Steroid-sparing immunosuppressive regimens are widely used after kidney transplantation as clinicians seek to minimize steroid-related adverse events.1 Rapid steroid discontinuation lowers the rates of cataracts, vascular necrosis and cytomegalovirus infection2 and may improve growth in children.3 Cardiovascular benefits are less well established, although positive effects have been observed for lipid profile and weight gain.4,5 Generally, it is difficult to demonstrate differences in randomized trials in unselected populations, and studies of steroid minimization regimens have often not reported steroid-related side effects, or were underpowered to demonstrate a significant difference. However, even if data remain unconvincing about reduced cardiovascular risk, the non-cardiovascular side effects of chronic steroid therapy require costly management6 and some, such as weight gain, can adversely affect patient compliance7,8 with potential consequences for graft outcomes.

Strategies for steroid minimization comprise either ‘steroid avoidance’ (or ‘steroid-free’) regimens, which will be defined here as no intravenous (i.v.) or oral steroids after the first 1–2 weeks post-transplant, or steroid withdrawal, i.e. steroid discontinuation after weeks 1–2.4,9,10

Steroid avoidance protocols have become increasingly popular in recent years. Steroid-free immunosuppression without induction therapy has been attempted after kidney transplantation using a calcineurin inhibitor (CNI)-based maintenance regimen, but randomized trials showed acute rejection to be significantly more frequent4,11,12 or more severe13 than with standard steroids. A series of randomized trials comparing steroid avoidance versus ongoing steroid therapy in selected populations receiving interleukin-2 receptor (IL-2R) induction have reported mixed results.14–18 High rates of biopsy-proven acute rejection (BPAR) were observed in the FREEDOM study for patients given basiliximab induction with cyclosporine (CsA), mycophenolic acid (MPA) and either a single i.v. dose of steroids (31.5% BPAR at 1 year) or steroids to day 7 (26.1%).19 Other researchers have found no significant effect on BPAR using steroid avoidance with IL-2R antagonist induction, especially with concomitant tacrolimus,15,16 although BPAR may occur earlier.16

Many centers, particularly in the US, instead employ lymphocyte-depleting induction when steroid-free therapy is planned, an approach that may be partly based on experience from conventional immunosuppressive regimens. Randomized trials of patients at low or moderate immunological risk given standard steroids have shown comparable rejection rates using induction with an IL-2R antagonist or rabbit antithymocyte globulin (rATG),19–22 but for high-immunologic risk populations both the overall rejection rate and the incidence of steroid-resistant rejection is lower with rATG.23,24 This may have encouraged more frequent selection of rATG or another depleting agent when attempting steroid avoidance: almost 90% of kidney transplantation patients who are discharged on a steroid-free regimen in the US receive a lymphocyte-depleting induction agent,25 with the remaining 10% given IL-2R antagonist induction.

Steroid withdrawal after the first week post-transplant tends to be regarded with more caution. A meta-analysis of eight randomized trials found that steroid withdrawal after week 1 significantly increased both acute rejection and graft loss.26 In a further meta-analysis, in which only those trials that included CNI therapy with mycophenolate mofetil (MMF) were analyzed, graft loss was minimized, but the risk of acute rejection was still higher following steroid withdrawal.27 The Kidney Disease: Improving Global Outcomes (KDIGO) recommendations suggest that in low-risk patients given induction therapy, steroids can be discontinued during week 1.28 If steroids are administered beyond the first week post-transplant, however, KDIGO suggests that they be maintained thereafter.28 This may, however, be an oversimplification. A recent Cochrane systematic review concluded that although late (month 3–6) steroid withdrawal is associated with increased rejection rates in CsA-treated patients (risk ratio 1.61, p=0.0018), this does not appear to be the case when tacrolimus is given with MMF (risk ratio 0.82, p=0.29).9,10,29,30 Equally, the safety of steroid withdrawal may be affected by the choice of induction therapy.

This article reviews the evidence relating to lymphocyte-depleting induction with rATG or alemtuzumab to support steroid avoidance or withdrawal in kidney transplant recipients. It should be borne in mind that alemtuzumab in solid organ transplantation is used off-label, and the indication is not likely to be forthcoming.

So far, head-to-head comparisons of steroid withdrawal with or without lymphocyte-depleting induction therapy have not been undertaken. Typically, studies of steroid withdrawal either before or after the first year post-transplant have not included induction therapy.4,9,10

One large study of late (month 3) steroid withdrawal allowed lymphocyte-depleting induction according to local practice. A third of patients (34% and 33% in the steroid-withdrawal and control groups, respectively) received induction, but this included both rATG and monomurab (OKT3).51 In this subgroup of 169 patients, the 12-month incidence of BPAR was 14% without steroids compared to 17% in the steroids group, i.e. steroid withdrawal was successfully achieved. In contrast, for the 331 patients without any induction therapy, BPAR was higher in the steroid-withdrawal patients (29% versus 15% in controls). Although only a subanalysis, these data suggest an advantage for lymphocyte depleting induction when undertaking steroid withdrawal after kidney transplantation.

In the SPIESSER study, 150 moderate-risk or low-risk kidney transplant patients underwent steroid withdrawal at the end of month 5 post-transplant.49 All patients received rATG induction.52 The intervention group was randomized to sirolimus and MMF in a CNI-free regimen, while the control group received CsA and MMF. In both arms, the 12-month incidence of BPAR was low (14% with sirolimus/MMF, 9% with CsA/MMF) and more than 80% of patients remained steroid-free at month 12. Steroid withdrawal around month 6 thus appears feasible under rATG induction with or without CNI therapy, when high-risk patients are excluded.

Steroid-free immunosuppression is less frequently attempted in high-risk patients, and data are limited (Table 3). Two retrospective single-center studies in African-American populations have assessed outcomes following rATG induction with i.v. steroids to day 4, both with a maintenance regimen of tacrolimus combined with either sirolimus or MMF.43,53 The one-year rates of BPAR in both series were acceptable for this high-risk group (13–14%), and graft and patient survival was excellent, but firm conclusions cannot be drawn in the absence of prospective trials. One small randomized trial (n=21) selectively recruited high-immunological risk patients (PRA >20% or repeat transplantation).55 The treatment regimens comprised alemtuzumab (a single dose of 30mg) with tacrolimus and no steroids, or rATG (total dose 6mg/kg) with tacrolimus, MMF, and steroids to day 5. In this small series, the rate of BPAR at a median of 12.4 months was 18.2% with alemtuzumab and 37.5% with rATG, while graft survival rates were 85.7% and 87.5%, respectively.55 These relatively high rates of rejection suggest that steroid avoidance may not be a favorable option in high-risk individuals unless there is a specific clinical imperative. None of these studies employed Luminex® technology to define the immunological risk pre-transplant, however, which makes any conclusion uncertain.

The advantages of steroid avoidance also have to be balanced against the risk of recurrence of certain glomerulopathies. In particular, several retrospective studies have shown that recurrence of IgA nephropathy is reduced in the presence of long-term steroid therapy. Among them, Clayton et al.58 analyzed data from the Australia and New Zealand Dialysis and Transplant Registry (ANZDATA) from 1521 adult recipients of a first kidney transplant in whom kidney failure was due to biopsy-proven IgA nephropathy. After adjustment for various confounding variables, steroid use was strongly associated with a reduced risk of IgA nephropathy recurrence (HR 0.50, 95% CI 0.30–0.84). Only 9% of patients in the study population received depleting induction therapy. Kukla et al. observed that steroid avoidance may be associated with a higher rate of recurrent glomerulopathy but found no apparent increase in the risk of graft loss in patients treated with rATG induction coupled with maintenance therapy comprising tacrolimus and an antimetabolite.59 Von Visger et al.60 recently showed similar results in a cohort of 124 patients, in which 91% of patients in the steroid-free group received lymphocyte-depleting induction. Although the mechanisms involved have not been elucidated, steroid avoidance would not appear advisable in kidney transplant patients in whom kidney failure was caused by IgA nephropathy.

In other types of glomerular diseases with risk of recurrence, such as focal segmental glomerulosclerosis, membranoproliferative glomerulonephritis or lupus nephritis, for which steroids are used as part of the treatment strategy in native kidneys, some centers tend to maintain steroids in the immunosuppressive protocol after transplantation even in the absence of relevant studies.

Steroid-sparing strategies are of particular interest in children. Randomized trials of steroid avoidance in pediatric populations under lymphocyte-depleting induction have not, however, been carried out to date. In a single-center prospective study, Chavers et al. used rATG induction at a relatively high dose (1.5mg/kg for between five and seven doses) with i.v. steroids to day 5, with CsA and MMF as maintenance therapy3 (Table 3). The rate of BPAR (19% at two years) was acceptable and, importantly, children given the rATG/steroid-free regimen showed significantly improved growth compared to matched controls given oral steroid therapy (mean standard height deviation score −0.9 versus −1.9 in controls; p=0.003). Other steroid-related effects (body mass index, lipid profile, blood pressure) were unaffected although leukopenia was reduced (p=0.02) and there were fewer Epstein-Barr infections in the steroid-free arm (p=0.04).3 Li et al., in a single-center retrospective study, achieved a 0% rate of BPAR at a mean of 13 months post-transplant when a relatively high dose of rATG was employed (total dose 9mg/kg) with no oral steroids, tacrolimus and MMF, but growth data were not reported and there was no comparison versus a steroid-treated cohort. Lastly, a retrospective single-center study by Wittenhagen et al. has described a low rate of BPAR (9% at one year) in children given lymphocyte-depleting induction (primarily rATG) with different maintenance regimens.57 Use of depleting induction in a steroid-avoidance regimen using standard maintenance therapy seems adequate to prevent rejection in children following kidney transplantation, and merits a more rigorous examination.

A small number of trials have investigated novel immunosuppressive regimens that avoid both steroids and CNI therapy.61–63 Ferguson and colleagues randomized 89 kidney transplant recipients at low or moderate immunological risk to one of three groups: the costimulation blocker belatacept with MMF, belatacept with sirolimus, or tacrolimus with MMF.62 All patients received rATG induction (total dose 6mg/kg) with i.v. steroids to day 4 post-transplant and no oral steroids. At one year, rates of BPAR were acceptable with belatacept and MMF (14%), but lower and similar in the belatacept/sirolimus group (4%) and the control arm (3%). More patients in the conventional tacrolimus/MMF group remained steroid-free at month 12, however (93% versus 73% or 77% in the belatacept group). So while promising, a CNI-free strategy may not be optimal for achieving steroid-free immunosuppression even with lymphocyte-depleting induction. Other small trials have also pointed to low or acceptable rates of BPAR when alemtuzumab induction is used to support CNI-free and steroid-free therapy,61,63 one of which achieved highly minimized immunosuppression by giving bone marrow infusions,63 but such regimens are not currently used in routine practice.

The literature contains few direct comparisons of steroid minimization regimens versus conventional steroid treatment in kidney transplant patients receiving lymphocyte-depleting induction. Studies are typically relatively small, with limited follow-up periods, and to date, the occurrence of de novo antibodies and antibody-mediated rejection is undocumented. Nevertheless, the available data show that the risk of acute rejection is low when rATG induction is given to support steroid-free regimens after kidney transplantation, at least for the first one to two years post-transplant. In the US, it is common practice to use rATG and stop steroids by day 7, and this appears to be a reasonable approach that does not increase the risk of rejection. In contrast, rejection may be more likely under IL-2R antagonist induction using a steroid-avoidance regimen. Induction with alemtuzumab appears effective in supporting steroid-free immunosuppression, although it is not licensed for use in solid organ transplant recipients. For patients at high immunological risk, the limited data available suggest that steroid minimization may not be advisable even with depleting induction therapy. The availability of single-antigen bead technology, permitting more precise measurement of anti-HLA antibodies, may enable more accurate identification of high-risk individuals and feasibly, expand the pool of patients considered candidates for steroid minimization.

In children, non-randomized studies indicate that depleting induction with standard maintenance therapy effectively prevents acute rejection with steroid avoidance, an encouraging finding given the growth inhibition associated with chronic steroid administration. More ambitious regimens have been attempted which include both steroid avoidance and CNI avoidance, of which depleting induction with maintenance therapy comprising belatacept and sirolimus appears the most promising, but currently can be regarded as an experimental approach.

There is some suggestion that steroid-associated side effects such as dyslipidemia and weight gain are lower with steroid-free immunosuppression, but this has not been shown to translate to an improvement in hard clinical endpoints. Given the lack of strong evidence in this field, no strong conclusions can be made.

Early steroid therapy with subsequent withdrawal (i.e. later than the first one or two weeks post-transplant) may be possible without additional risk of BPAR in patients who have received rATG induction, but while encouraging the data are too sparse for firm conclusions. Where late steroid withdrawal is attempted (months 3 to 6 post-transplant), it seems preferable to employ tacrolimus-based immunosuppression over CsA. Overall, the effectiveness of steroid avoidance regimens in this setting is better documented than subsequent withdrawal strategies.

In summary, there is a suggestion that steroid avoidance is beneficial for patients after renal transplantation, with the potential to avoid or reduce steroid-related comorbidities. Whilst the limited literature seems to indicate that depleting induction therapy could be the treatment of choice for steroid minimization and avoidance in kidney transplantation, the results of prospective randomized, controlled studies are eagerly awaited.50

All authors actively contributed to the content of the manuscript during discussions, critically reviewed draft material, and approved the final manuscript for submission.

The authors attended a meeting at which the data for inclusion in the paper were discussed and presentations made by the authors. The meeting was coordinated by Sanofi but the authors received no remuneration or funding for travel expenses. Julio Pascual and Marta Crespo are supported by grants FISPI13/00598 and RedinrenRD12/0021/0024.

Maarten Naesens has received travel support, research grant or consultancy fees from Sanofi, Novartis, Astellas and Roche. Stefan Berger has received travel support from Astellas and consultancy fees form Sanofi, Novartis and Astellas. Luigi Biancone has received travel support, research grant or consultancy fees from Sanofi, Novartis, and Astellas. Marta Crespo has received public funding for research from the Spanish Ministry of Health and travel support, research grant, consultancy fees or speaker's honoraria from Novartis, Astellas, Sanofi and Abbvie. Arjang Djamali is a consultant to Sanofi and has received grants from BMS and Takeda. Alexandre Hertig has received consultancy fees from Sanofi, Bristol-Myers-Squibb, travel grants from Novartis and CSL Behring, and research grant from Astellas and CSL Behring. Robert Öllinger has no conflicts of interest to declare. José Portolés has received travel support, research grant or consultancy fees from Sanofi, Novartis, and Astellas. Andreas Zuckermann has received research grants from Astellas, Roche, Novartis and Sanofi, is a member of the speakers’ bureau for Novartis and Sanofi, and is a member of advisory boards for Sanofi and Sandoz. Julio Pascual has received travel support, research grant or consultancy fees from Sanofi, Novartis and Astellas.