Introduction

Aggressive antidiabetic therapy and rapid glycemic control may result in a diabetic neuropathy called treatment-induced neuropathy of diabetes (TIND) [1, 2]. On the other hand, Charcot neuroarthropathy is a chronic, devastating and destructive disease of the bone structure and joints in patients with neuropathy. It is characterised by painful or painless bone and joint destruction in limbs that have lost sensory innervation [3, 4]; we have reported a case developed following rapid glycemic control [5]. In this project, we accessed databases of three diabetes centres, extracted a clinically significant number of cases of Charcot neuroarthropathy, and investigated whether its onset was associated with rapid glycemic control. We aimed to investigate this association for better understanding of the physiology of this complication.

Research design and methods

We retrospectively analyzed medical records of persons with diabetes and acute Charcot neuroarthropathy, who had been referred to three hospital centres (Department of Diabetes, Sud-Francilien Hospital, Corbeil-Essonnes, France; Podology Unit, Cochin Hospital, Paris, France; and Department of Diabetes, Pitie-Salpetriere Hospital, Paris) from 2008 to 2018. Patients with suspected active Charcot neuroarthropathy were referred by general practitioners, diabetologists or emergency services of these centres.

Results

A total of 44 patients diagnosed with diabetes and acute Charcot neuroarthropathy were included in the study. Patients median age was 60 years (interquartile range = 50–71 years); 29 were males and 15 females living with diabetes for 16±3,2 years. Among patients with a foot deformity, 29 had a distortion in the tarsometatarsal joints and 13 in the metatarsophalangeal joints; the remaining two had knee or ankle disease. All subjects had diabetic sensitive neuropathy; only three patients had reported a previous autonomic neuropathy diagnosed before the onset of acute Charcot neuroarthropathy.

Discussion

In our study, HbA1c levels significantly declined during the six months before the onset of acute Charcot neuroarthropathy. HbA1c levels decreased more in patients with T1DM compared to patients with T2DM (S2 Fig). Significant heterogeneity was found in the magnitude of HbA1c reduction with treatment intensification (S3 Fig). Taking a cut-off of 2 points for the decrease in HbA1c% in six months, seventeen participants (two with T1DM and fifteen with T2DM) had an HbA1c reduction equal to or higher than the cut-off.

We included 44 diabetic patients with acute Charcot neuroarthropathy, a large sample size for an unusual disorder (its prevalence in patients with diabetes is 0.1% and 0.4% [10,11]). As expected, all included subjects had diabetic sensitive neuropathy. The distribution of HbA1c reduction among the included patients was bimodal. Therefore, we used a cut-off point (HbA1c reduction of more than 2 points in six months) similar to the one used by Gibbons et al. [1, 2] to define TIND (HbA1c reduction of more than 2 points in three months) [1]. In our study, we have a similar reduction, but in over six months, the clinical presentation of CN is longer than that of TNID.

Interestingly, the prevalence of TIND reported by Gibbons et al. (up to 10%) [1, 2] is similar to ours (11.4% of treatment-associated acute Charcot neuroarthropathy).

The Charcot neuroarthropathy (CN) is a rare and devastating disorder presenting with peripheral and/or autonomic neuropathy that affects people with diabetes [3, 4], [in our study only three patients had reported a diagnosis of autonomic neuropathy]. A late diagnosis of CN can have serious consequences, including amputation. The physiopathology of this disorder is poorly known. A new series of experiments was carried on the evolution of bone modelling factors in the appearance of Charcot neuroarthropathy with the implication of the receptor activator of nuclear factor-B ligand (RANKL) and its natural antagonist, osteoprotegerin (OPG) [12,13], it has already been described the value of OPG can vary after insulin treatment in patients living with type 1 diabetes [14], this may refine the hypotheses and the results of our study, however, the RANKL value was not evaluated in that study [14].

More, a diagnosis of neuroarthropathy in particular circumstances which can be linked to rapid correction of the glycemic balance without treating the subject directly has been described, as the onset of Charcot neuroarthropathy during weight loss after bariatric surgery [15] and after kidney and pancreas transplantation [16], it is for the first time we can clearly describe an association type relationship and not a direct causality between the onset of acute CN with the rapid glycemic control, a study tackled the level of HbA1C as a predictive factor in the onset of an OAN, but without conclusive results [17], in other words, one thinks that this rapid reduction in HbA1C levels contributes to the month in a partial way to the activation of the inflammatory phenomena which are currently the triggers of OAN [18]

Study limitations include: (i) The single-group design often employed in retrospective studies limits the researchers’ ability to determine cause and effect. Although it is usually impossible to include a control group in a retrospective study, whenever possible a control group should be included to help establish the cause-and-effect relationship; (ii) data extraction was centre-based, potentially ignoring site-specific factors.

In conclusion, HbA1c levels significantly declined during the six months preceding the onset of acute Charcot neuroarthropathy. This prevalence is similar to the prevalence of up to 10% reported by Gibbons et al. [1, 2] for TIND. Our results suggest that rapid glycemic control in chronically unbalanced patients with diabetic neuropathy is an associated factor for acute Charcot neuroarthropathy.

Supporting information

Acknowledgments

The authors acknowledge Ricardo P. Garay (Craven, Villemoisson-sur-Orge, France) for editorial support.