Review question
This review evaluated how common (the prevalence), and what might cause (the risk factors), early and late adverse effects (side effects), of the kidney in childhood cancer survivors (CCS).
Background
Improvements in diagnostics and treatment for childhood cancer has resulted in a major increase in survival. However, CCS are at risk of developing adverse effects as a result of their cancer treatment, including adverse effects on the kidneys. Little is known about the prevalence and risk factors of kidney function problems in long-term CCS. The kidneys can compensate for problems in their functioning, however, some symptoms can be experienced over time, depending on which kidney functions are affected.
Study characteristics
The evidence is current to March 2017. We included 61 studies; 46 on prevalence, six for both prevalence and risk factors, and nine studies that did not meet all the requirements for this review, but evaluated risk factors (non-eligible studies). Participants in the studies had been treated before the age of 21 years with chemotherapy (i.e. cisplatin, carboplatin, ifosfamide), radiation, or surgery involving the kidneys, or a combination of these treatments. The studies took place at least one year after the participants had finished their treatment. The 52 studies that evaluated prevalence of adverse kidney effects included 13,327 participants, of whom 4499 underwent kidney function tests. The studies were very different from each other, in the types of participants and treatments, length of follow-up and how they measured treatment results, and their methods were of variable quality.
Key results
The percentage of CCS with kidney problems ranged from 0% to 84%. Reported risk factors were often inconsistent among studies.
The prevalence of chronic kidney disease ranged from 2.4% to 32% in 244 participants (7/52 studies).
Thirty-six out of 52 studies, including at least 432 participants, carried out a kidney function test called glomerular filtration rate (GFR). An abnormal GFR was found to be present in 0% to 73.7% of participants. One eligible study found an increased risk of abnormal GFR in participants who had been treated with total body irradiation (TBI) and received certain types of antibiotics (aminoglycosides and vancomycin). Four non-eligible studies reported an increased risk of abnormal GFR for participants treated with surgery of the kidney and ifosfamide. Some studies also reported that cisplatin and long follow-up duration were risk factors.
Twenty-two out of 52 studies, including 851 participants, assessed an abnormal amount of proteins in the urine, which they found in 3.5% to 84% of participants. Risk factors, evaluated by three non-eligible studies, included cisplatin, ifosfamide, TBI, and a combination of surgery and radiation involving the kidney. However, the results of these studies did not agree, and we could not analyse their results together because they used different definitions.
Eleven out of 52 studies looked at a low level of phosphate in the blood (hypophosphataemia), or problems with the reabsorption of phosphate by the kidneys in 246 participants. Prevalence of hypophosphataemia ranged between 0% and 36.8% in 287 participants. The studies found problems with the reabsorption of phosphate by the kidneys in 0% to 62.5% of participants. One non-eligible study investigated risk factors, but could not find any association with hypophosphataemia.
Four out of 52 studies, including 128 CCS, evaluated a low level of magnesium in the blood (hypomagnesaemia). Prevalence ranged between 13.2% and 28.6%. Two non-eligible studies identified cisplatin as risk factor for hypomagnesaemia. Other reported risk factors were carboplatin, surgery of the kidney, and follow-up time. However, studies were contradictory.
The prevalence of high blood pressure ranged from 0% to 50% in 2464 participants (30/52 studies). Risk factors reported by one eligible study were older age at screening and radiation involving the kidney. A high body mass index was reported as a risk factor by three non-eligible studies. Other reported risk factors included follow-up time, and radiation involving the kidney or TBI. However, studies were contradictory.
Quality of evidence
All studies showed problems that could affect our confidence in their results. More, and especially higher-quality research is needed to gain better insight into kidney adverse effects and related risk factors.
The prevalence of adverse renal effects after treatment with cisplatin, carboplatin, ifosfamide, radiation therapy involving the kidney region, nephrectomy, or any combination of these, ranged from 0% to 84% depending on the study population, received treatment combination, reported outcome measure, follow-up duration and methodological quality. With currently available evidence, it was not possible to draw solid conclusions regarding the prevalence of, and treatment-related risk factors for, specific adverse renal effects. Future studies should focus on adequate study designs and reporting, including large prospective cohort studies with adequate control groups when possible. In addition, these studies should deploy multivariable risk factor analyses to correct for possible confounding. Next to research concerning known nephrotoxic therapies, exploring nephrotoxicity after new therapeutic agents is advised for future studies. Until more evidence becomes available, CCS should preferably be enrolled into long-term follow-up programmes to monitor their renal function and blood pressure.
Improvements in diagnostics and treatment for paediatric malignancies resulted in a major increase in survival. However, childhood cancer survivors (CCS) are at risk of developing adverse effects caused by multimodal treatment for their malignancy. Nephrotoxicity is a known side effect of several treatments, including cisplatin, carboplatin, ifosfamide, radiotherapy and nephrectomy, and can cause glomerular filtration rate (GFR) impairment, proteinuria, tubulopathy, and hypertension. Evidence about the long-term effects of these treatments on renal function remains inconclusive. It is important to know the risk of, and risk factors for, early and late adverse renal effects, so that ultimately treatment and screening protocols can be adjusted. This review is an update of a previously published Cochrane Review.
To evaluate existing evidence on the effects of potentially nephrotoxic treatment modalities on the prevalence of renal dysfunction in survivors treated for childhood cancer with a median or mean survival of at least one year after cessation of treatment, where possible in comparison with the general population or CCS treated without potentially nephrotoxic treatment. In addition, to evaluate evidence on associated risk factors, such as follow-up duration, age at time of diagnosis and treatment combinations, as well as the effect of doses.
On 31 March 2017 we searched the following electronic databases: CENTRAL, MEDLINE and Embase. In addition, we screened reference lists of relevant studies and we searched the congress proceedings of the International Society of Pediatric Oncology (SIOP) and The American Society of Pediatric Hematology/Oncology (ASPHO) from 2010 to 2016/2017.
Except for case reports, case series and studies including fewer than 20 participants, we included studies with all study designs that reported on renal function (one year or longer after cessation of treatment), in CCS treated before the age of 21 years with cisplatin, carboplatin, ifosfamide, radiation involving the kidney region, a nephrectomy, or a combination of two or more of these treatments. When not all treatment modalities were described or the study group of interest was unclear, a study was not eligible for the evaluation of prevalence. We still included it for the assessment of risk factors if it had performed a multivariable analysis.
Two review authors independently performed study selection, 'Risk of bias' assessment and data extraction using standardised data collection forms. We performed analyses according to the guidelines of the Cochrane Handbook for Systematic Reviews of Interventions.
Apart from the remaining 37 studies included from the original review, the search resulted in the inclusion of 24 new studies. In total, we included 61 studies; 46 for prevalence, six for both prevalence and risk factors, and nine not meeting the inclusion criteria, but assessing risk factors. The 52 studies evaluating the prevalence of renal dysfunction included 13,327 participants of interest, of whom at least 4499 underwent renal function testing. The prevalence of adverse renal effects ranged from 0% to 84%. This variation may be due to diversity of included malignancies, received treatments, reported outcome measures, follow-up duration and the methodological quality of available evidence.
Seven out of 52 studies, including 244 participants, reported the prevalence of chronic kidney disease, which ranged from 2.4% to 32%.
Of these 52 studies, 36 studied a decreased (estimated) GFR, including at least 432 CCS, and found it was present in 0% to 73.7% of participants. One eligible study reported an increased risk of glomerular dysfunction after concomitant treatment with aminoglycosides and vancomycin in CCS receiving total body irradiation (TBI). Four non-eligible studies assessing a total cohort of CCS, found nephrectomy and (high-dose (HD)) ifosfamide as risk factors for decreased GFR. The majority also reported cisplatin as a risk factor. In addition, two non-eligible studies showed an association of a longer follow-up period with glomerular dysfunction.
Twenty-two out of 52 studies, including 851 participants, studied proteinuria, which was present in 3.5% to 84% of participants. Risk factors, analysed by three non-eligible studies, included HD cisplatin, (HD) ifosfamide, TBI, and a combination of nephrectomy and abdominal radiotherapy. However, studies were contradictory and incomparable.
Eleven out of 52 studies assessed hypophosphataemia or tubular phosphate reabsorption (TPR), or both. Prevalence ranged between 0% and 36.8% for hypophosphataemia in 287 participants, and from 0% to 62.5% for impaired TPR in 246 participants. One non-eligible study investigated risk factors for hypophosphataemia, but could not find any association.
Four out of 52 studies, including 128 CCS, assessed the prevalence of hypomagnesaemia, which ranged between 13.2% and 28.6%. Both non-eligible studies investigating risk factors identified cisplatin as a risk factor. Carboplatin, nephrectomy and follow-up time were other reported risk factors.
The prevalence of hypertension ranged from 0% to 50% in 2464 participants (30/52 studies). Risk factors reported by one eligible study were older age at screening and abdominal radiotherapy. A non-eligible study also found long follow-up time as risk factor. Three non-eligible studies showed that a higher body mass index increased the risk of hypertension. Treatment-related risk factors were abdominal radiotherapy and TBI, but studies were inconsistent.
Because of the profound heterogeneity of the studies, it was not possible to perform meta-analyses. Risk of bias was present in all studies.