Advice

Appendix

Appendix

Contents

Data tables

Table 1: Summary of the Bamonti et al. (2010) study

Table 2: Summary of results from the Bamonti et al. (2010) study

Table 3: Summary of the Heil (2012) study

Table 4: Summary of results from the Heil (2012) study

Table 5: Summary of the Lee et al. (2009) study

Table 6: Summary of results from the Lee et al. (2009) study

Table 7: Summary of the Obeid and Herrmann (2007) study

Table 8: Summary of results from the Obeid and Herrmann (2007) study

Table 9: Summary of the Remacha et al. (2014) study

Table 10: Summary of results from the Remacha et al. (2014) study

Table 11: Summary of the Sobsczynska‑Malefora et al. (2014) study

Table 12: Summary of results from the Sobsczynska‑Malefora et al. (2014) study

Table 13: Summary of the Valente et al. (2011) study

Table 14: Summary of results from the Valente et al. (2011) study

Table 15: Summary of the Woo et al. (2010) study

Table 16: Summary of results from the Woo et al. (2010) study

Table 1 Summary of the Bamonti et al. (2010) study

Study component

Description

Objectives/hypotheses

To establish a cut‑off threshold for holoTC for identifying vitamin B12 deficiency using the Active‑B12 assay (AxSYM analyser) and to evaluate the analytical performance of the Active‑B12 assay.

Study design

Cross‑sectional.

Setting

Italy (precise clinical setting unclear).

Inclusion/exclusion criteria

Routine blood specimens with serum total vitamin B12 concentration ≤221 pmol/l.

Primary outcomes

Cut‑off point for holoTC.

Sensitivity, specificity and area under the curve for this cut‑off threshold against a reference standard of total vitamin B12.

Statistical methods

Analytical imprecision.

AUC, sensitivity and specificity of assay using derived cut‑off point of ≤40 pmol/l.

Conclusions

Results confirmed the reliability of the Active‑B12 assay and that it is adequate for routine use in assessing vitamin B12 deficiency in populations with reduced total vitamin B12 values.

Abbreviations: AUC, area under the curve; holoTC, holotranscobalamin.

Table 2 Summary of results from the Bamonti et al. (2010) study

Patients included

  • 250 'routine blood specimens'

  • 107 men (mean age 59 years, SD 18.8)

  • 143 women (mean age 54.2 years, SD 23.1).

No further details reported about sampling or sample characteristics.

Primary outcome results

Mean recovery for spiked specimens 95% (95% CI 90–100%).

Detection limit 0.07 pmol/l.

Cut‑off threshold for holoTC ≤0.40 pmol/l.

Compared against a total vitamin B12 cut‑off threshold of <139 pmol/l.

Sensitivity 0.74 (95%CI 0.62–0.86).

Specificity 0.52 (95%CI 0.38–0.66).

AUC 0.75 (0.63–0.87).

AUC for other markers of vitamin B12 deficiency:

  • folate 0.61 (95%CI 0.47–0.75)

  • homocysteine 0.32 (95%CI 0.19–0.45)

  • creatinine 0.42 (95%CI 0.28–0.56).

Abbreviations:AUC, area under the curve; CI, confidence interval; holoTC, holotranscobalamin; SD, standard deviation.

Table 3 Summary of the Heil et al. (2012) study

Study component

Description

Objectives/hypotheses

To validate the diagnostic accuracy of holoTC as a screening test for metabolic vitamin B12 deficiency. The Abbott AxSYM assay was used for testing holoTC.

Study design

Cross‑sectional.

Setting

Multi‑centre, 5 clinical laboratories in the Netherlands.

Inclusion/exclusion criteria

Samples from patients aged ≥18 years, with normal renal function, when vitamin B12 testing has been clinically requested.

Primary outcomes

AUC for holoTC and vitamin B12 levels.

Reference standard for vitamin B12 deficiency was elevated MMA levels at 3 thresholds.

Statistical methods

ROC curve analysis

Conclusions

HoloTC has better test performance than vitamin B12 and can replace vitamin B12 testing in detecting vitamin B12 deficiency. Suggested cut‑off value 32 pmol/l.

Abbreviations: AUC, area under the curve; holoTC, holotranscobalamin; MMA, methylmalonic acid; ROC, receiver operating characteristics.

Table 4 Summary of results from the Heil et al. (2012) study

Patients included

n=360

Mean age (range) 59 (19–100.5) years.

Prevalence of elevated MMA (>0.45 μmol/l) 13%.

Primary outcome results

AUC results:

MMA threshold of >0.32 μmol/l

  • holoTC 0.70 (95%CI 0.64–0.87)

  • vitamin B12 0.63 (95%CI 0.56–0.70)

  • p=0.01.

MMA threshold of >0.45 μmol/l

  • holoTC 0.78 (95%CI 0.69–0.87)

  • vitamin B12 0.70 (95%CI 0.61–0.79)

  • p=0.06.

MMA threshold of >0.77 μmol/l

  • holoTC 0.92 (95%CI 0.85–0.98)

  • vitamin B12 0.73 (95%CI 0.60–0.87)

  • p=0.01.

Abbreviations:AUC, area under the curve; holoTC, holotranscobalamin; MMA, methlymalonic acid; n, number of patients.

Table 5 Summary of the Lee et al. (2009) study

Study component

Description

Objectives/hypotheses

To compare the diagnostic performance of the Active‑B12 assay (AxSYM analyser) for holoTC with total vitamin B12 levels in patients after gastrectomy.

Study design

Cross‑sectional

Setting

Department of laboratory medicine, South Korea.

Inclusion/exclusion criteria

Included patients who have had a gastrectomy.

Excluded people with estimated glomerular filtration rate <30 ml/min/1.73 m2.

Primary outcomes

Serum holoTC

Serum total vitamin B12

Hcy

MCV

Statistical methods

Total precision

Lower limit of reference range for holoTC

% agreement of classifications

Correlation between outcomes

Conclusions

The authors concluded that serum holoTC is more sensitive than total serum vitamin B12 for detecting vitamin B12 deficiency and is therefore a more effective marker.

Abbreviations: holoTC, holotranscobalamin; Hcy, homocysteine; MCV, mean cell volume.

Table 6 Summary of results from the Lee et al. (2009) study

Patients included

128 patients after gastrectomy (55 female, 73 male), mean age 60.3 (range 29–86) years, mean time since gastrectomy 30.3 (range 1–100) months.

100 healthy people as controls (58 male, 42 female), age range 24–74 years.

Primary outcome results

Coefficients of variation for total precision of Active‑B12 assay:

  • 9.4% for low‑level (mean 23.05 pmol/l) controls

  • 7.9% for high‑level (mean 54.56 pmol/l) controls.

From healthy reference group:

Lower limit of normal 42.48 pmol/l (derived from the lower 95% confidence interval).

In the sample of patients:

Serum holoTC: low in 32 (25%) patients.

Total vitamin B12: low in 10 (7.8%) patients, borderline in 50 (39.1%) patients.

Serum Hcy: high in 40 (30.4%) patients.

MCV: high in 35 (27.3%) patients.

HoloTC was low in all patients with low total vitamin B12 and normal in all patients with normal total vitamin B12.

In the 50 patients with borderline total vitamin B12, 44% were classified as low for holoTC.

Of the 22 patients classified as having low holoTC and borderline total B12, 9 were classified as having high Hcy and 7 were clinically suspected of having vitamin B12 deficiency (clinical criteria were not reported).

Patients with both low holoTC and total vitamin B12 had significantly higher Hcy levels than those with normal values for either total vitamin B12, holoTC or both(p<0.001).

Correlations:

Total vitamin B12 correlation with holoTC: r=0.6591; p<0.001.

HoloTC correlation with Hcy: r=0.4407; p<0.001.

Total vitamin B12 with Hcy: r=3599; p<0.001.

Neither holoTC nor total vitamin B12 was correlated with MCV.

Abbreviations:holoTC, holotranscobalamin; Hcy, homocysteine; MCV, mean cell volume.

Table 7 Summary of the Obeid and Herrmann (2007) study

Study component

Description

Objectives/hypotheses

To test the utility of holoTC in assessing vitamin B12 status.

Study design

Case series.

The authors were not aware of the patients' underlying diseases, medications or vitamin supplements.

Assays:

  • holoTC – a microparticle enzyme immunoassay assay (AxSYM, Abbott)

  • MMA – gas chromatography‑mass spectrometry

  • cobalamin – a chemiluminescence immunoassay (ADVIA Centaur System, Bayer).

Setting

1018 serum samples that were referred to the authors' laboratory in a university hospital, Germany for total cobalamin testing, between January and August 2006.

Inclusion/exclusion criteria

Not specified. The study population consisted of non‑selected patients.

Primary outcomes

Levels of total cobalamin, MMA, and holoTC; correlation between these markers.

Statistical methods

ANOVA and post‑hoc Tamhane tests were applied. Log transformations were applied on each variable before ANOVA test. Spearman test was used to detect correlations between different markers.

Conclusions

Compared to total cobalamin, a better performance of the holoTC assay was seen in detecting elevated concentrations of MMA in patients with normal renal function. Most patients with combined low holoTC and elevated MMA had normal concentrations of total cobalamin. HoloTC can be used as a first‑line parameter in detecting cobalamin deficiency.

Abbreviations: ANOVA, analysis of variance; holoTC, holotranscobalamin; MMA, methylmalonic acid.

Table 8 Summary of results from the Obeid and Herrmann (2007) study

Patients included

n=1018 samples from 1018 patients (500 females and 518 males).

Characteristics ( median [5th–95th percentiles]):

  • age: 58 (24–83) years

  • creatinine: 79.6 (4.2–795.6) µmol/l

  • total cobalamin: 296 (147–858) pmol/l

  • MMA: 249 (111–1290) nmol/l

  • holoTC: 49 (14–142) pmol/l.

Primary outcome results

Patients with higher serum creatinine levels had higher levels of MMA and holoTC compared with those who had lower serum creatinine. There was no difference in total cobalamin according to creatinine levels.

In patients (number unclear) with normal creatinine levels, those identified with MMA >291 nmol/l (the 4th quartile) showed statistically significant differences in holoTC and cobalamin compared with patients with MMA <291 nmol/l.

The association between serum holoTC and serum MMA was not significant in the total group (r=‑0.051; p=0.103). However, this was attributed to renal function, because only patients with increased serum creatinine (number unclear) showed no correlation between serum holoTC and MMA (r=‑0.36; p<0.001).

In patients with MMA ≥300 nmol/l and holoTC ≤35 pmol/l (n=448), total cobalamin was within normal range (median 212; 25th/75th percentiles 171/272 pmol/l).

In patients (number unclear) with normal renal function, most patients with cobalamin deficiency (holoTC ≤35 pmol/l and MMA ≥300 nmol/l) had cobalamin between 165 and 400 pmol/l. Few patients (n=4) had low cobalamin and normal MMA and holoTC (false positive).

ROC analysis in detecting a MMA level of ≥300 nmol/l in patients (number unclear) with normal renal function: the AUC was 71% for holoTC and 60% for vitamin B12, indicating a better diagnostic sensitivity and specificity for holoTC compared with total cobalamin (confidence intervals and statistical significance test were not reported). A sensitivity of 72% could be expected by using a cut‑off of 35 pmol/l for holoTC and 243 pmol/l for total cobalamin.

Abbreviations:holoTC, holotranscobalamin; MMA, methylmalonic acid; ROC, receiver operating characteristic.

Table 9 Summary of the Remacha et al. (2014) study

Study component

Description

Objectives/hypotheses

To evaluate holoTC levels measured using the Abbott AxSYM Active‑B12 assay and conduct a concordance analysis with MMA and Hcy levels in patients with low or borderline levels of serum cobalamin.

Study design

Cross‑sectional

Setting

Unclear. Haematology department, Barcelona.

Inclusion/exclusion criteria

Reference group: healthy individuals without anaemia.

Low levels of vitamin B12 group: cobalamin ≤200 pmol/l.

Folate deficiency: RCF <500 nmol/l and cobalamin >200 pmol/l.

Primary outcomes

Mean (SD) holoTC levels by group.

Kappa index between low holoTC and elevated MMA/ Hcy.

Statistical methods

Kappa index

Conclusions

HoloTC levels are decreased in patients with low vitamin B12 and folate levels.

Concordance of holoTC with MMA and Hcy levels in this group is poor. Around half of samples (n=13) with normal holoTC had elevated MMA or Hcy. This does not support holoTC as the earliest marker of vitamin B12 deficiency.

Abbreviations: holoTC, holotranscobalamin; Hcy, homocysteine; MMA, methlymalonic acid; RCF, red cell folate; SD, standard deviation.

Table 10 Summary of results from the Remacha et al. (2014) study

Patients included

  • 45 healthy individuals without anaemia.

  • 106 patients with low levels of serum vitamin B12 (≤200 pmol/l. In this group clinical characteristics were recorded to determine their relationship with holoTC levels).

  • 27 patients with folate deficiency (RCF <500 nmol/l and vitamin B12 >200 pmol/L).

No further details reported on the people in the study.

Primary outcome results

In patients with low vitamin B12, the mean of serum holoTC was 25.1 pmol/l (range 0.18–109.4).

In 75 (71%) of this group serum, holoTC was below the cut‑off value of 33.5 pmol/l. Of 31 samples with low cobalamin but normal holoTC, MMA or Hcy levels were elevated in 13, indicating likely cobalamin deficiency.

Mean difference of holoTC compared to healthy group:

  • Low vitamin B12 group: 52.6 pmol/l (95%CI 34.3–70.9, p<0.001).

  • Folate deficiency group: 39.6 pmol/l (range 10.8–75.5).

Concordance analysis:

Low vitamin B12 cut‑off of ≤200 pmol/l:

  • MMA with holoTC: concordance 55.6% (Kappa index non‑significant).

  • Hcy with holoTC: concordance 62% (Kappa index 0.245, p=0.006).

Low vitamin B12 cut‑off of ≤150pmol/l:

  • MMA levels with holoTC: concordance not reported (Kappa index 0.104, non‑significant).

  • Hcy with holoTC: concordance 74% (Kappa index 0.215, p=0.08).

Abbreviations:holoTC, holotranscobalamin; Hcy, homocysteine; MMA methlymalonic acid; RCF, red cell folate.

Table 11 Summary of the Sobczynska‑Malefora et al. (2014) study

Study component

Description

Objectives/hypotheses

To evaluate the pattern of MMA results from samples with holoTC in the indeterminate range, as measured using the Abbott Active‑B12 assay (AxSYM analyser).

Study design

Cross‑sectional

Setting

London NHS hospital laboratory.

Inclusion/exclusion criteria

All samples referred for vitamin B12 status evaluation in a 4‑month period.

Primary outcomes

HoloTC (Active‑B12)

MMA

Statistical methods

Spearman's correlations

Conclusions

Indeterminate holoTC (25–50 pmol/l) is a poor predictor of MMA levels according to the thresholds for elevation used in this study.

Abbreviations: holoTC, holotranscobalamin; MMA, methlymalonic acid.

Table 12 Summary of results from the Sobczynska‑Malefora et al. (2014) study

Patients included

4175 samples routinely referred for analysis of vitamin B12 status; median age 56 (range 0–101) years; 55% female.

Primary outcome results

205(5%) samples had low holoTC (<25 pmol/l).

1019 (24%) were in the indeterminate range.

802 had MMA analysis because renal function was normal or unknown.

244 samples had elevated MMA and 534 did not show elevated MMA.

Incidence of elevated MMA according to holoTC level:

holoTC 25–29 pmol/l: 41%

holoTC 30–34 pmol/l: 32%

holoTC 35–39 pmol/l: 33%

holoTC 40–44 pmol/l: 30%

holoTC 45–50 pmol/l: 26%

holoTC was not correlated with MMA levels.

Abbreviations:holoTC, holotranscobalamin; Hcy, homocysteine; MMA methlymalonic acid.

Table 13 Summary of the Valente et al. (2011) study

Study component

Description

Objectives/hypotheses

To investigated the ability of holoTC (measured using the Abbott AxSYM Active‑B12 assay), total Hcy, MMA, serum and erythrocyte folate and other haematological variables to discriminate cobalamin deficiency in an older population.

Study design

Cross‑sectional

Setting

Geriatric outpatient memory clinic. Clinical laboratory.

Inclusion/exclusion criteria

700 consecutive patients attending a geriatric outpatient memory clinic.

120 healthy volunteers from manufacturer employees and medical students.

Primary outcomes

holoTC (measured using the Active‑B12 assay), total Hcy, MMA, serum and erythrocyte folate.

Reference standard: red cell cobalamin <33 pmol/l.

Statistical methods

ROC plots; sensitivity, specificity and proportion of results correctly allocated; PPV; NPV.

Conclusions

HoloTC performed significantly better than all other indicators in this group of older people.

Abbreviations: Hcy, homocysteine; holoTC, holotranscobalamin; MMA, methlymalonic acid; NPV, negative predictive value; PPV, positive predictive value; RCF, red cell folate; ROC, receiver operating characteristic.

Table 14 Summary of results from the Valente et al. (2011) study

Patients included

Outpatients: n=700; mean age 81 (2.5th to 97.5th percentile 69–92) years; 490 (70%) female.

Reference population: n=120; median age 31 (range 18–62) years, 53% female.

Primary outcome results:

HoloTC (pmol/l) (mean and 95% CI)

Vitamin B12 deficient

Vitamin B12 non‑deficient

Analysis

18.2 (15.6–21.2)

51.4 (49–53.9)

p=0.0001

Prevalence

Prevalence:

  • low holoTC: 8.1%

  • low total cobalamin: 8%

  • low red cell cobalamin: 9.6%

  • elevated MMA: 41.7%

  • elevated Hcy: 52.2%.

ROC analysis–AUC a

  • holoTC 0.90

  • serum cobalamin 0.80

  • MMA 0.78

  • total Hcy 0.75.

ROC analysis– diagnostic performance at a single cut‑off a

Threshold 19.6 pmol/l

Threshold 29.9 pmol/l

  • Sensitivity (%)

holoTC

50.7

80.6

Serum cobalamin

13.4

86.6

MMA

17.9

88.1

  • Specificity (%)

holoTC

96.4

84.5

Serum cobalamin

99.1

57.3

MMA

98.7

17.9

  • PPV (%)

holoTC

60

36

Serum cobalamin

60

18

MMA

60

16

  • NPV (%)

holoTC

95

98

Serum cobalamin

92

98

MMA

92

98

Abbreviations: AUC, area under the curve; CI, confidence interval; holoTC, holotranscobalamin; Hcy, homocysteine; MMA methlymalonic acid; PPV, positive predictive value; ROC, receiver operating characteristic; NPV, negative predictive value.

a 95% CIs not reported.

Table 15 Summary of the Woo et al. (2010) study

Study component

Description

Objectives/hypotheses

To test the association between levels of holoTC measured using the Abbott AxSYM Active‑B12 assay with serum vitamin B12 levels, perform a precision test of the assay and determine the diagnostic value of the assay.

Study design

Cross‑sectional

Setting

Clinical laboratory and a South Korean Hospital.

Inclusion/exclusion criteria

Samples from patients, for whom a vitamin B12 test had been requested because of conditions such as Alzheimer's disease, dementia, Parkinson's disease, cancer, unstable angina and infarction.

Samples from patients with normocytic or macrocytic anaemia, who were admitted to a hospital in South Korea between August 2007 and March 2008.

Primary outcomes

holoTC, cut‑off 35 pmol/l; vitamin B12, cut‑off 150 pmol/l; folate, cut‑off 3 µg/l; Hcy, cut‑off 12 µmol/l.

Statistical methods

Assay imprecision

Chi‑square test for association between holoTC and other markers.

Conclusions

HoloTC was more sensitive than serum vitamin B12 (no formal reference standard applied).

Abbreviations: holoTC, holotranscobalamin; Hcy, homocysteine.

Table 16 Summary of results from the Woo et al. (2010) study

Patients included

n=184 (90 female, 94 male).

45 samples from patients, for whom a serum vitamin B12 test had been requested because of conditions such as dementia with Alzheimer's disease, Parkinson's disease, cancer (including adenocarcinoma of the stomach), unstable angina, and infarction.

139 samples from patients with normocytic or macrocytic anaemia, who were admitted to Dong‑A University Hospital between August 2007 and March 2008.

Primary outcome results

Within‑run and between‑run imprecision values all less than 3.5%.

Low holoTC (<35 pmol/l) was seen in 7 (4%) of samples. Of these, only 1 had both low holoTC and low serum vitamin B12; 4 showed borderline and 2 normal vitamin B12 levels.

In 2 patients with low holoTC, Hcy was elevated (>12.0 nmol/l) indicating vitamin B12 deficiency, while folate levels were normal.

In 10 samples with low folate levels (<3.0 µg/l), holoTC levels were within normal range.

Chi‑square test to determine relationship between low holoTC and vitamin B12 levels: p=0.0001.

Abbreviations:holoTC, holotranscobalamin; Hcy, homocysteine.