Maternally inherited diabetes and deafness complicated by mesangial galactose-deficient IgA1 deposits: a case report

Maternally inherited diabetes and deafness (MIDD) result from genetic abnormalities in mitochondrial DNA, with an A to G substitution at position 3243 (m.3243A > G) being the most common point mutation [1]. Kidneys are susceptible to mitochondrial dysfunction, and MIDD patients have a high incidence of end-stage renal disease. The prevalence of MIDD is high in Japan, and MIDD accounts for 0.9 to 5.9% of patients with diabetes who receive dialysis [2, 3].

Immunoglobulin (Ig)A nephropathy (IgAN) is the most common cause of glomerulonephritis worldwide. Previous multi-ethnic genome-wide association studies have identified risk loci resulting in a predisposition to IgAN [4‐7]. Furthermore, although IgAN is prevalent worldwide, Japanese individuals carry a genetic predisposition to IgAN [6]. Therefore, nephrologists could encounter patients with MIDD, which is complicated by IgAN. However, diagnosing MIDD in patients with IgAN is challenging, particularly when renal manifestations precede diabetes or deafness. Here, we demonstrate the clinical and histological course of MIDD, complicated by IgAN, through repeated renal biopsies.

Our case demonstrates the difficulties in diagnosing mitochondrial nephropathy during the early stages of the disease, especially when it is complicated by other glomerular diseases and lacks the clinical key features such as diabetes and deafness.

An A to G substitution at position 3243 (m.3243A > G) of mitochondrial DNA affects the mitochondrial tRNA^Leu tertiary structure and leads to defects in the activities of complexes 1, and 4 of the respiratory chain within the mitochondria [10, 11]. Therefore, MIDD typically affects metabolically active organs such as the endocrine pancreas and cochlea, and in some cases, also the retina, muscles, kidneys, and brain. Renal manifestation sometimes precedes the diagnosis of either diabetes or deafness and can even be the sole manifestation of MIDD [12‐14]. Proteinuria is a common presentation of the disease.

Focal segmental glomerular sclerotic (FSGS) lesions or tubular damage complicated by mitochondrial cytopathies are prevalent findings in the renal biopsy specimens of patients with MIDD. Electron microscopy findings of abundant and morphologically abnormal mitochondria in the cytoplasm can facilitate diagnosing mitochondrial nephropathy. However, heteroplasmy associated with mitochondrial DNA mutations often hampers the diagnosis of mitochondrial diseases through sampling errors.

The utility of GSECs for diagnosing MIDD was first described by Kobayashi et al. in 2010 who reported that GSECs were frequently observed in the collecting ducts or tubules of the kidneys of patients with MIDD [15]. The ultrastructure of GSECs is defined by abnormally shaped or increased numbers of mitochondria per cell. GSECs are observed not only in genetic mitochondrial disorders but also in miscellaneous conditions, including ageing, low birth weight, various drug toxicities, and heavy metal poisoning [16, 17]. In our case, neither FSGS lesions nor tubular damages were obvious in the first biopsy specimen. Mesangial widening and IgA deposition led us to diagnose the patient with IgAN at the time, as this also explained her persistent proteinuria. GSECs were observed in the collecting ducts in the first biopsy specimen, but the specimens for electron microscopy did not contain cells with abnormally shaped mitochondria; moreover, diabetes and deafness were not evident at the time. Six years after her initial renal biopsy, the patient underwent a second biopsy for persisting proteinuria. This second biopsy revealed a disappearance of mesangial IgA deposits with increasing numbers of global scleroses. No FSGS lesions were evident. Moderate tubular damage was observed, and there were fewer GSECs when compared to the first biopsy. The disappearance of IgA deposits in patients with IgAN has been occasionally observed after immunosuppressive treatment alone, or with tonsillectomy [18‐20]. The patient in our case had undergone renin-angiotensin system (RAS) blockade therapy and tonsillectomy. Several recent studies reported that patients with IgAN had significantly higher levels of serum IgA1 levels with galactose-deficient O-linked glycan moieties in the hinge region (Gd-IgA1) when compared to healthy individuals. Furthermore, Gd-IgA1 has been reported to be the predominant variant of IgA1 in the mesangium of patients with IgAN [21, 22]. Using a newly developed antibody, we also confirmed that the IgA deposits in our patient contained Gd-IgA1 [9]. The origin of Gd-IgA1 has yet to be determined, but some reports suggested that the mucosal immune systems, including the tonsils, are a potential source of Gd-IgA1 [23, 24]. In this report, we could not address how serum Gd-IgA1 levels of our patient changed after tonsillectomy due to a lack of serum sample at the first biopsy. Further studies are needed to clarify whether tonsillectomy can contribute to a decrease in Gd-IgA1 production and its disappearance from the glomeruli.

In a literature search, we found four case reports (three from Japan) of mitochondrial nephropathy complicated by glomerular IgA deposition [25‐28]. All patients underwent renal biopsies for persistent proteinuria before a diagnosis of mitochondrial disease was made. According to previous studies, IgA deposition was observed in 4 to 10% of consecutive necropsies without clinical evidence of renal disease [29‐31] and 10 to 30% of renal allografts at transplantation [32‐34]. Therefore, IgA deposition can occur in the general population without clinical signs of renal disease. All previous cases discussed the possibility of mitochondrial nephropathies being accompanied by IgA deposits. It is difficult to clearly differentiate latent IgA deposits from IgAN when the condition is complicated by other renal pathologies. Haematuria or histological findings such as mesangial cell proliferation or C3 deposition might be helpful to distinguish between the two conditions. In our case, haematuria was negative, and the pathological changes were not severe enough to explain the cause of proteinuria at the first biopsy. However, our immunofluorescence analysis using the anti-Gd-IgA1 antibody supported the possibility of an IgAN diagnosis.

In this case, we illustrate the clinical and histological course of MIDD complicated by IgAN through repeated renal biopsies. Moreover, using a newly developed antibody against Gd-IgA1 revealed that the mesangial IgA deposits consisted of Gd-IgA1. This antibody might be useful to distinguish IgAN from latent IgA deposits.