Abstract

Background.

The safety and efficacy of early steroid withdrawal or avoidance in patients receiving a kidney transplant (KT) are controversial.

Methods.

We performed a systematic review and a meta-analysis of the randomized controlled studies about steroid avoidance or withdrawal after a few days in patients receiving a KT and treated with antibody induction and cyclosporine (CsA) or tacrolimus (Tac) plus mycophenolate mofetil (MMF) (nine available studies and 1934 participants).

Results.

Death and graft loss (including or excluding death with function) were similar in steroid avoidance and control patients, with no differences between CsA and Tac studies. After steroid avoidance, acute rejection was more frequent than conventional steroid use in CsA trials [risk ratios (RR) 1.59, 95% confidence intervals (95% CI) 1.01–2.49] but not when Tac was used (RR 1.06, 95% CI 0.79–1.42). Steroid avoidance was associated with less frequent new-onset diabetes mellitus, but this decrease was only evident with CsA (RR 0.54, 95% CI 0.30–0.98), whereas this difference was not significant analysing Tac studies (RR 0.75, 95% CI 0.32–1.77). Despite this trend, the corresponding interaction tests were not statistically significant (P = 0.140 and P = 0.535, for acute rejection and new-onset diabetes mellitus, respectively). Serum creatinine, creatinine clearance, mean blood pressure, serum cholesterol and serum triglycerides were similar in both groups.

Conclusions.

Steroid avoidance or early withdrawal within the first 2 weeks is safe in KT recipients receiving induction with anti-interleukin-2 receptor antibodies or thymoglobulin and a drug regimen based on calcineurin inhibitor and MMF. However, the real benefits remain unclear.

Introduction

Renal transplantation is the treatment of choice for end-stage renal disease, with demonstrated improvement in survival and quality of life with respect to the treatment with dialysis [ 1 ]. Efficacy in acute rejection control and excellent short-term results has shifted the focus of investigators to the assessment of better low-toxicity protocols that could maintain good long-term graft function and patient survival. Two drug families have been perceived as essential in immunosuppressive regimens for decades: calcineurin inhibitors (CNI) and corticosteroids. The CNI cyclosporine (CsA) and tacrolimus (Tac) increase renal graft fibrosis and vascular damage [ 2 , 3 ], but their avoidance is not safe [ 4 ]. Corticosteroids are associated with the development of hypertension, glucose intolerance, hyperlipidaemia, bone and ocular diseases, growth retardation, obesity and infection. Despite steroid dose is usually very low from the early phases after transplantation, these complications may impair quality of life and survival. Randomized controlled trials (RCTs) have shown that the corticosteroid withdrawal some months after transplantation is associated with a high risk of acute rejection, even when maintenance therapy is based in the potent CNI and mycophenolic acid combination [ 5 , 6 ]. Rather than late steroid withdrawal, early withdrawal or avoidance have been increasingly popular for many transplant physicians, especially after a large observational study was available [ 7 ]. The only available systematic review on steroid avoidance or early withdrawal was published in 1993, when potent Tac and mycophenolic acid derivatives were not available yet [ 8 ]. More controlled studies have been performed during the last decade but have not been systematically reviewed.

Materials and methods

Inclusion criteria

We included all RCTs comparing steroid avoidance or very early withdrawal (steroid use limited to <10 days) with conventional steroid maintenance. To avoid the inclusion of outdated protocols without modern potent immunosuppressive agents, we only included trials with the combination of drugs currently used in clinical practice: CsA or Tac and mycophenolic acid derivatives or mammalian target of rapamycin (mTOR) inhibitors in the intervention and control arms. Only trials in which intention-to-treat (ITT) rates of acute rejection and graft failure were clearly established were included. RCTs without a steroid avoidance or withdrawal group were excluded as well as studies in which other interventions in addition to steroid-sparing strategy were attempted. Some RCTs not fulfilling these strict criteria were excluded. The participants were adult kidney transplant (KT) recipients, including first and retransplantation and excluding multiorgan transplant recipients. Observational studies were excluded.

Literature search

Relevant studies were obtained from the Cochrane Renal Group’s Specialized Register and the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE and EMBASE. We also used the list of references in recent important publications as well as the abstracts from international transplantation society scientific meetings. With duplicate reports of the same study, the most updated complete publication was selected and the duplicates discarded.

Outcome measures

Primary outcomes considered were death, graft loss (both including death with function and censored for death), ITT acute rejection, ITT biopsy-proven acute rejection, serum creatinine (SCr, milligrams per decilitre), creatinine clearance (CrCl, millilitres per minute), mean blood pressure (millimetres of mercury), serum cholesterol (milligrams per decilitre), low-density lipoprotein cholesterol (milligrams per decilitre), patients on lipid-lowering treatment, triglycerides (milligrams per decilitre), new-onset diabetes after transplant (NODAT), infections and malignant neoplasia.

Data extraction was carried out by two authors independently using standardized data extraction forms. Disagreements were minimal and resolved in consultation with the other authors. The quality of included trials was assessed independently by two authors. The items assessed were allocation concealment, blinding, ITT and completeness of follow-up.

Statistical analysis

Studies were grouped and analysed according to how steroids were used within the immunosuppressive algorithm: steroid avoidance or early withdrawal versus steroid maintenance. Subanalyses were undertaken, stratifying results by CNI co-intervention and by anti-metabolite (mycophenolate mofetil—MMF-, enteric-coated mycophenolic sodium—Myf- or an mTOR inhibitor) co-intervention. Interaction analyses to test for different effects in selected subgroups were undertaken as suggested by Altman and Bland [ 9 ].

For dichotomous outcomes, results were expressed as risk ratios (RR) with 95% confidence intervals (CI). When studies showed 0 events, the RR could not be estimated and hence, we used relative difference (RD) for these analyses to check whether the overall conclusions were robust, as the RD could be calculated using all studies. Where continuous scales of measurement were used to assess the effects of treatment (blood pressure, serum cholesterol, kidney function measured by SCr, CrCl, bone density), the mean difference (MD) was used or the standardized mean difference (SMD) if different scales had been used. When SD was not available, it was estimated using standard error (if provided). Otherwise, the particular outcome was excluded from the analysis. Results of unfavourable dichotomous outcomes were expressed so that the left part in any graph indicates that steroid-sparing strategy is better than control. The same convention was applied for MD and SMD. Heterogeneity was analysed using a chi-square test with a 10% used for the limit of statistical significance and we computed the I2 statistic to measure the proportion of total variation in the estimates of treatment effect that was due to heterogeneity beyond chance [ 10 ]. Analysis was undertaken using both random and fixed effects models, and no important differences were observed. Results reported here used the random effects model, as this is more conservative in the presence of heterogeneity. As a sensitive analysis, we restricted analyses to selected subgroups according to study quality (study methodological quality, publication type and the length of follow-up), to explore possible sources of heterogeneity and to demonstrate robustness of our findings. In addition, the effect of trial size in the results was estimated calculating the Log[OR], with a P-value <0.05 showing significant effect. Cochrane RevMan (v 4.3) was used as software for the analysis.

Results

Description of studies

RCTs in which a steroid-sparing strategy was compared with steroid maintenance in adult patients were analysed in 34 reports of 29 studies; a total of 5675 randomized participants. Among them nine RCTs (1934 patients) assessed steroid avoidance studies with MMF/Myf/mTOR [ 11–19 ] ( Figure 1 ). Five were based on CsA [ 11–15 ] and four on Tac [ 16–19 ] ( Table 1 ). An additional trial compared steroid early withdrawal within a few days and steroid withdrawal after some months and was excluded from the analysis [ 20 ]. Most studies had two arms, but two studies had three arms [ 15 , 18 ]. In one of them [ 15 ], patients were randomized to Group 1, steroid avoidance, Group 2, steroid very early withdrawal (at 7 days) and Group 3, conventional steroid maintenance. In view of our steroid avoidance definition (<14 days steroids) and the similar results between Group 1 and Group 2 in this study, we joined both groups in one steroid avoidance unique group ( n = 224). In another trial [ 18 ], patients were randomized to Group 1 steroid avoidance, induction with basiliximab and Tac + steroids, Group 2 steroid avoidance with Tac + MMF or Group 3 Tac + MMF and conventional steroid maintenance. Consequently, we only included data from Groups 1 and 3.

Fig. 1.

Flowchart showing process of identification of RCTs for inclusion in systematic review.

Fig. 1.

Flowchart showing process of identification of RCTs for inclusion in systematic review.

Table 1.

RCTs comparing steroid avoidance or very early withdrawal and steroid conventional use and maintenance included in this review a

Trials n Multicentre trial CNI Anti-metabolite Antibody induction treatment Conventional therapy Follow-up period (months) Completeness of follow-up (%) 
Kim et al. [ 11 ]  24 Yes CsA MMF IL-2 receptor blocker (basiliximab) The same but with steroids 24 Not available 
Kumar et al. [ 12 ]  77 No CsA MMF/SRL IL-2 receptor blocker (basiliximab) The same but with steroids 24 100 
Vincenti et al. [ 13 ]  83 Yes CsA MMF IL-2 receptor blocker (basiliximab) The same but with steroids 12 100 
Montagnino et al. [ 14 ]  133 No CsA Everolimus IL-2 receptor blocker (basiliximab) The same but with steroids 24 Not available 
Vincenti et al. [ 15 ]  224 Yes CsA Myf IL-2 receptor blocker (basiliximab) The same but with steroids 12 Not available 
Laftavi et al. [ 16 ]  60 No Tac MMF Anti-lymphocytic depletive antibodies The same but with steroids 12 Not available 
Rostaing et al. [ 17 ]  538 Yes Tac MMF IL-2 receptor blocker (daclizumab) Tac + MMF and steroids 82 
Vitko et al. [ 18 ]  298 Yes Tac MMF No The same but with steroids 93 
Woodle et al. [ 19 ]  386 Yes Tac MMF Anti-lymphocytic depletive antibodies or IL-2 receptor blocker The same but with steroids 60 100 
Trials n Multicentre trial CNI Anti-metabolite Antibody induction treatment Conventional therapy Follow-up period (months) Completeness of follow-up (%) 
Kim et al. [ 11 ]  24 Yes CsA MMF IL-2 receptor blocker (basiliximab) The same but with steroids 24 Not available 
Kumar et al. [ 12 ]  77 No CsA MMF/SRL IL-2 receptor blocker (basiliximab) The same but with steroids 24 100 
Vincenti et al. [ 13 ]  83 Yes CsA MMF IL-2 receptor blocker (basiliximab) The same but with steroids 12 100 
Montagnino et al. [ 14 ]  133 No CsA Everolimus IL-2 receptor blocker (basiliximab) The same but with steroids 24 Not available 
Vincenti et al. [ 15 ]  224 Yes CsA Myf IL-2 receptor blocker (basiliximab) The same but with steroids 12 Not available 
Laftavi et al. [ 16 ]  60 No Tac MMF Anti-lymphocytic depletive antibodies The same but with steroids 12 Not available 
Rostaing et al. [ 17 ]  538 Yes Tac MMF IL-2 receptor blocker (daclizumab) Tac + MMF and steroids 82 
Vitko et al. [ 18 ]  298 Yes Tac MMF No The same but with steroids 93 
Woodle et al. [ 19 ]  386 Yes Tac MMF Anti-lymphocytic depletive antibodies or IL-2 receptor blocker The same but with steroids 60 100 
a

SRL, sirolimus.

The number of participants was very variable, from 24 patients [ 11 ] to 538 patients [ 17 ]. One small study was available only in abstract form [ 11 ]. The controlled follow-up was 6 months in two trials [ 17 , 18 ], 12 months in three trials [ 13 , 15 , 16 ], 24 months in three trials [ 11 , 13 , 14 ] and 60 months in one trial [ 19 ]. The review included data from 3294 patient-years of follow-up.

Only one trial [ 14 ] included an mTOR (everolimus) as the third drug, the other eight RCTs included MMF or Myf. Hence, the value of mTORs in steroid-sparing strategies was not adequately reviewed. However, given the similar results obtained in this trial and MMF or Myf trials, this trial was kept in the analysis.

Antibody induction was used in all but one steroid avoidance RCTs [ 18 ]. Consequently, it was not possible to compare steroid withdrawal or avoidance with or without induction. Among the steroid avoidance RCTs using antibody induction, in six RCTs the antibody was an interleukin-2 (IL-2) receptor blocker [ 11–15 , 17 ] and one used anti-lymphocytic depletive antibodies [ 16 ]. In another one, the type of antibody (IL-2 receptor blocker or anti-lymphocytic depletive) was chosen at the investigator’s discretion [ 19 ]. Consequently, a meta-analysis comparing the type of induction was not possible.

Risk of bias in included studies

Four RCTs reported adequate methods of allocation concealment [ 12 , 17–19 ], one RCT used inadequate methods [ 16 ] and the remaining four RCTs were randomized but gave no clear indication of the allocation method used [ 11 , 13 , 14 , 15 ]. Only one RCT reported blinding of both investigators and participants [ 19 ]. ITT analysis was confirmed for six studies and was unclear for the remaining three RCTs [ 11 , 13 , 16 ]. Completeness of follow-up was not reported and could not be calculated with the data provided in four RCTs [ 11 , 14 , 16 , 18 ] and was 82–100% in five RCTs where it could be calculated.

Outcomes

Death and graft loss (including or excluding death with function) were similar in steroid avoidance and control patients, with no differences between CsA and Tac studies ( Table 3 and Figure 2A ). After steroid avoidance, acute rejection was more frequent than conventional steroid use in CsA studies (RR 1.59, 95% CI 1.01–2.49) and not Tac studies (RR 1.06, 95% CI 0.79–1.42) ( Figure 2B ), although this different effect was not statistically significant (P interaction = 0.140). This trend is also observed considering only biopsy-proven cases, but the difference between both subgroups according to CNI is less important (CsA RR 1.89, 95% CI 1.33–2.69 and Tac RR 1.69, 95% CI 0.87–3.25)(P interaction = 0.769). Steroid avoidance was associated with less frequent NODAT requiring any treatment, but this decrease was only evident with CsA (RR 0.54, 95% CI 0.30–0.98), whereas this difference was not significant analysing Tac studies (RR 0.75, 95% CI 0.32–1.77) ( Figure 2C ). The interaction test was not statistically significant (P interaction = 0.535).

Infections and malignant neoplasia were seen with similar incidence in steroid sparing and control groups ( Table 2 ). Serum creatinine, creatinine clearance, mean blood pressure, serum cholesterol and serum triglycerides were similar in both groups ( Table 3 ).

Table 2.

Results obtained from the meta-analysis stratified by CNI: dichotomous outcomes a

   No. of participants
 
Effect size
 
Tests for heterogeneity
 
Outcome, by CNI No. of trials Avoidance Maintenance  RR (95% CI) b P-value P-value I2 (%)  
Death 
    All 24/988 21/886 1.02 (0.53–1.96) 0.95 0.80 
    CsA 7/386 5/266 0.88 (0.27–2.87) 0.83 0.63 
    Tac 17/602 16/620 1.09 (0.50–2.37) 0.83 0.54 
Graft loss including death 
    All 66/988 53/886 1.12 (0.73, 1.73) 0.59 0.35 10 
    CsA 17/386 15/266 0.79 (0.38–1.67) 0.54 0.52 
    Tac 49/602 38/620 1.29 (0.71–2.34) 0.40 0.21 36 
Graft loss excluding death 
    All 43/988 32/886 1.21 (0.70–2.07) 0.49 0.33 14 
    CsA 10/386 10/266 0.72 (0.28–1.84) 0.49 0.51 
    Tac 33/602 22/620 1.50 (0.78–2.88) 0.23 0.28 21 
ITT acute rejection* 
    All 123/542 91/475 1.20 (0.93–1.53) 0.16 0.57 
    CsA 56/250 24/169 1.59 (1.01–2.49) 0.044 0.64 
    Tac 67/292 67/306 1.06 (0.79–1.42) 0.71 0.89 
ITT biopsy-proven acute rejection** 
    All 226/975 118/882 1.75 (1.23–2.50) 0.002 0.03 56 
    CsA 99/341 36/234 1.89 (1.33–2.69) 0.0003 0.78 
    Tac 127/634 82/648 1.69 (0.87–3.25) 0.12 0.003 78 
Patients on lipid-lowering therapy 
    All 120/375 84/258 0.84 (0.68–1.05) 0.12 0.40 
    CsA 101/224 57/111 0.88 (0.70–1.11) 0.27 NA NA 
    Tac 19/151 27/147 0.69 (0.40–1.18) 0.17 NA NA 
NODAT requiring any treatment*** 
    All 57/798 66/678 0.65 (0.38–1.14) 0.13 0.07 50 
    CsA 19/319 20/202 0.54 (0.30–0.98) 0.043 0.52 
    Tac 38/479 46/476 0.75 (0.32–1.77) 0.52 0.02 73 
Infections 
    All 225/931 180/842 1.15 (0.87–1.51) 0.34 0.09 48 
    CsA 60/329 30/222 1.63 (0.62–4.33) 0.32 0.11 55 
    Tac 165/602 150/620 1.13 (0.82–1.57) 0.46 0.08 61 
Malignant neoplasia 
    All 3/635 8/536 0.49 (0.13–1.88) 0.30 0.40 
    CsA 3/224 1/111 1.49 (0.16–14.13) 0.73 NA NA 
    Tac 0/411 7/425 0.27 (0.05–1.43) 0.12 0.49 
   No. of participants
 
Effect size
 
Tests for heterogeneity
 
Outcome, by CNI No. of trials Avoidance Maintenance  RR (95% CI) b P-value P-value I2 (%)  
Death 
    All 24/988 21/886 1.02 (0.53–1.96) 0.95 0.80 
    CsA 7/386 5/266 0.88 (0.27–2.87) 0.83 0.63 
    Tac 17/602 16/620 1.09 (0.50–2.37) 0.83 0.54 
Graft loss including death 
    All 66/988 53/886 1.12 (0.73, 1.73) 0.59 0.35 10 
    CsA 17/386 15/266 0.79 (0.38–1.67) 0.54 0.52 
    Tac 49/602 38/620 1.29 (0.71–2.34) 0.40 0.21 36 
Graft loss excluding death 
    All 43/988 32/886 1.21 (0.70–2.07) 0.49 0.33 14 
    CsA 10/386 10/266 0.72 (0.28–1.84) 0.49 0.51 
    Tac 33/602 22/620 1.50 (0.78–2.88) 0.23 0.28 21 
ITT acute rejection* 
    All 123/542 91/475 1.20 (0.93–1.53) 0.16 0.57 
    CsA 56/250 24/169 1.59 (1.01–2.49) 0.044 0.64 
    Tac 67/292 67/306 1.06 (0.79–1.42) 0.71 0.89 
ITT biopsy-proven acute rejection** 
    All 226/975 118/882 1.75 (1.23–2.50) 0.002 0.03 56 
    CsA 99/341 36/234 1.89 (1.33–2.69) 0.0003 0.78 
    Tac 127/634 82/648 1.69 (0.87–3.25) 0.12 0.003 78 
Patients on lipid-lowering therapy 
    All 120/375 84/258 0.84 (0.68–1.05) 0.12 0.40 
    CsA 101/224 57/111 0.88 (0.70–1.11) 0.27 NA NA 
    Tac 19/151 27/147 0.69 (0.40–1.18) 0.17 NA NA 
NODAT requiring any treatment*** 
    All 57/798 66/678 0.65 (0.38–1.14) 0.13 0.07 50 
    CsA 19/319 20/202 0.54 (0.30–0.98) 0.043 0.52 
    Tac 38/479 46/476 0.75 (0.32–1.77) 0.52 0.02 73 
Infections 
    All 225/931 180/842 1.15 (0.87–1.51) 0.34 0.09 48 
    CsA 60/329 30/222 1.63 (0.62–4.33) 0.32 0.11 55 
    Tac 165/602 150/620 1.13 (0.82–1.57) 0.46 0.08 61 
Malignant neoplasia 
    All 3/635 8/536 0.49 (0.13–1.88) 0.30 0.40 
    CsA 3/224 1/111 1.49 (0.16–14.13) 0.73 NA NA 
    Tac 0/411 7/425 0.27 (0.05–1.43) 0.12 0.49 
a

ITT, intention to treat population; NODAT, new-onset diabetes mellitus; Bold values, significant values.

b

RR values <1 favour steroid withdrawal.

*P-value for interaction = 0.140.

**P-value for interaction = 0.769.

***P-value for interaction = 0.535.

Table 3.

Results obtained from the meta-analysis stratified by CNI: continuous variables

    Effect size
 
Tests for heterogeneity
 
Outcome, by CNI No. of trials No. of participants  Mean difference (95% CI) a P-value P-value I2 (%)  
Serum creatinine (mg/dL) 
    All 1406 0.02 (−0.05 to 0.10) 0.57 0.52 
    CsA 184 −0.10 (−0.29 to 0.09) 0.31 1.00 
    Tac 1222 0.04 (−0.04 to 0.12) 0.30 0.51 
Creatinine clearance (mL/min) 
    All 1672 −1.06 (−2.95 to 0.83) 0.27 0.90 
    CsA 390 0.16 (−3.74 to 4.05) 0.94 0.49 
    Tac 1282 −1.44 (−3.61 to 0.73) 0.19 0.95 
Mean blood pressure (mmHg) 
    All 933 0.09 (−0.12 to 0.30) 0.40 0.76 
    CsA 335 0.08 (−0.15 to 0.30) 0.52 NA NA 
    Tac 598 0.16 (−0.35 to 0.67) 0.53 1.00 
Serum cholesterol (mg/dL) 
    All 1364 −8.52 (−21.69 to 4.65) 0.20 0.15 47 
    CsA 896 −20.00 (−47.24 to 7.24) 0.15 1.00 
    Tac 468 −4.96 (−23.45 to 13.53) 0.60 0.07 69 
Serum triglycerides (mg/dL) 
    All 1006 −12.36 (−37.92 to 13.20) 0.34 0.21 35 
    CsA 478 −34.00 (−77.13 to 9.13) 0.12 1.00 
    Tac 538 −4.50 (−21.77 to 12.77) 0.61 NA NA 
    Effect size
 
Tests for heterogeneity
 
Outcome, by CNI No. of trials No. of participants  Mean difference (95% CI) a P-value P-value I2 (%)  
Serum creatinine (mg/dL) 
    All 1406 0.02 (−0.05 to 0.10) 0.57 0.52 
    CsA 184 −0.10 (−0.29 to 0.09) 0.31 1.00 
    Tac 1222 0.04 (−0.04 to 0.12) 0.30 0.51 
Creatinine clearance (mL/min) 
    All 1672 −1.06 (−2.95 to 0.83) 0.27 0.90 
    CsA 390 0.16 (−3.74 to 4.05) 0.94 0.49 
    Tac 1282 −1.44 (−3.61 to 0.73) 0.19 0.95 
Mean blood pressure (mmHg) 
    All 933 0.09 (−0.12 to 0.30) 0.40 0.76 
    CsA 335 0.08 (−0.15 to 0.30) 0.52 NA NA 
    Tac 598 0.16 (−0.35 to 0.67) 0.53 1.00 
Serum cholesterol (mg/dL) 
    All 1364 −8.52 (−21.69 to 4.65) 0.20 0.15 47 
    CsA 896 −20.00 (−47.24 to 7.24) 0.15 1.00 
    Tac 468 −4.96 (−23.45 to 13.53) 0.60 0.07 69 
Serum triglycerides (mg/dL) 
    All 1006 −12.36 (−37.92 to 13.20) 0.34 0.21 35 
    CsA 478 −34.00 (−77.13 to 9.13) 0.12 1.00 
    Tac 538 −4.50 (−21.77 to 12.77) 0.61 NA NA 
a

Mean difference values <0 favour steroid withdrawal. For mean blood pressure, standard mean difference.

Three studies were considered to be low-quality studies, defined as those failing three of four high-quality characteristics (allocation concealment, blinding, ITT and completeness of follow-up) [ 11 , 13 , 16 ]. We repeated the analyses excluding these three studies for the main outcomes death, graft loss including and excluding death, ITT acute rejection, ITT biopsy-proven acute rejection and NODAT requiring any treatment. The effect of early steroid avoidance or withdrawal was estimated to be similar after these exclusions (Supplementary form). However, as expected, 95% CIs for RR are now slightly wider and hence, those marginal statistical significances now vanished (i.e. ITT acute rejection and NODAT requiring any treatment in CsA subgroup).

We have checked for the effect of trial size on effect size (as measured as Log[OR]) for the following outcomes: death, graft loss including death, graft loss excluding death, ITT acute rejection, ITT biopsy-proven acute rejection and NODAT. The P-values shown test the hypothesis of no effect of trial size (P > 0.05) except for the outcome graft loss including death (P = 0.052) in which the greater the size the greater the estimate of the effect of steroids avoidance as compared to maintenance.

Discussion

Two parameters seem to be very relevant when assessing steroid-sparing strategies: (i) the moment of steroid withdrawal and (ii) the duration of follow-up to observe kidney evolution without steroid treatment during adequate periods of time. Two main steroid-sparing strategies have been studied over the last 20 years in these studies: steroid avoidance or very early withdrawal and late steroid withdrawal. We defined steroid avoidance as the use of steroids for ≤2 weeks and steroid withdrawal as the stopping of steroids after >2 weeks of treatment. The majority of the nine studies analysing steroid avoidance used steroids for ≤7 days. Most studies had 12–24 months of follow-up after either steroid avoidance or early withdrawal, and only one of them reported up to 5 years of controlled evaluation.

This review confirms that steroid avoidance or very early withdrawal strategies in KT are not associated with increased mortality or graft loss. Only steroid withdrawal without MMF/Myf/mTOR using CsA was associated with higher rates of graft loss excluding death [ 21 ]. As no Tac studies without MMF/Myf/mTOR are available to date, we do not know whether or not steroid-sparing strategies are safe with Tac-based immunosuppression not including MMF/Myf/mTOR. Most of the steroid avoidance studies were designed with antibody induction in the steroid avoidance arm, so the results obtained cannot be confidently extrapolated for steroid avoidance protocols without adequate induction treatment at least with an anti-IL-2 receptor antibody. This is in agreement with a very recent observation from the US Registry showing that de novo steroid-free immunosuppression as currently practiced in the USA appears to carry no increased risk of adverse clinical outcomes in the intermediate term [ 22 ].

Overall, acute rejection was more frequent with steroid-sparing strategies than with conventional steroid use. When we analysed studies comparing steroid withdrawal after 3–6 months with maintenance, an increase in acute rejection rates with ITT analysis or biopsy-proven rejections since the first day after transplantation was observed [ 6 ]. Steroid avoidance patients showed increased acute rejection rates in comparison with conventional steroid use and maintenance, but this increase was only significant when including CsA-based protocols and was not seen in Tac-treated patients. However, the formal statistical test of interaction did not reach statistical significance perhaps given the low number of studies included in each of the treatment strata (CsA and Tac). It appears that acute rejection is less frequent when steroids are completely avoided or withdrawn very early than when they are stopped after >3 months of treatment. A possible mechanistic explanation may be an increase in the synthesis of chemokines in the setting of up-regulated chemokine receptor expression mediated by steroids [ 23 ]. Although the question remains whether the possible deleterious effects of reversible rejection in a very low percentage of patients outweigh the possible beneficial effects of steroid avoidance, the vast majority of patients do benefit from being without steroids early after transplantation without immediate risk of rejection. Despite a significant increase in acute rejection risk in CsA trials, the very low absolute number of rejections might be enough to argue for a positive recommendation. Steroid avoidance or very early withdrawal was particularly safe in patients receiving induction treatment with antibodies and the combination Tac–MMF.

The benefits of steroid-sparing strategies were not easily determined in this review because of frequent under-reporting of relevant data in many studies. Many adverse events classically related to steroid use were not significantly reduced with a well-defined steroid-sparing strategy, or at least, such benefits were not adequately reported. The effects on blood pressure were scarcely reported. The reduction in total cholesterol and anti-hyperlipidaemic drug need was important in steroid withdrawal trials [ 6 ], but such benefit was not evident in steroid avoidance studies. NODAT rate was decreased in steroid avoidance or early steroid withdrawal patients, but this benefit was only evident in patients treated with CsA. In Tac-treated allograft recipients, it seems that the more diabetogenic effect of Tac outweighted the potential benefit of avoiding steroids in the development of NODAT. Although the reduction in NODAT was very significant in one European Tac study with steroid avoidance [ 17 ], the significance disappeared when joined with the US double-blind steroid avoidance study with Tac, which did not show such an effect [ 19 ]. It is possible that high Tac levels were responsible for the absence of such a beneficial effect in this last study. Finally, no relevant impact could be observed in the infection rate and cancer development, thus suggesting that the increased rates of such events in KT recipients are not strongly related to steroid use.

Only one previous systematic review has addressed this topic, including four steroid avoidance studies in patients on CsA with or without azathioprine (AZA) [ 8 ]. The incidence of acute rejection was higher in steroid avoidance patients, but no deleterious effects on graft survival were seen. The relative short-term follow-up meant the authors were unable to definitively support the conclusion that an increase in acute rejection rate after steroid avoidance did not imply long-term graft function deterioration.

The main strength of this review is that it has identified a good number of RCTs assessing steroid avoidance or almost avoidance strategies, currently attempted by the transplant community. It also analyses different profiles in patients receiving CsA versus Tac, but excluding trials using AZA and including only protocols with MMF, or its variant Myf. Our methodology was robust, including all possible studies published, even in abstract form, in any language and with assessment of data quality. Consequently, this review provides information to guide treatment decisions on steroid sparing in adult KT, particularly the absence of harmful consequences, better than the existence of clear benefits.

Nonetheless, this review has some relevant limitations. Only one study of steroid avoidance included an mTOR inhibitor as a third drug (everolimus) and consequently, we cannot extrapolate the safety of steroid avoidance or withdrawal to protocols including mTOR inhibitors instead of MMF or Myf as a third drug. Another limitation is that, despite some studies have extended their follow-up periods to >2 years (one study described 5 years), the extension periods beyond the first year were frequently retrospective in nature and not prospectively designed. It is well known that the possible harmful effects of steroid minimization may remain hidden until >4 or 5 years of steroid-free treatment [ 21 ]. We do not know the deleterious effects that the absence of steroids may have in kidney grafts maintained steroid free for >5 years.

Conclusions

Regarding safety of steroid-sparing strategies in KT, we did not find enough evidence demonstrating an increased high risk of early graft failure after steroid avoidance or almost avoidance in patients receiving CsA and MMF despite an increased acute rejection rate. The combination Tac–MMF is particularly safe to avoid steroids after the first few days post-transplantation. Our results may support that this potent immunosuppression allows safe steroid avoidance or elimination a few days after KT if antibody induction treatment is prescribed. The strength of the evidence was less when reviewing the potential benefits of steroid-sparing strategies. A significant reduction in NODAT incidence was not found in the overall result. Although a reduction was observed after steroid avoidance if the CNI used was CsA, the interaction between CNI and NODAT was not significant. A long-term RCT that has adequate statistical power to detect differences in acute rejection and major adverse events is needed to determine whether the benefits of steroid avoidance outweigh the harm.

Contributions—Study concept and design: J.P., J.Z.; acquisition of data: J.P., C.G., A.R., J.Z.; analysis and interpretation of data: J.P., M.C., J.Z.; drafting of the manuscript: J.P., A.R., J.Z.; critical revision of the manuscript: J.P., C.G., A.R., M.C., J.Z.; statistical analysis: A.R., J.Z.

Conflict of interest statement . None declared.

References

1.
Wolfe
RA
Ashby
VB
Milford
EL
, et al.  . 
Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation and recipients of a first cadaveric transplant
N Engl J Med
 , 
1999
, vol. 
341
 (pg. 
1725
-
1730
)
2.
Nankivell
BJ
Borrows
RJ
Fung
CL
, et al.  . 
Calcineurin inhibitor nephrotoxicity: longitudinal assessment by protocol histology
Transplantation
 , 
2004
, vol. 
78
 (pg. 
557
-
565
)
3.
Naesens
M
Lerut
E
Damme
BV
, et al.  . 
Tacrolimus exposure and evolution of renal allograft histology in the first year after transplantation
Am J Transplant
 , 
2007
, vol. 
7
 (pg. 
2114
-
2123
)
4.
Ekberg
H
Tedesco-Silva
H
Demirbas
A
, et al.  . 
Reduced exposure to calcineurin inhibitors in renal transplantation
N Engl J Med
 , 
2007
, vol. 
357
 (pg. 
2562
-
2575
)
5.
Kasiske
BL
Chakkera
HA
Louis
TA
, et al.  . 
A meta-analysis of immunosuppression withdrawal trials in renal transplantation
J Am Soc Nephrol
 , 
2000
, vol. 
11
 (pg. 
1910
-
1917
)
6.
Pascual
J
Quereda
C
Zamora
J
, et al.  . 
Steroid withdrawal in renal transplant patients on triple therapy with a calcineurin inhibitor and mycophenolate mofetil: a meta-analysis of randomized controlled trials
Transplantation
 , 
2004
, vol. 
78
 (pg. 
1548
-
1556
)
7.
Matas
AJ
Kandaswamy
R
Gillingham
KJ
, et al.  . 
Prednisone-free maintenance immunosuppression-a 5-year experience
Am J Transplant
 , 
2005
, vol. 
5
 (pg. 
2473
-
2478
)
8.
Hricik
DE
O'Toole
MA
Schulak
JA
, et al.  . 
Steroid-free immunosuppression in cyclosporine-treated renal transplant recipients: a meta-analysis
J Am Soc Nephrol
 , 
1993
, vol. 
4
 (pg. 
1300
-
1305
)
9.
Altman
DG
Bland
JM
Interaction revisited: the difference between two estimates
BMJ
 , 
2003
, vol. 
326
 pg. 
219
 
10.
Higgins
JP
Thompson
SG
Deeks
JJ
, et al.  . 
Measuring inconsistency in meta-analyses
BMJ
 , 
2003
, vol. 
327
 (pg. 
557
-
560
)
11.
Kim
EH
Gohh
R
Morrissey
P
, et al.  . 
Rapid steroid withdrawal versus standard steroid treatment in patients treated with basiliximab, cyclosporine, and mycophenolate mofetil for the prevention of acute rejection in kidney transplantation: a 2-year follow-up
Am J Transplant
 , 
2002
, vol. 
2
 
Suppl 3
pg. 
397
 
12.
Kumar
MS
Xiao
SG
Fyfe
B
, et al.  . 
Steroid avoidance in renal transplantation using basiliximab induction, cyclosporine-based immunosuppression and protocol biopsies
Clin Transplant
 , 
2005
, vol. 
19
 (pg. 
61
-
69
)
13.
Vincenti
F
Monaco
A
Grinyo
J
, et al.  . 
Multicenter randomized prospective trial of steroid withdrawal in renal transplant recipients receiving basiliximab, cyclosporine microemulsion and mycophenolate mofetil
Am J Transplant
 , 
2003
, vol. 
3
 (pg. 
306
-
311
)
14.
Montagnino
G
Sandrini
S
Casciani
C
, et al.  . 
A randomized trial of steroid avoidance in renal transplant patients treated with everolimus and cyclosporine
Transplant Proc
 , 
2005
, vol. 
37
 (pg. 
788
-
790
)
15.
Vincenti
F
Schena
FP
Paraskevas
S
, et al.  . 
A randomized, multicenter study of steroid avoidance, early steroid withdrawal or standard steroid therapy in kidney transplant recipients
Am J Transplant
 , 
2008
, vol. 
8
 (pg. 
307
-
316
)
16.
Laftavi
MR
Stephan
R
Stefanick
B
, et al.  . 
Randomized prospective trial of early steroid withdrawal compared with low-dose steroids in renal transplant recipients using serial protocol biopsies to assess efficacy and safety
Surgery
 , 
2005
, vol. 
137
 (pg. 
364
-
371
)
17.
Rostaing
L
Cantarovich
D
Mourad
G
, et al.  . 
Corticosteroid-free immunosuppression with tacrolimus, mycophenolate mofetil, and daclizumab induction in renal transplantation
Transplantation
 , 
2005
, vol. 
79
 (pg. 
807
-
814
)
18.
Vitko
S
Klinger
M
Salmela
K
, et al.  . 
Two corticosteroid-free regimens-tacrolimus monotherapy after basiliximab administration and tacrolimus/mycophenolate mofetil-in comparison with a standard triple regimen in renal transplantation: results of the Atlas study
Transplantation
 , 
2005
, vol. 
80
 (pg. 
1734
-
1741
)
19.
Woodle
ES
First
MR
Pirsch
J
, et al.  . 
A prospective, randomized, double-blind, placebo-controlled multicenter trial comparing early (7 day) corticosteroid cessation versus long-term, low-dose corticosteroid therapy
Ann Surg
 , 
2008
, vol. 
248
 (pg. 
564
-
577
)
20.
Ter Meulen
CG
van Riemsdijk
I
Hene
RJ
, et al.  . 
Steroid-withdrawal at 3 days after renal transplantation with anti-IL-2 receptor alpha therapy: a prospective, randomized, multicenter study
Am J Transplant
 , 
2004
, vol. 
4
 (pg. 
803
-
810
)
21.
Sinclair
NR
Low-dose steroid therapy in cyclosporine-treated renal transplant recipients with well-functioning grafts. The Canadian Multicentre Transplant Study Group
CMAJ
 , 
1992
, vol. 
147
 (pg. 
645
-
657
)
22.
Luan
FL
Steffick
DE
Gadegbeku
C
, et al.  . 
Graft and patient survival in kidney transplant recipients selected for de novo steroid-free maintenance immunosuppression
Am J Transplant
 , 
2009
, vol. 
9
 (pg. 
160
-
168
)
23.
Almawi
WY
Melemedjian
OK
Rieder
MJ
An alternate mechanism of glucocorticoid anti-proliferative effect: promotion of a Th2 cytokine-secreting profile
Clin Transplant
 , 
1999
, vol. 
13
 (pg. 
365
-
374
)

Author notes

*
Some parts published in extended form (458 pages) in Pascual J, Zamora J, Galeano C, Royuela A, Quereda C. Steroid avoidance or withdrawal for kidney transplant recipients. Cochrane Database of Systematic Reviews 2009, Issue 1. Art. No.: CD005632. DOI: 10.1002/14651858.CD005632.pub2. Copyright © 2009 The Cochrane Collaboration. Published by JohnWiley & Sons, Ltd.

Comments

0 Comments