Serum Homocysteine and Methylmalonic Acid Levels in Cobalamin and Folate Deficiencies

The combined use of homocysteine and methylmalonic acid (MMA) levels can differentiate cobalamin from folate deficiency, because most patients with folate deficiency have normal MMA levels, and the remainder have only mild elevations.² These two tests are useful diagnostically. The abnormally high levels of metabolites return to normal only when the patient receives replacement with the appropriate (deficient) vitamin. A positive response to cobalamin, documented by falling levels of homocysteine and MMA, is evidence of cobalamin deficiency. Conversely, therapy with folate results in a decrease in the isolated homocysteine level if folate deficiency is present.² Indeed, because several variables that are not related to vitamin deficiency (such as age, mild renal dysfunction) can falsely elevate serum homocysteine and MMA levels, if there is ambiguity, proof of vitamin deficiency would require clear-cut demonstration of a reduction in metabolite levels after specific vitamin supplementation.^2,3

Modified Therapeutic Trials

The traditional therapeutic trial using physiologic doses of vitamins (100 mcg of folate or 1 mcg of cobalamin given daily while monitoring the reticulocyte response)¹ has given way to a modified therapeutic trial. Rather than making the diagnosis of a deficiency, the intention is often to confirm the clinical suspicion that the patient does not have deficiency. This can be demonstrated by lack of response to full replacement doses of both vitamins (1 mg of folic acid orally for 10 days and 1 mg of cobalamin intramuscularly or subcutaneously daily for 10 days). Clinical scenarios in which such trials may be applicable (after drawing blood for serum cobalamin and folate levels) are as follows:

In all therapeutic trials, if there is no evidence of response within 10 days, bone marrow aspiration is indicated to identify another primary hematologic disease.

Table 12-2 Serum Cobalamin: False-Positive and False-Negative Test Results

IF, Intrinsic factor; TC, transcobalamin.

^*Although a low serum cobalamin level is not synonymous with cobalamin deficiency, 5% of patients with true cobalamin deficiency have low-normal cobalamin levels, a potentially serious problem because the patient’s underlying cobalamin deficiency will progress if uncorrected.

Diagnosing Folate Deficiency

When combined with a clinical picture of megaloblastic anemia and additional results of cobalamin levels, the serum folate concentration is the cheapest and most useful initial biochemical test to diagnose folate deficiency² (see Table 12-1). The serum folate level is highly sensitive to folate intake, and a single hospital meal may normalize it in a patient with true folate deficiency. Rapidly developing nutritional folate deficiency first leads to a decline in the serum folate level below normal (less than 2 ng/mL) in about 3 weeks; it is a sensitive indicator of negative folate balance.¹ However, isolated reduction of serum folate level in the absence of megaloblastosis (i.e., false-positive result) occurs in one-third of hospitalized patients with anorexia, after acute alcohol consumption, during normal pregnancy, and in patients on anticonvulsants²; unfortunately, these are the very groups at high risk for folate deficiency and the people who exhibit low serum folate levels when they become folate deficient.¹ Conversely, in 25% to 50% of cases (predominantly alcoholics) with folate-deficient megaloblastosis, the serum folate levels may be low normal or borderline (2 to 4 ng/mL).² The serum folate level alone should never dictate therapy. It is important to consider the clinical picture, peripheral smear, and bone marrow morphology and also to rule out underlying cobalamin deficiency.

Summary of the Clinical Usefulness of Tests for Cobalamin and Folate Deficiencies

Within the clinical context of hematologic or neurologic features that suggest the diagnosis of cobalamin deficiency, if the cobalamin levels are suggestive but not definitive, then the MMA and homocysteine tests are an excellent gold standard test to confirm a clinical diagnosis. Patients with clinical cobalamin deficiency usually have MMA values over 1000 nM and homocysteine values that are over 25 µM. The MMA and homocysteine test results are much more sensitive than cobalamin levels and progressively increase much earlier than the drop in cobalamin levels; one or both metabolites was increased in 99.8% of more than 400 patients with proven cobalamin deficiency.²

Based on the lower costs of serum cobalamin and folate compared with serum MMA and homocysteine levels, it is recommended (see Table 12-1) to first use the cheaper tests that can assist in the diagnosis of cobalamin and folate deficiency.² Clinicians should also restrict use of serum MMA and homocysteine to patients with borderline cobalamin and folate levels; to patients with existing conditions associated with difficulties in the interpretation of test results; to situations in which cobalamin and folate levels are low, when a high MMA level is useful in confirming cobalamin deficiency (rather than attributing the condition to folate deficiency alone); and to patients with clearly low serum levels but for whom there is an alternative explanation for the findings that caused an unusual serum cobalamin level to be obtained (e.g., a diabetic or alcoholic with peripheral neuropathy, an alcoholic with a high MCV and a low serum cobalamin without anemia). In these cases, serum levels of metabolites can assist in the diagnosis of vitamin deficiency.

Diagnostic algorithms consistently stress the value of clinical data to improve the pretest probability of serum cobalamin and serum folate tests.² Without detailed clinical information, the combined test results for serum cobalamin, folate, and metabolite (homocysteine and MMA) are not sufficiently unambiguous to diagnose and distinguish cobalamin deficiency from combined cobalamin-plus-folate deficiency. In combined cobalamin-plus-folate deficiency, both vitamins would be needed to restore baseline values, particularly of homocysteine.²

Table 12-3 Clinical Conditions Not to Be Confused With Megaloblastosis

MCV, Mean corpuscular volume.

^*The central pallor that normally occupies about one-third of the normal red blood cell is decreased in macro-ovalocytes. This contrasts with the finding of thin macrocytes, in which the central pallor is increased.

^†Although megaloblastosis implies that a bone marrow test has been performed, with the addition of highly sensitive tests for the specific diagnosis of cobalamin and folate deficiency, the need for a bone marrow test is often dictated by the urgency to make the diagnosis.

^‡When the Coulter counter readings of a high MCV are not confirmed by looking at the peripheral smear.

Table 12-4 Beneficial Effects of Homocysteine-Lowering Therapy on Nonhematopoietic Systems

NTD, Neural tube defect.

^*Paper with randomized controlled trial data; grade A studies.

Miscellaneous Megaloblastic Anemias Not Caused by Cobalamin or Folate Deficiency

Drugs That Perturb Folate Metabolism

Ethanol. Although beer has a higher folate content than other alcoholic beverages, alcoholism may lead to neglect of healthy dietary practices in favor of alcohol. Patients who have one nutritious meal each day tend to stave off the eventual development of folate deficiency. Alcohol consumption leads to a relatively rapid (2- to 4-day) fall in serum folate levels. Excess alcohol consumption is possibly the most common cause of folate deficiency in the United States.¹

Trimethoprim and pyrimethamine bind to bacterial and parasitic dihydrofolate reductase with much greater affinity than to human dihydrofolate reductase, but patients with underlying folate deficiency appear to be more susceptible to the effects of these drugs. The megaloblastosis can be reversed by folinic acid (5-formyl-tetrahydrofolate [5-formyl-THF]; leucovorin).

Methotrexate binds with high affinity to human dihydrofolate reductase and leads to trapping of folate as a metabolically inert form (dihydrofolate). This leads to a true depletion of THF within hours and consequently to functional deficiency of 5,10-methylene-THF and reduced thymidylate synthesis. Although megaloblastosis can develop rapidly, the toxic effects of methotrexate can be avoided by rescue with 5-formyl-THF (leucovorin).

Sulfasalazine produces megaloblastosis in up to two-thirds of patients taking full doses (over 2 g/day) by decreasing absorption of folates and induction of Heinz body hemolytic anemia (i.e., increased requirements).

Anticonvulsants can induce NTD, and consensus guidelines have stressed the importance of ensuring that pregnant women² and children^22–25 with epilepsy be prescribed folates together with anticonvulsants. Whereas folates protect against spontaneous abortion,²⁶ folic acid supplementation of women receiving antiepileptic drugs, which are known to interfere with folate absorption, also led to a significant reduction of spontaneous abortion.²⁷ Now there is new clinical data on phenytoin-induced gingival hyperplasia, which is a cosmetically undesirable side effect that affects a large percentage of patients, usually between 2 and 6 months of initiating therapy. A recent randomized controlled trial among children 6 to 15 years of age who were initiated on phenytoin has provided incontrovertible evidence that taking folic acid 0.5 mg daily can largely prevent phenytoin-induced gingival hyperplasia¹⁴; whereas 88% in the placebo group developed gingival hyperplasia, only 21% in the folic acid group developed this side effect. The data from this paper provides more “ammunition” to encourage young women on antiepileptic drugs to keep taking folic acid to prevent them from getting cosmetically unsightly gingival hyperplasia (particularly if reducing the risk for having a baby with NTD is too nebulous a concept for them). The only caveat is that before initiating long-term folic acid supplements, the cobalamin status must be normalized.

Although antineoplastics and antiretroviral antinucleosides such as azidothymidine lead to megaloblastosis, the temporal sequence and investigations to rule out cobalamin or folate deficiency should easily lead to a correct causal assignment.

Diagnostic Bone Marrow Aspiration

Is bone marrow aspiration always necessary to diagnose cobalamin- or folate-deficient megaloblastosis? With the addition of highly sensitive serum tests for the specific diagnosis of cobalamin and folate deficiency, the need for a bone marrow test is often dictated by the urgency to diagnose megaloblastosis (with results available in an hour). For example, in the case of florid hematologic disease with or without neurologic disease suggestive of cobalamin or folate deficiency, bone marrow aspiration carried out as soon as possible is invaluable in assisting the rapid diagnosis of megaloblastosis. However, in the outpatient setting, when the patient has a characteristic peripheral smear, or for a patient with a primary neuropsychiatric presentation, a case can be made to initiate the sequence of diagnostic tests without bone marrow aspiration by proceeding with measurement of serum levels of vitamins or metabolites (see Table 12-1). In a pregnant patient with pancytopenia with macro-ovalocytes, hypersegmented PMNs, and reticulocytopenia with a history of noncompliance with prenatal supplements (and no neurologic findings suggestive of cobalamin deficiency), bone marrow aspiration may not be necessary to initiate therapy for a strong presumptive diagnosis of folate deficiency. If there is no evidence of response within 10 days, bone marrow aspiration is indicated.

Practicing Classical Medicine Without the (Classical) Schilling Test

The Schilling test was used to identify the locus of cobalamin malabsorption and, in some instances such as pernicious anemia or bacterial overgrowth, the cause of cobalamin deficiency. However, this test has been unavailable since 2003 in the United States. Because only 50% of patients with pernicious anemia have serum anti–intrinsic factor (IF) antibodies, there are limitations in differentiating the other one-half without measurable anti-IF antibodies from those with food-bound cobalamin malabsorption and others with an intestinal cause for cobalamin malabsorption. A minimalist approach is to work around the nonavailability of the Schilling test and use a classical clinical approach to rule out potential differential diagnoses of cobalamin deficiency. For example, most conditions predisposing to cobalamin deficiency should be clinically manifest by the time cobalamin deficiency is evident. It should therefore be possible to identify several conditions through a detailed dietary history or past medical history, travel history, and drug history to suggest a dietary cause, esophagogastroduodenal disease, pancreatic insufficiency, impaired bowel motility, or other autoimmune diseases. The history and physical examination could provide further leads and suggest additional focused laboratory testing for rare conditions (stool for ova, anti–tissue transglutaminase antibodies, lipase, gastrin, intestinal biopsy, or radiographic contrast studies for stasis, strictures, fistulas). With no further leads, and therefore by default, one can assume that the diagnosis is either pernicious anemia or food-cobalamin malabsorption, which are both treated with similar replacement doses of cobalamin. For the younger patient with megaloblastic anemia, differentiating juvenile pernicious anemia and congenital IF deficiency would warrant measurement of gastric juice for IF and achlorhydria, deoxyribonucleic acid (DNA) for gastric IF can identify hereditary megaloblastic anemia, and DNA for mutations in cubam receptor (amnionless/cubulin genes) could identify Imerslund-Gräsbeck syndrome.

Thus the history, physical findings, and focused laboratory tests with careful clinical follow-up can potentially identify the cause of the majority of cases of cobalamin deficiency and bypass the need for a Schilling test.