Long-term cisplatin nephrotoxicity after childhood cancer: a systematic review and meta-analysis

Cisplatin is a chemotherapeutic drug commonly used in the treatment of many childhood solid malignancies. It is known to cause long-term nephrotoxicity, most commonly manifesting as reduced glomerular filtration rate and hypomagnesaemia. Existing literature regarding the epidemiology of long-term nephrotoxicity in childhood cancer describes large variation in prevalence and risk factors.

This study is to evaluate the prevalence of, and risk factors for, long-term cisplatin nephrotoxicity after treatment for childhood cancer.

Studies were eligible for inclusion if they: (i) evaluated participants treated with cisplatin who were diagnosed with cancer < 18 years of age; (ii) investigated any author-defined measure of nephrotoxicity; and (iii) performed this evaluation 3 or more months after cisplatin cessation. Studies whose scope was broader than this were included if appropriate subgroup analysis was performed.

Prevalence of reduced glomerular filtration rate (GFR) ranged between 5.9 and 48.1%. Pooled prevalence of reduced GFR using studies with a modern consensus threshold of 90 ml/min/1.73 m² was 29% (95% CI 0.0-58%). Prevalence of hypomagnesaemia ranged between 8.0 and 71.4%. Pooled prevalence of hypomagnesaemia was 37% (95% CI 22-51%). Substantial heterogeneity was present, with I² statistics of 94% and 73% for reduced GFR and hypomagnesaemia respectively. All large, long-term follow-up studies described increased risk of reduced GFR with increasing cumulative cisplatin dose. Included studies varied as to whether cisplatin was a risk factor for proteinuria, and whether age was a risk factor for cisplatin nephrotoxicity.

A wide range of study methodologies were noted which impeded analysis. No studies yielded data from developing health-care settings. No non-English studies were included, further limiting generalisability.

Both of the most common manifestations of long-term cisplatin nephrotoxicity have a prevalence of approximately a third, with increasing cumulative dose conferring increased risk of nephrotoxicity. Further work is needed to characterise the relationship between reduced GFR and hypomagnesaemia, investigate other risk factors and understand the interindividual variation in susceptibility to nephrotoxicity.