Renal transplantation in human immunodeficiency virus (HIV)-positive children
Renal transplantation is being performed in adult human immunodeficiency virus (HIV)-positive patients and increasingly in paediatric patients as well. A multidisciplinary team involving an infectious disease professional is required to assist with HIV viral-load monitoring and in choosing the most appropriate highly active antiretroviral therapy (HAART). Drug interactions complicate immunosuppressant therapy and require careful management. The acute rejection rates appear to be similar in adults to those in noninfective transplant recipients. Induction with basiliximab and calcineurin-based immunosuppression appears to be safe and effective in these recipients. Prophylaxis is advised for a variety of infections and may need life-long administration, especially in children. Organ shortage remains a significant problem, and kidneys from deceased HIV-positive donors have been used successfully in a small study population. Overall, with careful planning and close follow-up, successful renal transplantation for paediatric HIV-infected recipients is possible.
KeywordsPaediatric HIV-infected renal transplantationAcute kidney injuryHAARTImmunosuppression
Renal transplantation with the use of immunosuppression was initially thought to be contraindicated in patients with human immunodeficiency virus (HIV) infection. However, with the advancements in treating HIV infection, this has now become possible with careful planning in well-resourced regions . Increasingly, adult studies are reporting good outcomes with renal transplantation in patients who are well controlled on highly active antiretroviral treatment (HAART) [2-4]. This treatment regimen requires involvement of a multidisciplinary team to monitor renal function and HIV status following transplant. There have been isolated paediatric renal transplants in HIV-infected patients worldwide (a few in the USA; none in Africa as yet), with the UK recently performing the first such paediatric transplant. For this reason, there is very little evidence to support any paediatric management using this regimen; thus, recommendations have had to be extrapolated from the adult world.
HIV prevalence in paediatrics
In addition, there are concerns that HIV-positive [infection confirmed by enzyme-linked immunosorbent assay (ELISA) or polymerase chain reaction (PCR)] or HIV-exposed children (mothers are HIV positive, and child may or may not be infected) with poor nutrition may have worse surgical outcomes than noninfected patients . There are genetic concerns in individuals of African descent that variations in MYH9 (encoding nonmuscle myosin IIA heavy chain) may predispose to forms of focal segmental glomerulosclerosis (FSGS), including HIV-associated nephropathy (HIVAN) . However, newer studies suggest that two coding risk variants within the apolipoprotein L-1 (APOL1) gene on chromosome 22 in a larger interval have a stronger genetic association with FSGS than the neighbouring MYH9 gene. The APOL1 risk alleles for renal disease occur in >30 % of African American chromosomes, and further studies of molecular mechanisms are required given this high frequency and disease risk [11, 12].
Pathogenesis of renal disease in HIV-infected children
1. Urinary tract infections
2. Acute kidney injury – dehydration, sepsis
3. Thrombotic angiopathies
a. HIV-associated nephropathy (HIVAN)
b. HIV-associated immune complex disease (HIVICK)
c. Non-HIV related including:
i. Proliferative postinfectious glomerulonephritis (GN)
ii. Lupus-like nephritis
iii. Mesangial proliferative IgA nephropathy
iv. Membranous and membranoproliferative nephropathy
a. Hepatitis B and C
b. Tuberculosis (TB)
c. Opportunistic infections
7. Nephrotoxic drugs
a. Protease inhibitors – crystals and stones
b. Tenofovir – proximal renal tubular/Fanconi syndrome
c. Antibiotics – aminoglycosides
Criteria for diagnosis of HIV-associated nephropathy (HIVAN) (based on information in )
1. Persistent proteinuria (>1 + Albustix or protein/creatinine ratio of >0.1 mg/l for ≥2 months)
2. Abnormal urinary microscopic examination
3. Enlarged echogenic kidneys on ultrasound on two separate studies done 2 months apart
4. Black race with nephrotic-range proteinuria
In addition, paediatric HIVAN appears to respond well to HAART if detected in the early stages. However, these patients often only present later with established renal dysfunction.
Screening for renal disease
In adults, HIVAN is seen as a consequence of advanced HIV disease. As a result, it has been recommended that patients with HIV disease should be screened for proteinuria and renal dysfunction . Early detection and diagnosis is essential for preventing and slowing decline of renal function in HIV-infected paediatric patients. In addition, patients on nephrotoxic drugs such as Tenofovir should also be monitored for renal involvement.
In studies from Africa, where screening of HIV-positive children has been performed, the first Nigerian study (Port Harcourt) showed that 18 % of those tested had significant proteinuria ; a study in Lagos, Nigeria, compared 88 HIV-infected to 50 non-HIV-infected children matched for age and sex, showing 20.5 % of HIV children with proteinuria [protein/creatinine ratio (Pr/Cr) > 0.2)]compared with only 6 % proteinuria in noninfected children (p < 0.026) . A study in Johannesburg, South Africa, of 180 children who underwent more specific screening for microalbuminuria showed 27/110 (25 %) HIV-positive children with microalbuminuria compared with 1/70 (1 %) noninfected children (p = 0.00003) .
Paediatric patients with HIV infection and significant persistent proteinuria may benefit from a renal biopsy if facilities are available. Two paediatric studies with significant numbers were performed in South Africa: In 49 biopsies reported by Ramsuran et al. , 65.3 % had FSGS, of whom 26.5 % had collapsing GN, 16.3 % had mesangial proliferation, 10.2 % had HIVICK and 8.2 % had minimal-change nephrotic syndrome (MCNS). A similar study with larger numbers from Kala (presented at the African Paediatric Nephrology Association Congress 2013, Ghana, unpublished data) showed that in 90 paediatric HIV-infected patient biopsies, 43 % showed HIVICK, 37 % FSGS, 8.9 % severe interstitial nephritis, 3.3 % each mesangioproliferative, MCNS and mesangiocapillary disease. Even though FSGS is prominent, other renal pathology, especially immune-complex-related diseases, are also seen in paediatric HIV renal disease. Electronic Supplementary Material Fig. 1 provides an example of histology stainings for mesangial proliferation, collapsing FSGS, tubular dilatation and interstitial inflammation. Many paediatric HIV-infected patients have coexistent tuberculosis (TB). Disseminated TB has also been documented to present with interstitial nephritis and nephrotic-range proteinuria, which resolves on TB treatment .
Serum-creatinine-based estimates of glomerular filtration rate (GFR) using the Schwartz formula are most widely used, but body mass abnormalities and exposure to multiple medications can affect renal creatinine clearance, and as a result, some centres in developed countries are increasingly using serum cystatin C. However, this test is inaccessible in parts of the world such as Africa, where HIV is most prevalent.
Novel biomarkers tested in HIV-infected children and adults for use in CKD screening include:
1. Neutrophil-gelatinase-associated lipocalin (NGAL), a protein markedly up-regulated in renal tubules and urine (uNGAL) of patients with HIVAN, may be useful to distinguish cases of HIVAN from other proteinuric glomerulopathies in HIV-infected patients. Considering the specific expression of NGAL messenger RNA (mRNA) in dilated microcysts, uNGAL level may be useful for following progression of these lesions, which are characteristic of HIVAN [29, 30]. One paediatric study demonstrated progressive decreases in uNGAL levels with the initiation of HAART treatment in a child with biopsy-proven HIVAN, in association with improvement in renal function and reduced viral load . A further study showed that serum NGAL levels are decreased in HIV-infected patients, in correlation with diminished neutrophil numbers and function, and that NGAL in the serum returns to normal values in response to HAART .
2. Single urine biomarkers may not be sufficient, and tubular injury may also be suggested by β2-microglobulin and retinol-binding protein (RBP) and other markers of renal tubular injury in HIV-infected children [31, 33]. This includes the role of urinary β2 microglobulin as a marker for tenofovir-induced renal tubular injury in children .
3. Urinary levels of iron and iron-related binding proteins transferrin, hemopexin, haptoglobin and lactoferrin, may all assist in identifying HIV-infected children at risk of developing HIVAN .
Choices in renal replacement therapy
Peritoneal dialysis is often the preferred modality for dialysis in children. In the adult European Renal Association Dialysis and Transplant Registry, the prevalence of HIV-infected patients on dialysis was only 0.12 % , with survival rates being significantly lower than matched HIV-negative controls on dialysis. In an adult Spanish study, 48/8,744 patients were infected with HIV, of whom 81.3 % received haemodialysis, 16.7 % transplantation and only 2 % peritoneal dialysis. However, dialysis modality does not seem to be a determining factor in survival of these patients . In view of the fact that most children with HIV infection are from developing countries, the choice of renal replacement therapy (RRT) depends on locally available dialysis modality.
Criteria for transplanting HIV-infected patients with end-stage kidney disease
Patients should have CD4+ T-cell counts of at least 200 cells/cubic mm. Absolute T-cell counts in children < 6 years of age should be higher and the percentage of CD4 cells >25 (personal communication; Brian Eley, July 2013)
Undetectable viral load: plasma HIV type 1 RNA levels (<50 copies/ml)
Stable and compliant on HAART for >3–6 months
If coinfected with hepatitis B or C (HBV or HCV) (less likely in children): no evidence of cirrhosis on liver biopsy
Previous opportunistic infections are not necessarily a contraindication .
Deciding which HAART to commence
The type of HAART to institute is best determined by a multidisciplinary approach in conjunction with an infectious disease team, which can advise on specific drug therapy. Adult studies to date continue patients on their pretransplantation antiretroviral regimen with no restrictions to date; most of this knowledge was extrapolated from a large adult study of 150 patients and smaller European studies from France and Spain [37-40]. The recommendation to start HAART in children is different to that in adults, and most developed countries would suggest starting HAART with much higher CD4 counts and even as early as the diagnosis of HIV is made in children. These recommendations are constantly under review.
An important principle to remember is that drug interactions between tacrolimus and protease inhibitors due to inhibition of both P-glycoprotein and the cytochrome-P450-metabolising enzyme CYP 3A4 lead to extremely high blood levels of tacrolimus, requiring significant dosage adjustments . The initial recommendation was to increase dosing interval up to five times normal. However, with the availability of tacrolimus suspension in paediatrics, titration with dosing as little as 0.005 mg/kg once daily or on alternate days is possible whilst performing blood-level measurements to achieve therapeutic ranges. Newer protocols with effective HAART strategies may replace protease inhibitors altogether and make drug monitoring easier. An integrase inhibitor—Raltegravir—has been suggested as a new, non-protease-inhibitor alternative . The specific protocol used must be reviewed with the infectious disease team involved in patient management, who will also need to continue monitoring the child’s HIV viral load following the transplant.
HAART therapy in itself may also have nephrotoxic effects, and the protease inhibitors Indinavir and Tenofovir have received the most negative renal publicity. Indinavir has a tendency towards sludging of crystalline aggregates both in the renal tract, resulting in stones, as well as the renal parenchyma, with resultant tubulointerstitial injury. Increased fluid intake has been recommended to counteract these problems [21, 42, 43]. Atazanavir is a newer, once-daily drug that appears to have a slightly lower incidence of renal calculi but still requires close monitoring. Tenofovir is a nucleotide reverse-transcriptase inhibitor (NRTI) now used increasingly in HAART protocols and in paediatrics often as first-line therapy but has shown an association with mitochondrial injury (enlarged dysmorphic mitochondria seen on biopsy) and resultant proximal renal tubular dysfunction and Fanconi syndrome [44, 45]. Clinically, this can be detected by normoglycaemic glycosuria, hypophosphataemia and low-grade proteinuria, which is of low molecular weight and tubular in origin. Investigations including urine PR/Cr ratio are preferable to dipstick analysis or albuminuria levels. Tubular function assessment with measurement of retinol-binding protein (RBP) and tubular reabsorption of phosphate is also helpful.
Routine vaccinations are as per usual pretransplant protocol of the transplant unit. In addition, vaccination against HAV and HBV (if negative antibodies), influenza, as well as booster doses of pneumococcal and meningococcal vaccines if appropriate, should be administered (Guy’s and St. Thomas’ Hospital Renal Unit Transplant Protocol; courtesy of Rachel Hilton, personal communication, July 2013). However live vaccines such as measles and varicella may pose a problem, especially in nonimmune children, and is controversial. TB immunisation in the form of Bacillus Calmette–Guérin (BCG) is also problematic, as it is a live vaccine and can result in TB disease; thus, many units exclude this vaccine and prefer to start anti-TB prophylaxis posttransplant if in a high risk area. The duration of this prophylaxis is debatable but may be life long if living in a high-risk TB area.
Immunosuppression use in children
Induction therapy was used with 51 % of patients receiving basiliximab/daclizumab and 32 % antithymocyte globulin (ATG), showing a marginally higher risk of patient death in the ATG group compared with the basiliximab group in Stock et al.’s paper . In children, the current recommendation is to use basiliximab, in keeping with National Institute for Health and Care Excellence (NICE) guidelines for paediatric non-HIV transplant recipients (www.nice.org.uk) until further studies are done.
Stock et al.  used a calcineurin inhibitor (CNI)-based immunosuppressant protocol, with 66 % using tacrolimus and 22 % cyclosporine. Use of tacrolimus as well as a higher tacrolimus trough level (8.6 vs 9.4 ng/ml) was associated with a decreased risk of first acute allograft rejection (p = 0.04). In view of these findings with better results, despite there being no trials in children as yet, it would be recommended to use tacrolimus as the CNI of choice and avoid the cosmetic side effects seen with use of cyclosporine. In living related donors, in which protease inhibitors are the mainstay of HAART, starting tacrolimus some time prior to the transplant can be useful to establish appropriate dosing and serum trough levels. There is no strong evidence to support this regimen, but it is widely practised in paediatric units, even in HIV noninfected living related transplant recipients. A purine antagonist in the form of mycophenolate mofetil was used in 87 % of patients. It is worth mentioning anecdotally that in an HIV-infected paediatric patient with a significantly reduced white cell count and recurrent infections undergoing a renal transplant, a switch to azathioprine was well tolerated. Steroid therapy was used in all adult studies, and this may be an area in paediatrics that requires closer study to determine whether steroid-free/rapid steroid withdrawal protocols could be used. An alternative may be to wean to alternate-day low-dose steroids as per the protocol practised in many European units.
Rejection rates in HIV-infected recipients
The initial concern for renal transplantation in HIV-infected patients was that rejection rates would be significantly higher than in the healthy population. In the adult studies published in 2010, one study showed more acute rejection, with delayed graft function and graft survival being lower but not statistically significant . The cumulative incidence of rejection in a large adult study with 150 patients was 31 % at 1 year and 41 % at 3 years . The rates of patient and graft survival at 3 years were generally between the reported rates for those >65 years of age and the overall rates for all kidney transplants in the US database. To date, no database provides sufficient numbers in paediatrics world wide to assess outcomes, but anecdotally, the few cases that have been done appear to be in keeping with adult results.
Progression of HIV disease
Theoretically, this has been a significant concern, but there is no evidence of accelerated HIV disease progression in adult studies despite initial decline in CD4+ T-cell count . HIV viraemia remains well controlled.
Patients who received ATG as induction had twice as many serious infections per follow-up year as those who did not . In children, this may be a good reason not to use this agent. Of 150 adult patients, 57(38 %) had 140 infections: bacterial 69 %, fungal 9 %, viral 6 % and protozoal 1 %. The most common sites were genitourinary tract (26 %), respiratory tract (20 %) and blood (19 %). About 60 % of serious infections occurred within the first 6 months after transplantation. Viral infections specifically included five cases of polyomavirus nephropathy (BK). Patients who tested positive for HCV infection had a higher rate of serious infections. Children generally have a lower rate of HCV; thus, this is not currently a significant clinical problem. Adult studies have documented that cytomegalovirus (CMV) IgG positivity ranges between 80 % and 96 % of recipients [38, 39]. In paediatric recipients, this number will be significantly lower and thus may need increased use of anti-CMV prophylaxis. CMV viraemia seen in two adults responded well to orally administered valganciclovir . CMV PCR should be monitored on a regular basis; this will depend on test availability and protocol of the transplant unit. Epstein–Barr virus (EBV) infections were not specifically commented on in adult studies, which will have more significance in children. However, no cases of posttransplant lymphoproliferative disease related to EBV were documented in adult studies (the French adult study reported one non-EBV B-cell lymphoma ) but will require close monitoring in children, and it is beneficial to document donor and recipient EBV status at the time of transplant and then monitor EBV PCR levels regularly in the same way as CMV during follow-up. A high level of suspicion for opportunistic infections needs to be maintained, including for Kaposi’s sarcoma (KS) in older paediatric patients (KS is the most common malignancy in adult renal transplant recipients, with incidence varying from 47.7 % to 79.1 % ), Mycobacterium tuberculosis in high-risk areas, candidal oesophagitis, cryptosporidiosis and Pneumocystis jirovecii pneumonia.
There is currently insufficient knowledge regarding HIVAN and HIVICK to predict recurrence in the new graft in the same way as conventional focal segmental sclerosis, which can be as high as 30 % in first grafts .
Many children with HIV infection may already be on antifungal prophylaxis prior to transplantation, and there may be a need to continue this after surgery. It is important to take into account drug interactions, such as fluconazole with tacrolimus, in addition to protease inhibitors making drug level monitoring difficult. Topical agents, such as nystatin, may not be sufficient if patients have preceding fungal problems, such as lung infections. The newer azoles may have less interaction but still need close monitoring. Lifelong P. Jirovecii pneumonia prophylaxis has been used in adult studies with good effect (http://aidsinfo.nih.gov/guidelines). This is important in children who already have underlying chronic lung disease. Valganciclovir is also indicated as prophylaxis, especially in cases of CMV mismatch (donor positive with recipient negative), thus illustrating the importance of both donor and recipient status. There may also be a role for CMV prophylaxis to prevent reactivation of CMV disease in CMV-positive recipients . TB prophylaxis to prevent M. tuberculosis should be a priority in areas of the world where there is a high prevalence of TB and where HIV and TB coexist. The choice of agent is controversial in view of developing resistance, as well as once again being responsible for drug interactions along the p450 cytochrome system. Isoniazid (INH) is widely used in developing countries and is effective when resistance is not a major problem. In the USA, macrolide antibiotics have been recommended to prevent M. avium infection in adult studies .
What action has the shortage of organs prompted?
The biggest challenge currently affecting HIV-infected patients is the availability of organs, and this is a particular problem in resource-constrained areas where the rate of HIV is high but access to RRT is low, resulting in many patients being sent home to die. Once it was established that the outcome of renal transplantation was similar in HIV-infected and noninfected patients using HIV-negative donor kidneys, the focus was turned to unused HIV-positive deceased donor kidneys. In Cape Town, South Africa, four adult renal transplantations were performed from September to November 2008 using HIV-positive recipients and HIV-positive donors, with good outcomes [49, 50]. HIV-positive recipients were all receiving HAART and had stable disease with no evidence of opportunistic infections. To date, this group has performed 22 such transplants with success, having had two kidneys lost to rejection in sensitised patients. No evidence of accelerating HIV disease has been documented (personal communication, Elmi Muller). This regimen may be an alternative therapeutic approach where resources are severely limited. It has not, however, been performed in a paediatric patient but may be considered if circumstances require it.
Conclusion: What does the future hold?
The United Nations Children’s Fund (UNICEF) Millennium Development Goals plan to eliminate new HIV infections among children by 2015, which seems unlikely in the little time left. In the longer term, hopefully, this condition will be eradicated. However, until such time, there have been small numbers of paediatric HIV-infected patients receiving renal transplantation successfully in developed countries, as well as the first paediatric case in the UK. Paediatric transplantation for HIV-infected patients in developing areas such as Africa is mostly unavailable, and thus, it is unlikely that there will be large studies in children. As an alternative strategy, we can continue to learn from adult studies and extrapolate principles in order to perform safe and successful kidney transplantations in paediatric HIV-infected recipients.