Methylmalonic acid as a marker for cobalamin deficiency: fact or fantasy? Elucidations from the cobalamin-deficient rat.

Abstract

It was with considerable interest that we read the annotation ‘Diagnosis of cobalamin deficiency: the old and the new’ ( 3). We would like to raise some points for discussion, mainly concerning the validity of some biochemical markers (methylmalonic acid (MMA) and homocysteine (HCYS)) for cobalamin (Cbl) deficiency. We realize that the annotation was addressed mainly to haematologists, but we think that the question as to the reliability of ascertaining a Cbl-deficient status concerns not only haematologists but also neurologists, gastroenterologists and even oncologists and therefore is of considerable importance. Over the last few years ( 8, 7) we have worked with a new experimental model, the totally gastrectomized (TGX) rat, for reproducing a permanent Cbl-deficient status, which eventually brings about a severe anaemia and a severe spongiosus myeloneuropathy, similar to the human subacute combined degeneration (SCD). We have also carried out similar studies on rats which had been chronically fed a Cbl-deficient (Cbl-D) diet ( 8, 7). More recently we have collaborated with Dr Lindenbaum and his group in investigating whether or not the biochemical markers for Cbl deficiency could also be meaningful in the TGX-rat, paying particular attention to a possible involvement of MMA in the pathogenesis of SCD ( 1). Although we realize that it is hazardous to extrapolate from the rat to humans and/or vice versa we can conclude from published data ( 6) that MMA may be a much more reliable marker than HCYS for Cbl deficiency, in agreement with that demonstrated in human patients ( 2). To further support this statement, we now report previously unpublished data (obtained in collaboration with Dr Lindenbaum) on serum levels of MMA and HCYS of rats chronically fed a Cbl-D diet (see Table I). In this case, serum MMA levels increased in a constant manner, but when the experimental period was prolonged, the magnitude of increase did not alter. Furthermore, in TGX-rats we found: (a) increased MMA levels even in some organs having a L-methlymalonilCoA (MMCoA) mutase (E.C. 5.4.99.2) activity; (b) a fall in serum levels of MMA and HYCS after repeated Cbl injections ( 6). These responses to postoperative Cbl treatment ( 6) were not ‘partial’ and certainly not ‘trivial’, as Drs Chanarin and Metz judged the MMA levels to Cbl therapy to be. Furthermore, we can specify that among the TGX-rats only one out of 28 and only two of the rats chronically fed a Cbl-D diet had a low serum Cbl level together with a serum MMA level within the normal range of values. We did not ignore the well-known possibility of a rise in serum MMA levels not specifically linked to the Cbl deficiency ( 5). We demonstrated that increases in serum MMA levels of TGX-rats are partially due to an overproduction of propionic acid by gut flora. Indeed, the postoperative treatment of TGX-rats with lincomycin markedly decreased serum MMA levels, which, however, still remained significantly higher than those observed in the controls ( 6). In keeping with this is the fact that the magnitude of the increase of serum MMA levels in TGX-rats (in whose intestine a bacterial overgrowth is known to occur) ( 6) is far greater than that observed in rats chronically fed a Cbl-D diet (where a bacterial overgrowth in the intestine can be reasonably excluded) (see Table I). Finally it is quite surprising that Drs Chanarin and Metz do not consider the behaviour of serum MMA level in the Cbl-deficient patients before and after Cbl therapy as ‘evidence of a Cbl deficiency’. If this were so, we should rediscuss the principle of causality in epistemology and in biology! This by no means necessarily signifies that MMA enhancement is due only to Cbl deficiency and that it might not occur in other diseases without Cbl deficiency ( 4). It is just that we think one cannot deny that Cbl deficiency brings about an impairment in L-MMCoA mutase reaction, which leads to an enhancement in MMA levels in physiological fluids, such as serum, urine and cerebrospinal fluid ( 1). This should not lead to the conclusion that considering serum MMA levels alone constitutes a definitive test, but simply that it can be one of the tests for Cbl deficiency, in agreement with the final statement in the annotation by Drs Chanarin and Metz.