CORRESPONDENCE

URINARY METHYLMALONlC ACID AND COBALAMIN E;WX;CY IN THE

To the Editor: The study by Drs. Norman and

Morrison made very interesting and informat.ive reading.’ The high prevalence and availability of therapy for cobalamin deficiency in the elderly make any noninva- sive diagnostic index valuable. In this context, I would like to draw your attention to some additional data on this subject.

Yamauchi and coworkersZ studied the relation of urinary methylmalonic acid (MMA) ex- cretion and Vit,amin B,, defi- ciency. They found that patients with neurologic disturbances ex- creted larger amounts of MMA than those without neurologic disorders. They also concluded that MMA could be a useful ad- junct to distinguish megaloblastic anemia from myelodysplasia with niegaloblastosis.

As 2% of the general population are carriers of inherited methyl- malonic acidemia, Rasmussen and Nathan’ examined the urinary ex- cretion of MMA in patients with methylmalonic acidemia. They found urinary MMA levels to range from 1.18 to 2.48 mm01 per mol of creatinine (reference range 0.58 to 3.56). This reveals that het,erozy- gotes for inherited methylmalonic acidemia would not give false-pos- itive results for cobalamin defi- ciency, and hence testing for uri- nary MMA remains useful in this subgroup.

Finally, Rasmussen1 has de- scribed the use of another method for the measurement of urinary MMA, anion exchange extraction using formate and formic acid. This technique was found to be both reliable and convenient for the evaluation of cobalamin de& criency, particularly in patients with normal or moderately re-

duced serum cobalamin levels. The above data, in conjunction

with the authors’ study,’ promise to attenuate an important treatable cause of permanent neurologic disability in the elderly.

Ajay Anand, MD Division of Hematology

and Oncology New England Deaconess

Hospital

1. Norman EJ, Morrison JA. Screening elderly pattents for cobalamin (Vitamin 13,~) deficiency using the urinary methylmalonic acid assay by gas chromatography mass spectrometry. Am J Med. 1993;94:58%594. 2. Yamaucht H, Omine M, Tsukamoto N, et al. Urinary methylmalonic acrd excretion and clrnical features in megaloblastic anemia due to Vitamin B!, deficiency. (Japanese). Jon J Chn Hematol. 1989:30:835-839. 3. Rasmussen K. Nathan E. The clinrcal evaluation of cobalamin deflclency by determlnatron of methylmafonic acid in serum or urine IS not invalldated by the presence of heterozygous methylmalonic acldemra. J Cl/n Chem Clin Biochem. 1990;28:419-421. 4. Rasmussen K. Solrdphase sample extraction for rapid determinatton of methylmalonic acrd in serum and urine by a stable-isotope-drlutron method. Urn hem. 1989;35:260-264.

Manuscript submItted June 23, 1993 and accepted July 6, 1993.

MORE ON HDL SUBFRACTIONS

To the Editor: I have read with interest the re-

view on the HDL subfractions by Silverman et al. ’ Essentially, the article deals with t,he two funda- mental subclasses of HDL: HDLZ (density 1.063 to 1.125 g/mL) and HDL3(dl.l25tol210g/mL).Two other subclasses are frequently referred to in the review: Lp(A-I with A-II) and Lp(A-I without A- II), based on the presence or ab- sence of apoA-II in the HDL par- ticle; both types occur within the HDL2 and HDL3 density ranges. The authors also mention a num- ber of isolation methodologies other than ultracentrifugation and their disadvantages. One major

disadvantage is the alteration of the lipoprotein particle, so that the related classification of HDL may not reflect the physiologic one.’ Quite appropriately, they state that “Measuring subfrac- tions. . . may provide information not, only about atherosclerotic risk but also about the underlying physiologic mechanisms respon- sible for that risk.“’ A novel tech- nique (isotachophoresis) resolves HDL in at least six subpopula- tions, which differ for lipid and apolipoprotein composition as well as for their affinity of bind- ing to the cell-surface HDL recep- tors.” Indeed, the same apolipo- protein contained in different HDL particles assumes different conformations which profoundly affect its function(s).” This con- sideration leads us to one as- pect-function-dealt with in the first part of the review.

The range of functions exerted by apolipoproteins is wider than appearing from Table I.’ Among these functions, of clinical rele- vance can be considered the process of repair and nerve re- generation3 and the stimulated release of placental lactogen hor- mone.” Another property of HDL apolipoproteins is interaction with thyroid hormones (T4, T3) via specific and medium affinity sites (one per molecule). At least for apoA-I and apoE, the two apos studied in greater detail, the re- spective thyroid hormone binding domain is located in the exon-3 coded region, as opposed to the exon-4 coded lipid binding do- main.‘j,7 As a result of this inter- action, circulating HDL are the major lipoprotein carrier for thy- roid hormones. For instance, they carry 2% to 4% of total plasma T4 versus about 0.2% carried by low- density lipoproteins (LDL) and 0.03% by very low-density lipopro- teins (VLDL).N The LDLT4 com- plex is internalized by LDL-recep- tor competent human fibroblasts, indicating that LDL delivers to

514 May 1995 The American Journal of Medicine” Volume 98