Paraneoplastic syndromes are clinical profiles that can favour early diagnosis of the associated tumour, which makes recognising them very important.1,2 Anti-Ma antibodies are included in the group known as ‘well characterised onconeuronal antibodies’. As such, they allow doctors to diagnose ‘definite paraneoplastic syndrome’, even when no tumour has been detected.3 The anti-Ma2 antibody is present in men with testicular tumours and isolated or combined limbic encephalitis (LE), diencephalic encephalitis (DE), or brainstem encephalitis (BE). Before the discovery of anti-Ma2 antibodies, anti-Ma antibodies had been reported in 4 patients with associated tumours of diverse aetiology and location (parotid gland, breast, lung [non-small-cell], and colon), and with pancerebellar syndromes or brainstem involvement.4 Both types of antibody recognise Ma family proteins, which are exclusively found in neurons and testicular germ cells, 2 locations considered immunologically privileged. While anti-Ma2 antibodies recognise Ma2 protein (PNMA2), anti-Ma antibodies recognise Ma1 and Ma2 (PNMA1 and PNMA2).5

The interest sparked by the identification of a link between anti-Ma2, LE/DE/BE, and testicular cancer6 led to classifying patients with anti-Ma antibodies exclusively as having Ma2, or to analysing only Ma2 reactivity in other cases. In this context, it was not clear whether the patient presented anti-Ma2 or anti-Ma antibodies. Differentiating between the 2 antibodies is important, since neurological symptoms seem to be associated with different tumours. In order to clarify this topic, we have reviewed our series of patients with anti-Ma and anti-Ma2 antibodies to better define associated neurological symptoms and tumour types, as well as responses to treatment.

We gathered a total of 36 patients aged 18 years and older diagnosed with paraneoplastic syndrome associated with antibodies against Ma proteins. Diagnosis was performed at the neuroimmunological research laboratory at Institut d’Investigacions Biomèdiques August Pi i Sunyer (IDIBAPS), Hospital Clínic de Barcelona. We excluded 4 patients due to lack of clinical information, resulting in a final sample of 32 patients. In all cases, we initially performed immunohistochemical screening on frozen sections of rat brain perfusion-fixed with 2% paraformaldehyde. Subsequently, results indicating characteristic immunoreactivity (Fig. 1) were confirmed by commercial immunoblot (Ravo Diagnostika, Friburgo, Germany) (Fig. 2). Patients’ clinical data were gathered from the medical histories and examinations at that hospital, or else provided by neurologists responsible for their care at other hospitals. Additionally, we conducted a literature review of new cases dating back to 2008, beginning immediately after the last published review on this topic,7 using a PubMed database search of the following keywords: “anti-Ma”, “anti-Ta”, “anti-Ma2”, “Ma1”, “Ma2”, and “paraneoplastic neurological syndromes”.

Figure 1.

Immunohistochemical study of rat cerebellum incubated with serum from a patient with anti-Ma2 antibodies using the immunoperoxidase technique with avidin-biotin-peroxidase complex. We observed reactivity in the shape of large intracytoplasmic dots in neurons of the dentate nucleus.

(0.09MB).

Figure 2.

Immunoblot strip showing reactivity of different sera with anti-Ma (column 1), anti-Ma1 (column 2), or anti-Ma2 (column 3), or negative control (column 4).

(0.08MB).

The clinical analysis of our series of patients with paraneoplastic neurological syndromes and anti-Ma antibodies differs from other previously published series in some areas.7,31 In our series, age of onset was not associated with antibody type and at 70%, the percentage of men was significantly higher than the previously described 43%.31 The most frequent presentation was the combination of LE, DE, and BE, but some patients presented different neurological symptoms, such as anti-Ma-associated pure cerebellar syndromes, or anti-Ma2-associated isolated involvement of the peripheral nervous system. In our literature review, we identified 5 similar cases, including 2 with cerebellar syndrome (one isolated case with anti-Ma antibodies and another case with polyneuropathy associated with anti-Ma219), one with myeloradiculopathy,11 and one case of multiple mononeuropathy23 (this patient also presented asymptomatic temporal lobe lesions in MRI scans). The latter 2 cases presented anti-Ma2 antibodies. The fifth was a possible case of primary lateral sclerosis-like phenotype and progressive spastic paraparesis associated with anti-Ma2 antibodies.22

The first series of anti-Ma2- and anti-Ma-associated paraneoplastic neurological syndromes31 included 38 patients. Thirty-four presented classical symptoms of LE, DE, and BE and only 4 had atypical syndromes: 2 with cerebellar syndrome, one with myelopathy, and the fourth patient with myelopathy and radiculo-plexopathy, although these last 2 entities were associated with LE and BE. Cerebellar involvement was detected in 11 patients (9 with anti-Ma); it was associated with classical symptoms except in 2 cases. The second series examining this entity included 22 patients (14 with anti-Ma antibodies and 8 with anti-Ma2 antibodies).7 Of the total, 19 developed typical syndromes, and only 3 showed atypical symptoms: one patient had isolated polyneuropathy, one had upper and lower motor neuron disease (both anti-Ma2), and the third patient displayed cerebellar syndrome associated with parkinsonism and anti-Ma. Cerebellar involvement, although associated with typical clinical symptoms in this series, was identified in 8 patients (7 with anti-Ma).

In the mentioned studies, the majority of patients with anti-Ma2 were young men with testicular cancer. However, in this series and in other recent ones, the percentage of testicular cancer (40%) does not account for half of the total despite being predominant in the anti-Ma2 group. The comparison between the anti-Ma2 and the anti-Ma groups did not reveal significant differences in the associations with different types of tumours (germ cell testicular cancer, lung and pleural cancer, gastrointestinal cancer, and other less frequent tumours). All idiopathic cases in our sample were associated with anti-Ma2 antibodies, and the literature review returned 4 idiopathic cases associated with anti-Ma2 (although these reports did not specify Ma1 reactivity).8,17,19,22

Clinical outcomes seem favourable for Ma2 patients, with improvement or stabilisation of all symptoms; in contrast, Ma patients showed poorer clinical outcomes.7,31 These differences may be influenced by the varying percentages of tumour types in each group, as well as by unknown but probably immunological mechanisms. Although Ma2 patients were more likely to have testicular tumours (a neoplasia with a high cure rate after orchiectomy), no significant differences were detected. In other types of tumours presenting greater infiltration or a larger volume, and in those that are more difficult to treat, such as lung cancer, persistence of cancer may contribute to a patient's severe neurological progression.

In conclusion, our series suggests that both anti-Ma2 and anti-Ma1 antibodies are linked to neurological symptoms characterised by variable degrees of association with LE, DE, and BE. However, some patients present isolated cerebellar symptoms or peripheral nervous system syndromes. Determining the reactivity of anti-Ma1 and anti-Ma2 antibodies is important since neurological syndromes associated with anti-Ma2 antibodies respond better to treatment. Finally, we can confirm the existence of cases with antibodies reacting only against Ma1. We will need to study more patients in order to characterise the clinical symptoms associated with anti-Ma1.

This study received no funding of any kind.

The authors have no conflicts of interest to declare.