Association of elevated vitamin B12 with oncohematological diseases in a cohort of 79,524 patients from Latvia
Summary. Aim: Currently there are some large-scale studies of elevated total vitamin B12 in relation to diseases and their prognosis. Aim of this retrospective study was to determine association of increased B12 as an additional diagnostic marker of oncohematological diseases by a statistical analysis of clinical data of 79,524 patients. Materials and Methods: Overall Latvian population representative data on B12 testing in 79,524 patients were obtained from laboratory database. The following exclusion criteria were applied: fluctuating B12 results within a three-month period, elevated (> 100 U/L) alanine transaminase or aspartate transaminase, hepatitis (HAV, HBV, and HCV) infection, reduced glomerular filtration rate (< 45 mL/min/1.73 m2). As a control group, individuals with normal B12 level and any oncologic diagnosis (solid cancer or hematological malignancies) were selected. Results: After application of step-by-step exclusion filters, 1,373 patients were left with significantly increased level of plasma B12 (> 1,700 pg/mL). Odds ratios for oncohematological diseases in total and myeloid leukemia (including acute, chronic and unspecified) in patient group with elevated B12 were found to be 6.0 (95% CI 4.7–7.6; p < 0.0001) and 19.2 (95% CI 13.1–28.0; p <0.0001), respectively, as compared to the control group. Conclusion: Elevated total B12 could be considered as a potential marker for oncohematological disorders.
Submitted: June 14, 2019.
*Correspondence: E-mail: firstname.lastname@example.org
Abbreviations used: ALT — alanine transaminase; AST — aspartate transaminase; CBC — complete blood count; GFR — glomerular filtration rate; HC — haptocorrin; OR — odds ratio; SE — standard error; SIR — standardized incidence ratio; TCB — transcobalamin.
Clinical laboratory practice often encounters patients with significantly increased plasma levels of total vitamin B12. The etiology of this phenomenon has been understudied. Generally, elevated B12 is associated with liver and kidney damage, as well as with oncohematological processes.
It is not currently clear whether total B12 level can serve as a diagnostic and/or prognostic marker in oncohematology. It is necessary to clarify which onchematological diseases are associated with an increased level of B12, applying exclusion criteria (e.g. liver and kidney damage) in selected patient group for the further research.
Vitamin B12, also referred as cobalamin, is an organometallic water-soluble compound in which a cobalt atom is situated within a corrin ring. Vitamin B12 cannot be synthesized by the human body cells, though it is synthesized by a microbiota of the small intestine. The main external sources of the vitamin B12 are meat, fish, eggs and dairy products, seldom it is found in the products of plant origin. Vitamin B12 is the cofactor for two enzymes, methionine synthase and methylmalonyl-CoA-mutase.
Methionine synthase locates in the cytoplasm and requires vitamin B12 in the form of methylcobalamin to catalyze the conversion of homocysteine to an essential amino acid methionine. Thus, vitamin B12 is important for DNA synthesis, regenerating methionine for protein synthesis and methylation, as well as for the development and initial myelination of the central nervous system and for the maintenance of normal central nervous system function .
In the blood, B12 circulates in two forms — metabolically active and inactive. Transcobalamins (TCBs) are required for B12 transportation in the organism and for the liver uptake. TCB type I (TCBI) and III (TCBIII) ensure binding of 80% of the circulating vitamin B12. These proteins belong to haptocorrin (HC) superfamily .
TCB type II (TCBII) plays essential role in the key processes of tissue and hepatic uptake of vitamin B12. Metabolically active form consists of B12 bound to TCBII — holotranscobalamin II. Metabolically inactive B12 is bound to the HC, in this form B12 is accumulated in the liver. In cases of liver damage, inactive form increases and the level of total B12 also increases proportionally. HCs are produced by the granulocyte cell line, and along with myeloproliferative processes B12 plasma levels are also increased. Metabolically active form holotranscobalamin delivers B12 to the cells thus engaging it into biochemical pathways .
Currently there are multiple large-scale studies of elevated B12 in relation to the diseases and their prognosis in different patient populations [4, 5]. Arendt et al.  analyzed large cohort of Danish population for incidence of cancer in patients with elevated B12. They found that elevated B12 was associated with the risk of subsequently diagnosed cancer. Ryg et al.  in their study used threshold of elevated B12 as 1,626 pg/mL (converted originally in article — 1,200 pmol/L). Standard incidence ratio of developing hematologic malignancies within one-year interval was found to be 24.14 (95% CI 19.51–29.54) . Andres et al.  in their review article discusses the association of significantly high levels of B12 (> 9,000 pg/mL) with oncohematologic pathologies, including myeloproliferative disorders. It is hypothesized that B12 increase in myeloproliferative disorders is due to HCs release by granules of myeoloid cell lineage . Chicke et al.  in their study including 2,064 patients report association of the elevated B12 (> 1,275 pg/mL) with blood malignancies (p < 0.05).
Unrelated to health condition B12 elevation in many cases could be explained as an excess intake of vitamin B12 supplements. Liberation from the internal reservoirs, overproduction of TCBs or disturbances in the process of the clearance also has influence on B12 elevation in blood [7, 9].
To estimate whether elevated B12 level could be considered as a diagnostic and prognostic marker for oncohematological disorders, a large-scale patient data statistical analysis is required. It is also important to identify and exclude unrelated to cancer cases of elevated B12. For example, excessive B12 intake is relatively easy to identify in the anamnesis due to a fluctuation of the plasma B12 levels . Liver diseases increase plasma B12 level during damage of hepatocyte therefore releasing B12 bounded to haptocorins [8, 10, 11]. Significant increase of plasma B12 might be related even to the kidney diseases. Carmel et al.  hypothesized that cellular uptake of cobalamin by the abundant TCBII receptors in the kidney may be involved in the pathophysiology of elevated B12 as well.
Vast majority of the studies on elevated B12 content are done with an application of simplified statistical methods usually by associating one parameter with the disease. The problem needs to be assessed from different aspects, using different algorithms and incorporating as many patients’ data sets as possible in to the statistical analysis.
In this study, we applied multi-criteria statistical algorithm with critically considered exclusion criteria to a 79,524 unique total B12 patient data (79,524 unique patients with at least one total B12 measurement). Our goal was to analyze correlation of elevated B12 with oncohematological diseases in large Latvian population representative data.
MATERIALS AND METHODS
Patients. Data from all adult (≥ 18 years old) patients requesting B12 analysis in the E. Gulbis Laboratory (Riga, Latvia) from January 2004 to January 2018 were included in the study (n = 79,524). All patients’ data was pseudo-depersonalized. During that time laboratory performed more than 1,800,000 unique patient analysis that practically represents overall Latvian population (1,934,379 in 2017, Latvian statistical agency). Overall, each year the laboratory tests 1,200,000–1,300,000 patients.
The following exclusion criteria were defined:
1) fluctuating B12 value within 3 months from the first measurement;
2) elevated (> 100 U/L) alanine transaminase (ALT) or aspartate transaminase (AST);
3) decreased (< 45 mL/min/1.73 m2) glomerular filtration rate (GFR);
4) positive viral hepatitis (HAV, HBV, HCV).
Three months interval was chosen as clinically relevant period to exclude B12 elevation unrelated to cancer (hepatotoxicity, nephrotoxicity) — for biochemical marker normalization.
Threshold of > 771 pg/mL (from Roche Diagnostics B12 test manual) was defined for the elevated B12, and we defined > 1,700 pg/mL for the significantly elevated B12. This high value was chosen to exclude elevated B12 unrelated to health conditions cases since B12 prescription as supplement and medication effects drastically blood B12 level.
For the control group (statistical threshold) we selected patients with normal B12 and any cancer diagnosis (C00–C97). Information about cancer diagnosis was derived from State registry for oncologic patients.
Laboratory methods. All analysis was performed in E. Gulbis Laboratory — a certified (ISO/IEC 17025, ISO 15189:2013) clinical laboratory (Riga, Latvia), which provides laboratory services in all regions of Latvia. All the testing was performed according to the reagent manufacturers’ user manuals.
Total B12 was measured using electrochemiluminescence immunoassay method ECLIA on Cobas e411 analyzer (Roche Diagnostics, Switzerland). Holotranscobalamine II was measured using chemiluminescent immunoassay method on ADVIA Centaur XP system (Siemens Healthcare, Germany). Alanine transaminase and aspartate transaminase were measured on ADVIA 1200 analyzer (Siemens Healthcare, Germany). Folic acid was measured using electrochemiluminescence binding assay on Cobas e801 immunoassay analyzer (Roche Diagnostics, Switzerland). Homocysteine was measured using competitive immunoassay on ADVIA Centaur System (Siemens Healthcare, Germany).
Complete blood cells count (CBC) was detected by multiple commercially available assays with 5-part differentiation. Creatinine was measured using enzymatic reaction on ADVIA 1200 (Siemens Healthcare, Germany). GFR was determined using different formulas depending on patient’s age, body weight and/or height data (protocols are available upon request).
HAV IgM was performed using IgM capture immunoassay on ADVIA Centaur XP system (Siemens Healthcare, Germany). HAV IgG was performed using competitive immunoassay on ADVIA Centaur XP System (Siemens Healthcare, Germany). Anti-HBs and HBsAg was performed using sandwich immunoassay on ADVIA Centaur XP System (Siemens Healthcare, Germany). Anti-HCV was performed using indirect sandwich immunoassay on ADVIA Centaur XP System (Siemens Healthcare, Germany).
Statistical analysis. For B12 association with oncologic diseases odds ratio (OR) was calculated by formula given in Fig. 1. OR, standard error (SE) and confidence interval (95% CI) were calculated according to a practical medical statistics . Test of significance (р-value) was calculated as the area of the normal distribution that falls outside ± z . OR analysis was performed for all oncologic diseases listed in the laboratory database. In this article, we show only statistically significant results.
Initial group included data on 79,524 patients with at least one B12 measurement. These data were analyzed applying step-by-step exclusion algorithm. Patient group selection and analysis is shown in the flow chart (see Fig. 1). At first, the patients with fluctuating B12 analysis were excluded, n = 8,082. The number of patients with stable elevated B12 level is comparable to the number of patients with rapidly fluctuating B12 levels. Hypothetically, reason for such cases could be supplementary intake and B12 medication. For this reason, only the patients with stable elevated B12 values for more than 3-month period were included in the further analysis.
Fig. 1. Flow chart depicting selection of the patients of interest and OR calculation
Further filters excluded patients with elevated ALT or AST within 3 months from B12 measurement, n = 2,263; decreased GFR within 3 months from B12 measurement, n = 1,106; ever positive for any viral hepatitis (HAV, HBV, HCV). Groups of interests for OR calculation were as follows: a) patients with significantly elevated B12 (significantly elevated B12 > 1700 pg/mL) and a cancer diagnosis (n = 162), b) patients with significantly elevated B12 without a cancer diagnosis (n = 698), c) patients with normal (197–771 pg/mL) B12 and a cancer diagnosis (n = 4,655), d) patients with normal B12 without a cancer diagnosis (n = 56,972). Flow chart shows OR calculation for all types of malignant tumors (solid and blood cancer). Patients’ data with other cancer diagnosis are summarized in the Table.
Table. Multifactorial statistical analysis of B12 association with oncohematological disorders
Note: C00–C97 — malignant tumors (solid and blood cancer); C92 — myeloid leukemia; C81–C96 — malignant tumors of the hematopoietic and lymphoid tissues. Diagnosis abbreviations are used according to WHO international classification of diseases (ICD-10). The Table represents total patients’ database divided in three major groups: with significantly elevated B12 > 1,700 pg/mL, with elevated B12 (717–1,700 pg/mL) and with fluctuating B12 results.
Data of 79,524 patients were obtained with at least one B12 measurements during the study period. It was observed that the number of patients performing B12 test was increasing constantly over the study time, reflecting the increasing interest of B12 in the clinical diagnostics (Fig. 2). This increase is representative for the whole country, as the market share of E. Gulbis Laboratory in Latvia largely remained unchanged.
Fig. 2. Montlhy distribution of B12 tests during study
The number of 82,184 differs from 79,524, because it is not «processed», it includes all the results (inconclusive, repeated). Data analysis revealed the big number of patients with rapidly fluctuating B12 values within short period (n = 8,082); the fluctuation period between significantly elevated (> 1,700 pg/mL), normal (197–771 pg/mL) and reduced (< 197 pg/mL) B12 values was less than 3 month. A possible explanation of the rapid fluctuations of B12 level can be related to an irregular intake or therapeutic medication of supplementary B12.
To check the importance of the excluded and other parameter changes, which are related to B12 metabolic pathway and kinetics, we performed different test results being out of normal reference range comparison to the B12 (Fig. 3).
Fig. 3. Frequency of variety clinical tests being out of normal reference range in relation to the B12 values. Y-axis depicts frequency of different clinical analysis results being out of (lower and higher) normal reference values (%), X-axis shows B12 measurement units (pg/mL). Vertical lines indicate reference ranges — lowest normal (norm-low), highest normal (norm-high) and significantly high defined in this study. All clinical tests for each patient were performed in the 3-month interval of B12 measurement. GFR — glomerular filtration rate, FERRIT — ferritin, FOLAT — folic acid, ESR — erythrocyte sedimentation rate, CBC — complete blood count, ABC — absolute count of basophils, AMC — absolute count of monocytes, ALC — absolute count of lymphocytes, AEC — absolute count of eosinophils
Increased frequency of analysis being out of normal reference values (abnormal results) in the clinical testing are observed in hematologic, liver and kidney parameters when B12 is out for reference range (see Fig. 3). This observation shows importance of B12 parameter influencing/or resulting from other essential clinical parameters. Statistical fluctuations are seen due to a small part of patients performed multiple tests.
Abnormalities in hematologic parameters (CBC) show increased frequency when B12 is elevated or reduced in 30–65% of the cases (CBC, see Fig. 3). Folic acid has shown high abnormality frequency in more than 40% of the cases when B12 level is more than 3,000 pg/mL (FOLAT, see Fig. 3). Folic acid and B12 share the same metabolic pathway and abnormality of one of them influences the other.
OR was calculated to analyze whether B12 level could be considered as a potential diagnostic marker for oncologic diseases (see the Table). In cases when B12 is significantly elevated, the diagnosis of myeloid leukemia occurs 19 times more often. Oncohematological disorders occur 6 times more often, oncologic diagnosis — 2.8 times more often when B12 is significantly elevated. Assuming that cancer diagnosis for some reason correlates with B12 > 1,700 pg/ml, oncohematological diagnosis correlation is significantly stronger, i.e., at elevated levels of B12 in blood plasma (> 1,700 pg/ml) oncohematological diagnosis is more frequent.
In our study, we report strong association of significantly elevated B12 with oncohematological disorders of myeloid cell lineage origin. Our retrospective data are enough to suggest B12 as a potential diagnostic marker for myeloid leukemia when liver and kidney diseases are excluded.
Patients with elevated B12 level in the blood (> 3,000 pg/mL, reference value 119–663 pg/mL) not receiving B12 as a supplement to medication or food, are found quite often in our and other studies. In different studies elevated blood levels of B12 (> 820 pg/mL) were reported in 12–18% of hospitalized patients [8, 10, 12, 15]. In the available literature, explanations for such cases are incomplete burdening decision-making and patients’ management. Increased levels of B12 related to a renal impairment, liver damage, and oncological processes have been reported in various sources [3, 7, 11, 16]. To understand B12 metabolism in general Nexo and Hoffmann-Lücke discusses the need for testing of holotranscobalamin, which is metabolically active form of B12 .
There are some studies on B12 usage as a prognostic marker for chronic liver disease  and hepatocellular carcinoma . Different sources of literature associate elevation of HCs (partial fraction of B12) with autoimmune lymphoproliferative disease  and hepatocellular carcinoma .
Chiche et al.  in their research including 2,064 patients report association of elevated elevated B12 (> 1,275 pg/mL) with blood malignancies (p < 0.05). Arendt et al.  analyzed large cohort of Danish population for incidence of different cancer types for cases with elevated B12 and found strong association of developing cancers of hematological origin within one year interval from elevation of B12 > 1,084 pg/mL (converted, originally in article 800 pmol/L) — standardized incidence ratio (SIR) 24.14 (95% CI 19.51–29.54), with overall oncohematological disease SIR of 4.52 (95% CI 4.23–4.82).
If we manipulate with these data, as SIR of elevated B12 division with SIR of disease incidence in overall population — 24.14/4.52 we get 5.3, what is comparable to our OR of 6.0. We accent that we analyzed overall population representative data, and excluded a lot of unrelated to cancer cases. We also could lose part of data, due to toxic effect of chemotherapy performed in oncohematological patients (exclusion criteria — elevation of ALT or AST, and decrease of GFR).
Our results including 79,524 patients confirm potential of B12 level as a diagnostic or prognostic marker for oncohematological disorders. Significantly elevated B12 level association with oncohematological diagnosis are shown, especially with myeloid leukemia. Further studies are necessary to understand B12 elevation mechanisms in oncohematological patients. Theoretically, elevation of B12 could be linked to HCs release by the granules of myeloid lineage cells in cases of myeloid leukemia, as it was discussed in multiple studies.
Investigations on elevated B12 in association with myeloid leukemia firstly appeared in literature in the middle of last century [20, 21]. Importance of B12 use as diagnostic tool for cancer has been actualized during the past ten years, but there are still some questions.
The number of patients with stable elevated B12 level (n = 8,082) is comparable to the number of patients with rapidly fluctuating B12 levels (n = 8,442). We hypothesized reason for such cases could be supplementary intake and B12 medication. Traditionally vitamin supplements contain B12 at much more than 100% of recommended day intake. Studies for B12 elevation in blood due to supplementary intake are required.
It is unclear, when B12 level increases in oncohematological patients — before or after CBC changes. In order to understand possible correlations between vitamin B12 level and oncohematological patients, in-depth analysis of patient’s medical history and other results has to be performed in the next study phase.
We have identified an important group of primary diagnosed chronic myeloid leukemia patients for whom further investigations have to be continued in order to determine what kind of B12 form is elevated. For diagnostic purposes it is important to clarify whether B12 itself or any of its forms could serve as diagnostic marker of myeloid leukemias. It would be beneficial to address a correlation between fluctuations of B12 in response to the therapies of myeloid leukemia and their role in hepatotoxic effect.
We would like to express our thanks to Ludmila Volozonoka for her valuable help in preparing the text of our paper.
No Comments » Add comments