Monoclonal IgM and neuropathy: not always anti-MAG

In this issue of Blood, Le Cann et al reported a series of 45 patients with the very rare CANOMAD syndrome (chronic ataxic neuropathy, ophthalmoplegia, immunoglobulin M [IgM] paraprotein, cold agglutinins, and disialosyl antibodies) that almost doubled the number of published patients.¹  The only other series with 18 patients was published in 2001 with no data on treatment.²  This study was an excellent example of collaboration between neurologists and hematologists that took place in 17 French centers. This number of patients allowed them to complete the description of CANOMAD, and particularly, of the associated monoclonal B-cell diseases, and more importantly, to provide a precise guideline on treatment.

Coexistence of neuropathy and monoclonal immunoglobulin (MIg) represents a common but complex problem that requires neurological and hematological collaboration for a precise diagnosis. Because of the frequency of MIg, it can be a coincidental cooccurrence, in the majority of cases for IgG and IgA isotypes but not for IgM isotype associated with a neuropathy in one-third of patients.³  The more frequent autoantigen in IgM MIg–associated neuropathy is the myelin-associated glycoprotein (MAG). IgM MIgs with anti-MAG antibody activity are associated with a characteristic distal, slowly progressive, demyelinating symmetric sensory and motor polyneuropathy.⁴ 

The antigens that are the target of the monoclonal IgM in CANOMAD are different; they are gangliosides that all harbor disialosyl groups, predominantly GD1b, GD3, GT1b, and GQ1b. Interestingly, such gangliosides are localized in the neurons of dorsal root ganglia and within the oculomotor nerves. When these antigens are the targets of polyclonal antibodies, usually of IgG idiotype, the patient develops a transient syndrome with ophthalmoplegia (Miller Fisher syndrome), or an acute ataxic neuropathy (see table), illustrating 2 features of monoclonal gammopathy of clinical significance (MGCS), a recently described group of diseases to which the CANOMAD syndrome belongs.⁵  The first one is that in MGCS the clinical presentation is less dependent on the underlying B-cell proliferation than on the MIg characteristics and in the case of an autoimmune mechanism on the antibodies’ target. In CANOMAD, the neurological manifestations (chronic ataxic neuropathy and ophthalmoplegia) are in line with the target antigens’ localization.

The second one is that you can often find a polyclonal counterpart of MGCS, such as mixed cryoglobulins, type II, because of a monoclonal IgM, whereas type III is completely polyclonal. In cryoglobulinemia, it is the constant exposure to a stimulating antigen (fragment crystallizable of IgG in immune complexes, most of the time with hepatitis C virus) that leads to the apparition of a monoclonal IgM with the same rheumatoid activity. Miller Fisher syndrome and Guillain Barre syndrome are often preceded by an infection, Campylobacter jejuni being the most frequent agent. We may postulate that a chronic infection by these bacteria could subsequently convert from polyclonal antigen-dependent B-cell proliferation to monoclonal IgM against gangliosides.^6,7 

In the present series and in the literature, not all the patients harbored all the features of the syndrome. In this series, only 13% of the patients presented ophthalmoplegia at diagnosis and only 44% during the course of the disease, whereas cold agglutinins, usually due to an anti-PR activity, were detectable in only 34% of patients. The predominant neurologic sign is clearly chronic sensory ataxia, and Yuki and Uncini, considering the acronym CANOMAD too restrictive, have proposed the term CANDA for chronic ataxic neuropathy with disialosyl antibodies.⁷  The key to the diagnosis of CANOMAD/CANDA is the search of anti–ganglioside antibodies with IgM specificity, and the clinicians should be aware that they must be searched in a patient with an IgM gammopathy and a sensory ataxia even if ophthalmoplegia and cold agglutinins are leaking. A prompt diagnosis is, as usual in MGCS, of paramount importance because, without efficient treatment, organ injuries can worsen, as in this series where almost one-third of patients have severe disability.

MGCS are a group of diseases where tissue injuries are not due to the expansion of clonal malignant ce"lls but to the pathogenic activities of the MIg.⁵  As in other MGCS, for example, AL amyloidosis, where 40% of patients have a smoldering myeloma, a significant proportion (36%) of patients in this series have a hematologic malignancy, mainly Waldenström macroglobulinemia or low-grade B-cell lymphoma, a fact not described in the 2001 series of 18 patients.²  These hematologic diseases usually do not have a high tumor mass and are diagnosed at a younger age than in patients with the same hematologic malignancy but without MGCS. This early diagnosis is probably due to the symptoms caused by the MGCS. In CANOMAD/CANDA, the median age is 62 years, younger than the 70 found in Waldenström macroglobulinemia. As mentioned earlier, the neuropathy does not seem to be more severe in patients with hematologic malignancies.

The most important part of this paper is probably the data on management. It clearly demonstrates that IV immunoglobulins (IVIGs) and rituximab-based regimens are effective with 50% to 60% clinical responses and that corticosteroids and immunosuppressive therapies are useless. IVIGs seem to be effective in MGCS because of a proved or putative autoantibody mechanism like in necrobiotic xanthogranuloma, systemic capillary leak syndrome, or scleromyxedema.⁵  IVIGs are efficient in CANOMAD/CANDA but not in anti-MAG neuropathy, where intramyelin IgM deposits are a hallmark of the disease.

However, IVIGs usually have only a temporary action, and as in other MGCS (Schnitzler syndrome being the exception because of the impressive efficacy of anti–IL-1 antibodies), a secondary treatment, which targets the B-cell clone producing the toxic MIg, is often necessary.^5,8  CANOMAD/CANDA is associated with a monoclonal IgM, and the use of rituximab alone or associated with other agents is logical, assuming that the B-cell clone expresses CD20, even if, as in other MGCS, the nature of the responsible clone must, whenever possible, be precisely characterized. Le Cann et al proposed a functional algorithm with first IVIG and then a rituximab-based regimen for nonresponding or early relapsing patients. Several questions remain on the best time to introduce rituximab, on the role of the new treatments targeting B-cell proliferation, such as BTK and BCL2 inhibitors, and also on whether the intensity of treatment should be adapted to the disease’s severity.