Article Text


Antiviral agents and corticosteroids in the treatment of severe acute respiratory syndrome (SARS)
  1. W C Yu1,
  2. D S C Hui2,
  3. M Chan-Yeung3
  1. 1Department of Medicine & Geriatrics, Princess Margaret Hospital, Hong Kong Special Administrative Region, China
  2. 2Faculty of Medicine & Therapeutics, The Chinese University of Hong Kong, Hong Kong Special Administrative Region, China
  3. 3Faculty of Medicine, The University of Hong Kong, Hong Kong Special Administrative Region, China
  1. Correspondence to:
    Dr W C Yu
    Department of Medicine, Princess Margaret Hospital, Lai King, Hong Kong SAR, China;

Statistics from

Systematic evaluation of treatment modalities for SARS is still needed

The epidemic of severe acute respiratory syndrome (SARS) of 2003 caught the medical profession by surprise. The accumulated global total number of cases was 8098 with 774 deaths, a case-fatality ratio of 9.6%.1 Although the novel coronavirus (SARS-CoV) was discovered within weeks,2 treatment was inevitably empirical as controlled clinical trials were not possible during the epidemic of this new and serious illness. Many antiviral and immunomodulatory drugs, as well as other treatments such as convalescent patient plasma and traditional Chinese medicines, have been tried. Ribavirin and corticosteroids are by far the most widely used treatments for SARS. In the later phase of the epidemic lopinavir and ritonavir in combination were also used in Hong Kong.



Ribavirin is used extensively for the treatment of SARS and was given to over 90% of patients in Hong Kong. It is a nucleoside analogue that has activity against a number of DNA and RNA viruses in vitro.3 The mechanism of action of ribavirin has been studied for decades and is still under active debate.4 In early March 2003, before the isolation of the SARS-CoV, many experts believed that the mysterious severe illness was due to an unknown virus and ribavirin was empirically given because of its broad spectrum antiviral activity. Furthermore, corticosteroids were increasingly prescribed for the treatment of SARS and some believed that such treatment would be dangerous if not covered with an antiviral agent. The published reports on the effectiveness of ribavirin were mostly retrospective case series with intrinsic methodological issues and it is difficult to draw conclusions. The major side effect of ribavirin is anaemia which occurs in 27–59% of patients.5–9 Anaemia reduces oxygen transport and potentiates the existing problem of oxygenation and tissue hypoxia. Other significant side effects include raised transaminases and bradycardia,5 as well as hypocalcaemia, hypomagnesaemia, and risk of teratogenicity.10 In a detailed study on the clinical course and viral load, Peiris et al11 reported that 14 patients given a standard regimen of ribavirin and steroids showed a peak viral load at day 10 from onset of illness. This study, although involving a small number of subjects, clearly indicated the inability of ribavirin to clear SARS-CoV from patients with SARS. The result of this study also explained why patients treated with ribavirin early in the illness were able to infect healthcare workers when they subsequently required endotracheal intubation. The lack of in vitro activity of the drug against SARS-CoV12–14 cast further doubts on the usefulness of ribavirin in SARS. The use of ribavirin in SARS has been reviewed elsewhere.15,16

Lopinavir and ritonavir

Lopinavir and ritonavir (LPV/r) are protease inhibitors which, in combination, have been licensed for the treatment of HIV disease. Ritonavir has little antiviral activity and its role is to inhibit CYP3A mediated metabolism of lopinavir, thus increasing the serum concentration of lopinavir. In the laboratory lopinavir and ribavirin have significant synergism in inhibiting SARS-CoV6 and, on that basis, this combination—together with steroids—have been used in some centres in Hong Kong since mid April 2003. In this retrospective study the authors found that the 12 patients who received early treatment with LPV/r together with ribavirin and steroids had significantly fewer 21 day adverse clinical outcomes (acute respiratory distress syndrome or death) than 111 historical controls receiving ribavirin and steroids. Other benefits of the LPV/r group included favourable viral load profiles (in six patients), early rise of lymphocyte counts, and a reduced need for “rescue” pulse steroid doses. Adverse events attributable to LPV/r were minimal. Similar findings were reported in a case controlled study involving more patients from Hong Kong.17 Randomised controlled trials are being planned in Hong Kong to confirm these results should SARS re-emerge.


Corticosteroids have been used widely to treat SARS, first in mainland China and then in Hong Kong. The main rationale for their use in SARS is that, in acute viral respiratory infections, early response cytokines such as interferon gamma (IFN-γ), tumour necrosis factor, interleukin 1 (IL-1), and interleukin 6 (IL-6) contribute to tissue injury,18,19 and corticosteroid treatment may suppress the “cytokine storm”.20 Peiris et al hypothesised that the clinical worsening often observed during the second phase of illness is the result of immunopathological damage from an overexuberant host response.13 In a newly published report Wong et al21 showed in 20 consecutive adults with SARS that there was a marked increase in the Th1 cytokine IFN-γ, inflammatory cytokines IL-1, IL-6, and IL-12 for at least 2 weeks after disease onset. The chemokine profile showed a significant increase in IL-8, monocyte chemoatttractant protein-1 (MCP-1), and IFN-γ inducible protein-10 (IP-10). Corticosteroids significantly reduce IL-8, MCP-1, and IP-10 concentrations 5–8 days after treatment. The data confirmed the Th1 cell mediated immunity and hyperinnate inflammatory response in SARS through the accumulation of monocytes/macrophages and neutrophils. Another rationale for use of steroids in SARS is the necroscopic finding of features of acute respiratory distress syndrome (ARDS),22,23 and there have been reports of successful use of steroids in the treatment of ARDS24 and septic shock.25 In addition, systemic steroids have been used in the treatment of some infections with variable success.26–29 On the other hand, the potential for corticosteroids to suppress the innate host defence against SARS-CoV resulting in increased viral replication has to be considered. Chu et al reported an increase in viral load in one patient following pulse methylprednisolone therapy.6 Increased replication of other respiratory viruses has also been reported following steroid therapy.26,30–32

Whereas “low dose” steroids at 0.5–1.0 mg/kg/day prednisolone (or equivalent) have been used in infections, ARDS and septic shock, “pulse doses” at 0.5–1.0 g/day methylprednisolone have generally not been recommended for these conditions but were used extensively in SARS, particularly in the second week of illness when patients often show acute clinical deterioration. The efficacy of pulse steroids in SARS remains to be determined, but it is conceivable that higher steroid doses will result in a higher incidence and severity of side effects.

Published case series examining the clinical efficacy of steroid treatment in SARS7,9,33–40 suffer the same methodological problems as those of ribavirin. In addition, there is a wide variety of steroid dosing schedules making retrospective analysis of steroid efficacy exceptionally difficult. There is so far no systematic review of the efficacy of corticosteroid treatment in SARS based on the numerous published studies. Some investigators do feel that judicious use of corticosteroids is beneficial, but randomised controlled studies are needed to confirm the beneficial effects as well as to give insight into the optimal regimen. The possible beneficial effects, however, have to be balanced against the significant side effects including nosocomial infections,7,9,40,41 hyperglycaemia, hypokalaemia, hypertension, and gastrointestinal haemorrhage.7–9 Avascular necrosis of bone (AVN) is perhaps the most distressing medium term side effect of steroids in patients with SARS. Preliminary data on a cohort of 330 adult patients from Princess Margaret Hospital, Hong Kong who received various doses of steroids and in whom magnetic resonance imaging was performed at an average of 7.5 months from illness onset showed that AVN was present in 48 of them (14.5%, (unpublished data). Of the 48, 16 (33%) had unilateral involvement of the femoral head and 19 (40%) had bilateral involvement of the femoral head. Univariate analysis showed that the total steroid dose was significantly associated with development of AVN (unpublished data).


As SARS has only recently appeared and a limited number of patients have been managed in different locations, it is understandable that there has been a lack of systematic and critical evaluation of treatment in the form of randomised controlled trials. Nonetheless, the enormous effort that researchers put into looking for effective treatments for SARS is highly commended. The recent re-emergence of SARS did not result in secondary spread, but is nevertheless a reminder that it could strike again. What may be even more threatening is the deadly avian influenza A (H5N1) which has repeatedly demonstrated its ability to infect humans, and may acquire the ability for efficient human to human transmission in the future. It is hoped that, when epidemics of new disease strikes, a systematic way of evaluating treatment modalities would be in place to provide answers to important questions in the shortest possible time.

Systematic evaluation of treatment modalities for SARS is still needed


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.