The main result of our study is that lactate under non acid conditions is most effective in aluminum liberation with uncoated heating devices. It was confirmed, that sodium chloride solution itself has only very limited capacity of aluminum liberation. However, aluminum concentrations measured with the use of acetate buffered sodium solutions or commercial balanced electrolyte solutions are reasonable.
Aluminium is one of the most studied toxic metals and associated with many diseases [6] like encephalopathies such as Alzheimers [7], impaired neurologic development of (preterm) infants [8] and osteomalacia [9]. Although soluble aluminum in plasma underlies renal clearance, a deposition of up to 20% is possible [10].
One possible safety threshold is the recommended threshold for parenteral nutrition as this threshold reflects an intravenous administration. For preterm infants with need of parenteral nutrition a repeated intravenous aluminum supply of > 4–5 µg kg− 1 d− 1 is a risk for encephalopathy, impaired neurologic development and osteomalacia [8, 9]. The American Society for Clinical Nutrition (ASCN) and the American Society for Parenteral and Enteral Nutrition (ASPEN) therefore recommended a threshold of 25 µg l− 1 [11] which is also noted by the United States FDA [12].
A different approach is to apply the oral minimum risk level for aluminum, derived by the Agency for Toxic Substances and Disease Registry to be at 1 mg.kg− 1 d− 1 [13] [14]. Aluminum shows a poor bioavailability with only 0.1% resorption after oral administration [15]. If the calculated threshold with a correction factor of 1000 based on the bioavailability is applied, the proposed calculations lead to a tolerable parenteral minimal risk level of not more than 1 µg kg− 1 day− 1 or 70 µg kg− 1 day− 1 for a 70 kg adult. However, the nonrecurring use of a fluid warmer is different to long term use of parenteral nutrition or a daily oral dose and therefore the limits cannot be transferred directly to the occasional use of fluid warming. A third way of estimating the maximum safe aluminum exposure is to apply the threshold for aluminum salt in vaccines of 850 µg dose− 1 [16]. To date there are no regulations for maximal tolerable aluminium concentrations of iv fluids available.
Even if these values are inconsistent, it can be assumed that any kind of aluminum infusion is inadvertent and a dosage of more than 1000 µg for adults is unsafe. Regarding the results of this investigation not only the concentration of aluminum in the warmed fluid alone is relevant. For the estimation of safety, the concentration has to be multiplied with the administrated volume. Fluid warming is indicated and recommended fluid rates of more than 500 ml h− 1. For bleeding patients e.g. after major trauma high volume fluid demands of more than 5 l.h− 1 are reasonable. In consequence the observed concentrations of aluminum liberated by uncoated fluid warming devices as enFlow® and Level1® are both not safe and bear a potential risk of aluminum intoxication and maybe long term effects like enzephalopathia [17] [18]. Our study confirms the observation of aluminum liberation by the enFlow® device [3] [4] and the Level1 [19]. Higher observed aluminum concentrations in this study compared to the investigation of Cabrera et al. [19] may be explained by the use of a high flow disposable (Level1® DI-300) in this investigation rather than the moderate flow disposable (Level1® DI-100) [19]. The difference between these disposables is a higher efficacy for the DI-300, presumably by enlarged surface of the heat exchanging element.
In the previous study [3] the increase of temperature was associated with an increase in aluminum liberation by an uncoated device. Temperature is influential and heat may theoretically enhance both chemical reactions and dissolution rates. It was also demonstrated previously that the flow rate has an effect on aluminum concentration. Higher flow rates are accompanied with lower aluminum concentrations, but of course the amount of aluminum infused will be less affected. The results from the actual investigation describe a strong effect of pH on aluminum liberation mediated by organic acids (lactate, acetate or malate) and their salts. At a pH of 9, aluminum hydroxide is formed and this is coordinated to the lactic acid or the corresponding salt [20]. At a high lactate concentration, the aluminum complex is also formed to a greater extent. But unlike acetic acid, lactic acid forms a stable complex with aluminum. This finding is in accordance with Fig. 1 and Fig. 2. This effect may also explain the results of an investigation of the enFlow® device using blood products [4] reporting rather low levels of aluminum in expired packed red blood cells because these blood products do not only contain high levels of lactate but also have a low pH.
The strength of a bench investigation is the standardized examination of aluminum liberation. However, a problem of any bench investigation is that no resulting plasma concentrations of aluminum can be reported. We can’t also assess any patient outcomes such as cognitive dysfunction induced by the infused aluminium. However, the exposition of patients resulting from uncoated fluid warming devices using aluminum as heating element can be estimated and bear a potential risk of toxic levels.
In conclusion, uncoated fluid warming devices using aluminum as heating element bear the risk of aluminum liberation in a potential toxic amount. Organic acids and their salts like lactate and acetate are compounds provoking the reaction, especially under non acid conditions. Test protocols for leaching of aluminum from fluid warming devices should apply lactate spiked saline better than saline solution.