Introduction
When the first renal transplants were being performed in the early 1960s the immunosuppressive properties of corticosteroids were already noted, and steroids were successfully used to reverse episodes of acute rejection. In his 1964 book, Starzl described the use of steroids as a “pretreatment” alongside azathioprine as prophylaxis against rejection, based on the premise that rejection was almost inevitable with azathioprine alone. This work was soon followed by groups from Virginia, London, and Cleveland, with the use of steroids in combination with azathioprine from the time of transplantation and increased doses during episodes of acute rejection. This combination of azathioprine and steroids became the mainstay of transplant immunosuppression for almost 20 years.
From the beginning of this so-called azathioprine era, arbitrarily large doses of steroids were given from the time of transplantation with a gradual reduction over 6 to 12 months to maintenance levels. The high doses of steroids used with azathioprine were responsible for most of the morbidity of transplantation (discussed later). It was not until the 1970s that a series of randomized trials as well as observational studies slowly led to the realization that low-dose steroids were as effective as high-dose steroids in preventing rejection and that there was a major reduction in steroid complications of transplantation with low-dose regimens.
With the introduction of newer, more potent induction and maintenance agents, there has been a great deal of interest over recent years in further reducing steroid doses and either withdrawing them from maintenance immunosuppressive regimens or avoiding them altogether.
Mechanism of Action
Steroids are administered as prednisone or prednisolone. These agents are absorbed rapidly from the gut, and peak plasma concentrations occur 1 to 3 hours after administration. The mechanism of action of steroids is extremely complex and is still not understood fully. Steroids are antiinflammatory as well as being immunosuppressive. It was first noted by Billingham et al . that cortisone would produce a modest prolongation of the life of skin allografts in the rabbit. In the treatment of acute rejection, it is probably the antiinflammatory activity that produces the immediate response, whereas when used prophylactically it is the immunosuppressive activity that is predominant. A small randomized trial comparing prednisolone with a nonsteroidal antiinflammatory agent (ibuprofen) showed a higher rate of rejection in the patients receiving the nonsteroidal agent, confirming that the antiinflammatory effect of steroids is not its major role in renal transplantation.
Steroids are metabolized in the liver, where prednisone is converted to prednisolone. Although it has been estimated that the bioavailability of prednisone is approximately 80% of that achieved by prednisolone, no evidence exists in practice that there is a difference in outcome between prednisone (used most commonly in the US) or prednisolone (used most commonly in Europe). The half-life of steroids is short—about 60 minutes for prednisone and 200 minutes for prednisolone. These half-lives are increased substantially in the presence of hepatic dysfunction and are shorter in the presence of drugs such as phenytoin and rifampicin that induce hepatic enzymes. There is no evidence that these interactions have produced significant problems in clinical practice. It has also been shown that the clearance of prednisolone is slower in patients on cyclosporine compared with patients on azathioprine. A later study suggested, however, that cyclosporine did not influence the metabolism of methylprednisolone, but the authors noted a considerable variation of the metabolism of methylprednisolone among patients. The time- and dose-dependent induction of UDP-glucuronosyltransferase activity by steroids may increase the clearance of mycophenolic acid, reducing exposure to mycophenolate. Cattaneo et al . have demonstrated that as steroids are tapered over the postoperative period, the mycophenolic acid area-under-the-curve increases. The pharmacokinetics of prednisolone during sirolimus therapy have also been studied, with some evidence for a minor interaction between sirolimus and prednisolone in some patients.
Steroids do have a significant effect in vitro on T cell proliferation, blocking interleukin-2 production. A variety of other actions may augment their immunosuppressive activity (e.g., preventing the induction of interleukin-1 and interleukin-6 genes in macrophages). Its antiinflammatory activity is perhaps mediated by the inhibition of migration of monocytes to areas of inflammation, and this same antiinflammatory activity has a marked deleterious effect on wound healing.
Steroid Resistance
The sensitivity of individuals to steroid therapy is highly variable. A study in healthy volunteers demonstrated a wide interindividual variation in the inhibition of lymphocyte proliferation by steroids. Steroid resistance is frequently seen in patients with inflammatory conditions and has been shown to correlate well with in vitro measurements of lymphocyte steroid sensitivity in patients suffering with rheumatoid arthritis, ulcerative colitis, asthma, and systemic lupus erythmatosus.
In vitro studies of lymphocyte steroid sensitivity have demonstrated a higher incidence of resistance in patients with chronic renal failure than in healthy volunteers (52.9% vs. 3.8%). In renal transplant recipients, Langhoff and colleagues have demonstrated that pretransplant in vitro measurements of lymphocyte sensitivity are predictive of graft survival at 1 year in patients coadministered azathioprine, but less so in those receiving cyclosporine. These results have been confirmed in vivo, with significantly higher sensitivity to methylprednisolone seen in those patients with graft function at 6 months compared with patients with graft failure. This difference in sensitivity is smaller in cyclosporine-treated patients than those receiving azathioprine, suggesting that this effect may in part be offset by the use of newer, more potent immunosuppressant agents. Dialysis patients demonstrating steroid resistance have an increased risk of acute rejection and chronic allograft nephropathy posttransplant. Interestingly, reduced lymphocyte prednisolone sensitivity correlates with impaired sensitivity to cyclosporine and tacrolimus, which may play a role in the high risk of allograft rejection in these patients.
A number of potential mechanisms for this resistance to steroids have been suggested, including receptor downregulation by glucocorticoid exposure, negative inhibition by the beta-isoform of the glucocorticoid receptor, or inhibition of the alpha isoform of the receptor by the proinflammatory transcription factor NK-κB in inflammatory conditions.
Dosage
When used prophylactically with azathioprine, steroids were used initially in high doses (e.g., 100 mg/day), reducing to a maintenance dose of 20 mg/day over 6 to 9 months. McGeown and coworkers were the first to report consistently excellent graft survival with a low incidence of steroid-related complications using a lower prednisolone dose of 20 mg/day given orally as a single morning dose, with a further reduction occurring at 6 months to a baseline maintenance dose of 10 mg/day. Because most of the Belfast patients had had bilateral nephrectomies and all had had more than 100 blood transfusions before transplantation, it was not clear whether the excellent results were related to the low dosage of steroids or to a transfusion effect, which was recognized widely as an important factor in improving graft outcome in the azathioprine era.
Trials of low-dose steroids versus high-dose steroids were carried out in Oxford, then in many other centers, all of which showed not only that low-dose steroids were as effective as high-dose steroids in preventing rejection but also that there was a significant reduction in steroid-related complications in patients receiving low-dose steroids. The results of these trials led quickly to the wide adoption of low-dose steroid regimens with azathioprine. The results of a large multicenter trial reported by D’Apice et al., demonstrated that low-dose steroids were only equally effective as high-dose steroids in preventing rejection if therapeutic doses of azathioprine were used (i.e., >2 mg/kg/day).
With the introduction of cyclosporine, steroids remained in use with or without azathioprine. In general, low-dose steroid protocols were continued, although there was a tendency, particularly in North America, to go back toward higher steroid dosage regimens in the first few weeks posttransplantation. This was a relatively transient practice and now, with modern immunosuppressive protocols, not only are low-dose steroids the norm but indeed discontinuation of steroids is becoming increasingly possible (see later).
Whether steroids should be given as a single daily dose in the morning or in divided doses has not been resolved. Because of the short half-life of prednisone and prednisolone, divided doses may be more rational, but it could be argued that a single morning daily dose would be more appropriate, taking into account the diurnal rhythm of glucocorticoid metabolism. There is no clinical evidence that one or the other protocol is more effective or less likely to produce side effects.
For many years, maintenance doses of prednisone or prednisolone of 10 mg/day were standard therapy in association with azathioprine. In patients with long-surviving grafts with good function, steroid dosages have been reduced to 5 or 6 mg/day. It is unlikely, however, that many patients who are on long-term azathioprine and steroids would be able to have their steroid dosage reduced to much less than 5 mg/day. Attempts in the past to withdraw steroids have often led to the onset of rejection when doses of less than 5 mg/day are reached. This is important to note as there are many long surviving patients still on azathioprine and steroids. It should also be remembered that when patients have been on steroids for many years, their adrenocortical function may not recover from the long-standing suppression as the steroid dose is reduced (see later).
Alternate-day steroid therapy for maintenance has also been used widely, especially in children, in an attempt to reduce side effects, particularly growth retardation. In children, alternate-day therapy may be associated with a greater incidence of rejection, but this is probably not the case in adults. A small randomized trial of alternate-day therapy failed to show any benefit over daily steroids, however. Alternate-day therapy may lead to greater problems with respect to compliance, in contrast to a daily regimen of steroids.
It has been and still is common practice to administer a bolus of methylprednisolone prophylactically during the transplant operation with the aim of increasing immunosuppression and perhaps preventing delayed graft function, but a randomized prospective trial of bolus methylprednisolone versus placebo at the time of surgery did not show any benefit of the high perioperative intravenous dose of methylprednisolone. Nevertheless it remains standard practice whatever the immunosuppressive regimen is to be.
Treatment of Acute Rejection
Steroids in high doses are the first approach to the treatment of an acute rejection episode. Early experience involved either increasing the oral dosage of steroids to high levels (e.g., 200 mg/day for 3 days), with a rapid reduction over 10 days to the dosage levels of steroids being given before the acute rejection episode, or boluses of intravenous methylprednisolone (e.g., 0.5–1 g/day for 3–5 days). Probably both approaches are equally effective. In an early randomized prospective trial in Oxford, however, high intravenous doses were as effective as high oral doses in reversing rejection, but there was a definite suggestion that steroid-related complications were lower in those who received intravenous therapy. In a randomized study in children, a high intravenous dose of methylprednisolone (600 mg/m 2 daily for 3 days) was no more effective than low oral doses of prednisolone, reversing rejection in 70% as opposed to 72% of episodes.
The most common form of high-dose intravenous therapy to treat acute rejection has been 1 g of methylprednisolone given intravenously as a single bolus daily for 3 days. The intravenous bolus should be administered slowly over 5 minutes because the sudden injection of the bolus can lead to cardiac arrhythmias. It is probable that 1 g of methylprednisolone is a much greater dose than required; in Oxford for many years now we have used 0.5 g of methylprednisolone daily intravenously for 3 days, whereas the Stockholm unit has used 0.25 g daily intravenously for 3 days. The lower intravenous doses do not appear to be associated with any greater incidence of steroid-resistant rejection, as originally suggested by a prospective trial of high-dose versus low-dose intravenous steroids to treat rejection. Similarly in a small double-blind, randomized trial, Stromstrad et al . failed to show any therapeutic benefit of a 30 mg/kg bolus over a 3 mg/kg bolus, and similarly Lui et al . failed to show any benefit of a bolus of 15 mg/kg of body weight over a bolus of 3 mg/kg.
Side Effects
The side effects of continuous steroid therapy are numerous ( Table 16.1 ). High-dose steroids were responsible for most complications of renal transplantation in the azathioprine era, which have reduced markedly with the more recent widespread use of low-dose steroids. In a study of the cost of steroid side effects over 10 years in a cohort of 50 patients, the additional cost per patient attributable to a steroid complication was assessed at $5300 (US dollars). Cost analysis in a randomized controlled trial of early steroid withdrawal from Egypt has shown a 2.9-fold increase in management costs for steroid-related morbidities in the steroid maintenance arm.
| Cushingoid facies | Hypertension |
|---|---|
| Wound healing | Psychiatric disturbance |
| Growth retardation | Cataracts |
| Diabetes | Pancreatitis |
| Hyperlipidemia | Skin changes |
| Bone disease | Peptic ulceration |
| Obesity |
Cushingoid Facies
Cushingoid facies used to be the hallmark of a renal transplant patient: a moon face, buffalo hump, acne, obese torso, and thin, easily bruised skin, all representing the cumulative effect of high-dose steroids. With lower-dose steroids, Cushingoid facies is seen much less often, although most patients show modest changes in their facies in the early months posttransplantation. Most patients on low-dose steroids, which is the normal practice now with cyclosporine or tacrolimus, have relatively minimal facial changes related to steroids.
Wound Healing
The antiinflammatory activity of steroids leads to poor wound healing. In the days of high-dose steroids, this was a major problem, influencing the healing not only of the incision, but also of the ureterovesical reconstruction. With low-dose steroids, poor wound healing is no longer a major problem.
Growth Retardation
Growth retardation is of particular concern in children after renal transplantation. With modern lower doses of steroids growth retardation is less of a problem. As already discussed, the use of alternate-day steroids may reduce growth retardation, and further steroid reduction and withdrawal have also been successful in allowing catch-up growth in some studies (see later).
Diabetes
Glycosuria and insulin-dependent and non–insulin-dependent diabetes are common after transplantation. The occurrence of diabetes is related, in part, to steroid usage but it has become more common with the concomitant use of cyclosporine and tacrolimus, both of which can induce diabetes independently of steroids. In the presence of these two agents, the use of steroids augments the potential for diabetes, and often patients who become diabetic on cyclosporine or tacrolimus have a regression of the diabetes when steroid therapy is discontinued.
Hyperlipidemia
Hypercholesterolemia and hypertriglyceridemia are associated with steroid use. Hyperlipidemia has become a greater problem in the calcineurin inhibitor (CNI) era as cyclosporine and tacrolimus both lead to an increased incidence, although this may be less of a problem with tacrolimus. Withdrawal of steroids leads to improvements in the lipid profile (see later).
Bone Disease
Bone disease is a common and major problem posttransplantation, especially in the postmenopausal woman. In the days of high-dose steroid therapy posttransplantation, avascular necrosis of bones, particularly of the head of the femur, was common, occurring with an incidence of approximately 10% to 15% within 2 years of transplantation ( Fig. 16.1 ). All the evidence suggests that this incidence was due to a cumulative effect of steroid dosage. As low-dose steroid protocols were introduced, the incidence of avascular necrosis decreased dramatically. However, the cumulative dose of steroids received by a patient on a high-dose steroid regimen, as opposed to a low-dose regimen, is not that much higher after 6 months.
Osteoporosis is associated with steroid therapy. In a randomized study, Hollander and colleagues showed that vertebral bone density was increased significantly in patients discontinuing steroids. Similar evidence was reported by Aroldi et al. in a randomized study of three different immunosuppressive protocols and vertebral bone density. These investigators showed that lumbar bone density decreased significantly in patients receiving cyclosporine and steroids but increased significantly in patients receiving cyclosporine alone without steroids. Meta-analysis of studies in transplant recipients has demonstrated a significant reduction in the risk of fractures and increase in bone mineral density with the use of bisphosphonates or vitamin D analogs, suggesting that the use of these agents should be considered, particularly in recipients on long-term steroids.
Many patients coming to renal transplantation have a degree of secondary hyperparathyroidism, and bone changes related to the hyperparathyroidism are enhanced by steroid therapy. Much more aggressive approaches to parathyroidectomy in patients with renal failure are being taken by most units now before transplantation. In patients after transplantation with raised parathormone levels, early parathyroidectomy also should be considered.
Obesity
Steroid therapy leads to a marked increase in appetite, and without any dietary restrictions after transplantation, all patients tend to gain weight, which is in addition to a weight increase resulting from salt and water retention. Many patients become obese (body mass index >30), and this adds to the risks of poor survival. Every attempt should be made to advise patients from the time of transplantation to restrict calorie intake carefully because once patients have gained weight in the presence of steroid therapy, it is extremely difficult for them to reduce their weight.
Hypertension
Hypertension after transplantation is common and is related in part to steroids as well as CNI use. In steroid withdrawal protocols, hypertension improves once steroids are discontinued (see later).
Psychiatric Disturbance
Psychiatric disturbance is evident in patients on steroids in two ways. In the early days posttransplantation, particularly with the need for high-dose steroids to treat rejection, significant psychiatric mood changes may be observed. Later, when steroids are being withdrawn or reduced to low doses, psychiatric mood changes, especially depression, may occur.
Cataracts
Steroid-related cataracts are common after renal transplantation, occurring in approximately 25% of patients.
Pancreatitis
Acute pancreatitis occurs with a much greater incidence after renal transplantation than would be expected. Azathioprine and steroids have been associated with acute pancreatitis. The pancreatitis is probably related to overall immunosuppression and is often severe. The clinical features of acute pancreatitis can be masked to some extent by steroids.
Skin Changes
Long-term steroids produce typical skin changes in renal transplant patients, the skin being thin, atrophic, easily bruised, and susceptible to knocks. A syndrome known as transplant leg is associated with long-term steroid usage; this occurs, for example, when a patient bumps into a chair or a table (a trivial injury), and a flap of skin is stripped or elevated from the lower leg.
Peptic Ulceration
Although it is debatable whether steroids lead to development of peptic ulceration, most units use prophylactic H 2 antagonists or proton pump inhibitors in the early months posttransplantation, when steroid doses are at their highest. The advent of low-dose steroid therapy has been associated with a dramatic diminution in the incidence of peptic ulceration after transplantation.
Acute Abdomen
In all renal transplant patients who present with an acute abdomen, steroids may mask the symptoms noted by the patient. If this fact is not remembered, diagnosis of diverticulitis or a perforated peptic ulcer may be delayed with disastrous results.
Adrenal Insufficiency
As in other conditions in which long-term corticosteroids are used, there is a risk of adrenal insufficiency with cessation of therapy or during periods of stress, trauma, surgery, or acute illness. A recent meta-analysis suggests that the absolute risk of adrenal insufficiency in renal transplant recipients is 56.2% (95% confidence interval [CI] 42.9%–68.6%). Although higher doses and longer-term use increase the risk, the effect is not entirely predictable, and the risk of insufficiency should always be considered in patients taking maintenance steroids.
Steroid Withdrawal and Avoidance
The side effects of steroids as outlined previously have generated a great deal of interest in withdrawal or even complete avoidance of these agents after renal transplantation. One of the major causes of death with a functioning graft after renal transplantation is cardiovascular disease, so the effect of long-term steroid use on cardiovascular risk factors such as hypertension, hyperlipidemia, and posttransplant diabetes is likely to be important. Any advantage seen in terms of cardiovascular risk must be balanced against the reduction in immunosuppression and potential detriment to graft function, and such concerns have led to a large number of clinical trials investigating the risk and benefit of steroid withdrawal protocols.
Early Experience
In early patients treated with steroids and azathioprine, attempts to withdraw steroids resulted in almost inevitable acute rejection when the dose fell below 5 to 6 mg/day. The introduction of cyclosporine led to renewed interest in such protocols, and spawned a number of clinical trials investigating steroid withdrawal at varying times posttransplant. An early meta-analysis of seven studies in cyclosporine-treated patients by Hricik and colleagues demonstrated an increased risk of acute rejection, but with no detriment to patient or graft survival. Five of the seven included studies did not utilize triple therapy with azathioprine (cyclosporine alone) and in four studies steroids were avoided completely from the time of transplantation. Furthermore, only one study reported follow-up beyond 2 years posttransplant. This was the Canadian Multicentre Cyclosporine trial, which demonstrated superior long-term graft survival in patients who continued taking steroids.
The results of this meta-analysis led to concerns about the early withdrawal of steroids from cyclosporine-based regimens, leading to further randomized controlled trials investigating late withdrawal in patients with stable renal function at 1 year posttransplant. In the trial from Oxford, patients on triple therapy (cyclosporine, azathioprine, and steroids) randomized to stop steroids demonstrated a 10% deterioration in renal function at 1 year, which then stabilized with no increase in graft loss. Both of these studies demonstrated beneficial effects in cardiovascular risk factors such as hypertension and blood lipids.
Contemporary Immunosuppression
After the introduction of tacrolimus and mycophenolate mofetil (MMF) to maintenance immunosuppressive regimens, a large number of randomized trials with various combinations of induction and maintenance agents, as well as varying times of steroid withdrawal, have been published in the literature. In general the trend has been back toward earlier tapering and withdrawal of steroids, with withdrawal as early as 1 week posttransplant in a number of studies. One of the major concerns, given the excess acute rejection seen in the early steroid withdrawal trials with cyclosporine, is the effect of the reduction of immunosuppression on long-term graft and patient outcomes. Perhaps the most impressive study to date addressing this issue is the Astellas Steroid Withdrawal study from Woodle and colleagues, in which patients receiving tacrolimus, MMF, and steroids were randomized at day 7 to withdraw steroids or continue a long-term maintenance dose of 5 mg/day from month 6. Randomization and blinding were maintained for the duration of the 5-year follow-up. Although the study demonstrated an excess of mild, steroid-sensitive acute rejection in the withdrawal arm, 5-year patient and graft survival were unaffected. Benefits were seen in cardiovascular risk factors such as triglycerides, insulin requirements, and weight gain, leading the authors to conclude that early steroid withdrawal is safe in the longer term in patients treated with a tacrolimus and MMF regimen.
It was first noted by Pescovitz and colleagues that sirolimus may aid in the withdrawal of steroids from a calcineurin-based regimen. Further observational studies appear to support this, and the use of sirolimus may also allow steroid withdrawal alongside reduction in the exposure to CNIs.
The use of newer immunosuppressive agents such as MMF or sirolimus may also allow the safe withdrawal of steroids earlier than previously seen with cyclosporine-based regimens. Kumar and colleagues have reported a 3-year analysis of a large trial of 300 patients receiving basiliximab induction, a CNI, and MMF or sirolimus, in which patients were randomized to have steroids withdrawn on day 2 or to continue steroids. There was no difference in graft function, patient and graft survival, biopsy proven acute rejection, or chronic allograft nephropathy. The incidence of new-onset diabetes was lower in the steroid-free group.
A study by Gelens and colleagues attempted to combine early steroid withdrawal with a CNI-free maintenance regimen. Patients were randomized to receive tacrolimus and sirolimus; tacrolimus and MMF; or daclizumab induction, sirolimus, and MMF. Steroids were withdrawn after 2 days in all patients. The trial was halted after an interim analysis demonstrated an unacceptably high incidence of acute rejection in the CNI-free group. Thus it would appear that even with modern immunosuppressant agents and antibody induction, it is not possible to combine the complete withdrawal of CNIs with steroid withdrawal. Recent attempts to combine belatacept-based immunosuppression with steroid avoidance have resulted in unacceptably high rates of acute rejection, even in the context of depleting antibody induction. Until further evidence is generated, steroid avoidance or withdrawal cannot be recommended in combination with belatacept-based, CNI-free regimens.
A number of meta-analyses of steroid withdrawal protocols have been published to attempt to draw conclusions from the large body of randomized trial evidence in this area. Pascual and colleagues performed a meta-analysis of six such trials, four with MMF and cyclosporine and two with MMF and tacrolimus. Although the risk of acute rejection was increased just over twofold when steroids were withdrawn, there was no significant difference in the incidence of graft failure. Unfortunately, this meta-analysis did not differentiate the relative steroid-sparing potential of cyclosporine and tacrolimus. A similar meta-analysis from Tan and colleagues also demonstrated an increase in the risk of acute rejection, balanced by a reduction in the risk of opportunistic and urinary tract infections in a small subset of the included trials.
A comprehensive meta-analysis from our own group included all steroid withdrawal and avoidance studies over a 27-year period, incorporating 34 studies and 5637 patients. Overall analysis confirmed an increase in acute rejection with steroid avoidance or withdrawal ( Fig. 16.2 ; relative risk [RR] 1.56, P < 0.0001), although no difference in steroid-resistant acute rejection was found, suggesting that these are mainly steroid-sensitive mild rejection episodes. Despite the increased rate of rejection and a small (clinically insignificant) increase in serum creatinine, there was no difference in graft or patient survival. Significant benefits in cardiovascular risk factors were identified with steroid avoidance or withdrawal, with the incidence of hypertension (RR 0.90, P < 0.0001), new onset diabetes (RR 0.64, P = 0.0006), and hypercholesterolemia (RR 0.76, P < 0.0001) all reduced. Interestingly, no interaction was found between the use of induction immunosuppression, maintenance immunosuppression, or the time of steroid withdrawal on the ability to withdraw steroids. Post hoc analysis has, however, demonstrated a dose effect, with the cardiovascular benefit seen reduced when withdrawing lower doses of steroids.
