Anti-interleukin-2 receptor antibodies for the prevention of rejection in liver transplant recipients: a systematic review and meta-analysis

Abstract

Background: Antibody induction therapy aims at preventing acute cellular rejection by reducing T-cell proliferation and activation. We evaluated the efficacy and side effects of two anti-interleukin-2 receptor antibodies (IL2RAs), basiliximab and daclizumab, for prevention of liver transplant rejection in adult patients.

Methods: Randomized controlled trials (RCTs) on basiliximab or daclizumab were identified by searching multiple databases and reference lists published up to July, 2015. Endpoints included acute rejection events and mortality rates. Risk ratio (RR) and 95% confidence interval (CI) were calculated and pooled for a meta-analysis.

Results: Patients treated with IL2RA-based therapy were less likely to suffer acute rejection compared to control group (steroid or steroid-free). Patients in all groups had similar mortality rate. In the subgroup analysis, basiliximab and daclizumab-based therapies did not reduced acute rejection rate. No significant difference was found in mortality rate between both types of IL-2RA treatment groups and control groups. In the subgroup analysis regarding experimental design, no significant difference in the acute rejection and mortality rates were found between “steroid plus IL2RA versus steroid” and “IL2RA versus steroid” groups.

Conclusion: IL2RA-based induction therapy reduces rate of acute rejection events but does not reduce mortality. However, optimal regimen relating to IL2RA-based induction therapy remains undetermined.

KEY MESSAGES

• IL2RA-based induction therapy was effective in reduction of acute rejection events but it did not reduce mortality rate.

• Basiliximab-based induction therapy might be more effective than daclizumab-based induction therapy in reduction of acute rejection.

• No significant difference in acute rejection and mortality rate was found between types of IL2RAs or IL2RA-steroid combined therapy.

Keywords:

• Liver transplantation
• antibodies
• humanized
• immunosuppressive agents
• basiliximab
• daclizumab

Introduction

The purpose of immunosuppressive therapy administered immediately before and/or during liver transplantation is to block the development of acute rejection caused predominantly by host T-cells reacting against the major and minor histocompatibility antigens (HLA, etc.). Most immunosuppressive drugs disrupt T-cell activation pathways and T-cell proliferation. For example, calcineurin inhibitors (tacrolimus, cyclosporine) block production of IL-2 and subsequent T-cell proliferation, but it often takes several weeks for these drugs to reach sufficient therapeutic concentrations (1). Other anti-proliferative agents act through inhibition of a rate-limiting enzyme, inosine monophosphate dehydrogenase, used for nucleotide synthesis in B and T lymphocytes (mycophenolate mofetil [MMF]), or compete for nucleotide uptake (azathioprine). After a therapeutic concentration is reached, they halt T-cell proliferation (1). Inhibitors of mammalian target of rapamycin (mTOR) (e.g., sirolimus and everolimus) block the ability of IL-2 and other cytokines to activate T-cells through the interruption of IL-2R post-receptor signaling pathways (2). Currently, a standard immunosuppressive treatment includes early post transplantation administration of calcineurin inhibitors (e.g., tacrolimus, cyclosporine) alone or with an anti-proliferative agent (e.g., MMF, azathioprine) in combination with corticosteroids. Tacrolimus and corticosteroids, however, are associated with serious side effects, including nephrotoxicity (tacrolimus), hypertension, and metabolic effects (corticosteroids) (1). In addition, corticosteroids are associated with increased risks of infection, diabetes mellitus, hypertension, obesity, and prolonged use of steroids increases risks of osteoporosis and fractures (3).

Antibodies can be used as a quick and potent immunosuppressive therapy to prevent the acute rejection reactions. Treatment with antibodies specific for T-lymphocytes reduces the number of circulating T-cells in the liver transplant recipient and is called “antibody” induction therapy. Antibody induction therapy helps prevent a fulminant T-cell response before calcineurin inhibitors, anti-proliferative agents, or mTOR inhibitors reach therapeutic concentrations (3). In addition, antibody induction allows delayed introduction or dose reduction of calcineurin inhibitors, and it has been used to replace corticosteroids in immunosuppressive protocols (4–6). Currently, a variety of both polyclonal and monoclonal antibodies are being regularly utilized to prevent and treat organ rejection (2). Monoclonal antibodies against T-cells have a single specificity for proteins presented on T-cells, such as CD3 (muromonab-CD3), CD52 surface protein (Alemtuzumab), or the activated IL-2 receptor, CD25 (basiliximab, daclizumab) (2,7). Basiliximab and daclizumab bind the IL-2a receptor’s α chain, leading to prevention of the IL-2 induced clonal expansion of activated lymphocytes. Both drugs are being used in the prophylaxis of the acute organ rejection after transplantation (2,7,8) T-cell specific antibodies are very potent drugs which cause profound immunosuppression (9). This may cause increased risks of infection and neoplastic complications (10). Therefore, the question remains whether T-cell specific antibodies should be used for induction of immunosuppression in solid organ transplants (11).

The aim of this study was to evaluate the efficacy of IL2RA induction therapy for adult liver transplant patients by comparing outcomes of IL2RA-based induction therapy with steroid or steroid-free therapy. In addition, we evaluated relative efficacy of different IL2RAs (basiliximab, daclizumab) and different induction therapy protocols (i.e., “steroid plus IL2RA versus steroid” and “IL2RA versus steroid”).

Materials and methods

Search strategy

This study was based on a systematic review and meta-analysis according to Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines (12). The Medline, EMBASE, Cochrane Library (CENTRAL, London, UK), PsycINFO, CINAHL databases were searched by two independent reviewers for articles published up to July, 2015 with various combinations of following keywords: anti-interleukin-2, anti-interleukin-2 receptor antibody, basiliximab, daclizumab, liver transplantation. The reference lists of the included studies were searched for additional articles. We included randomized controlled trials (RCTs) that evaluated adult patients (>18 years) who received a liver transplantation and were treated with either IL-2RA induction therapy (basiliximab or daclizumab) or control therapy (steroid alone or steroid-free) and published in English.

Single-arm studies, case series, letters, comments, editorials, proceeding, personal communication, as well as studies that were conducted on pediatric or adolescent patients or studies with a control group that received other types of antibody-based therapy (e.g., anti-thyroglobulin, alemtuzumab) were excluded from the analysis.

Data extraction

Data was extracted independently by two reviewers. We extracted data on study population (number, age, and gender of subjects in each group), details of intervention (CNI, steroid, anti-proliferative agents, and mTOR inhibitors), length of follow-ups, data on outcomes of interest (acute rejection and mortality), and adverse events.

Assessment of risk of bias

The study quality was assessed using the Cochrane Risk of Bias Tool (13).

The quality assessment was performed by two independent reviewers; the third reviewer was consulted if no consensus could be reached.

Statistical analysis

The primary and secondary outcomes for this meta-analysis were the acute rejection rate and mortality rate, respectively. Odds ratio (OR) with 95% confidence interval (CI) were calculated for each individual study and for all studies combined. A χ²-based test of homogeneity was performed, and the Cochran’s Q inconsistency index (I²) statistics were determined. I²> 50% indicated that heterogeneity existed among the studies. We applied both fixed effect and random effects models, and reported only the results of the random effects model unless there was a significant difference between the results of two models. An OR >1 indicated that patients who underwent corticosteroid treatment had lower rate of acute rejection; whereas, an OR >1 indicated that patients who underwent IL2-RA therapy had lower mortality rate Pooled effects were calculated and a 2-sided p value <0.05 indicated statistical significance. Sensitivity analysis was carried out using the leave one-out approach. Publication bias was assessed by running Egger’s test and constructing a funnel plot. Data points forming a symmetric funnel-shaped distribution and one-tailed significance level of p >0.05 on Egger’s test indicated the absence of publication bias. Subgroup analyses were performed according to type of IL2-RA (basiliximab or daclizumab) and induction treatment protocol (steroid plus IL2RA or steroid-free IL2RA induction therapy). All statistical analyses were performed using the statistical software Comprehensive Meta-Analysis, version 2.0 (Biostat, Englewood, NJ).

Results

Literature search

Using the keyword-based search, we initially identified 121 articles. Eighty-seven articles did not satisfy inclusion criteria and were excluded from further analysis. Full texts of 34 studies were reviewed. Of these, five were non-RCTs, three recruited pediatric liver transplant recipients, nine reported unrelated outcomes, four assessed different interventions, and therefore were excluded from the analysis. Thus, we were able to identify 13 eligible publications (4–6,14–23). The flow chart describing selection of the trials for the analysis is presented in Figure 1.

Figure 1. PRISMA flowchart.

Study characteristics

Characteristics of the included studies are summarized in Table 1. Four trials evaluated basiliximab (6,14,18,22) and nine trials evaluated daclizumab (4,5,15–17,19–21,23). The number of patients in the studies ranged from 30 to 844 (total =4241). Mean age ranged from 46 to 56.4 and medium age ranged from 49 to 55. Length of follow-up ranged from 3 to 114 months. All patients in the included studies received calcineurin inhibitors (e.g., tacrolimus, cyclosporine A) alone or with MMF at the early post-transplantation stage. Three major types of experimental design were found among the studies (Table 1): Type I, IL2RA induction therapy was compared with steroid-free therapy (4,6); Type II, IL2RA induction therapy was compared to steroid therapy (5,12,14–17); Type III, steroid plus IL2RA therapy was compared to steroid therapy (11,13,18–20). The 12 studies reported a wide range of adverse events (Table 2). Overall, IL2RA induction therapy arm and control arm had a similar percentage of patients with adverse events. In general, the common adverse events included infection (bacterial infection and viral infection), hepatitis C virus (HCV) recurrence, renal failure, leucopenia, post-transplant diabetes mellitus (PTDM), hypertension, and hyperglycemia.

Table 1. Summary of basic characteristics of selected studies for meta-analysis.

First author (year)	Number of patients	Age (years)	Experimental design	Type of design	Length of follow-ups (months)
Trunecka (6)	Arm 1: 283	Arm 1: 54.3 (9.1)	Arm 1: Tacrolimus (0.2 mg/kg/day; Day 1) + MMF	I	6
	Arm 2: 287	Arm 2: 54.0 (9.7)	Arm 2: Tacrolimus (0.15 mg–0.175 mg/kg/day; Day 1) + MMF + basiliximab^a
	Arm 3: 274	Arm 3: 53.7 (10.6)	Arm 3: Tacrolimus (0.2 mg/kg/day; Day 5) + MMF + basiliximab^a
Neumann (17)	Arm 1: 68	Arm 1: 55.3 (6.5)	Arm 1: Tacrolimus (0.1–0.15 mg/mL) + corticosteroid	II	12
	Arm 2: 67	Arm 2: 53.1 (9.6)	Arm 2: Tacrolimus (0.1–0.15 mg/mL) + daclizumab^a
Klintmalm (20)	Arm 1: 181	Arm 1: 51.3 (6.7)	Arm 1: Tacrolimus (0.08–0.12 mg/kg/day) + corticosteroid	II	20.9
	Arm 2: 168	Arm 2: 51.6 (7.8)	Arm 2: Tacrolimus (0.08–0.12 mg/kg/day) + MMF + corticosteroid
	Arm 3: 168	Arm 3: 51.3 (7.4)	Arm 3: Tacrolimus (0.08–0.12 mg/kg/day) + MMF + daclizumab^a
Neuberger (16)	Arm 1: 181	Arm 1: 53 (19–69)	Arm 1: Tacrolimus (0.1–0.15 mg/kg/day) + corticosteroid	III	12
	Arm 2: 168	Arm 2: 54 (18–73)	Arm 2: Tacrolimus (0.05–0.1 mg/kg/day) + corticosteroid + MMF
	Arm 3: 168	Arm 3: 55 (26–70)	Arm 3: Tacrolimus (0.05–0.1 mg/kg/day) + corticosteroid + MMF + daclizumab^a
Otero (15)	Arm 1: 79	Arm 1: 54 (18–67)	Arm 1: Tacrolimus (0.05 mg/kg/day) + corticosteroid	II	12
	Arm 2: 78	Arm 2: 55 (22–69)	Arm 2: Tacrolimus (0.05 mg/kg/day) + MMF + daclizumab^a
Becker (4)	Arm 1: 297	Arm 1: 53.1 (9.5)	Arm 1: Tacrolimus (0.1–0.15 mg/kg/day) + MMF	I	3
	Arm 2: 305	Arm 2: 53.7 (9.4)	Arm 2: Tacrolimus (0.1–0.15 mg/kg/day) + daclizumab^a
Lupo (18)	Arm 1: 21	Arm 1: 53.7 (30–65)	Arm 1: Cyclosporine A (10 mg/kg) + corticosteroid (S group)	II	21.6
	Arm 2: 26	Arm 2: 50.5 (34–64)	Arm 2: Cyclosporine A (10 mg/kg) + basiliximab^a (B group)
Schmeding (14)	Arm 1: 48	Arm 1: 49.4	Arm 1: Tacrolimus (10–15 ng/dL) + corticosteroid	III	76–114
	Arm 2: 51	Arm 2: 49.6	Arm 2: Tacrolimus (10–15 ng/dL) + corticosteroid + basiliximab^a
Boillot (5)	Arm 1: 347	Arm 1: 51.0 (9.8)	Arm 1: Tacrolimus (0.075 mg/kg/day) + corticosteroid (500 mg) + corticosteroid (15–20 mg/day)	II	3
	Arm 2: 351	Arm 2: 50.9 (10.4)	Arm 2: Tacrolimus (0.075 mg/kg/day) + corticosteroid (500 mg) + daclizumab^a
Kato (19)	Arm 1: 20	Arm 1: 50 (44–67)	Arm 1: Tacrolimus (8–12 ng/mL) + MMF + corticosteroid	II	15
	Arm 2: 19	Arm 2: 52 (42–68)	Arm 2: Tacrolimus (8–12 ng/mL) + MMF + daclizumab^a
Yoshida (21)	Arm 1: 76	Arm 1: 52.4 (8.04)	Arm 1: Tacrolimus (4–8 ng/mL) + MMF + corticosteroid	III	12
	Arm 2: 72	Arm 2: 53.0 (8.78)	Arm 2: Tacrolimus (4–8 ng/mL) + MMF + corticosteroid + daclizumab^a
Neuhaus (22)	Arm 1: 193	Arm 1: 50.2 (20–72)	Arm 1: Cyclosporine A (15 mg/kg) + corticosteroid + placebo	III	12
	Arm 2: 188	Arm 2: 49.0 (20–68)	Arm 2: Cyclosporine A (15 mg/kg) + corticosteroid + basiliximab^a
Washburn (23)	Arm 1: 15	Arm 1: 56.4 (8.5)	Arm 1: Tacrolimus (0.10 mg/kg/day) + MMF + corticosteroid	III	15–21
	Arm 2: 15	Arm 2: 48.1 (8.6)	Arm 2: Tacrolimus (0.10 mg/kg/day) + MMF + corticosteroid (day 1 only) + daclizumab^a

Data presented as mean, mean (SD) or medium (range); MMF: mycophenolate mofetil.

^a IL2RA (Interleukin-2 receptor antibodies).

Type I, IL2RA induction therapy versus steroid-free therapy; Type II, IL2RA induction therapy versus steroid therapy; Type III, steroid plus IL2RA therapy versus steroid therapy.

Table 2. Summary of adverse events in studies selected for meta-analysis.

	Trunecka (6)	Neumann (14)	Klintmalm (20)	Neuberger (16)	Otero (15)	Becker (4)	Lupo (18)	Schmeding (14)	Boillot (5)	Kato (19)	Yoshida (21)	Neuhaus (22)
Infection	NA	75%	35% versus 36%	NA	19% versus 11%	3% versus 8%	23% versus 19%	58% versus 58%	11% versus 11%	14% versus 37%	17% versus 28%	80% versus 83%
Viral infection	NA		NA	NA	NA	6% versus 8%	15% versus 24%	41% versus 32%	5% versus 12%	NA	NA	11% versus 11%
HCV recurrence	NA	42% versus 63%	68% versus 73%	10% versus 4%	23% versus 27%	7% versus 5%	46% versus 38%	2% versus 2%	NA	NA	NA	7% versus 13%
Renal failure	20% versus 26%	30% versus 34%	NA	8% versus 17%	19% versus 28%	26% versus 22%	NA	NA	28% versus 24%	NA	NA	NA
Anemia	30% versus 34%	33% versus 31%	NA	NA	31% versus 29%	16% versus 21%	NA	NA	NA	NA	NA	NA
Leucopenia	18 versus 17%	15% versus 22%	NA	16% versus 9%	35% versus 8%	4% versus 10%	NA	NA	NA	NA	NA	NA
Thrombocytopenia	16% versus 17%	28% versus 9%	NA	10% versus 9%	13% versus 15%	NA	NA	NA	NA	NA	NA	NA
PTDM	NA	2% versus 6%	19% versus 33%	39% versus 43%	9% versus 14%	10% versus 11%	4% versus14%	10% versus 21%	6% versus 15%	7% versus 43%	NA	NA
Hypertension	NA	25% versus 29%	67% versus 75%	27% versus 35%	21% versus 27%	42% versus 43%	NA	NA	13% versus 14%	36% versus 58%	NA	NA
Hyperglycemia	16% versus 17%	13% versus 19%	NA	NA	27% versus 24%	18% versus 19%	15% versus 19%	NA	12% versus 13%	NA	NA	NA

Data presented as mean percentage in arm(s) with IL2RA versus arm(s) without IL2RA.

IL2RA: anti-interleukin-2 receptor antibodies; HCV: hepatitis C virus; SVA: supraventricular arrhythmia; PTDM: posttransplant diabetes mellitus; NA: not available.

Meta-analysis of overall studies

There was no significant heterogeneity in the rate of acute rejection events between the studies (Q = 23.986, I²=49.97%, p = 0.020). The overall analysis revealed that patients treated with IL2-RA-based induction therapy were less likely to suffer acute rejection compared to patients in the control arm (steroid or steroid-free) [pooled OR_{random effects: =}0.75 (95%CI: 0.57–0.98), p = 0.038; pooled OR_{fixed effect} = 0.80] (Figure 2(A)). Twelve trials with a total of 3435 patients reported mortality rate. No significant heterogeneity in the mortality rate was found between the studies (Q = 15.587, I²=29.43%, p = 0.157). Overall, no significant difference in mortality rate was found between the IL2-RA-based induction group (IL-2RA arm) and the control arm (steroid or steroid-free) [pooled OR_{random effects}= 0.90 (95%CI: 0.68–1.20), p = 0.493; pooled OR_{fixed effect} = 0.91] (Figure 3(A)).

Figure 2. Meta-analysis for acute rejection events: (A) Subgroup analysis according to different type of IL2-RA (basiliximab or daclizumab) versus control arm (steroid or steroid-free); (B) Subgroup analysis of “steroid plus IL2RA versus steroid” (Type III) or “IL2RA versus steroid” (Type II) induction therapy.

Figure 3. Meta-analysis for mortality: (A) Subgroup analysis according to different type of IL2-RA (basiliximab or daclizumab) versus control arm (steroid or steroid-free). (B) Subgroup analysis of “steroid plus IL2RA versus steroid” (Type III) or “IL2RA versus steroid” (Type II) induction therapy.

Subgroup analysis for different IL2-RAs

The subgroup of patients treated with basiliximab-based induction therapy (IL-2RA arm) was analyzed in four studies (6,14,18,22). No significant difference was found in the rate of acute rejection events between basiliximab-based induction therapy (IL-2RA arm) and control (pooled OR_{random effects}= 0.77, 95%CI: 0.49–1.22, p = 0.265) (Figure 2(A)). In addition, patients treated with basiliximab-based induction therapy showed no significant difference in the mortality rate compared to patients in the control group [pooled OR_{random effects}= 0.74 (95%CI: 0.49–1.13), p = 0.167; pooled OR_{fixed effect} = 0.78] (Figure 3(A)).

The subgroup of patients treated with daclizumab-based induction therapy was compared to the control arm (steroid-free or steroid) (4,5,15–17,19–21,23). The results indicated that there was no significant difference in the rate of acute rejections between patients treated with daclizumab-based induction therapy and control arm (steroid-free or steroid) [pooled OR_{random effects}= 0.73 (95%CI: 0.52–1.03), p = 0.074; pooled OR_{fixed effect} = 0.82] (Figure 2(A)). In addition, no significant difference was found in the mortality rate between patients treated with daclizumab and control arm [pooled OR_{random effects}=1.07 (95%CI: 0.73–1.59), p = 0.722; pooled OR_{fixed effect} = 1.03] (Figure 3(A)).

Subgroup analysis for different experimental designs

For patients treated with daclizumab-based induction therapy, two subgroups with different experimental design, “steroid plus IL2RA versus steroid” (Type III) (16,21,23) and “IL2RA versus steroid” (Type II) (5,15,17,19,20), were analyzed for acute rejection and mortality rates. There was no significant difference in the acute rejection rate between patients treated with steroids and IL2RA induction therapy (IL-2RA arm) and steroids only [pooled OR_{random effects}=0.82 (95% CI: 0.36–1.89), p = 0.642; pooled OR_{fixed effect} = 0.67, Figure 2(B) and mortality [pooled OR_{random effects}=0.95 (95% CI: 0.34–2.63), p =0.920; pooled OR_{fixed effect} = 0.84, Figure 3(B). Similarly, there was no significant difference in the acute rejection rates between patients treated with IL2RA induction therapy (IL-2RA arm) and steroids [pooled OR_{random effects}=0.56 (95%CI: 0.30–1.06), p = 0.074, Figure 2(B) and mortality [pooled OR_{random effects}=1.29 (95% CI: 0.67–2.48), p =0.442; pooled OR_{fixed effect} = 1.16, Figure 3(B).

Sensitivity analysis and publication bias

Sensitivity analyses were performed using the leave-one-out approach (Table 3). The direction of combined estimates did not vary markedly with the removal of individual studies. Thus, the meta-analysis had good reliability, and the data were not overly influenced by each study. The results of Egger’s test showed no publication bias (Figure 4).

Figure 4. Funnel plot for the assessment of potential publication bias: (A) acute rejection events; (B) mortality.

Table 3. Sensitivity-analysis.

	Statistics with study removed
First author (year)	Points	Lower limit	Upper limit	Z value	p value
Acute rejection
Trunecka (6)	0.76	0.57	1.02	−1.84	0.066
Neumann (14)	0.78	0.60	1.02	−1.84	0.066
Klintmalm (20)	0.73	0.56	0.95	−2.33	0.020
Neuberger (16)	0.78	0.59	1.03	−1.78	0.074
Otero (15)	0.79	0.61	1.02	−1.81	0.071
Becker (4)	0.70	0.54	0.93	−2.51	0.012
Lupo (18)	0.76	0.58	0.99	−2.01	0.045
Schmeding (14)	0.72	0.55	0.94	−2.39	0.017
Boillot (5)	0.72	0.53	0.97	−2.16	0.031
Kato (19)	0.79	0.63	0.99	−2.02	0.043
Yoshida (21)	0.74	0.56	0.98	−2.07	0.039
Neuhaus (22)	0.74	0.55	1.00	−1.96	0.050
Washburn (23)	0.74	0.57	0.95	−2.36	0.018
Mortality
Trunecka (6)	0.90	0.64	1.26	−0.62	0.535
Neumann (14)	0.87	0.69	1.09	−1.20	0.230
Klintmalm (20)	0.94	0.69	1.29	−0.36	0.720
Neuberger (16)	0.96	0.72	1.30	−0.24	0.810
Otero (15)	0.89	0.66	1.21	−0.74	0.458
Becker (4)	0.91	0.65	1.26	−0.57	0.566
Lupo (18)	0.95	0.72	1.25	−0.36	0.717
Schmeding (14)	0.96	0.71	1.28	−0.29	0.772
Boillot (5)	0.89	0.64	1.22	−0.74	0.459
Yoshida (21)	0.87	0.66	1.16	−0.94	0.349
Neuhaus (22)	0.94	0.67	1.30	−0.40	0.691
Washburn (23)	0.92	0.68	1.24	−0.58	0.563

Assessment of risk of bias

Risk of potential bias of each individual study and the risk of bias of all included studies are shown in Figures 5(A) and (B), respectively. The major sources of bias were performance and selection biases, since none of the included studies could successfully blind the patients or outcome assessors.

Figure 5. Summary of quality assessment. (A) Risk of potential bias of individual studies. (B) Risk of bias of all included studies.

Discussion

Many transplant teams are reducing or avoiding corticosteroids in the induction and maintenance immunosuppressive therapies due to their side effects (24). Antibody induction therapy plays a major role in preventing acute rejection within the first few weeks after liver transplantation in corticosteroid-lacking immunosuppressive regimens. Although two similar systematic reviews had been published in 2014 (3,8), this systematic review and meta-analysis includes two recent randomized controlled clinical trials conducted in European countries (6,17). We showed that treatment with IL-2RA-based therapy significantly decreased the incidence of acute rejection but it did not improve overall survival. Overall, our results are in agreement with results of previous meta-analyses. Similar to our results, it was shown that IL2-RA therapy substantially reduces the risk of acute rejection but does not reduce mortality (8,25). Lack of effect on mortality with antibody induction treatment can be explained by multiple causes of death in patients undergoing liver transplantation. Acute rejection and infection remain the greatest factors affecting survival after organ transplantation. Although acute rejection is a critical factor that affects graft survival, some cases of mortality after daclizumab treatment appeared to be related to higher incidence of severe infections (26). Therefore, reduction in the risk of acute rejection may not necessarily be associated with reduced mortality.

Basiliximab and daclizumab were proven to be effective for prophylaxis of acute rejection after renal transplantation, and they were well tolerated in clinical trials (27). Efficacy of basiliximab and daclizumab had been compared in previous RCTs, but results are controversial. In our study, subgroup analysis for two types of IL-2RAs showed that patients who received either basiliximab-based induction therapy or daclizumab-based induction therapy did not have significantly lower acute rejection rate compared with patients in the control arm. Our finding is not consistent with results of meta-analysis reported by Wang et al. (28). They found that basiliximab did not significantly reduce the incidence of acute rejection events within one-year post-transplantation, but daclizumab did. The discrepancy between Wang’s report and our study may be due to the difference in experimental designs of included studies. We included primary research articles that described RCTs recruiting patients who received liver transplantation and were treated with IL-2RA-based induction therapy (with or without steroids) or control therapy (steroid or steroid-free). Wang et al. selected all RCTs described in published articles or abstracts that had one arm evaluating basiliximab or daclizumab treatments (29). In addition, the treatment protocols and efficacious dosages of basiliximab and daclizumab were different among the trials. It was reported that two doses induction therapy with basiliximab or five dose induction therapy with daclizumab reduces the incidence and severity of acute rejection, but use of less than five daclizumab doses was not supported by phase III trials (30,31). Interestingly, basiliximab and daclizumab seem to be equally effective in preventing acute rejection in patients receiving renal transplantation (relative risk= 0.67 versus 0.66 for basiliximab versus daclizumab) (32).

Although daclizumab was widely used for more than a decade, it was discontinued in the USA in 2009 due to the availability of more popular alternatives and decreasing demand (26). Despite discontinuation, the clinical reports on daclizumab can still be found in 2014 and 2015 (29,33,34). Recently, daclizumab was proposed as a possible treatment agent for multiple sclerosis and may return to the market (35).

We included clinical trials that allowed adjustments of the maintenance therapy (tacrolimus or cyclosporine, alone or with MMF). In addition, there were substantial differences in the regimens of IL-2RA-based induction therapy (experimental arm) and control therapy (control arm) in the included studies. To determine the efficiency of either steroid plus IL2RA therapy or IL2RA therapy alone compared to control steroid therapy, we performed the subgroup analysis. No significant difference in the rates of acute rejection events and mortality was found in both subgroups. However, the small number of articles included in subgroup analysis limits our conclusions.

The incidence of adverse events varied among the studies. For example, PTDM was reported in eight trials with incidence of less than 10% in antibody induction therapy arm. In contrast, one study reported the diabetes incidence of up to 39% in the antibody induction therapy arm (16). Factors that contribute to this difference in incidence of PTDM warrant further investigation. One of the possible causes can be attributed to the increased steroid dose received by patients in this study. Specifically, participants of this trial received a cumulative steroid dose that rose five fold from 13.1 mg/kg at one-month post-transplantation to 49.4 mg/kg at 9–12 months post-transplantation (16). In comparison, the studies with less than 10% incidence of PTDM reported the use of lower dosages of corticosteroids or no steroids. For example, a bolus of 0.5 mg or 1 g corticosteroid dosage was administered to patients at transplantation and a maintenance dose of 15–20 mg was continued (5,15); and no further corticosteroids were administered in the daclizumab arm (4,5,19). For the trial reported by Kato et al., control arm (tacrolimus + corticosteroids + MMF) had 43% of PTDM comparing to 7% of PTDM in antibody induction therapy arm (tacrolimus + daclizumab + MMF). Taken together, these results strongly suggest that high doses of corticosteroids for extended time periods play a major role in development of PTDM.

Infections were reported in 9 out of 11 trials. The rate of infections was much higher in three studies (14,20,22), ranging from 35% to 80%. Klintmalm et al. (20) reported infection in 35% of patients in steroid-free antibody induction therapy arm (tacrolimus + MMF + daclizumab), suggesting that steroid treatment was not the likely cause of infection at least in this group of patients. On the other hand, patients in the Klintmalm et al. trial received liver transplantation for hepatitis C, and were expected to have higher rate of complications.

This meta-analysis should be interpreted in the context of several limitations. The included studies used different immunosuppressive regimens (tacrolimus versus cyclosporine A), different drug dosages (tacrolimus and corticosteroid), different IL-2RA-based induction therapy protocols (IL-2RA + steroid versus IL-2RA alone), different length of follow-up (ranged from 3 months to 114 months) and different control arms (steroid versus steroid-free). Although the direction of combined estimates did not vary markedly with the removal of individual studies during the sensitivity analysis, the effect of outlier cannot be ignored. For example, overall analysis indicated that daclizumab-based induction therapy was not associated with reduction of acute rejection events. However, in the trial reported by Washburn et al. (23), higher rate of acute rejections was observed in the antibody induction therapy arm (tacrolimus + MMF + daclizumab) (25%) compared with control arm (tacrolimus + MMF + corticosteroid) (6.7%), which might affect the statistical outcome for the daclizumab-based induction therapy. In addition, inadequate blinding of participants and personnel in some of the included clinical trials contributed to bias that should be accounted for during interpretation of the results.

In conclusion, this meta-analysis found that treatment with IL2RA-based induction therapy reduced acute rejection events but did not reduce overall mortality compared to no induction treatment (steroid or steroid-free therapy). In the subgroup analysis, Basiliximab-based therapy and daclizumab-based therapy were not associated with reduced acute rejection and mortality rates. In the subgroup analysis regarding experimental design, no significant difference in the acute rejection and mortality rates was found between “steroid plus IL2RA versus steroid” and “IL2RA versus steroid” groups. Further trials are warranted to determine the optimal regimen of IL2RA-based induction therapy.

Disclosure statement

The authors report no conflicts of interest.