Fibroblast growth factor receptors as diagnostic markers of acquired sensory neuronopathies

Abstract

The present invention relates to the diagnosis of acquired sensory neuronopathies (SNN), and to the treatment of these disorders.

Claims

1. A method for treating a patient suffering from an acquired sensory neuronopathy who shows immunoreactivity toward a protein of the tyrosine kinase receptor family chosen from FGFR1, FGFR2 or FGFR3 and/or toward a GRB10 protein, which method comprises a) detecting in a biological sample of the patient immunoreactivity toward a protein of the tyrosine kinase receptor family chosen from FGFR1, FGFR2 or FGFR3 and/or toward a GRB10 protein; and b) administering to the patient at least an immunosuppressant and/or immunomodulator compound(s), or a pharmaceutical composition comprising said compound(s).

2. The method of claim 1, wherein the immunosuppressant compound is chosen from the group consisting of tacrolimus (CAS number 104987-11-3), cyclosporine (CAS number 59865-13-3), methotrexate (CAS number 59-05-2), glucocorticoids, cyclophosphamide (CAS number 50-18-0), azathioprine (CAS number 446-86-6), mycophenolate mofetil (CAS number 24280-93-1), anti-CD20 antibodies and fingolimod (CAS number 162359-55-9).

3. The method of claim 1, wherein the protein of the tyrosine kinase receptor family is chosen from the group consisting of i) FGFR3, ii) a protein which comprises or consists of the intracellular kinase domain of FGFR3, iii) a protein which comprises or consists of the TRK1 subunit of the intracellular kinase domain of FGFR3, and iv) a protein which comprises or consists of the TRK2 subunit of the intracellular kinase domain of FGFR3.

4. The method of claim 1, wherein the proteins toward which immunoreactivity is detected are i) a protein chosen from the group consisting of FGFR3, a protein which comprises or consists of the intracellular kinase domain of FGFR3, a protein which comprises or consists of the TRK1 subunit of the intracellular kinase domain of FGFR3, and a protein which comprises or consists of the TRK2 subunit of the intracellular kinase domain of FGFR3, and ii) GRB10.

5. An analytical method comprising i) obtaining a biological sample from a patient having an acquired sensory neuronopathy and ii) detecting, in the biological sample, antibodies toward a protein of the tyrosine kinase receptor family chosen from FGFR1, FGFR2 or FGFR3 and/or toward a GRB10 protein.

Description

DESCRIPTION OF THE FIGURES

(1) FIG. 1 illustrates the level of reactivity of 16 non paraneoplastic SNN and 30 controls on FGFR3. The dotted line indicates the cut-off level. Protoarray assay

(2) FIG. 2 illustrates the level of reactivity of 16 non paraneoplastic SNN and 30 controls on GRB10. The dotted line indicates the cut-off level. Protoarray assay.

(3) FIG. 3 is a histogram which illustrates the results of ELISA assays carried out on sera from patients suffering from autoimmune diseases or neuropathies, with the human FGFR3 399-806 aa peptide corresponding to the protein intracellular domain.

(4) Comparison of the mean OD (optical density) minus blank value in the groups (ANOVA test) is presented: D=blood donors; lupus/SGS=autoimmune diseases; other neurop=other neuropathies; SNN=sensory neuronopathies. The patient and control groups are on the x-axis and the OD values minus blank values are on the y-axis.

(5) FIG. 4 is a histogram which illustrates the results of ELISA assays carried out on sera from patients suffering from autoimmune diseases or neuropathies, with the human FGFR3 399-806 aa peptide corresponding to the protein intracellular domain.

(6) Comparison of the normalized OD in four groups (ANOVA test) is presented: D=blood donors; lupus/SGS=autoimmune diseases; other neurop=other neuropathies; SNN=sensory neuronopathies. The patient and control groups are on the x-axis and the normalised OD values are on the y-axis.

(7) FIG. 5 illustrates the results of ELISA assays with the human FGFR3 399-806 aa peptide corresponding to the protein intracellular domain.

(8) Part A: normalised OD.

(9) Part B: index.

(10) D=blood donors; lupus/SGS=autoimmune diseases; other neurop=other neuropathies; SNN=sensory neuronopathies.

(11) The dotted line indicates the cut-off level.

EXAMPLES

Example 1

Materials and Methods

(12) 1.1. Serum Samples

(13) Serum samples from 261 individuals were collected. Sera were snap frozen at −80° C. and stored until utilisation. Individuals were classified into four groups. Group I corresponded to 59 healthy blood donors. Group II included 56 patients with systemic autoimmune diseases and lupus or Sjögren syndrome antibodies, group III included 104 patients with a diagnosis of sensory neuropathy, and group IV included 42 patients with sensorimotor peripheral neuropathies. Among group III, the neuropathy was a probable or possible SNN according to published criteria in 85 patients (Camdessanche J. P. et al., Brain 2009; 132:1723-1733). Among the 19 who did not fill these criteria, 7 had small fiber SNN and 2 a sensory neuropathy that could not be distinguished between proximal demylinating polyneuropathy (P-CIDP) and SNN. In the other, the criteria were not fulfilled mainly because ENMG sensory nerve abnormalities were mild. A known dysimmune context was associated with the neuropathy in 72/104 patients including, Sjögren syndrome, lupus, lupus anticoagulant, unclassified connective disorders, monoclonal gammopathy, including CANOMAD syndrome, and HIV infection. Eight had paraneoplastic SNN with Hu antibodies. Patients in group IV had Guillain-Barré syndrome, chronic inflammatory demyelinating polyneuropathy, multifocal motor neuropathy, neuropathy with monoclonal gammopathy, mononeuritis multiplex with vasculitis, diabetic neuropathy, hereditary neuropathy or idiopathic length dependent axonal neuropathy. As a whole the neuropathy was dysimmune in 11/42.

(14) 1.2. Protein Arrays

(15) Sera were probed in the human ProtoArray v.4.2 (Invitrogen, Carlsbad, N. Mex., USA). These microarrays contained 9000 human GST-tagged proteins, expressed in Sf9 insect cells and spotted in duplicate. Protoarrays were used according to the recommendations of the manufacturer. Briefly, the slides were equilibrated at 4° C. for 15 min and then incubated with blocking buffer (50 mM Hepes, pH7.5, 200 mM Nacl, 0.08% Triton X-100, 25% Glycerol, 20 Mm Reduced glutathione, 1 mM DTT, 1X Roti-Block) for 1 h at 4° C. with gentle shaking. Then slides were incubated with human sera diluted 1:500 in washing buffer (1X PBS, 0.1% Tween 20, 1X Roti-Block) for 90 min at 4° C. with gentle shaking. Slides were washed five times with the washing buffer and incubated with 1 μg/ml of secondary antibodies (Alexa Fluor 647 goat anti-human IgG antibody; Invitrogen) to detect the Human bound antibodies. The arrays were washed and dried by centrifugation at 200 g for 1 min. An array used as a control was incubated with the secondary antibody for background determination. A protein gradient of purified human IgG printed on each subarray was used as a positive control. Finally, the slides were scanned using a Genepix 4000B scanner (Axon, Union City, Calif., USA) with the laser set at 635 mm, the laser power at 100%, and the photomultiplier gain at 800. GenePix Pro 3.0 image analysis software (Axon, Union City, Calif., USA) was used for the quantification. The ProtoArray Prospector v5.0 software was used to identify immunoreactivities. This software uses the Chebyshev inequality P value, which is derived by testing the null hypothesis. Two statistical tools were used to detect positive spots, CI p-value and z-score. They correspond respectively to the probability that a spot is similar to a negative control, and to the spot signal minus average (for all spots) divided by the standard deviation for all spots. The NNS serum samples were compared versus control serum samples to obtain significant protein hits. The software also calculates the z score for each printed spot's fluorescent intensity. The z score indicates the deviation of each protein's antibody reading from its distribution mean (SD).

(16) 1.3. Dot Blot

(17) A dot blot was used to assess the ability of target proteins to screen for SNN status in serum. Two to 10 μl of the purified recombinant peptide consisting of the 399-806 aa of the protein corresponding to the intracellular kinase domain of human FGFR3 (Invitrogen) used for the ProtoArray was spotted in nitrocellulose membrane. The membrane dried for 5 min and was blocked with 5% BSA in TBS-T for 1 h. Then serum samples (dilution 1/50 in blocking buffer) were incubated overnight at 4° C. After washing three times, biotinylated goat anti anti-human IgG antibody (Dako, Glostrup, Denmark) diluted 1:100 in blocking buffer, followed by streptavidin-peroxidise was used. Immunoreactive spots were visualized using diaminobenzidine (DAB) detection reagents. As control, a rabbit polyclonal anti-FGFR3 antibody directed against the 742-806 aminoacide sequence of human FGFR3 (GeneTex, San Antonio, Tex., USA) was used.

(18) 1.4. ELISA

(19) The ELISA method was used to assess the ability of target protein identified by protoarray analysis to screen for sensitive ganglionopathy status in serum. Briefly, microtiter plates (Maxisorp, Nunc) were coated overnight at 4° C. with 1 μg/ml of the purified recombinant intracellular domain of human FGFR3 (Invitrogen), the full length recombinant FGFR1 (NCBI Reference Sequence: NP_075594.1; SEQ ID NO: 2) or FGFR2 proteins (AAH39243; SEQ ID NO: 4 (Novus biologicals) in carbonate-bicarbonate 0.05 M pH 9.6 solution. After washing one time with washing buffer (PBS containing 0.1% Tween 20), plates were blocked with blocking buffer, washing buffer containing 3% SVF and 0.1% gelatine for 1 h at room temperature. Then, serum samples (dilution 1/50 in blocking buffer) were incubated for 1 h at room temperature. After washing four times with washing buffer, anti-human IgG peroxidase-labelled (dilution 1/3000 in blocking buffer) was added for 1 h à room temperature. Then the signal was developed with o-Phenylenediamine substrate for 15 min to 30 min (sigma) and read at 450 nm. Each serum was tested in duplicate and the mean optic density (OD) of the readings was taken into account for the analysis. Controls included blank wells containing the products of the reaction minus human sera and the secondary antibody. The rabbit polyclonal anti-FGFR3 antibody diluted at 1/1000 and the appropriate secondary antibody were used as control and to normalise readings among plates. The specific reactivity of patients or controls' samples was obtained by subtracting the readings of plates of a pool of sera from a panel of healthy blood donors from the readings of plates of tested sera (hereafter designated as the normalized OD). In addition to limit variation among tests an index was built as follows: (OD of the tested serum/OD of the blank wells)/(OD of the pool of control sera/OD of the blank wells). To be considered as positive, the serum must be positive both for the normalized OD and the index.

(20) 1.5. Expression of FGFR by Sensory Neurons.

(21) Purified cultures of dorsal root ganglia (DRG) neurons were established from embryonic day 18 (E18) rats as previously described (Seilheimer B. et al., J. Cell. Biol. 1988; 107:341-351). Briefly, DRG were collected in L-15 medium (Gibco Invitrogen) containing antibiotics, centrifuged, and incubated for 15 min in 0.25% EDTA-trypsin (Gibco Invitrogen), then centrifuged and resuspended in MCM. The pellets were mechanically dissociated by passage through a 21 gauge needle and the neurons plated on poly-L-lysine-coated coverslips in MCM containing 50 ng/mL of recombinant human b-nerve growth factor (NGF, Peprothech, Rocky Hill, N.J., USA) and 5 mg/mL of glucose. After two days of cultures cells were fixed for 3 minutes in 5% paraformaldehyde and tested by immunohistochemistry with either anti-FGFR3, FGFR1, FGFR2 (GenTex), CRMP5 (home made) antibodies diluted at 1/500 and revealed either with FITC-conjugated goat anti-rabbit IgG antibody (Sigma-Aldrich) at a 1:1000 dilution or rhodamine-conjugated goat anti-mouse IgG antibody (Sigma-Aldrich) at a 1:2000 dilution. Double-immunolabelling was performed by incubating the sample with the primary antibodies under the conditions described above, followed by the appropriate FITC- and rhodamine-conjugated antibodies.

(22) 1.6. Immunocytochemistry on HEK293 Cells.

(23) HEK293 cells were transfected with plasmids containing the full length (pEGFPN3-FGFR3 full lengh), the intracellular domain (i.e. fragment which spans amino acids 397 to 806 of FGFR3) (pEGFPN3-FGFR3 cytoplasmic domain), the TRK1 (pEGFPN3-FGFR3-TK1 domain) or the TRK2 (pEGFPN3-FGFR3-TK2 domain) subunits of the intracellular domain of human FGFR3 tagged with EGFP or plasmid without insert (control).

(24) The plasmids pEGFPN3-FGFR3 full lengh, pEGFPN3-FGFR3 cytoplasmic domain, pEGFPN3-FGFR3-TK1 domain and pEGFPN3-FGFR3-TK2 domain were amplified from pcDNA3-hFGFR3 (a generous gift from Dr. Vigdis Sorensen, University of Oslo, Norway) by the following primers: 5′ATGGGCGCCCCTGCCTGC3′ (SEQ ID NO: 14) and 5′CGTCCGCGAGCCCCCAC3′ (SEQ ID NO: 15) for pEGFPN3-FGFR3 full lengh. 5′ATGAAGAAAGGCCTGGGCTCC3′ (SEQ ID NO: 16) and 5′CGTCCGCGAGCCCCCAC3′ (SEQ ID NO: 17) for pEGFPN3-FGFR3 cytoplasmic domain 5′ATGAAGAAAGGCCTGGGCTCC3′ (SEQ ID NO: 18) and 5′CCACTTCACGGGCAGCC 3′ (SEQ ID NO: 19) for pEGFPN3-FGFR3-TK1 domain 5′ATGGCGCCTGAGGCCTTG3′ (SEQ ID NO: 20) and 5′CGTCCGCGAGCCCCCAC3′ (SEQ ID NO: 21) for pEGFPN3-FGFR3-TK2 domain

(25) The PCR product was cut with Hind III and BamH1 and ligated into EGFP-N3 vector (Clontech laboratories).

(26) HEK293 cells were cultured in Dulbecco's modified minimal essential medium (DMEM, Gibco) containing 10% fetal bovine serum (FBS, Gibco) and antibiotics (25 U/ml penicillin and 25 μg/ml streptomycin and amphotericin) at 37° C. with 5% CO2. The day before transfection, the HEK293 cells were seeded into 24 well culture plates (9.104 per well).

(27) HEK293 cells were transfected using lipofectamine LTX (Invitrogen). A 100 μl mix containing 0.5 μg plasmidic DNA and 1.5 μl lipofectamine in OPTI-MEM medium (Invitrogen) was added to each well for 48 h at 37° C.

(28) Cells were then fixed in 4% paraformaldehyde for 5 min, washed with PBS. The cells were then blocked and permeabilized with 0.2% gelatin buffer containing 0.1% triton X-100 at room temperature for 1 h before being incubated overnight at 4° C. with patient's sera (diluted 1/20). After incubation, cells were rinsed with PBS and incubated with TRITC-goat anti-human IgG (Interchim) diluted 1/2000 for 3 h at 4° C.

(29) Immunostaining was observed with Zeiss fluorescence microscope.

Example 2

Results

(30) 2.1. Protein Arrays

(31) 46 serum samples from 16 non paraneoplastic SNN (4 with associated dysimmune disorders) and 30 controls (15 healthy blood donors, 8 anti-Hu associated SNN and 7 non SNN neuropathies) were probed in the human ProtoArray v.4.2 (Invitrogen). Using the ProtoArray Prospector v5.0 software, CI p-value and z-score with the respective standard cut-off of 0.001 and 4, 442 immunoreactivities were identified as significantly different in the SNN group versus the control group. Using the following stringent criteria: reactivity restricted to the SNN group, level of reactivity >1010 in the SNN group and <900 in the control group and Z score >5, only two immunoreactivities distinguished significantly the SNN group from the control one: namely anti-FGFR3 (NCBI Reference Sequence: NP_000133.1) and GRB10 (NCBI Reference Sequence: NM_001001550.1) reactivity present in 7/16 patients respectively (cf. FIGS. 1 and 2). Six sera reacted with both FGFR3 and GRB10 and one with each of them respectively so that 8 sera reacted with at least one of them. The level of reactivity was wholly higher with FGFR3 than with GRB10.

(32) 2.2. ELISA with the FGFR3 Protein.

(33) 152 out of the 271 sera (patients or controls) were tested from 2 to 10 times, median 3, and used for adjusting the method.

(34) By an ANOVA test, the OD minus the blank value of sample sera was significantly higher in the sensory neuropathy and autoimmune disease groups comparatively to blood donors and patients with other peripheral neuropathy (FIG. 3). By the same test, patients with sensory neuropathy had higher normalized OD comparatively to the other groups (FIG. 4).

(35) ROC curves were used to determine the cut-off values of the normalised OD and the index discriminating the SNN group from blood donors and other peripheral neuropathies with 100% specificity. A value of 0.13 for the normalized OD and 1.55 for the index were determined as the respective cut-off values. To be considered as positive for anti-FGFR3 antibodies both the normalized OD and the index values of a given serum had to be superior to the cut-off levels.

(36) This allowed the identification of 14/104 patients with sensory neuropathy, 0/59 blood donors, 0/42 patients with other neuropathies and 5/56 patients with autoimmune diseases with FGFR3 antibodies (FIG. 5). The neuropathy in group III, was associated to an autoimmune context in 3/14 patients and diagnosed as possible or probable SNN in 13/14 cases the last patients having a form difficult to differentiate between SNN and P-CIDP. When going back to the clinical file of the 5 patients with autoimmune disease (group II), two of them proved to have trigeminal sensory neuronopathy, one sensory neuropathy in the four limbs, one chronic thoracic neuropathic pain and one no known neurological syndrome.

(37) 2.3. Dot Blot with FGFR3

(38) To confirm the Elisa result by another method, the commercial FGFR3 antibody and two sera with patients with anti-FGFR3 antibody reacted by dot blot with the FGFR3 protein while sera from a blood donor did not (data not shown).

(39) 2.4. Elisa with FGFR1 and FGFR2 Proteins

(40) As the FGFRs proteins share a high degree of homology, to test the possibility of serum cross immunoreactivity between FGFR3 and the other FGFRs, a sample of SNN, autoimmune disease and control sera were tested by Elisa with the FGFR1 and 2 recombinant proteins. FGFR4 was not tested as none of the SNN and control sera analysed by the protoarray method reacted with it. Four out of 13 sera positive for FGFR3 were found positive with FGFR1 while none of 14 sera negative for FGFR3 were positive for FGFR1. Concerning FRGF2, 2/7 sera positive for FGFR3 were positive for FGFR2 while none of 5 sera negative for FGFR3 were positive for FGFR2. Two sera reacted with the three proteins. These results show that although cross immunoreactivity may occur with FGFR 1 and 2, FGFR3 is more frequently recognized by autoantibodies in patients with sensory neuronopathy.

(41) 2.5. Expression of FGFR Proteins by Sensory Neurons In vitro.

(42) As the patients with anti FGFR3 immunoreactivity were thought to have primary sensory neuron involvement that might be mediated to an auto-immunoreactivity directed toward FGFR3, the inventors checked that FGFR3 was expressed by dorsal root ganglia sensory neurons. By immunohistochemistry the FGFR3, 2 and 1 antibodies immunolabelled the cytoplasm of sensory neurons double labelled with the CRMP5 antibody while the FGFR5 antibody immunolabelled neuron nuclei (data not shown)).

(43) 2.6. Clinical Characteristics of the Neuropathy in Patients with FGFR3 Immunoreactivity.

(44) As 17 of the 19 patients with FGFR3 antibody had sensory peripheral neuropathy (89%) multivariate logistic regression was used to compare the clinical and electrophysiological characteristics of their neuropathy with that of 36 patients without anti-FGFR3 antibody and non paraneoplastic sensory neuropathy and 31 patients with anti-Hu antibody and paraneoplastic SNN (see Table 1 below). The clinical and electrophysiological items analyzed in this study have been published elsewhere (Camdessanche J. P. et al., Brain 2009; 132:1723-1733).

(45) TABLE-US-00001 TABLE 1 comparative clinical manifestation in patients with FGFR3 positive, FGFR3 negative and anti-Hu positive sensory neuropathy. FGFR3 positive FGFR3 negative Hu positive Number 19 38 31 Age (M + SD) 49.1 ± 15.6 60.5 ± 13.9 61.6 ± 11.3 Sex (female) 70% 50% 19% Dysimmune contex 42% 26% — ONSET Acute 17% 12% 36% Subacute 22% 22% 48% Progressive 61% 66% 16% Face 28% 3% 0% Lower limbs 56% 71% 58% Upper limbs 39% 45% 80% FULL DEVEVELOPMENT Lower limbs 78% 95% 97% Upper limbs 78% 71% 94% Face 28% 11% 13% Trunk 17% 11% 10% Dysautonomia 22% 27% 23% Pain 55% 46% 58% Ataxia 50% 62% 74% Assymetry 56% 17% 48% CSF Normal 50% 44% 0%

(46) Comparatively to patients without FGFR3 antibody and no paraneoplastic sensory neuropathy, patients with FGFR3 antibody tended to be younger (OR 0.97:0.93-1.02 95% CI; p=0.06), had more frequent trigeminal nerve involvement at onset of the neuropathy (OR 27.6:2.3-331.1 95% CI; p=0.009) and asymmetrical distribution of sensory loss at full development of the disorder (OR 8.7:1.9-41.2 95% CI; p=0.006). Comparatively to anti-Hu paraneoplastic SNN, patients with FGFR3 antibody were younger (OR 0.94:0.90-0.99 95% CI; p=0.02) and less frequently male (OR 0, 13:0.03-0.50 95% CI; p=0.003). Their neuropathy tended to have a more frequent slow progressive course (OR 7.2:0.9-62.9 95% CI; p=0.06) and less frequent ataxia (OR 0.04:0.001-1.1 95% CI; p=0.06). Interestingly none of the electrophysiological criteria used in this study (number of abolished sensory action potentials or motor nerve conduction abnormalities) discriminated patients with FGFR3 antibody from those without whether their neuropathy was paraneoplastic or not. Nine patients with FGFR3 antibody underwent a spinal tap. A mild elevation of protein concentration was observed in 3, oligoclonal bands in 3 and cell reaction in 3. As a whole the CSF was abnormal in 5/9. A nerve biopsy was available in 4 patients (1 with Sjögren syndrome and 3 without known autoimmune context) and showed fiber loss without regenerating cluster consistent with a diagnosis of SNN in all of them. In one patient without autoimmune context, nerve vasculitis consisting of CD3+T lymphocytes was present in the epineurium.

(47) 2.7. Reactivity of Sera with HEK FGFR3 Transfected Cells.

(48) Serum samples from four patients with sensory neuronopathy and anti-FGFR3 antibody detected by ELISA and four healthy blood donors were tested on transfected cells. The results are summarized in the table 2 below. Different patterns of reactivity were observed: some sera reacted with the full length protein and intracellular domain while others also reacted with TRK1 and 2. As a whole, the four sera reacted with the intracellular domain of the FGFR3 protein in concordance with the ELISA method which used the intracellular domain of the protein.

(49) TABLE-US-00002 TABLE 2 immunoreactivity of the serum of four patients with anti-FGFR3 antibody by ELISA on HEK 293 cells transfected with the full length FGFR3, the intracytoplasmic domain, TRK1 or TRK2 subunit of the intracellular domain of FGFR3. FL- intracellular sera FGFR3 domain TRK1 TRK2 Patient 1 pos pos neg neg Patient 2 pos pos neg neg Patient 3 pos pos pos pos Patient 4 neg pos neg neg Control 1 neg neg neg neg Control 2 neg neg neg neg Control 3 neg neg neg neg Control 4 neg neg neg neg pos: positive. neg: negative.

(50) 2.8. Discussion

(51) The inventors have identified FGFRs and their associated GRB proteins as potential target of IgG autoantibodies in patients with sensory neuronopathy (SNN).

(52) Elisa and protein arrays using a peptide from the intracellular kinase domain of FGFR3 (amino acids 399-806) allowed identification of a group of patients with SSN (most of them without known associated autoimmune context) harbouring anti-FGFR3 antibodies. Immunocytochemistry performed on cells transfected with the full length FGFR3, the intracellular domain (amino acids 397-806), TRK1 or TRK2 confirms that the FGFR3 anti-sera reacted with the tyrosine kinase domain, although some sera also reacted with TRK1 or TRK2.

(53) As a whole, the FGFR3 antibody associated neuropathy follows the criteria of probable SNN in 82% of cases according to published criteria in patients with available information (Camdessanche J. P. et al., Brain 2009; 132:1723-1733), the others but one having sensory neurons involvement on clinical evaluation. In some patients, the sensory neuron disorder may present as small fibre neuropathy or trigeminal nerve neuropathy. Interestingly one patient for whom the diagnosis hesitated between CIDP and SNN and another one presenting with distal sensory neuropathy can be reallocated to a diagnosis of SNN because of the presence of FGFR3 antibody.

(54) SNN may occur with lupus or Sjögren syndrome but the neuropathy frequently appears months or years before the autoimmune disorder becomes apparent while many cases of SNN never develop any autoimmune context even after a very protracted course. It is not established that when present the autoimmune disease is responsible for the neuropathy and most probably their co-occurrence results from the association of two autonomous autoimmune disorders. In this study, 47% of patients with FGFR3 antibodies have an autoimmune context including lupus or lupus anticoagulant, Sjögren syndrome, sclerodermy or unclassified autoimmune disorder. One patient had HIV infection. In the others, there was no known autoimmune context. Interestingly, in one of them a nerve biopsy disclosed vasculitis in the epinerium confirming that the disorder involves inflammatory mechanisms.

(55) Therefore detection of FGFR3 antibodies early in the course of the neuropathy leads to ascribe the neuropathy to an otherwise undetected autoimmune disorder. There are to date no known biomarkers for non paraneoplastic SNN. The availability of such a diagnostic tool allows the identification of a group of patients who are currently considered as having an idiopathic disease and who can now be candidate to receive an immunological treatment.