The WHO (World Health Organisation) estimated the prevalence of diabetes to be 422 million people in 2014, compared to 108 million in 1980. This has led to an increasing number of diabetic patients referred to nephrologists for diagnostic purposes. Diabetic nephropathy is the most common renal disease in this population and is usually a presumptive diagnosis based on clinical and biological features although microscopic examination of a renal sample acquired through renal biopsy is the only way to be certain of this diagnosis. However, kidney biopsy is an invasive procedure carrying a low but incontestable risk of adverse event such as post-procedural pain and bleeding. Consequently, nephrologist around the world feel that renal biopsy should only be performed in patients with type 2 diabetes to detect non-diabetic renal disease, when the diagnosis of diabetic nephropathy is dubious or unlikely. This likeliness is based on the presence or absence of typical feature such as diabetic retinopathy, hematuria, progressive decline of renal function or increase of proteinuria, long duration of diabetes, nephrotic syndrome. These feature were identified by the comparison of patients with type 2 diabetes and non-diabetic renal disease (alone or associated to diabetic nephropathy) and isolated diabetic nephropathy.

However, it is not known if the presence (or absence) of these atypical features by themselves are indeed signs of non-diabetic renal disease and necessitate to perform renal biopsy. The aim of the study is to determine if these atypical features are relevant indications to perform renal biopsy. To answer this question, will be analyze the medical records of patients with type 2 diabetes who underwent renal biopsy in five French nephrology center to determine, in each case, the indication of the biopsy and if this latter benefitted the patients.

In addition, will be evaluate the prognosis value of the Renal Pathology Society classification of diabetic nephropathy in patients with type 2 diabetes and diabetic nephropathy.

The purpose of the study is to evaluate the accuracy of kidney biopsies indications in patients with type 2 diabetes to diagnose non-diabetic renal disease based on canonical atypical features (Absence of diabetic retinopathy, Low or rapidly decreasing GFR, Rapidly increasing proteinuria or nephrotic syndrome and Presence of active urinary sediment. This study is a retrospective observational case only study, recruiting patients over 18 years old with type 2 diabetes who underwent kidney biopsy in five French nephrology centers between 2006 and 2015.

Will be collected demographical, clinical and biological data at the time of the renal biopsy and at the last follow-up from the patients' medical charts.

Indications for renal biopsy will be categorized as

1. Atypical feature of etiological significance in the presence of any atypical feature not listed below (including acute kidney injury as defined by the stage 1 of KDIG guidelines)
2. Brutal nephrotic syndrome in the absence of the above criterion
3. Rapid decline of GFR (defined as 50 % eGFR decline over >1 week but < 1 year) in the absence of criteria 1-2
4. Absence of proteinuria in the absence of criteria 1-3
5. Rapid increase of proteinuria in the absence of criteria 1-4
6. Presence of hematuria in the absence of criteria 1-5
7. Absence of diabetic retinopathy in the absence of criteria 1-6 Will be evaluate the actual probability to reach a non-diabetic renal disease for each of these indications.

Will be also perform a classical analysis by assessing the association of clinical and biological feature such as HbA1c, diabetes duration, absence of diabetic retinopathy,… with the presence of a non-diabetic renal disease.

In addition, the biopsy samples will be scored according to the RPS diabetic nephropathy classification to evaluate the prognostic value of this classification.

Inclusion Criteria:

• Patients with type 2 diabetes
• Over 18 year old
• With a first non-transplant renal biopsy for non-tumoral indication between 01/01/2006 and 12/31/2015

Exclusion Criteria:

• Patients having expressed their opposition regarding the usage of their data in this research
• Patients with previous renal biopsy
• Patients with non-type 2 diabetes (type 1, post-pancreatitis, genetic, etc.)

All patients with type 2 diabetes who underwent a first renal biopsy in one of the participating centers (nephrology unit of Pontchaillou Hospital in Rennes, Conception hospital in Marseille, Bichat and Necker hospitals in Paris and Bretonneau hospital in Tours) will be included in the study unless they meet one of the exclusion criteria.