Prevalence and Spectrum of Neuropathies in a Cohort of 585 Patients With Immunoglobulin A Monoclonal Gammopathy

ABSTRACT

Background

Neuropathies associated with IgA monoclonal gammopathy are poorly understood, and the interpretation of the presence of such gammopathy in a patient with neuropathy may be challenging.

Methods

The neurological and hematological features of all patients newly diagnosed with IgA gammopathy by immunofixation in our center from January 2016 to December 2020 were retrospectively analyzed. Patients with neuropathy were identified through the medical records. The etiology was reviewed by two neurologists and classified into three groups: (i) IgA‐related neuropathies, (ii) IgA‐unrelated neuropathies with an identified alternative etiology, and (iii) neuropathies of uncertain relationship with IgA (NURIA) based on a negative extensive work‐up.

Results

Among 585 patients with IgA gammopathy, 79 had neuropathy (14%). Neuropathy was IgA‐related in 10 patients (13%): eight AL amyloidosis and two POEMS. In this group, the core features were neuropathic pain, autonomic dysfunction, fatigue or weight loss, and a lambda light chain. IgA‐unrelated neuropathies were more frequent (N = 64, 81%), encompassing mainly chemotherapy‐induced (N = 34) and diabetic (N = 15) neuropathies. Five patients (6%) were classified as NURIA: four had mild sensory‐predominant length‐dependent axonal neuropathy, and one had severe, progressive, motor‐predominant axonal neuropathy.

Conclusions

In patients with IgA gammopathy, neuropathies have a low prevalence and a wide etiological spectrum. AL amyloidosis and POEMS syndrome are rare but crucial to identify, as disease‐modifying treatments are available. Future studies should help better characterize the rare cases of neuropathy with an uncertain relationship to IgA gammopathy.

1. Introduction

Interpreting the presence of gammopathy in patients with neuropathy may be challenging. While the association of neuropathy with IgM gammopathy has been extensively studied in the past decade [1], little is known about the association with IgA gammopathy. Only a few case series have been reported more than 20 years ago, often grouping IgA and IgG gammopathies [2, 3, 4]. Thus, we aim to explore the frequency, clinical presentation, and etiologic distribution of peripheral neuropathies in a cohort of IgA gammopathies from a single French center.

2. Methods

This study was approved by the Institutional Review Board of Henri Mondor University Hospital (IRB #00011558). All patients who had newly diagnosed IgA monoclonal gammopathy by serum immunofixation in our center from January 2016 to December 2020 (Hydrasys 2 scan focusing, Sebia, France) were retrospectively included. Patients were identified from the exhaustive register of the Biology Department. Patients with multiple clonal gammopathies were included. Electronic medical records were reviewed by two peripheral nerve specialists (VP, TG) to collect demographic, clinical, and investigational data. Neuropathy was suspected when both clinical suggestive signs and symptoms were documented. Compressive mononeuropathy or radiculopathy were not included. Nerve conduction studies (NCS) were not needed to confirm typical diabetic sensory polyneuropathy or toxic distal neuropathy. In all other cases, large fiber neuropathy was confirmed on NCS and small fiber neuropathy on specific assessments (laser‐evoked potentials, warm detection thresholds, or electrochemical skin conductance) with autonomic dysfunction based on clinical assessment [5]. Patients were classified into three groups depending on the underlying neuropathy etiology: (i) IgA‐related neuropathies when a pathophysiological link between the neuropathy and gammopathy was proven (AL amyloidosis with amyloid deposition on a biopsy or POEMS syndrome using the international diagnostic criteria [6]), (ii) IgA‐unrelated neuropathies when an alternate cause was identified, including drug toxicity, diabetes mellitus, vitamin deficiency, chronic kidney disease, alcoholism, genetic disease, and vasculitis without any evidence of IgA involvement, and (iii) neuropathies of uncertain relationship with IgA (NURIA) when neuropathy remained of unknown cause after an extensive work‐up conducted by a peripheral nerve specialist.

3. Results

Six hundred and forty‐seven patients were diagnosed with IgA gammopathy. Of those, 62 patients were excluded due to missing data, and 585 patients were analyzed. The detailed flowchart is represented in Figure 1.

FIGURE 1.

Flowchart of the cohort and the etiologic distribution of neuropathies. *Five patients had mixed etiologies of neuropathy. **Etiology for IgA‐unrelated vasculitis were type II cryoglobulinemia (n = 1), eosinophilic granulomatosis with polyangiitis (n = 1) and periarteritis nodosa (n = 1). NURIA, neuropathy of uncertain relationship with IgA; POEMS, polyneuropathy organomegaly endocrinopathy monoclonal gammopathy skin changes; vATTR, hereditary transthyretin amyloidosis.

After data analysis, 79 patients (13.5%) had neuropathy: 47 were male (59.5%), the median age at onset of gammopathy was 65 years [IQR 59.5–73], and the mean follow‐up time from IgA diagnosis was 4 years (range 0–16). Gammopathy was associated with a lambda light chain in 50.6%. A myelogram was performed on 52 patients (10 with IgA‐related neuropathy, 42 with IgA‐unrelated neuropathy). Four patients had a bone marrow biopsy (one with POEMS syndrome, one with NURIA, and two with IgA‐unrelated neuropathy). Six patients had additional clonal IgG gammopathy (one patient with AL amyloidosis, four with an IgA‐unrelated cause, and one with NURIA).

Neuropathy was IgA‐related in 10 patients (12.7% of the neuropathy cohort and 1.7% of the entire cohort) including eight cases of AL amyloidosis and two cases of POEMS syndrome. The main clinical and hematological findings in these patients are shown in Table 1. IgA gammopathy was known prior to the onset of neuropathy in six patients. IgA was associated with a lambda light chain in all 10 patients. Median quantification of the monoclonal peak was 8.2 g/L [IQR 3.4–10.4]. Neuropathic pain and autonomic dysfunction, with mainly orthostatic hypotension and gastrointestinal involvement, were both reported in 70% of the patients. Eighty percent of patients experienced fatigue or weight loss (> 5% of usual weight or 5 kg over 6 months). AL amyloidosis was diagnosed in five patients through salivary gland biopsies and in three patients through endomyocardial biopsies. Diagnostic confirmation required more than one tissue sample in five cases and up to three tissue samples in three of these cases. No nerve biopsy was needed.

TABLE 1.

Clinical features of the 10 patients with an IgA‐related neuropathy.

Patient	Gender	IgA lambda level (g/L)	Age at IgA diagnosis	Follow‐up time from IgA diagnosis (years)	Hematological condition	Bone marrow plasma cells on initial myelogram	IgA detection relative to neuropathy diagnosis
1	F	8,8	57	15	AL amyloidosis then multiple myeloma	7%	Before
2	M	7,2	54	2	AL amyloidosis and multiple myeloma	23%	Before
3	F	5,4	73	1	AL amyloidosis	3%	During
4	F	10,2	51	4	AL amyloidosis and multiple myeloma	NA	Before
5 a	M	2	66	2	AL amyloidosis	7%	During
6	M	NQ	70	2	AL amyloidosis and multiple myeloma	8%	Before
7	F	16	53	6	AL amyloidosis and multiple myeloma	36%	Before
8	M	10,4	51	5	AL amyloidosis	5%	During
9	F	3,4	66	6	MGUS then POEMS syndrome	3%	Before
10	F	18,6	61	2	POEMS syndrome	8%	During

Patient	Age of neuropathy symptoms onset	Neuropathy cause	Diagnosis confirmation	Neurological phenotype	Impaired general condition c
1	Late 60s	AL amyloidosis	Endomyocardial biopsy	Painful small fiber neuropathy with autonomic dysfunction	Yes
2	Early 50s	AL amyloidosis	Endomyocardial biopsy	Painful small fiber neuropathy with autonomic dysfunction	Yes
3	Early 70s	AL amyloidosis	Salivary gland biopsy	Small fiber neuropathy with autonomic dysfunction	No
4	NA	AL amyloidosis	Salivary gland biopsy	Small fiber neuropathy with autonomic dysfunction	Yes
5 a	Early 60s	AL amyloidosis	Salivary gland biopsy	Painful sensorimotor non‐length‐dependent axonal neuropathy with autonomic dysfunction	No
6	Early 70s	AL amyloidosis	Salivary gland biopsy	Sensorimotor non‐length‐dependent axonal neuropathy with autonomic dysfunction	Yes
7	Late 50s	AL amyloidosis	Endomyocardial biopsy	Painful sensory predominant axonal length‐dependent neuropathy with autonomic dysfunction	Yes
8	Early 50s	AL amyloidosis	Salivary gland biopsy	Painful sensory predominant axonal length‐dependent neuropathy	Yes
9	Late 60s	POEMS	Fulfilled criteria b	Painful sensorimotor non‐length‐dependent demyelinating neuropathy	Yes
10	Early 60s	POEMS	Fulfilled criteria b	Painful sensorimotor non‐length‐dependent demyelinating neuropathy	Yes

Abbreviations: AAO, age at onset; diag, diagnosis; F, female; M, male; MGUS, monoclonal gammopathy of unknown significance; NA, non available; NQ, non‐quantifiable.

^a Patient #5 had both IgA lambda monoclonal gammopathy and IgG lambda monoclonal gammopathy.

^b POEMS was confirmed according to the diagnostic criteria in Patients 9 and 10⁶: Patient 9 had two mandatory major criteria, two other major criteria, and two minor criteria. Patient 10 had two mandatory major criteria and one other major criterion.

^c Impaired general condition includes fatigue and involuntary weight loss defined by more than 5% of the usual weight or more than 5 kg in 6 months.

Neuropathy was IgA‐unrelated in 64 patients (81% of the neuropathy cohort). Among these, 34 cases were chemotherapy‐induced, particularly in the context of multiple myeloma, while 15 were due to diabetes. Five patients had mixed etiologies (Figure 1). Nerve biopsy was performed for two cases of vasculitis.

Finally, five patients were classified as NURIA (6.3% of the neuropathy cohort). Of these, three had a kappa light chain associated with IgA. Four patients presented with an axonal length‐dependent sensory‐predominant polyneuropathy. Neuropathy in these cases was slowly progressive over the years (mean follow‐up of 5.8 years [range 1–12]). In contrast, the fifth patient presented with severe, progressive sensorimotor axonal neuropathy. An extensive work‐up was negative, including an autoimmune panel (antiganglioside, onconeural, and anti‐nodal/paranodal antibodies), CSF analysis, VEGF testing, and TTR gene sequencing. A nerve biopsy revealed no amyloid or IgA deposition. The patient did not respond to intravenous immunoglobulin therapy or corticosteroids. No alternative etiology has been identified to date.

4. Discussion

In this work, we estimated the prevalence of neuropathy at 13.5% in a cohort of more than 500 patients with newly diagnosed IgA gammopathy over a 5‐year period. However, neuropathy was rarely attributed to the presence of IgA (1.7% of the whole cohort).

The estimated rate of neuropathy is congruent with the literature [7], though previous reports were based on smaller series. Using a similar approach and from the same registry, our team has previously studied the prevalence of neuropathies in a cohort of 550 patients with IgM [1]. There were 115 cases of neuropathy (20.9%) of which 83 were IgM‐related (15.1%). These rates were significantly higher than those from our IgA gammopathy cohort. This difference may be explained by the occurrence of specific immune IgM‐related neuropathies, such as neuropathy associated with anti‐MAG antibodies or CANOMAD syndrome.

This present study offers valuable insights into the various etiologies of neuropathy in patients with IgA gammopathy. IgA‐related neuropathies were rare, representing less than 2% of the entire cohort. In contrast, unrelated causes were prevalent with chemotherapy‐induced neuropathy being the most frequent condition, primarily due to the neurotoxicity of multiple myeloma treatment. Other common etiologies of IgA‐unrelated neuropathies, such as diabetes, alcoholism, or vitamin deficiency, should not be neglected.

Although IgA‐related neuropathies were found to be rare, they are life‐threatening conditions that are difficult to diagnose, yet treatments are available [6, 8]. In the case of AL amyloidosis, the diagnostic delay may be partly due to the limited sensitivity in detecting amyloid deposition [9] as illustrated in this study with two‐thirds of patients requiring multiple biopsies. We went on to identify specific features in this group. First, all patients with IgA‐related neuropathies had a lambda light chain, which is consistent with previous research in both AL Amyloidosis and POEMS [10, 11]. In addition, patients with IgA‐related neuropathy had frequent neuropathic pain, autonomic dysfunction, and fatigue or weight loss, which aligns with previous reports [6, 10]. This work underscores the importance of these biological and clinical red flags, highlighting the need for close follow‐up and repeated investigations in these patients.

In clinical practice, the interpretation of the relationship between monoclonal gammopathy and neuropathy is difficult when no alternative etiology is found. Kelly et al. showed in 1981 [12] that in a cohort of patients with peripheral neuropathy of unknown cause, the presence of monoclonal gammopathy of unknown significance (MGUS) was 6–10 times more frequent than in the general population, revealing a possible link between both entities. A more recent wide population‐based study found peripheral neuropathy to be 2.7 times more frequent in patients with MGUS than without [13]. However, this included IgM gammopathy, and it lacked clinical data with possible undiagnosed AL amyloidosis or POEMS syndrome. Furthermore, both the incidence of neuropathy and monoclonal gammopathy increase with age, possibly making their co‐occurrence incidental in the elderly population. In this work, these rare cases (five patients in the entire cohort) were grouped under the term NURIA, which may represent multiple etiologies. Most patients had a homogeneous presentation characterized by late‐onset, non‐ or slowly progressive, sensory‐predominant, length‐dependent axonal neuropathy. This phenotype matches chronic idiopathic axonal polyneuropathy (CIAP) [14] which has an estimated prevalence of 4%–8% in the general population. These considerations challenge the significance of monoclonal gammopathy in neuropathies of unknown cause, and consequently, recently identified genetic causes and risk factors should be explored [15].

This study has several limitations. The primary limitation is its retrospective design, which depends on medical records. Clinical data may be incomplete for patients with neuropathy who were not evaluated by a neurologist. Additionally, there is a risk that our hospital‐based study design underestimated pauci‐symptomatic neuropathies. Furthermore, a direct comparative analysis with neuropathies related to other types of gammopathy was not performed. However, by utilizing a cohort from a hematology lab, our study mitigates the typical sampling bias associated with neurology centers, thereby highlighting the rare occurrence of neuropathy within a large population of patients with IgA gammopathy.

This study also delineates the broad spectrum of neuropathy etiologies in these patients, noting the predominance of common IgA‐unrelated causes. Future studies should focus on improving the identification of rare but treatable IgA‐related neuropathies. Finally, a better understanding of the pathophysiological factors and clinical progression of NURIA is essential.

Author Contributions

Valentine Perrain: conceptualization, investigation, writing – original draft, writing – review and editing, formal analysis. Valérie Molinier‐Frenkel: resources, formal analysis, validation, data curation. Jehan Dupuis: validation, resources. Alain Créange: validation, resources. Jean‐Pascal Lefaucheur: supervision, resources, validation. Thierry Gendre: conceptualization, investigation, writing – review and editing, supervision, methodology, validation.

Conflicts of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.