Dear Sir: We read with interest the study by Kor et al. [1]. In this retrospective study, statins were not associated with a reduction in pulmonary or non-pulmonary organ dysfunction in patients with acute lung injury (ALI). We would suggest that the use of oxygenation, as measured by the PaO2/FiO2 ratio, as a marker of pulmonary dysfunction in ALI may be limited. This is supported by the ARDSnet low tidal volume study where oxygenation was lower in the first 3 days in the low tidal volume group despite this group having increased ventilator-free days (VFDs) and reduced mortality. Interestingly, while not significant, the statin cohort had higher VFDs and lower hospital mortality, although the authors appropriately point out there may be a confounding effect related to a propensity for statin administration.

The anti-inflammatory effects of statins are consistently seen in vitro and in animal studies of ALI. We have recently demonstrated the pulmonary anti-inflammatory effects of simvastatin in a model of ALI in healthy human volunteers induced by inhaled lipopolysaccharide [2]. Beneficial effects of statins have been seen in most, but not all, observational studies of patients admitted with pneumonia. Finally, in a prospective observational study of 196 patients with ALI admitted to Irish intensive care units, patients receiving treatment with a statin during admission, after adjusting for potential confounding variables, had a trend towards reduced mortality [3].

While we agree with the conclusion that the use of statin therapy in patients with ALI cannot be advocated at present, we would argue the findings from this study, as well as the studies above, strongly support the urgent need for randomised clinical trials of statins in ALI.

This study raises several other issues. While the anti-inflammatory effects of statins are likely to be a class effect, there are differences in the pleiotropic effects between lipophilic and hydrophilic statins [4], but unfortunately no data are presented in this study on statin usage and dose to inform subsequent studies. Furthermore, the metabolism of statins varies between individuals because of genetic variations, and this may result in differing plasma levels between individuals for a given dose of statin. This may be further accentuated by altered metabolism of statins in the critically ill. Kruger et al. [5] have shown that blood levels of atorvastatin after a single dose in critically ill patients varied widely. This may affect the balance of beneficial immunomodulatory and adverse effects of statins, and is likely to be an important consideration in designing subsequent clinical trials, particularly as the adverse effects of statins are dose dependent.

We are currently conducting two small phase-2 randomised controlled studies investigating the effects of simvastatin in the prevention (ISRCTN56543987) and treatment (ISRCTN70127774) of ALI. These will be reported within the next 12 months. If the beneficial effects of statins are demonstrated in these studies, larger studies powered for important clinical outcomes such as VFDs and mortality will be required to add clarity to the role of statins in ALI. For these future studies, it will be crucial to choose the appropriate statin and dose.