American Journal of Respiratory and Critical Care Medicine

The treatment of atypical pneumonia, subsequently termed severe acute respiratory syndrome (SARS), is controversial, and the efficacy of corticosteroid therapy is unknown. We have evaluated the clinical and radiographic outcomes of 72 patients with probable SARS (median age 37 years, 30 M), who received ribavirin and different steroid regimens in two regional hospitals. Chest radiographs were scored according to the percentage of lung field involved. Seventeen patients initially received pulse steroid (PS) (methylprednisolone ⩾ 500 mg/day) and 55 patients initially received nonpulse steroid (NPS) (methylprednisolone < 500 mg/day) therapy. The cumulative steroid dosage; intensive care unit admission, mechanical ventilation, and mortality rates; and hematologic and biochemical parameters were similar in both groups after 21 days. However, patients in the PS group had less oxygen requirement, better radiographic outcome, and less likelihood of requiring rescue PS therapy than their counterparts. There was no significant difference between the two groups in hemolytic anemia, severe secondary infections, or hematemesis, but patients in the PS group had less hyperglycaemia. Initial use of pulse methylprednisolone therapy appears to be a more efficacious and an equally safe steroid regimen when compared with regimens with lower dosage and should be considered as the preferred steroid regimen in the treatment of SARS, pending data from future randomized controlled trials.

Since the first reported outbreak of atypical pneumonia (subsequently termed severe acute respiratory syndrome or SARS) in Guangdong Province of China in late 2002, successive similar outbreaks were widely reported from March 2003 onward in Hong Kong (1, 2), Canada (3), and around 30 countries in the world. At the time of writing of this article, there were more than 8,000 reported probable SARS cases worldwide, with around 700 deaths. In Hong Kong, as of May 27, 2003, there were 1,728 reported cases of SARS, with 269 deaths (4). It is now known that SARS is caused by a novel coronavirus (SARS-CoV) (57). On the basis of early anecdotal experience and its radiologic resemblance to acute respiratory distress syndrome and other immunologically mediated lung diseases, corticosteroid has been used for the treatment of SARS (8). Ribavirin, a broad-spectrum antiviral agent, has also been used to treat SARS, although this has drawn considerable skepticism (9, 10). The rapid emergence of the epidemic did not permit pulmonologists to conduct any randomized controlled trials on the treatment of SARS, although there is a wealth of data ready for retrospective analysis. Although the incidence of SARS is declining in Hong Kong, mainland China, and Singapore, further outbreaks in the near future are still anticipated.

The use of corticosteroids is no less controversial than ribavirin in the treatment of SARS, and there are three different regimens used by physicians in Hong Kong. These were the use of hydrocortisone or methylprednisolone at dosages approximating to those used for treatment of acute severe asthma and the use of high-dose pulse methylprednisolone therapy similar to those used in treatment of organ rejection or acute respiratory distress syndrome (1114). Early anecdotal experience in mainland China and Hong Kong also showed that patients with SARS could deteriorate to acute respiratory distress syndrome and that many of these patients had responded to treatment with high-dose pulse methylprednisolone. The effectiveness of this regimen in improving radiologic and clinical features of otherwise deteriorating patients, initially treated with nonpulse steroid (NPS) regimens, has prompted some clinicians to use it as the initial steroid regimen for treatment of SARS. However, the efficacy and clinical outcome of these different steroid regimens are unknown. In addition, it is unclear if the initially low-dosage regimens are associated with less cumulative steroid usage, which is obviously desirable, particularly in view of the poor understanding of the pathogenesis of SARS pneumonia (15, 16). To address these issues, we have performed a retrospective study on the short-term clinical and radiographic outcomes of treatment with ribavirin and different steroid regimens for a cohort of 72 patients with probable SARS.

Subject Recruitment

The Queen Mary and the Queen Elizabeth Hospitals are regional general teaching hospitals in Hong Kong. Since the outbreak of SARS in Hong Kong in March 2003, patients with probable SARS, based on World Health Organization and CDC definitions (17, 18), were admitted to both hospitals either through the Accident and Emergency Department or referred from other hospitals. In brief, patients with probable SARS presented with high fever (> 38°C), cough or dyspnea, contact history with patients with SARS, and pulmonary infiltrates consistent with pneumonia on chest radiographs or high-resolution computed tomography of thorax that could not be explained by an alternative diagnosis. Microbiologic tests, to exclude other respiratory tract infections, included blood culture, sputum bacterial culture, sputum/nasopharyngeal aspirate for common respiratory viral antigens (influenza, parainfluenza, respiratory syncytial virus, and adenovirus), and serology for Chlamydia pneumoniae, Mycoplasma pneumoniae, and Legionella pneumophila. Microbiologic identification of SARS-CoV infection was performed by using reverse transcriptase–polymerase chain reaction to detect SARS-CoV ribonucleic acid in clinical specimens (sputum, nasopharyngeal aspirate, stool, or urine) and/or presence of a fourfold increase in anti–SARS-CoV IgG after 21 days as described previously (5, 1719). We retrospectively reviewed the clinical charts and chest radiographs for all patients with probable SARS admitted from March 9, 2003 to April 17, 2003.

Treatment Protocol

Broad-spectrum antibiotics (cefepime 2 g, intravenously three times a day or ceftriaxone 2 g, intravenous two times a day plus clarithromycin 500 mg, orally two times a day, or levofloxacin 500 mg, intravenously daily alone in the presence of penicillin allergy) for 10 to 14 days were commenced for all patients, as per standard protocols for treatment of severe community-acquired pneumonia. The timing to commence corticosteroid and ribavirin therapy varied for each individual patient and in our units required the consensus of the two designated pulmonologists for each case. The general principles were the presence of continued clinical instability or deterioration (increasing oxygen requirement and/or worsening of cough or dyspnea); progressive radiographic or high-resolution computed tomography deterioration or lack of improvement; explicit contact history with a patient with probable SARS; persistent lymphopenia and rise in alanine transaminase; and confident exclusion of other mimicking conditions. As the results of reverse transcriptase–polymerase chain reaction identification of SARS-CoV from saliva, urine, or stool and serologic evidence of SARS-CoV infection usually took considerable time to return, these were not always awaited for rapidly deteriorating patients who otherwise fulfilled the criteria described previously.

Once the diagnosis of SARS was confirmed, specific anti-SARS treatment included ribavirin (8 mg/kg, intravenously three times a day for the 7 days and then orally at 1.2 g three times a day for altogether 10–14 days) and systemic corticosteroid (1, 2, 8). Three different steroid regimens were identified. These included hydrocortisone (2 mg/kg, intravenously four times a day or 4 mg/kg, intravenously three times a day for 3–5 days, followed by oral prednisolone at 2 mg/kg daily at reducing dosage) and methylprednisolone ( 2–3 mg/kg, intravenously once daily for 5 days, followed by oral prednisolone at 2 mg/kg daily at reducing dosage). For these two arms, the initial dose of corticosteroid was equivalent to approximately 2 to 3 mg/kg of methylprednisolone daily. The third steroid treatment regimen was pulse methylprednisolone (500 mg, intravenously once daily for 5–7 days or 1 g, intravenously once daily for 3 days, followed by maintenance oral prednisolone 50 mg two times a day reducing to 20–30 mg daily on Day 21). This “pulse steroid” (PS) treatment had become the standard steroid regimen since March 30, 2003 at Queen Mary Hospital, as our teams of physicians had acquired more anecdotal experience on the benefits of “rescue PS” therapy in severe SARS (Figures 1 and 2)

. Irrespective of the initial steroid treatment regimen, rescue pulse methylprednisolone therapy (500 mg, intravenously daily for 3–5 days) was given to patients who failed to improve clinically (i.e., persistent oxygen requirement of ⩾ 2 L/minute for ⩾ 2 days, and/or progressive dyspnea at rest for ⩾ 2 days) and radiographically (i.e., static radiographic score, described below, for ⩾ 2 days). This rescue pulse therapy was also administered to patients who had deteriorated clinically (i.e., worsening dyspnea for ⩾ 1 day, and/or increasing oxygen requirement by ⩾ 2 L/minute for ⩾ 1 day) and radiographically (i.e., increase in radiographic score by two compared with admission chest radiograph) for 1 day or more. All patients were hospitalized in isolation wards for at least 21 days after commencement of anti-SARS therapy, and thus Day 21 was chosen as the cut-off for assessment of short-term outcomes. Patients were treated with supplemental oxygen when their resting oxygen saturation was less than 96% while breathing room air.

Clinical, Laboratory, and Radiographic Parameters

Throughout the hospitalization period, clinical parameters (heart rate, oxygen saturation, oxygen requirement, and temperature), routine blood tests (complete blood counts and alanine transaminase), and chest radiographs were performed daily. A thoracic radiologist, who was unaware of the origin of the films, reviewed the latter. The chest radiographs were scored according to the percentage of area involved, manifested as ground glass opacification, consolidation or nodular shadow, in each lung with a maximal score of 10 (equivalent to 100% area involved). The overall chest radiographic score was the summation of scores from both lungs, with a total score of 20 (20). The oxygen requirement (L/minute) was determined as the maximal level when assessed thrice daily by providing oxygen to each resting patient via nasal cannulae to maintain an SaO2 of 97% or more. The number of patients requiring supplemental oxygen therapy and the number of days for such requirement were determined for each treatment group.

Statistical Analysis

The demographic data and clinical, laboratory, and radiographic parameters were summarized in median and range or interquartile range (IQR) where appropriate. Mann–Whitney U test or Fisher Exact test was used to compare data or outcomes between treatment groups whenever appropriate. Categorical variables, namely comorbidities and oxygen requirements, were compared between treatment groups using chi-square tests. General linear model analysis with repeated measures was performed to compare radiographic scores between patients who received pulse methylprednisolone as initial therapy and their counterparts over the 21-day study period. The analysis was performed using the SPSS 10.0 Version package (SPSS Inc., Chicago, IL). A p value less than 0.05 was taken as statistically significant.

Baseline Clinical Characteristics and Steroid Treatment Groups

The age, comorbidities, interval between symptom onset to commencement of anti-SARS therapy, and clinical parameters, laboratory parameters, and radiographic score for the 72 patients are shown in Table 1

TABLE 1. Demographic and clinical characteristics of 72 patients with severe acute respiratory syndrome on day 1 of anti–severe acute respiratory syndrome therapy

Steroid Treatment Group

Clinical Characteristics
All Patients (n = 72; 42 F)
PS (n = 17; 10 F)
NPS (n = 55; 32 F)
p Value
Age, yr37 (23–82)38 (25–82)36 (23–73)0.45
Number of comorbidities0.69
Ischemic heart disease101
Diabetes mellitus202
*Number of days from symptom onset to anti-SARS treatment
4 (2, 6)4 (3, 7)4 (2, 6)0.19
*Clinical parameters on Day 1 of anti-SARS treatment
Body temperature, °C38.6 (37.9, 39.3)38.7 (37.6, 39.3)38.6 (38.0, 39.3)0.91
Pulse rate, per min100 (90, 109)100 (93, 110)100 (84, 108)0.49
SaO2, %96 (95, 97)95 (94, 96)96 (96, 98)0.02
*Laboratory parameters on Day 1 of anti-SARS treatment
Total leukocyte count, × 109/L5.1 (3.8, 6.7)4.8 (3.5, 6.4)5.6 (3.9, 7.7)0.18
Neutrophil count, × 109/L4.2 (3.0, 6.1)3.6 (2.8, 5.1)4.5 (3.2, 6.6)0.15
Lymphocyte count, × 109/L0.8 (0.6, 1.1)0.9 (0.8, 1.1)0.8 (0.6, 1.2)0.39
Platelet count, × 109/L187 (135, 233)182 (135, 263)189 (135, 227)0.85
Alanine transaminase, U/L28 (16, 43.8)33 (22.5, 66.3)26 (15, 42.8)0.13
Chest radiographic score, median (IQR)
2 (1, 4)
2.5 (1.8, 5)
2 (1, 3.5)

Definition of abbreviations: F = female; IQR = interquartile range; NPS = nonpulse steroid; PS = pulse steroid; SARS = severe acute respiratory syndrome.

Unless otherwise stated, data are expressed in median (range or IQR) where appropriate. p Values were obtained by comparing data between the PS and NPS groups.

. None of the patients was on regular medications except one patient with ischemic heart disease whose aspirin was stopped on commencement of steroid therapy. Antibiotic administration was similar for the two treatment groups and included the use of cefepime (n = 31) or ceftriaxone (n = 19) in combination with clarithromycin (n = 50) and levofloxacin (n = 22).

All 72 patients received corticosteroid and ribavirin therapy (Figure 3)

. As the hydrocortisone and methylprednisolone arms had very similar dosages, steroid treatment regimens were categorized as PS (equivalent to methylprednisolone ⩾ 500 mg/day) (n = 17) or NPS (equivalent to methylprednisolone < 500 mg/day) (n = 55). Of the NPS group, 21 patients received methylprednisolone (median 150, IQR 150–180 mg/day) and 34 patients received hydrocortisone (median 480, IQR 320–480 mg/day) as initial steroid therapy. Microbiologic identification of SARS-CoV infection was demonstrated in 69 patients by reverse transcriptase–polymerase chain reaction in clinical specimens (sputum, nasopharyngeal aspirate, stool, or urine) (overall 47.2%, PS 64.7%, and NPS 41.8%; p = 0.17) and by CoV IgG seroconversion (overall 95.8%, PS 94.1%, and NPS 96.4%; p = 0.56). PS therapy was used as rescue therapy in the presence of clinical (deteriorating dyspnea and/or increasing oxygen requirement) and radiologic (progressive increase in pulmonary infiltrates) deterioration for PS (n = 4) and NPS (n = 45, p < 0.0001) groups. The median (IQR) interval between the end of initial PS and the subsequent rescue therapy was 4.5 (2.5, 5) days for the PS group. The median (IQR) time between commencement of initial steroid and such rescue therapy was 5 (3, 6.5) days for the NPS group. The baseline clinical parameters on commencement of anti-SARS treatment, namely body temperature and pulse rate, laboratory parameters and chest radiographic scores, were similar for both groups (p > 0.05), although patients in the PS group had significantly lower SaO2 than their counterparts (p = 0.02, Table 1). There was no significant difference in the cumulative steroid dosage on Day 21 between PS (median 4.7 g methylprednisolone, range 1.4–14.5) and NPS (median 4.4 g methylprednisolone, range 0.3–11.4) groups (p = 0.33).

Clinical and Radiographic Outcomes

The daily clinical and radiographic patterns during the 3-week anti-SARS treatment are shown on Figure 4

. In both treatment groups, fever resolved within the first two days of treatment. The laboratory parameters had similar trends in both groups. Clinical (temperature, heart rate, and SaO2) and laboratory (total leukocyte, lymphocyte and platelet counts, and alanine transaminase) parameters were not significantly different between the two treatment groups on Day 21 (Table 2)

TABLE 2. Comparison of clinical and radiographic outcome parameters in 72 patients with severe acute respiratory syndrome with different steroid regimens

Steroid Treatment Group

PS (n = 17)
NPS (n = 55)
p Value
Clinical parameters on Day 21 of anti-SARS treatment
Body temperature, °C36.7 (36.2, 37.1)36.8 (36.5, 37)0.53
Pulse rate, per min90 (82, 99)88 (80, 98)0.70
SaO2, %96.5 (96, 98)96 (95.8, 98)0.86
Laboratory parameters on Day 21 of anti-SARS treatment
Total leukocyte count, × 109/L10.8 (8.8, 12.5)11.1 (8.6, 12.8)0.91
Neutrophil count, × 109/L9.3 (8.1, 10.4)9.7 (7.6, 12.2)0.54
Lymphocyte count, × 109/L0.73 (0.6, 1.4)0.75 (0.5, 1.2)0.64
Platelet count, × 109/L263 (227, 318)267 (226, 318)0.91
Alanine transaminase, U/L81 (34, 104)37 (30, 75)0.34
Admission to intensive care unit (n)1110.27
Need of mechanical ventilation (n)151.00
Number of deaths

For definition of abbreviations see Table 1.

Unless otherwise stated, data are expressed in median (IQR), where appropriate. p Values were obtained by comparing data between the PS and the NPS groups.


Significantly more patients in the NPS group (n = 29 of 55) required supplemental oxygen therapy than their counterparts (n = 4 of 17, p = 0.04). Patients in the NPS group had significantly longer interval of using supplemental oxygen therapy than those in the PS group (2.0, IQR 0.0 and 10.0 days; 0.0, IQR 0.0 and 1.0 days; p = 0.01). However, there were no significant differences between patients in the NPS and PS groups in the median (1.0, IQR 0.0 and 2.6 L/minute; 0.0, IQR 0.0 and 1.0 L/minute, p = 0.36) or in the maximal concentrations (1.0, IQR 0.0 and 4.0 L/minute; 0, IQR 0.0 and 1.0 L/minute; p = 0.24) of supplemental oxygen therapy. Among the patients who required supplemental oxygen therapy, there were no significant differences between those in the NPS and PS groups in the median oxygen requirement (2.5, IQR 1.5 and 5.2 L/minute; 4.0, IQR 2.4 and 10.7 L/minute; p = 0.60), in the maximal oxygen requirement (4, IQR 2.0 and 10.0 L/minute; 5.5, IQR 2.8 and 12.8 L/minute; p = 0.69), or in the number of days for which such therapy was required (10.0, IQR 4.5 and 14.5 days; 8.0, IQR 2.5 and 12.0 days; p = 0.45).

Patients in the PS group appeared to have their worst median radiographic scores on Day 4 and a steady improving trend thereafter to virtually completely normal chest radiograph (i.e., score < 1) on Day 17 (Figure 4A). However, the radiographic score continued to deteriorate for patients in the NPS group until Day 10, i.e., 5 days after they received rescue pulse methylprednisolone therapy. The radiographic score for NPS showed a steady improvement thereafter (Figure 4B). With the improving trends for both treatment groups, the median radiographic score reduced to 1 or less (i.e., nearly normal) on Day 12 for the PS group and on Day 20 for the NPS group. General linear model analysis showed that the overall trend for the chest radiographic scores was significantly lower for the patients in the PS compared with the NPS group (p = 0.026). There was no significant difference between the radiographic scores on Day 7 (p = 0.13), although the patients in the PS group had significantly lower scores than did patients in the NPS group on Days 14 and 21 of treatment (p = 0.04 and p = 0.04, respectively). There was no statistical difference between the PS group and the NPS group in the need of intensive care, use of mechanical ventilation, and mortality during the 3 weeks of SARS treatment (Table 2).

Adverse Effects from Anti-SARS Treatment

In our cohort of 72 patients with SARS, there were no major life-threatening complications. Adverse effects attributable to drug treatment included hemolytic anemia (defined as a 1.5-fold increase in bilirubin; overall 33.3%, PS 47.1%, NPS 29.1%; p = 0.24), hyperglycemia (defined as random blood glucose ⩾ 11.0 mmol/L; overall 25%, PS 0%, NPS 32.7%; p = 0.004), serious secondary infections (defined as pyrexial or bacteremic illness; overall 4.2%, PS 5.9%, NPS 3.6%; p = 0.56), and hematemesis (frank blood or coffee ground vomiting; overall 4.2%, PS 5.9%, NPS 3.6%; p = 0.56).

Our data, the first study comparing the efficacy of different steroid regimens in the treatment of SARS, showed no significant differences in intubation and in intensive care unit admission or mortality rates between the two treatment groups although patients in the PS group had significantly less oxygen requirement, better radiographic outcome, and less likelihood of requiring rescue PS therapy than their counterparts. There were no significant differences on the adverse prognostic factors, including age, presence of diabetes and chronic diseases, high baseline neutrophil count and presence of hepatitis B, and lamivudine therapy, between the study groups, although peaked lactate dehydrogenase levels were not routinely measured in our centers (2, 9, 19). Serious side effects attributable to treatment, including hemolytic anemia, severe secondary infections, or hematemesis, were not significantly different between the two treatment groups. In fact, patients in the PS group were significantly less likely to develop hyperglycaemia than their counterparts. Using the same criteria for giving rescue PS therapy to the patients in the NPS group, it was apparent that 45 of 55 patients in the NPS group actually needed rescue PS therapy to deal with deteriorating clinical and radiologic course. Although this lack of difference could be explained by the high proportion of patients in the NPS group requiring PS rescue therapy, we could still conclude that initial adoption of PS therapy could provide better clinical and radiologic outcomes without actually using more steroid in the treatment of SARS. Nevertheless, caution has to be exercised when interpreting these potentially biased retrospective data. The difference in outcomes was unlikely to be the result of evolving clinical practice, which had remained the same on treatment protocol, diagnostic criteria, and timing for commencement of anti-SARS treatment within the 5-week period. In addition, more experienced clinicians were increasingly challenged with patients who had less clear-cut epidemiologic linkage, as SARS became more “community-acquired” in Hong Kong. The difference in sample size between PS and NPS groups arose as Queen Mary Hospital adopted the use of PS over NPS as initial steroid therapy on March 30, 2003, although the clinical efficacy of corticosteroid therapy remains unproven in SARS.

Although data are lacking, it is worrisome that PS therapy could be associated with more opportunistic and ventilator-associated infections among intubated patients. It, therefore, appears logical that PS therapy should be considered earlier (or not too late) on in the deteriorating course of a patient with SARS to avoid this scenario. The better efficacy of PS over NPS in achieving radiographic resolution, and the frequent need of rescue PS among patients initially treated with NPS therapy, suggests that the pulmonary inflammation and infiltration in SARS requires high-dose steroid therapy and provides an insight into the pathogenesis of SARS. Our findings should form a strong basis for the design of future controlled trials for the treatment of SARS, although the current outbreak appears to be under control at the time of writing of this article. Should there be a recurrence, trials, particularly those designed to evaluate the efficacy of PS and NPS versus placebo therapy on conventional clinical parameters and more sophisticated ones such as Acute Physiology and Chronic Health Evaluation II and quality-of-life assessment instruments, should be in place and ready to commence at a moment's notice.

In a recent report of a major SARS outbreak involving 75 residents in a housing estate in Hong Kong, the treatment regimen consisted of ribavirin and initially hydrocortisone 800 mg/day for 10 days followed by a tailing course of prednisolone, with pulsed methylprednisolone if clinically deteriorated (19). The requirement of intensive care and mechanical ventilation was 32 and 20%, respectively. The mortality was reported to be 7%. The short-term clinical outcomes of 144 patients with SARS from Toronto were also reported recently (9). The treatment consisted of ribavirin at high dose (loading dose of 2 g intravenously followed by 4 g/day for 4 days and then 1.5 g/day for 3 days) in 88% of patients and a low dose steroid (20–50 mg/day of hydrocortisone). Of the entire cohort, 20% of patients were admitted to intensive care unit, and 14% required mechanical ventilation. The 21-day mortality was 6.5%. Comparing with these reported series, the clinical outcomes of our cohort were similar, with 17% requiring intensive care, 8% requiring mechanical ventilation, and 5.6% 21-day mortality. However, the treatment regimens were quite different among all these series, and the baseline clinical characteristics or severity might be different, thus making direct comparisons difficult.

Although little is known on the pathogenesis of SARS-CoV infection of the lung, a recent autopsy study suggested that cytokine dysregulation could be an important pathogenic mechanism (21). The ideal anti-SARS treatment should comprise an effective anti–SARS-CoV drug and an immunomodulating agent to dampen down the excessive and harmful immunologic response. Ribavirin, a broad-spectrum antiviral agent (22, 23), was originally empirically used in SARS as a desperate measure during very uncertain times when physicians had virtually no knowledge on SARS. A recent report on the Toronto experience showed that 76% of the patients in the study experienced hemolysis, compared with 33% in this study, attributable to ribavirin therapy (9). The lower percentage in our cohort is likely to be related to the lower dosage we used, compared with the Toronto group. It is imperative, particularly in light of considerable skepticism on its in vitro and in vivo efficacy, that we should drastically reflect on the anti-SARS role of ribavirin (15) and should consider conducting controlled studies on its efficacy.

The use of high-dose corticosteroid has been the cornerstone in the treatment of some infective or immunologic lung diseases. A recent systematic review on 11 randomized clinical trials showed that adjunctive use of systemic steroid, in combination with standard antituberculosis chemotherapy, was associated with significantly more rapid radiographic resolution of pulmonary infiltrates, closure of cavities, and resolution of fever and constitutional symptoms, without any increase in bacteriologic relapse, in moderate to severe pulmonary tuberculosis (24, 25). Adjunctive steroid treatment is associated with improvement in duration of mechanical ventilation, intensive care admission, and supplemental oxygen requirement in severe Pneumocystis carinii pneumonia (26, 27), especially if given within 72 hours of disease onset. Bronchiolitis obliterans with organizing pneumonia, an immune-mediated lung disease, could result from drugs (28), malignancy (29), radiation (30), and infections, especially those caused by organisms commonly associated with development of atypical pneumonia including Coxiella burnetii (31), M. pneumoniae (32), and Chlamydial species (33). The use of high-dose systemic steroid resulted in dramatic clinical and radiologic improvement in the majority of patients with Bronchiolitis obliterans with organizing pneumonia, despite a subset presenting with a fulminant course leading to death or chronic severe fibrosis (34). Apart from the common occurrence of dry cough, fever, and chills, there are also radiologic similarities between Bronchiolitis obliterans with organizing pneumonia and SARS, including the peripheral and lower zone predominance of consolidative changes (35). Despite these similarities, our data suggest that higher dosage of steroid, in the range for treatment of rejection of transplanted heart, liver, and kidney (1113), rather than for severe tuberculosis, P. carinii pneumonia, and Bronchiolitis obliterans with organizing pneumonia, could be more effective than lower dosage in the treatment of SARS (2527, 34). The use of such high-dose steroid therapy is potentially risky, especially for intubated patients, although our data showed that only 4.2% of our patients developed serious secondary infections.

The authors thank all health care workers in the Queen Mary Hospital and Queen Elizabeth Hospital for their dedication and excellent care of our patients with SARS. The authors also thank Colin Ko for his expert statistical advice, Christina Yan and June Sun for their help in data collection, and Christine So for secretarial assistance.

1. Tsang KW, Ho PL, Ooi GC, Yee WK, Wang T, Chan-Yeung M, Lam WK, Seto WH, Yam LY, Cheung MT, et al. A cluster of cases of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348:1977–1985.
2. Lee N, Hui D, Wu A, Chan P, Cameron P, Joynt GM, Ahuja A, Yung MY, Leung CB, To KF, et al. A major outbreak of severe acute respiratory syndrome in Hong Kong. N Engl J Med 2003;348:1986–1994.
3. Poutanen SM, Low DE, Henry B, Finkelstein S, Rose D, Green K, Tellier R, Draker R, Adachi D, Ayers M, et al. Identification of severe acute respiratory syndrome in Canada. N Engl J Med 2003;348:1995–2005.
4. Severe acute respiratory syndrome: cumulative number of reported probable cases of SARS. Available from: URL: (accessed on May 27, 2003).
5. Peiris JS, Lai ST, Poon LL, Guan Y, Yam LY, Lim W, Nicholis J, Yee WK, Yan WW, Cheung MT, et al. Coronavirus as a possible cause of severe acute respiratory syndrome. Lancet 2003;361:1319–1325.
6. Ksiazek TG, Erdman D, Goldsmith CS, Zaki SR, Peret T, Emery S, Tong S, Urbani C, Comer JA, Lim W, et al. A novel coronavirus associated with severe acute respiratory syndrome. N Engl J Med 2003;348:1953–1966.
7. Drosten C, Gunther S, Preiser W, van der Werf S, Brodt HR, Becker S, Rabenau H, Panning M, Kolesnikova L, Fouchier R, et al. Identification of a novel coronavirus in patients with severe acute respiratory syndrome. N Engl J Med 2003;348:1967–1976.
8. So LK, Lau AC, Yam LY, Cheung TM, Poon E, Yung RW, Yuen KY. Development of a standard treatment protocol for severe acute respiratory syndrome. Lancet 2003;361:1615–1617.
9. Booth CM, Matukas LM, Tomlinson GA, Rachlis AR, Rose DB, Dwosh HA, Walmsley SL, Mazzulli T, Avendano M, Derkach P, et al. Clinical features and short-term outcomes of 144 patients with SARS in the Greater Toronto Area. JAMA 2003;289:2801–2809.
10. Cyranoski D. Critics slam treatment for SARS as ineffective and perhaps dangerous. Nature 2003 May 1;423(6935):4.
11. The Mycophenolate Mofetil Renal Refractory Rejection Study Group. Mycophenolate mofetil for the treatment of refractory, acute, cellular renal transplant rejection. Transplantation 1996;61:722–729.
12. Lonquist JL, Radovancevic B, Vega JD, Burnett CM, Birovljev S, Saade NG, Duncan JM, Frazier OH. Reevaluation of steroid tapering after steroid pulse therapy for heart rejection. J Heart Lung Transplant 1992;11:913–919.
13. Volpin R, Angeli P, Galioto A, Fasolato S, Neri D, Barbazza F, Merenda R, Del Piccolo F, Strazzabosco M, Casagrande F. Comparison between two high-dose methylprednisolone schedules in the treatment of acute hepatic cellular rejection in liver transplant recipients: a controlled clinical trial. Liver Transpl 2002;8:527–534.
14. Meduri GU, Headley AS, Golden E, Carson SJ, Umberger RA, Kelso T, Tolley EA. Effect of prolonged methylprednisolone therapy in unresolving acute respiratory distress syndrome: a randomized controlled trial. JAMA 1998;280:159–165.
15. Wenzel RP, Edmond MB. Managing SARS amidst uncertainty. N Engl J Med 2003;348:1947–1948.
16. Oba Y, Lee N, Sung J. The use of corticosteroids in SARS. N Engl J Med 2003;348:2034–2035.
17. Severe acute respiratory syndrome: case definitions for surveillance of SARS. Available from: URL: (accessed on May 28, 2003).
18. Centers for Disease Control and Prevention. Severe Acute Respiratory Syndrome (SARS): diagnosis/evaluation. Available from: URL: (accessed on May 20, 2003).
19. Peiris JS, Chu CM, Cheng VC, Chan KS, Hung IF, Poon LL, Law KI, Tang BS, Hon TY, Chan CS, et al. Clinical progression and viral load in a community outbreak of coronavirus-associated SARS pneumonia: a prospective study. Lancet 2003;361:1767–1772.
20. Ooi GC, Khong PL, Lam B, Ho JC, Yiu MW, Wong WM, Wang T, Ho PL, Wong PC, Lam WK, et al. Relationship between radiological and clinical parameters in SARS. Radiology (In press)
21. Nicholls JM, Poon LL, Lee KC, Ng WF, Lai ST, Leung CY, Leung CY, Chu CM, Hui PK, Mak KL, et al. Lung pathology of fatal severe acute respiratory syndrome. Lancet 2003;361:1773–1778.
22. Champlin RE, Whimbey E. Community respiratory virus infections in bone marrow transplant recipients: the M.D. Anderson Cancer Center experience. Biol Blood Marrow Transplant 2001;7:8S–10S.
23. Wyde PR, Wilson SZ, Petrella R, Gilbert BE. Efficacy of high dose-short duration ribavirin aerosol in the treatment of respiratory syncytial virus infected cotton rats and influenza B virus infected mice. Antiviral Res 1987;7:211–220.
24. Smego RA, Ahmed N. A systematic review of the adjunctive use of systemic corticosteroids for pulmonary tuberculosis. Int J Tuberc Lung Dis 2003;7:208–213.
25. Muthuswamy P, Hu TC, Carasso B, Antonio M, Dandamudi N. Prednisone as adjunctive therapy in the management of pulmonary tuberculosis: report of 12 cases and review of the literature. Chest 1995;107:1621–1630.
26. Pareja JG, Garland R, Koziel H. Use of adjunctive corticosteroids in severe adult non-HIV Pneumocystis carinii pneumonia. Chest 1998;113:1215–1224.
27. Sistek CJ, Wordell CJ, Hauptman SP. Adjuvant corticosteroid therapy for Pneumocystis carinii pneumonia in AIDS patients. Ann Pharmacother 1992;26:1127–1133.
28. Erasmus JJ, McAdams HP, Rossi SE. Drug-induced lung injury. Semin Roentgenol 2002;37:72–81.
29. Mokhtari M, Bach PB, Tietjen PA, Stover DE. Bronchiolitis obliterans organizing pneumonia in cancer: a case series. Respir Med 2002;96:280–286.
30. Stover DE, Milite F, Zakowski M. A newly recognized syndrome: radiation-related bronchiolitis obliterans and organizing pneumonia: a case report and literature review. Respiration (Herrlisheim) 2001;68:540–544.
31. Perez de Llano LA, Racamonde AV, Bande MJ, Piquer MO, Nieves FB, Feijoo AR. Bronchiolitis obliterans with organizing pneumonia associated with acute Coxietlla burnetii infection. Respiration (Herrlisheim) 2001;68:425–427.
32. Llibre JM, Urban A, Garcia E, Carrasco MA, Murcia C. Bronchiolitis obliterans organizing pneumonia associated with acute Mycoplasma pneumoniae infection. Clin Infect Dis 1997;25:1340–1342.
33. Diehl JL, Gisselbrecht M, Meyer G, Israel-Biet D, Sors H. Bronchiolitis obliterans organizing pneumonia associated with Chlamydial infection. Eur Respir J 1996;9:1320–1322.
34. Cohen AJ, King TE Jr, Downey GP. Rapidly progressive bronchiolitis obliterans with organizing pneumonia. Am J Respir Crit Care Med 1994;149:1670–1675.
35. Muller NL, Staples CA, Miller RR. Bronchiolitis obliterans organizing pneumonia: CT features in 14 patients. Am J Roentgenol 1990;154:983–987.
Correspondence and requests for reprints should be addressed to Kenneth W. Tsang, M.D. (Hons.), FRCP, FCCP, FCP, Division of Respiratory and Critical Care Medicine, University Department of Medicine, The University of Hong Kong, Queen Mary Hospital, Pokfulam, Hong Kong SAR, People's Republic of China. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record