Impaired Antibody Responses to Pneumococcal Polysaccharide in Elderly Patients with Low Serum Vitamin B₁₂ Levels

Methods

Patients

Eligible patients were older than 65 years of age and had not previously had pneumococcal vaccination. Hospitalized patients were identified after their serum vitamin B₁₂ levels were measured in the routine laboratory of Maimonides Medical Center. If the patient and the private physician agreed to participate, vitamin B₁₂ therapy was not initiated for at least 4 weeks after vaccination. For each study patient, a comparable age- and diagnosis-matched patient with a normal serum vitamin B₁₂ level was used as a control. We excluded nonambulatory and immunosuppressed patients, such as those with cancer or those receiving corticosteroids, and patients with renal or hepatic diseases. All patients and controls received a subcutaneous injection of 0.5 mL of the 23-polyvalent pneumococcal polysaccharide vaccine (Pneumovax 23, Merck, Sharp & Dohme, West Point, Pennsylvania). Serum samples obtained before and 4 weeks after vaccination were frozen at −20 °C. All patients gave informed consent, and the institutional review board approved the study.

Vitamin B₁₂ and Antipneumococcal Antibody Assays

Serum vitamin B₁₂ levels were measured using a fluorometric enzyme-linked assay kit (Baxter Diagnostics, Deerfield, Illinois). In our laboratory, values less than 167 pmol/L are considered deficient, those between 167 pmol/L and 200 pmol/L are considered indeterminate, and those between 200 pmol/L and 713 pmol/L are considered normal. For our study, we defined low serum vitamin B₁₂ levels as 200 pmol/L or less and normal levels as greater than 200 pmol/L. Serum folate levels were also measured using a fluorometric enzyme-linked assay kit (Baxter Diagnostics) (normal range, 4.5 nmol/L to 31.7 nmol/L). Each of 12 pneumococcal serotypes (1, 3, 4, 6A, 7F, 8, 9N, 12F, 14, 18C, 19F, and 23F) was assayed by uniformly labeled 14C capsular polysaccharides. These reagents were individually tested for their precipitability with type-specific antisera and for their inhibitability by only the type-specific unlabeled polysaccharide. No reagent showed inhibition with pneumococcal C-polysaccharide. The serum to be assayed (25 µL) was added to 0.5 mL of antigen (20000 counts per minute) and allowed to stand at 37 °C for 20 minutes. An equal volume of saturated ammonium sulfate solution was added to precipitate antigen-antibody complexes. The precipitate was collected and dissolved in 10% Triton X-100 (Rohm and Haas, Philadelphia, Pennsylvania) and measured in a liquid scintillation spectrometer. The nanograms of antibody nitrogen per milliliter were calculated by reference to a standard curve [6].

Statistical Analysis

We calculated Pearson correlation coefficients for vitamin B₁₂ and folate levels compared with antibody titers before and after vaccination and the difference between titers. We used the nonparametric rank-sum test to compare the differences in antibody titers and other variables between the two patient groups. For comparison of antibody titers, we calculated the geometric mean of all 12 serotypes and the mean titer of four individual serotypes (3, 7F, 9N, and 23F). We chose these four serotypes because they are considered to be immunogenic. We also analyzed data using a regression analysis in which the change in vaccination titers (the geometric mean titer of 12 measured pneumococcal serotypes) was the outcome variable. Both vitamin B₁₂ levels and changes in antibody titer were transformed to their natural logarithms to meet the assumption of equality of variance and normality in regression analysis. The results for the transformed data were the same as those for the original data. Calculations were done on the SPS computer program (Chicago, Illinois) using the regression module.

Results

Table 1 shows the clinical and hematologic variables and pneumococcal antibody titers before and after vaccination for both patient groups. Both groups had similar ages; leukocyte counts; folate, creatinine, and aspartate aminotransferase levels; erythrocyte mean corpuscular volumes; hematocrits; and antipneumococcal antibody titers before vaccination (Table 2).

Figure 1, Figure 2, and Figure 3 show scatter plots for vitamin B₁₂ levels compared with antibody titers before and after vaccination and the difference between titers (titer after vaccination minus titer before vaccination). A strong linear association can be seen between the difference in antibody titers and vitamin B₁₂ levels (r = 0.61; P < 0.001) (Figure 3). This relation results from a strong association between vitamin B₁₂ levels and antibody titers after vaccination (r = 0.55; P = 0.001) Figure 2 and the lack of association between vitamin B₁₂ levels and antibody titers before vaccination (r = −0.04; P > 0.2) (Figure 1). Serum folate levels did not correlate with titers before and after vaccination or with the difference in titers.

View larger version:

• In this window
• In a new window

Figure 1. Scatter plot showing no association between vitamin B₁₂ levels and antibody titers before vaccination.

View larger version:

• In this window
• In a new window

Figure 2. Scatter plot showing a linear association between vitamin B₁₂ levels and antibody titers after vaccination.

View larger version:

• In this window
• In a new window

Figure 3. Scatter plot showing a linear association of vitamin B₁₂ levels and the difference between antibody titers before and after vaccination.

For the 12 serotypes, the average difference in titers between patients in the two groups was almost 600 ng of antibody nitrogen/mL (P = 0.005) (Table 3). We also noted a difference in antibody titers for serotypes 3 (P = 0.02), 7F (P = 0.01), and 9N (P = 0.03), but not 23F (P > 0.2).

Table 4 shows the results of the regression analysis; the outcome was equal to the difference between antipneumococcal antibody titers before and after vaccination. Vitamin B₁₂ levels were an independent variable in predicting the change in pneumococcal antibody titers, regardless of whether the outcome was the actual change in titer or its logarithmic transformation or whether vitamin B₁₂ levels were grouped (low compared with normal) or treated as a continuous variable. After adjustment for age and mean corpuscular volume, titers in patients with normal vitamin B₁₂ levels were, on average, 616 ng of antibody nitrogen/mL higher than those in patients with low vitamin B₁₂ levels (P = 0.005). Erythrocyte mean corpuscular volume was also an independent predictor of change in antibody titer (P = 0.03). As the mean corpuscular volume increased, the change in titers decreased. Age was not associated with the difference in titers.

Discussion

The importance of low serum vitamin B₁₂ levels in hematologically and neurologically normal elderly patients has recently been appreciated. The model of the development of vitamin B₁₂ deficiency through pathophysiologic stages, including inadequate ingestion or malabsorption, serum and cell depletion, tissue deficiency characterized by biochemical abnormalities and eventually clinical deficiency, has provided greater insight into the process [7, 8].

Approximately 30% of elderly patients develop atrophic gastritis [9]. Malabsorption of food-bound cobalamin caused by hypochlorhydria is the most frequent initial cause of negative vitamin B₁₂ balance [10-12]. As hypochlorhydria progresses to achlorhydria, intrinsic factor secretion may also diminish to a level that is insufficient to maintain adequate vitamin B₁₂ absorption, resulting in pernicious anemia. Transcobalamin II, the physiologic vitamin B₁₂ transport protein that binds vitamin B₁₂ after intrinsic factor-mediated transfer to the ileal enterocyte, becomes depleted as less vitamin B₁₂ is absorbed into the ileum. Depletion of vitamin B₁₂ on transcobalamin II, that is, low holotranscobalamin II levels, may be the earliest marker of serum depletion [13-16]. Tissue deficiency is noted with biochemical abnormalities such as an elevated serum homocysteine or methylmalonic acid level or an abnormal deoxyuridine suppression test result [5, 17]. About 7 years after negative vitamin B₁₂ balance starts, clinical deficiency occurs. Because a low serum vitamin B₁₂ level represents a phase in an ongoing pathologic process, it should not be ignored [18].

Methionine synthase, for which methylcobalamin is the coenzyme, demethylates methyltetrahydrofolate, enabling it to form the coenzyme for thymidylate synthetase [19-21]. In vitamin B₁₂-deficient megaloblastosis, DNA synthesis is slowed proportionately less than are protein and RNA synthesis. Morphologically, the cell nuclei appear to be less mature compared with the development of the cytoplasm. Transcobalamin II delivers vitamin B₁₂ to all tissues, but its delivery is especially important to rapidly proliferating cells. Although hematologic cells are classically affected, studies have shown that lymphoid cell function may also be impaired in vitamin B₁₂ deficiency. Hitzig and colleagues [22, 23] described a neonate with transcobalamin II deficiency. This patient developed severe megaloblastic anemia and profound hypogammaglobulinemia. After high-dose vitamin B₁₂ therapy, both cell lines returned to normal. The researchers speculated that multiple doses of antigens given before vitamin B₁₂ therapy had led to the differentiation of antigen-specific memory cells but that subsequent clonal expansion and maturation into plasma cells necessitated vitamin B₁₂ administration. Hall and colleagues [24] later showed that vitamin B₁₂-deficient cells produced lower immunoglobulin levels in vitro than cells that were not vitamin B₁₂-deficient and that immunoglobulin synthesis was enhanced when vitamin B₁₂ was added to the media [24]. Van Dommelen and coworkers [25] and Kafetz [26] described elderly patients with hypogammaglobulinemia associated with vitamin B₁₂-deficient megaloblastic anemia. After vitamin B₁₂ therapy, immunoglobulin levels returned to normal. As holotranscobalamin II becomes depleted during progressive negative vitamin B₁₂ balance, proliferation of lymphoid and hematologic cells probably becomes impaired.

The immunologic defense against pneumococcal infection depends on a type-specific anticapsular antibody that interacts with complement to opsonize organisms [27], thereby enhancing phagocytosis and clearance. Although defects in the components of complement, neutrophils, the liver, or the spleen may also impair this overall process, deficiency of type-specific antibodies is the most commonly identified immunologic deficiency that predisposes patients to pneumococcal infection [1]. Pneumococcal polysaccharide directly triggers the activation of B cells, but T cells and other genetic factors influence the immunoglobulin class and the magnitude of the antibody response. Clinical and experimental observations have established the association between the presence of specific antibodies and the protection against infection by homologous organisms. However, the exact level of antibody nitrogen that is protective is still in doubt, because differences in techniques of measuring antibody by radioimmunoassay and by enzyme-linked immunosorbent assay have yielded differing results [3]. The radioimmunoassay in our study used reagents that do not measure C polysaccharides. Antibody responsiveness among adults younger than age 55 years appears to be protective, because prospective case–control studies have determined the clinical efficacy of the vaccine to be 93%. However, antibody responses are diminished in immunocompetent elderly patients, and case–control studies have estimated that the efficacy of the vaccine is 46% to 70% [28-30]. Factors associated with vaccine failure include infections with pneumococcal serotypes not present in the 23-polyvalent polysaccharide vaccines, which may occur in 10% of elderly patients with pneumonia, and the inability to achieve adequate antibody titers. With the rare exception of an underlying Ig2 or Ig4 subclass deficiency, factors associated with a failure to achieve adequate protective antibody titers among immunocompetent adults are unknown.

We assessed antibody responsiveness to pneumococcal vaccine in immunocompetent adults with low serum vitamin B₁₂ levels and found that such patients had lower antibody responses to the mean of 12 serotypes and to 3 of 4 individual serotypes than did controls. Only further study can determine whether the association between low serum vitamin B₁₂ levels and impaired antibody responsiveness applies to other polysaccharides or to protein immunogens. Vitamin B₁₂ levels also predicted the change in antibody level when the results for all patients, including those with normal vitamin B₁₂ levels, were analyzed together. Our findings suggest that a low serum vitamin B₁₂ level is associated with impaired humoral immunity in elderly immunocompetent adults. Although a low serum vitamin B₁₂ level in these patients seemed to be of relatively minor clinical significance in other respects, decreased availability of vitamin B₁₂ to rapidly proliferating B lymphocytes may have impaired clonal expansion and synthesis of specific immunoglobulins. Further studies are needed to determine whether vitamin B₁₂ treatment can enhance the synthesis of specific immunoglobulins and improve the clinical efficacy of the pneumococcal vaccine in patients with low vitamin B₁₂ levels. Until evidence suggests that vitamin B₁₂ therapy improves vaccine efficiency, however, there is insufficient justification for administering the vitamin to patients with subclinical vitamin B₁₂ deficiency on the grounds that such therapy increases antibody responsiveness.

Dr. Schiffman: Department of Microbiology and Immunology, State University of New York Health Science Center at Brooklyn, 450 Clarkson Avenue, Brooklyn, NY 11203.