Renal transplantation represents currently the best therapeutic alternative for end-stage renal failure, not only in terms of patient outcomes (better quality of life and longer survival), but also in terms of costs for the society.

Progress achieved in the last 20 years has resulted in a drastic reduction of the incidence of "classic" (i.e. clinically patent) acute cellular rejection episodes.

Unfortunately, and rather unexpectedly, this progress has had hardly any effect on the frequency of the loss of kidney transplants beyond the first year, as shown by the stagnation of grafts' half lives.

Furthermore, the use of immunosuppressant combinations that are more and more powerful has an impact on adverse effects in recipients, including an increased incidence of infections, cancers, but also metabolic complications (diabetes, osteoporosis, dyslipidemia, etc.), which are cause of significant morbi-mortality.

In an attempt to improve on these disappointing outcomes, some teams have offered to perform screening biopsies: i.e. routine biopsies at specific time points during the follow up, irrespective of graft function. Their primary interest is to allow a pathological analysis of the graft at an early stage, i.e. when potential histological lesions allow for a diagnosis but before these lesions impact on graft's function. Indeed, it has been clearly demonstrated that therapeutic adjustments intended to protect the grafts are most effective when introduced early. There is a fairly broad consensus to perform these biopsies three months and one year after the transplantation. Performing screening biopsies has led to the identification of "subclinical" forms of rejection, i.e. graft infiltration by recipient immune effectors meeting the Banff histological criteria, but without increase in creatininemia.

Assuming that about 10% of screening biopsies performed at 3 months reveal a subclinical rejection, which needs to be treated, the management strategy for the remaining 90% of patients, whose biopsies show either i) a mild inflammatory infiltrates: i.e. "borderline changes", or ii) the complete absence of immune effectors in the graft is, poorly standardized.

The investigators therefore propose to conduct a prospective randomized trial to answer these questions simultaneously by evaluating a strategy to optimize the immunosuppression of renal graft recipients based on the presence or absence of subclinical intragraft inflammatory infiltrates in the screening biopsy performed at 3 months post transplantation. Patients with borderline changes (sub-study A) will be randomized to receive a treatment for rejection (corticosteroid boluses). Patients without inflammation in their graft (sub-study B) will be randomized for corticosteroid withdrawal. Impact on graft function, progression of histological lesions and incidence of morbidity will be evaluated.

Inclusion Criteria:

1. Common to both sub-studies (A and B)

   • Renal transplant patient aged between 18 and 75.
   • Patient who received a first or second renal graft
   • Immunosuppressive treatment consisting of an anti-calcineurin [cyclosporine (trough levels: 150<T0<300)], or tacrolimus (trough levels: 8<T0<12), mycophenolate mofetil and corticosteroids.
   • Patient who benefited from a screening renal biopsy 3 months after the graft
   • Patient who gave their informed consent
   • Patient affiliated to a social security scheme or being a beneficiary of such a scheme

2. Specific to sub-study A

   • Presence of "borderline" inflammatory infiltrates on the screening biopsy at 3 months as defined by the Banff classification 2013:

   • Absence of vascular lesions (v0) and:

     • tubulitis regardless of its significance (t1-3) with minimum interstitial infiltrate (i0-i1) OR
     • interstitial infiltrates (i2-3) without significant tubulitis (≤ t1)

3. Specific to sub-study B Absence of significant inflammatory infiltrates (i0-1 and t0) on the screening biopsy at 3 months

Exclusion Criteria:

1. Common to both sub-studies (A and B)

   • Histological subclinical rejection criteria on the screening biopsy at 3 months (Banff 2009: > i2+t2)
   • Donor specific antibodies in historical serum or de novo appearance during the first 3 months
   • Humoral lesions on the 3-month biopsy (Banff score g+ptc>2)
   • "Classic" acute rejection episode proven by biopsy during the first 3 months
   • Multiorgan transplantation
   • 3rd (or subsequent) renal transplantation
   • BK virus-associated nephropathy on the screening biopsy
   • Contraindication to the 1-year screening biopsy

2. Specific to sub-study B Initial nephropathy with a high risk of recurrence on corticosteroid withdrawal: segmental and focal and segmental glomerulosclerosis, lupus nephritis, vasculitis, or membranous glomerulonephritis