Dose-response relationships in aluminium toxicity in humans.

Introduction: Aluminium exposure is associated with bone disease (an elevated bone content of aluminium and reduced bone formation on bone biopsy) and neurotoxicity (features of altered brain functions and/or typical spike and slow wave waveforms on electroencephalogram) in patients with elevated blood aluminium concentrations.

Objectives: To critically analyse the literature to determine the dose-toxicity relationships between aluminium exposure and related bone disease and aluminium neurotoxicity.

Methods: A systematic review of the literature with collation and analysis of individual data of human cases of aluminium exposure was conducted between 1 January 1966 and 30 December 2020. Embase, MEDLINE (OVID MEDLINE), PubMed and TOXNET were searched with the following strategies: "Aluminium AND toxicity OR aluminium AND poisoning OR aluminium AND dialysis OR aluminium AND chronic renal failure OR aluminium AND intravenous" limited to "(human)". Inclusion criteria required individual data relating to aluminium exposure in humans. Papers in which features of aluminium toxicity and analytical confirmation of aluminium exposure could not be determined in individual patients were excluded.

Results: Thirty-seven papers were identified, which included data on 179 individuals exposed to aluminium. The sources of aluminium exposure (median duration of exposure) were: dialysis fluid (48 months) in 110 cases; oral aluminium hydroxide (20 months) in 20 cases; plasma exchange (2 months) in 16 cases; infant formula feed (minimal duration of 2 weeks) in 14 cases; intravesical exposures (2 days) in 13 oncology patients and potable water exposure in six cases.

Exposure to dialysis fluid: Of the 110 patients exposed to dialysis fluid, 99 were adults and 11 children, who were analysed separated. Of the adults, 50 with aluminium neurotoxicity had a median aluminium concentration of 467 µg/L (IQR 230 - 752), 28 with aluminium bone disease had a median aluminium concentration of 142 µg/L (IQR 46-309) and 21 with asymptomatic aluminium overload had a median aluminium concentration of 35 µg/L (IQR 26-51). Median aluminium concentrations were significantly greater in patients with aluminium neurotoxicity compared to those with aluminium bone disease (p < 0.0001) or asymptomatic aluminium overload (p < 0.0001).

Oral aluminium hydroxide: Of the 20 cases, 11 were adults and nine were children. Of the 11 adults, eight with aluminium neurotoxicity had a median aluminium concentration of 682 µg/L (IQR 438-770) and three with aluminium bone disease had a median aluminium concentration of 100 µg/L (IQR 62-138) (p = 0.007). Of the nine children, five had aluminium neurotoxicity with a median aluminium concentration of 335 µg/L (IQR 229-601), one had aluminium bone disease and an aluminium concentration of 1030 µg/L and three had asymptomatic aluminium overload with a median aluminium concentration 98 µg/L (IQR 65-365).

Plasma exchange: Three patients with stage 5 chronic kidney disease developed aluminium bone disease during plasma exchange; their median blood or serum aluminium concentration was 73 µg/L (IQR 59-81). Asymptomatic aluminium overload was reported in six patients receiving outpatient plasma exchange who had a median creatinine clearance of 71 mL/min (IQR 40-106) and a median aluminium concentration of 49 µg/L (IQR 34-116), and in seven intensive care patients with acute kidney injury whose median aluminium concentration was 30 µg/L (IQR 17-35); (p = 0.02).

Intravesical exposures: All 13 intravesical exposures developed aluminium neurotoxicity and had a median aluminium concentration of 157 µg/L (IQR 45-276).

Potable water: All six patients developed aluminium bone disease and their median blood aluminium concentration was 17 µg/L (IQR 13-100).

Conclusions: Toxic aluminium exposure can result in neurotoxicity and bone disease, especially in patients with chronic kidney disease. Adults with stage 5 chronic kidney disease chronically exposed to aluminium developed aluminium neurotoxicity at higher concentrations than those with aluminium bone disease or with asymptomatic aluminium overload. Aluminium neurotoxicity was reported at lower concentrations following acute exposure to intravesical aluminium. Extrapolating the relevance of these concentrations to the general population is problematic in that the data were derived from oncology patients, however, the possibility that aluminium neurotoxicity may occur at concentrations lower that those reported historically in patients with stage 5 chronic kidney disease cannot be excluded.