing at -20oC Not stated (but can assume second generation) Not stated A delay in freezing of serum for 2 or 4 h caused a loss of PTH concentration of 4% and 8% respectively. Overall conclusion: Clotted blood samples taken for PTH estimation should be allowed to clot for 30 min and then the serum separated and frozen without delay