Do wound complications or lymphoceles occur more often in solid organ transplant recipients on mTOR inhibitors? A systematic review of randomized controlled trials

Authors

  • Liset H. M. Pengel,

    1. Centre for Evidence in Transplantation, Clinical Effectiveness Unit, Royal College of Surgeons of England and the London School of Hygiene and Tropical Medicine, University of London, London, UK
    Search for more papers by this author
  • Liang Q. Liu,

    1. Centre for Evidence in Transplantation, Clinical Effectiveness Unit, Royal College of Surgeons of England and the London School of Hygiene and Tropical Medicine, University of London, London, UK
    Search for more papers by this author
  • Peter J. Morris

    1. Centre for Evidence in Transplantation, Clinical Effectiveness Unit, Royal College of Surgeons of England and the London School of Hygiene and Tropical Medicine, University of London, London, UK
    Search for more papers by this author

  • Conflicts of Interest
    PJM chairs a data safety committee for Bristol-Meyers Squibb and has in the past received lecture fees from Novartis, Astellas, Roche and Genzyme.

Liset HM Pengel PhD, Centre for Evidence in Transplantation, Royal College of Surgeons of England, 35-43 Lincoln’s Inn Fields, London WC2A 3PE, UK. Tel.: +44 20 7869 6627; fax: +44 20 7869 6644; e-mail: lpengel@rcseng.ac.uk

Summary

mTOR inhibitors have been associated with wound complications and lymphoceles. We systematically reviewed randomized controlled trials (RCTs) to compare these outcomes for solid organ transplant recipients. Relevant medical databases were searched to identify RCTs in solid organ transplantation comparing mTOR inhibitors with an alternative therapy reporting on wound complications and/or lymphoceles. Methodological quality of RCTs was assessed. Pooled analyses were performed to calculate odds ratios (OR) and 95% confidence intervals (CI). Thirty-seven RCTs in kidney, heart, simultaneous pancreas-kidney and liver transplantation were included. Pooled analyses showed a higher incidence of wound complications (OR 1.77, CI 1.31–2.37) and lymphoceles (OR 2.07, CI 1.62–2.65) for kidney transplant recipients on mTOR inhibitors together with calcineurin inhibitors (CNIs). There was also a higher incidence of wound complications (OR 3.00, CI 1.61–5.59) and lymphoceles (OR 2.13, CI 1.57–2.90) for kidney transplant recipients on mTOR inhibitors together with antimetabolites. Heart transplant patients receiving mTOR inhibitors together with CNIs also reported more wound complications (OR 1.82, CI 1.15–2.87). We found a higher incidence of wound complications and lymphoceles after kidney transplantation and a higher incidence of wound complications after heart transplantation for immunosuppressive regimens that included mTOR inhibitors from the time of transplantation.

Introduction

The introduction of calcineurin inhibitors (CNIs) has significantly improved the outcome of solid organ transplantation. However, CNIs have been associated with nephrotoxicity and other side-effects [1]. One alternative to CNIs is a new class of immunosuppressants, sirolimus (SRL) and everolimus (EVL) that inhibit the mammalian target of rapamycin (mTOR inhibitors). Quite early after the introduction of mTOR inhibitors into immunosuppressive regimens, it became apparent that the antiproliferative actions of mTOR inhibitors might have an effect on healing as evident by poor wound healing and the occurrence of lymphoceles after renal transplantation [2–4]. This antiproliferative effect on fibroblasts in the healing wound is likely to be the explanation for complications of healing [5]. In particular, tracheal dehiscence was noted after lung transplantation [6].

In 2006, a Cochrane review by Webster et al. [7] evaluated the efficacy and safety of mTOR inhibitors for kidney transplant recipients in the immediate post-transplant period. The authors found that patients treated with mTOR inhibitors showed an increased risk of developing lymphoceles when compared with patients treated with CNIs or antimetabolites. They did not find an increased risk for developing wound complications nor did they find a difference when comparing lower versus higher dose mTOR inhibitors or when comparing lower dose mTOR inhibitors plus standard CNIs versus higher dose mTOR inhibitors plus lower dose CNIs. Their literature search was done up to July 2005 and since then some of the included conference abstracts have been published as full articles and a number of additional, large RCTs have been published that report on wound complications or lymphoceles.

Therefore, the primary aim of the study was to evaluate the occurrence of wound complications and lymphoceles in solid organ transplant recipients receiving mTOR inhibitors from the time of transplantation compared with patients not receiving mTOR inhibitors. We tested the null hypothesis that wound complications and lymphoceles do not occur more commonly in patients receiving mTOR inhibitors than in patients who do not receive mTOR inhibitors.

As it has been suggested that steroid avoidance may reduce the negative impact of mTOR inhibitors on wound healing and lymphocele formation, a secondary analysis was planned to review the effects of mTOR inhibitors plus steroids versus mTOR inhibitors without steroids on these outcomes [8,9].

Methods

Inclusion criteria

For the primary analysis, eligible studies included randomized controlled trials (RCTs) in solid organ transplantation that compared mTOR inhibitors given from the time of transplantation with at least one alternative non-mTOR inhibitor intervention arm. For the secondary analysis, we included RCTs with mTOR inhibitors in both arms and at least one steroid arm and one steroid-free arm. Studies had to report on wound complications or lymphoceles. Wound complications included (superficial or deep) wound infection, (superficial) wound dehiscence, fascial dehiscence, wound debridement, delayed/slow wound healing, abnormal wound healing, partial wound healing, haematoma, seroma, wound inflammation, increased wound drainage, wound haemorrhage, wound secretion or incisional hernia. Studies only reporting on peripheral oedema, fluid overload and oedema were excluded.

Identification of studies

Full reports of RCTs were identified through searches of the Transplant Library, Medline, Embase and the Cochrane Central Register of Controlled Trials up to 24 March 2011 without language restrictions. Search terms in Medline and Cochrane included all MeSH terms for solid organ transplantation and other generic transplantation MeSH terms. The Cochrane highly sensitive search strategy was used to identify RCTs in Medline. Other specific search terms included sirolimus, rapamycin, rapamune, everolimus, ay 22-989, SDZ RAD and Certican. When there was more than one report of the same trial, all reports that reported on wound complications or lymphoceles were included in the review. The reference lists of identified RCTs or reviews were inspected for additional references. Conference abstracts were not included.

Data extractions and methodological quality

The following data were extracted from eligible articles by one reviewer: type of organ, intervention arms, induction therapy, mTOR inhibitor dose, steroid dose, number of participants, follow-up period, description and incidence of wound complications and/or lymphoceles. Methodological quality was assessed independently by two reviewers using both the Jadad score and the items allocation concealment and intention to treat [10,11]. The Jadad score addresses the items randomization, blinding and description of withdrawals and dropouts. The total Jadad score ranges from 0 to 5 with RCTs scoring at least 3 of 5 being considered to be consistent with sound methodological quality. However, also the use of allocation concealment and intention to treat were part of the overall assessment of quality. Intention to treat was defined as an analysis including all randomized participants, that was based on the groups to which participants were originally randomly assigned regardless of whether they satisfied the entry criteria, the treatment actually received and subsequent withdrawal or deviation from the protocol.

Analysis

The RCTs were analysed according to organ and concomitant therapy, i.e. mTOR inhibitors together with CNIs or antimetabolites. If there was more than one RCT with similar interventions, a meta-analysis was performed for the incidence of wound complications or lymphoceles. Data were extracted by two reviewers using a spreadsheet. Review Manager (RevMan) [Computer program] version 5.1.1 (Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2011) was used to pool the incidence of wound complications or lymphoceles among studies. We pooled all data irrespective of the length of follow-up, as most complications would develop within the first few months after transplantation. A subgroup analysis was also performed for studies considered to be of good methodological quality, i.e. studies that used adequate allocation concealment or scored at least 3 points on the Jadad scale. The Mantel–Haenszel random effects model was used to calculate odds ratios (OR). For trials with more than two arms, the incidences of wound complications or lymphoceles for the mTOR inhibitor or control arms were grouped and compared collectively with the comparator. If there were no events in both arms, the RCT was excluded from the quantitative analysis. Statistical heterogeneity was tested using the I2 statistic. A secondary analysis was performed to assess whether complications occur more commonly in patients receiving concomitant steroids compared with no steroids. To assess possible publication bias, funnel plots were created for comparisons that included at least 10 trials [12].

Results

Included studies

The literature search identified a total of 518 unique references of which 37 RCTs met the inclusion criteria for the qualitative analysis (Fig. 1). One study was omitted from the quantitative primary and secondary analyses, as it was reported that no events occurred in either study arms [13]. Thirty-one RCTs evaluated sirolimus whereas six RCTs evaluated everolimus. For the primary quantitative analysis, a total of 28 RCTs were published on kidney transplantation reporting on 8916 patients (Tables 1 and 2, Table S1), [13–43] four RCTs were published on heart transplantation reporting on 1289 patients (Table S2), [44–47] one RCT [48] was published on liver transplantation reporting on 78 patients (Table S3) and one RCT [49] was published on simultaneous kidney and pancreas transplantation reporting on 123 patients (Table S4). In kidney transplantation, 14 RCTs evaluated mTOR inhibitors together with CNIs, 13 RCTs evaluated mTOR inhibitors together with antimetabolites and 1 RCT evaluated an mTOR inhibitor together with belatacept. For all RCTs in heart transplantation, mTOR inhibitors were given together with CNIs. For the secondary quantitative analysis, i.e. evaluation of the possible impact of mTOR inhibitors combined with steroids versus no steroids, two RCTs on kidney transplantation were identified (Table S5) [50,51]. Most trials on kidney transplantation reported on both wound complications and lymphoceles (Table S6).

Figure 1.

 Flowchart of the selection of articles. SPK, simultaneous pancreas-kidney; mTOR-I, mTOR inhibitors.

Table 1.   Studies evaluating mTOR inhibitors plus calcineurin inhibitors for kidney transplantation (n = 15).
ReferencenComparisonInductionmTOR-I dosePred maintenance doseStudy periodOutcomesMethodological Quality
Loading doseMaintenanceJadad (0–5)ACITT
  1. SRL, sirolimus; Tac, tacrolimus; CsA, cyclosporine A; MMF, mycophenolate mofetil; MPA, mycophenolic acid; AZA, azathioprine; Pred, prednisolone; AC, allocation concealment; ITT, intention to treat.

  2. *Data regarding wound complications and lymphoceles were not included in the article but were provided by the author upon request. All wound complications were reported as serious adverse events.

  3. †For the first 21 months of this trial, a loading dose of 10 mg was given. Thereafter, no loading dose was given.

  4. ‡SRL groups were randomized a second time (1:1) after 3 months to discontinue SRL at either the end of month 3 or the end of month 5.

  5. §It was stated that there were no reports for lymphoceles and therefore the study was excluded from the quantitative meta-analysis.

Van Gurp [40]659I. SRL, Tac, pred
II. Tac, MMF, pred
None6 mg2 mg/day for 28 days, 1 mg thereafter125 mg bolus on day 1. From day 2 20 mg/day tapered to 5 mg/day by day 90. Discontinued thereafter.6 monthsWound complications, lymphoceles*2YesNo
Tedesco Silva [39]833I: 1.5 mg everolimus, reduced CsA, pred
II. 3 mg everolimus, reduced CsA, pred
III. MPA, standard CsA, pred
BasiliximabNoneI: 3–8 ng/ml
II: 6–12 ng/ml
According to local practice.12 monthsWound complications, lymphoceles3YesYes
Sampaio [38]100I. SRL, Tac, pred
II. MMF, Tac, pred
None15 mg2 mg/dayStarted at 30 mg/day with target of 10 mg/day by 3 months; target of 5 mg/day at 6 months.12 monthsWound complications, lymphoceles3NoYes
Anil Kumar [15]200I. CsA, MMF
II. CsA, SRL
III. Tac, MMF
IV. Tac, SRL
Basiliximab, methylprednisoloneNoneInitiated on day 4 at 2 mg/day. Target was 5–10 ng/mlNone5 yearsWound complications3NoYes
Vitko [42]977I. 0.5 mg SRL, Tac, pred
II. 2 mg SRL, Tac, pred
III. MMF, Tac, pred
NoneI: 1.5 mg; II: 6 mgI: 0.5 mg; II: 2 mgDay 1: 125 mg; subsequently tapered to 20 mg/day (day 14), 15 mg/day (day 28), 10 mg/day (day 42); 5 mg/day thereafter.6 monthsWound complications, lymphoceles2YesYes
Vitko [41]588I: 1.5 mg everolimus, CsA, pred
II. 3 mg everolimus, CsA, pred
III. MMF, CsA, pred
NoneNoneI: 0.75 mg everolimus b.i.d; II 1.5 mg everolimus b.i.d.Tapered from 20 mg/day on day 1 to ≥5 mg/day at 6 months.3 yearsLymphoceles3NoYes
Lorber [33]583I: 1.5 mg everolimus, CsA, pred
II. 3 mg everolimus, CsA, pred
III. MMF, CsA, pred
NoneNoneI: 0.75 mg everolimus b.i.d; II 1.5 mg everolimus b.i.d.Day 1: 500 mg. Tapered to 20 mg/day by day 30 and ≥5 mg/day for at least 6 months.3 yearsLymphoceles3NoYes
Kandaswamy [31]239I. CsA, MMF, pred (0–5 days)
II. High Tac, low SRL, pred (0–5 days)
III. Low Tac, high SRL, pred (0–5 days)
Thymoglobulin10 mg/None†I: 3–7 ng/ml; II: 8–12 ng/mlDay 1: 1 mg/kg; days 2–3 0.5 mg/kg; day 4–5 0.25 mg/kg. Discontinued thereafter.Mean 16 monthsWound complications2YesYes
Anil Kumar [14]150I: Tac, MMF
II: Tac, SRL
Basiliximab/methylprednisoloneNoneInitiated on day 4 at 2 mg/day. Target was 6–10 ng/mlNoneMedian 400 daysWound complications, lymphoceles3NoNo
Machado [35]70I: SRL, CsA, pred
II. AZA, CsA, pred
None6 mg2 mg/day0–30 days: 0.5 mg/kg/day (max. 30 mg/day). Tapered to 20 mg/day by 2 months and 10 mg/day between 3–6 months.12 monthsWound complications, lymphoceles1NoYes
Ciancio [18]150I. SRL, Tac, pred
II. SRL, CsA, pred
III. MMF, Tac, pred
Daclizumab4 mg8 ng/mlDays 1–3: 500 mg/day. Tapered to 0.3 mg/kg at 1 month and 0.15 mg/kg at 3 months.12 monthsWound complications, lymphoceles2NoYes
Van Hooff [13]104I: Tac, pred
II: 0.5 mg SRL, Tac, pred‡
III: 1 mg SRL, Tac, pred‡
IV: 2 mg SRL, Tac, pred‡
NoneI: 1.5 mg; II: 3 mg; III: 6 mgI: 0.5 mg; II: 1 mg; III: 2 mgDay 1: 125 mg; subsequently tapered to 20 mg/day (day 14), 15 mg/day (day 28), 10 mg/day (day 42); 5 mg/day thereafter.6 monthsLymphoceles§2NoNo
MacDonald [28,29,34]576I: 2 mg/day SRL, CsA, pred
II: 5 mg/day SRL, CsA, pred
III: placebo, CsA, pred
NoneI: 6 mg; II: 15 mgI: 2 mg/day; II: 5 mg/dayDay 0: 250 mg. Tapered to 30 mg/day by day 7, 10 mg/day by month 6 and 5–10 mg/day thereafter.12 monthsWound complications, lymphoceles5YesYes
Kahan [27]719I. 2 mg SRL, CsA, Pred
II. 5 mg SRL, CsA, Pred
III: AZA, CsA, Pred
None6 ml2 mg/day; 5 mg/dayDay 0: 500 mg. Tapered to 30 mg/day by day 6, 10 mg/day by month 6, 5–10 mg/day thereafter.12 monthsWound complications, lymphoceles5YesYes
Kahan [30]149I. 1 mg/m2/day SRL, high CsA, pred
II. 3 mg/m2/day SRL, high CsA, pred
III. 1 mg/m2/day SRL, low CsA, pred
IV. 3 mg/m2/day SRL, low CsA, pred
V. 5 mg/m2/day SRL, low CsA, pred
VI. high CsA, pred
None3 times the study doseI. 1 mg/m2/day II. 3 mg/m2/day
III. 1 mg/m2/day
IV. 3 mg/m2/day
V. 5 mg/m2/day
Day 0 500 mg. Tapered to 40 mg/day by day 6, 20–30 mg/day for weeks 1–6 and ≤10 mg/day thereafter.12 monthsWound complications2NoNo
Table 2.   Studies evaluating mTOR inhibitors plus antimetabolites for kidney transplantation (n = 13).
ReferencenComparison (n)InductionmTOR-I dosePred maintenance doseStudy periodOutcomesMethodological Quality
LoadingMaintenanceJadad (0–5)ACITT
  1. ATG, antithymocyte globulin; MMF, mycophenolate mofetil; SLR, sirolimus; AZA, azathioprine; Tac, tacrolimus; Pred, prednisolone; AC, allocation concealment; ITT, intention to treat.

Franz [23]127I. SRL, MMF, pred
II. CsA, MMF, pred
None30 mg/day for 3 daysStarted at 16 mg to reach 10–20 ng/ml during months 1–3; months 4–6: 8–15 ng/ml1000, 500, 250 mg on days 0, 1, 2 respectively. Then 0.5 mg/kg/day, tapered every other week by 5 mg to 15 mg, then tapered by 2.5 mg to 7.5 mg6 monthsWound complications, lymphoceles2NoYes
Glotz [24]141I. SRL, MMF, pred
II. Tac, MMF, pred
Group I: ATG15 mg/day for 2 daysStarted at 10 mg/day for 5 days. Target 12–20 ng/ml from day 8 to 1 year250 mg on day 0. Days 1–10: 0.3–0.5 mg/kg/day, days 11–30: 0.15–0.3 mg/kg/day, days 31–90: 0.1–0.15 mg/kg/day, 0.1 mg/kg/day thereafter12 monthsWound complications, lymphoceles2NoYes
Guba [26]141I. SRL, MMF, pred
II. CsA, MMF, pred
ATG0.1 mg/kgStarted at 2–4 mg/day. Target 8–12 ng/ml for months 0–3 and 5–10 ng/ml thereafterDay 0: 500 mg. Maintenance according to centre practice. Minimum 20 mg for weeks 0–2, 15 mg for weeks 3–8, 10 mg for months 2–4 and 2 mg thereafter12 monthsWound complications, lymphoceles3YesYes
Durrbach [20]72I. SRL, MMF, pred
II. CsA, MMF, pred
ATG30 mg/day for 2 daysStarted at 10 mg/day to reach 10–20 ng/mlAccording to centre practice6 monthsWound complications, lymphoceles2YesNo
Pescovitz [37]60I. SRL, MMF, pred
II. CsA, MMF, pred
Daclizumab15 mg/day for 3 days0–2 months: 10–25 ng/ml; 8–15 ng/ml thereafterAccording to centre practice.6 monthsWound complications2NoYes
Ekberg [21]1645I: low dose SRL, MMF, pred
II: standard CsA, MMF, pred
III: low dose CsA, MMF, pred
IV: low dose Tac, MMF, pred
Groups I, III, IV: Daclizumab 0–2 months post-transplant9 mg/day for 3 days4–8 ng/mlAccording to centre practice. Minimum at 0–2 weeks: 20 mg, 3–8 weeks: 15 mg, 2–4 months: 10 mg, 5 mg thereafter12 monthsWound complications, lymphoceles3YesNo
Barsoum [16]113I. Pred, CsA (0–3 mo), MMF (from 3 mo), SRL
II. Pred, CsA, MMF
None9 mg0–3 months: 5–10 ng/ml; 10–15 ng/ml thereafter0–1 month: 20–50 mg/day; 1–3 months: 10–20 mg/day; 3–6 months:10 mg/day; 6–12 months:5–10 mg/day; 12–24 months 0–5 mg/day24 monthsWound complications, lymphoceles0NoNo
Büchler [17]145I. SRL, MMF, pred
II: CsA, MMF, pred
ATG15 mg for 2 days10–15 ng/mlDay 0: 500 mg. Day 1–7: 1 mg/kg/day; days 8–14: 0.5 mg/kg/day. From day 14 progressive decrease to complete discontinuation at end of month 512 monthsWound complications, lymphoceles3NoYes
Flechner [22]61I.: SRL, MMF, pred
II. CsA,, MMF, pred
Basiliximab15 mg0–6 months: 10–12 ng/ml; 5–10 ng/ml thereafter500 mg/day for 3 days. Tapered from 120 mg to 30 mg by day 8, 27.5 mg by day 21 and 25 mg by day 30. Then tapered by 2.5 mg/month to 7.5 mg daily5 yearsWound complications, lymphoceles3NoYes
Martinez-Mier [36]41I. SRL, MMF, pred
II. CsA, MMF, pred
Basiliximab if HLA match was less than three.10 mg0–6 months: 10–15 ng/ml; 5–10 ng/ml thereafterStarted at 1 g intraoperatively. Tapered to 20 mg/day by day 6 and 5 mg/day at 6 months12 monthsWound complications2NoYes
Dean [19]123I. SRL, MMF, pred
II. Tac, MMF, pred
ATGStarted on day 4: 10 mg/day for 2 daysStarted at 5 mg/day to target trough levels of 15–20 ng/ml during months 0–4. 10–15 ng/ml thereafterDay 0: 500 mg. 0–1 month: 20 mg/day, from 3 months 5 mg/dayMean 21 monthsWound complications2NoNo
Kreis [32]78I. SRL, MMF (0–6 months), pred
II. CsA, MMF (0–6 months), pred
None24 mg/m2 for days 0–30–2 months: 30 ng/ml, 15 ng/ml thereafterDay 0: 500 mg. Day 1: 200 mg/day, tapered to 30 mg/day by day 7 and 10 mg/day by month 6; 6–12 months: 5–10 mg/day12 monthsWound complications2NoYes
Groth [25]83I. SRL, AZA, pred
II: CsA, AZA, pred
None16–24 mg/m2/dayStarted at 8–12 mg/m2/day. Adjusted to reach 30 ng/ml during months 0–2; 15 ng/ml thereafterDay 0: 500 mg. From day 1: 200 mg/day. Tapered to 30 mg/day by day 7 and 10 mg/day by month 612 monthsWound complications3YesYes

Methodological quality

For the primary analysis, nearly half of all RCTs (46%) were considered to be of good methodological quality according to the Jadad scale. Of the 35 RCTs, 14 trials described an appropriate method to generate the randomization sequence. Five trials were double-blinded, four of which adequately described the method of double-blinding. An adequate description of withdrawals and dropouts was given for 31 of 35 trials. More than half of all RCTs (63%) used intention to treat to analyse data and about one-third (37%) of trials adequately described allocation concealment. For the secondary analysis, the Jadad score of the two trials was inadequate. Only Montagnino et al. [50] adequately described allocation concealment and analysed the data according to intention to treat.

Wound complications and lymphoceles

Kidney transplantation

Pooled analyses showed that kidney transplant patients receiving mTOR inhibitors together with CNIs reported more wound complications (12 trials, n = 4787; OR 1.77, 95% confidence interval (CI) 1.31–2.37) and more lymphoceles (11 trials, n = 5370; OR 2.07, CI 1.62–2.65) than patients not receiving mTOR inhibitors (Figs 2 and 3). The heterogeneity was minimal (I= 0% for both analyses). A subgroup analysis of RCTs considered to be of good methodological quality also showed a higher incidence of wound complications and lymphoceles for patients on mTOR inhibitors (Table S7). The funnel plot for both analyses showed asymmetry suggesting that smaller studies with and without significant effects could have remained unpublished.

Figure 2.

 Forest plot indicating the odds ratio of the occurrence of wound complications in kidney transplant recipients on mTOR inhibitors plus calcineurin inhibitors.

Figure 3.

 Forest plot indicating the odds ratio of the occurrence of lymphoceles in kidney transplant recipients on mTOR inhibitors plus calcineurin inhibitors.

Kidney transplant patients receiving mTOR inhibitors together with antimetabolites reported more wound complications (13 trials, n = 2757; OR 3.00, CI 1.61–5.59) and lymphoceles (8 trials, n = 2372; OR 2.13, CI 1.57–2.90) than patients not receiving mTOR inhibitors (Figs 3 and 4 & Figure S1). The heterogeneity for the incidence of wound complications was substantial (I= 59%) but not for the incidence of lymphoceles (I= 0%). The subgroup analysis of RCTs considered to be of good methodological quality also showed a higher incidence of wound complications for patients on mTOR inhibitors (Table S7). The substantial heterogeneity of 60% was thereby reduced to 0% in the subgroup analysis indicating that heterogeneity could have been due to differences in methodological quality. The subgroup analysis for the incidence of lymphoceles found a higher incidence for patients on mTOR inhibitors when the Jadad score was at least 3, however, this was not found for studies using concealed allocation. The funnel plot for the analysis of wound complications showed asymmetry suggesting that small studies without significant effects could have remained unpublished.

Figure 4.

 Forest plot indicating the odds ratio of the occurrence of wound complications in kidney transplant recipients on mTOR inhibitors plus antimetabolites.

For both the analyses, i.e. mTORs together with either CNIs or antimetabolites, there was no increase or decrease of the incidence of wound complications or lymphoceles over time.

One RCT evaluated wound complications and lymphoceles in 89 kidney transplant recipients comparing mTOR inhibitors together with belatacept versus belatacept and MMF versus tacrolimus and MMF [43]. There were no cases of wound dehiscence in the mTOR group versus three cases of wound dehiscence in the non-mTOR groups and one case of lymphoceles in the mTOR group versus one case of lymphoceles in the non-mTOR groups.

Heart transplantation

Heart transplant patients receiving mTOR inhibitors together with CNIs reported more wound complications (four trials, n = 1278; OR 1.82, CI 1.15–2.87) than patients not receiving mTOR inhibitors (Supplemental Figure S2) [44–46]. Heterogeneity was minimal (I= 5%).

Liver transplantation

One RCT evaluated wound complications in 78 liver transplantation receiving either everolimus with early CsA withdrawal versus standard CsA and MMF [48]. Forty-six per cent of patients in the everolimus arm versus 27% of patients in the CsA arm experienced incisional hernias (P = 0.16) and 21% versus 31% experienced biliary complications (P = 0.51).

SPK transplantation

One trial evaluated wound complications in 123 SPK transplant patients receiving sirolimus together with tacrolimus versus MMF together with tacrolimus [49]. Impaired wound healing occurred in 13 patients of the sirolimus group versus 10 patients in the MMF group. There were four cases of lymphoceles in the sirolimus group versus three cases of lymphoceles in the MMF group.

Subgroup analysis: Steroid avoidance

Sandrini et al. [51] compared early steroid withdrawal at day 5 with late steroid withdrawal at month 6. Montagnino et al. [50] compared early steroid withdrawal at day 7 with continued low dose steroid use. Sandrini et al. found a higher incidence of wound healing complications in the late steroid withdrawal group compared with the early steroid withdrawal group (21% vs. 4%, P = 0.02). A pooled analysis of the two RCTs for the incidence of lymphoceles showed that patients in the early steroid withdrawal groups experienced less lymphoceles compared with the late steroid withdrawal and continued low dose steroid groups (n = 229; OR 0.19, CI 0.04–0.88). Heterogeneity was minimal (I= 0%).

Discussion

Meta-analysis of the available data showed a higher incidence of wound complications and lymphoceles after kidney transplantation and a higher incidence of wound complications after heart transplantation for immunosuppressive regimens that included mTOR inhibitors versus regimens that did not include mTOR inhibitors from the time of transplantation.

Wound complications and lymphoceles typically appear within the first few months after transplantation and thus, mTOR inhibitors should be avoided from the time of transplantation for probably 3 months to prevent such problems. However, one RCT comparing immediate introduction of everolimus with delayed introduction after 4 weeks in kidney transplantation found no differences in wound healing complications at 3 months or 12 months [52,53]. A number of RCTs have evaluated the late conversion to sirolimus at 3 months post-transplant but none of these studies report on the occurrence of wound complications or lymphoceles possibly because these problems are unlikely to occur after such time period [54–56].

As a result of its anti-inflammatory effect, steroid use can lead to poor wound healing [57]. This systematic review identified one RCT that showed a higher incidence of poor wound healing for late steroid withdrawal compared with early steroid withdrawal in kidney transplantation. A pooled analysis of two RCTs showed less lymphoceles for early steroid withdrawal compared with late steroid withdrawal or continued use of low dose steroids in kidney transplantation. A nonrandomized study on kidney transplantation comparing sirolimus without steroids with historical controls on long-term maintenance steroids also showed a lower incidence of lymphoceles for the steroid avoidance group but no differences between groups for wound hernia or wound dehiscence [8]. However, because of insufficient level 1 evidence, we cannot draw a definitive conclusion regarding the impact of steroids on wound healing and lymphoceles.

Several studies have evaluated risk factors for impaired wound healing. Knight et al. [58] conducted a retrospective review of wound complications in kidney transplantation. They found that older recipient age, obesity, Caucasian race, thymoglobulin induction and cumulative use of at least 35 mg sirolimus within 4 days post-transplant were independent risk factors for wound complications, which also included lymphoceles. Flechner et al. [59] divided a cohort of 513 consecutive patients into three groups according to their immunosuppression. Multivariate analysis showed that body mass index (BMI) and delayed graft function were risk factors for wound complications. Tiong et al. [60] aimed to develop a systematic approach to reduce wound complications in sirolimus-treated kidney transplant recipients. They concluded that for wound complications or wound complications needing surgery, a BMI > 32 was the strongest independent predictor. For lymphocele formation and lymphoceles needing treatment, a BMI > 32 and acute rejection were independent risk factors. Thus, even though none of these risk factors were identified from randomized controlled trials, these studies indicate that patient characteristics could also contribute to wound complications and lymphocele formation.

In conclusion, immediate use of mTOR inhibitors leads to a higher incidence of wound complications and lymphoceles. Therefore, mTOR inhibitors should be avoided in the first few months after transplantation.

Authorship

LHMP: designed the study, rated methodological quality, extracted and analysed data and wrote the paper. LL: extracted data and rated methodological quality. PJM: designed the study, analysed data and wrote the paper.

Funding sources

No funding was received for this study.

Ancillary