Drug treatment of scrub typhus

Introduction

Scrub typhus is a vector-borne zoonosis caused by the organism Orientia tsutsugamushi. ¹ This often overlooked and under-diagnosed infection is endemic in the south eastern and far eastern countries of Asia (India, Sri Lanka, Pakistan, Japan, Thailand and Korea) and extends to Afghanistan in the west and further south as far as northern Australia. This geographical area is loosely termed the tsutsugamushi triangle. ² The natural hosts of the pathogen are trombiculid mites (Leptotrombidium deliense, L. palladium, etc.). ¹ Infected larval stages of the mites (chiggers) inoculate humans (accidental hosts) while feeding. The pathogens multiply at the site of entry which later develops into an eschar. ³

The term scrub typhus is descriptive of the typical habitat endemic to the vector, although its existence is not limited to such niches. It is estimated that there are one million new cases each year and that one billion people are at risk of infection. ⁴ Often a simple febrile illness, scrub typhus is a disease with a potentially fatal outcome, causing multi-organ dysfunction in severe cases. ⁵ Left untreated, mortality rates can be as high as 30%. ⁶ Early diagnosis and treatment is, therefore, important. We review the current state of evidence on drug treatment of scrub typhus.

Methods

We searched MEDLINE using the search terms ‘scrub typhus’ or ‘Orientia tsutsugamushi’ with ‘treatment’, ‘antibiotics’ and the names of all standard antibiotics. There were 33 citations in the original search with these restrictions. Bibliographies of the cited literature were also searched. All abstracts were read through independently by the three authors and relevant papers were identified for full review. Eighteen relevant papers were identified.

Results

Relatively few antibiotics have been evaluated for the treatment of scrub typhus in clinical trials. Drugs which have shown efficacy are chloramphenicol, the tetracycline group antibiotics, macrolides, quinolones and rifampicin (Table 1). Other antibiotics have been shown to be ineffective.

One of the earliest antibiotics used in treatment was chloramphenicol. ^7–9 However, an early study showed that tetracycline was more effective ¹⁰ and, in view of the risks of aplastic anaemia, chloramphenicol is now seldom used to treat scrub typhus.

Tetracyclines are the most widely used drugs for the treatment of scrub typhus. In one study, a single dose of doxycycline 200 mg was shown to be as effective as a seven-day course of tetracycline. ¹¹ A three-day course of doxycycline was shown to be as effective as seven days of tetracycline in a multicentre randomized trial ¹² (n = 116); a meta-analysis confirmed the efficacy of doxycycline. ¹³ Treatment with doxycycline is associated with a rapid abatement of fever and this effect has even been considered almost diagnostic. ¹⁴ Chung et al. ¹⁵ demonstrate that this defervescence correlates with a rapid reduction of cytokines following treatment with doxycycline. In this study, serum levels of interleukin-10, tumour necrosis factor (TNF)-α, and interferon (IFN)-γ declined in patients within 24 h after the start of treatment. Furthermore, reductions in expression of messenger RNA for these cytokines were seen within the first week of treatment. Similarly, organism clearance also occurs within the first week of treatment; Kim et al. ¹⁶ assessed the organism's genomic products by nested polymerase chain reaction (PCR) before and after administration of doxycycline or rifampicin (n = 129) and demonstrated that the organism detection rate was reduced to 10% by day four compared to 90% before the initiation of antibiotics.

Macrolides such as azithromycin are also effective treatments for scrub typhus. The potential use of macrolides was suggested by in vitro studies, where azithromycin was shown to be superior to doxycycline in infected mouse fibroblast cells ⁹ . In a randomized trial in South Korea by Kim et al., ¹⁷ the efficacy of a single dose of 500 mg of azithromycin (n = 47) was compared to that of a seven-day course of doxycycline 200 mg daily (n = 46). No statistically significant difference was observed in time to defervescence, although it appeared to be slightly longer in the doxycycline-treated group (21 h versus 29 h). Cure (defined in this study as defervescence within five days of commencing treatment) was achieved in all patients in the first group and 93.5% in the latter group: the remainder in the doxycycline group responded after seven days of treatment and were treated for an extra two days. There were no relapses in either group, indicating a good post-antibiotic effect with even a single dose of azithromycin. In another open randomized study by Phimda et al., ¹⁸ no differences in efficacy were seen between doxycycline for seven days and azithromycin 500 mg daily for three days, although in this study the median time to defervescence was longer in the group treated with azithromycin.

These studies clearly show that azithromycin and doxycycline are equally effective. Gastrointestinal manifestations were significantly less frequent with azithromycin. ^17,18 Transient elevation of transaminases has been noted to occur with both drugs. ¹⁷ In a more recent prospective, open labelled randomized trial, a five-day course of telithromycin 800 mg daily was shown to be as effective as a course of doxycycline 100 mg twice daily for the same duration with regard to defervescence, reduction in other symptoms and adverse effects. ¹⁹

Although there were no statistically significant differences in efficacy between docycycline and azithromycin, azithromycin is easier to administer (500 mg daily for three days) and is better tolerated. However, it is significantly more expensive (doxycycline hyclate non-proprietary 200 mg daily for seven days costs US$ 3.5 while azithromycin non-proprietary 500 mg for three days costs US$ 11.25).

Resistance to drugs is relatively uncommon. However, resistance to both chloramphenicol and doxycycline has been reported from Chiangrai in northern Thailand. ²⁰ There is some in vitro evidence, ⁹ as well as evidence from animal studies, ²¹ that azithromycin is effective in these resistant strains. In the study by Kim et al., ¹⁷ the delayed defervescence with doxycycline could have been due to a partial resistance to the drug, although this is not definitely known. There is some evidence that rifampicin may be an effective alternative in such resistant cases of scrub typhus. Watt et al. ²² conducted a randomized controlled trial of doxycycline 200 mg daily for one week versus rifampicin 600 mg or 900 mg daily in two groups for mild scrub typhus in Chiangrai Regional Hospital, northern Thailand, the geographical area where doxycycline resistance was known to occur. Rifampicin proved superior to doxycycline in terms of faster defervescence (P = 0.01), prevention of relapses and less adverse effects. A combination of doxycycline and rifampicin was not found to be more effective than either of the monotherapy regimens. Despite the lack of adverse effects with rifampicin in this study, there are two potential problems with its use. First, rifampicin is well known to cause serious adverse effects. Secondly, indiscriminate use of rifampicin can increase the risk of development of rifampicin resistance in TB, which also has a high incidence in areas where scrub typhus is prevalent.

Limited evidence is available with regards to quinolones. Ciprofloxacin has been shown to be effective in animal studies. ²³ In a retrospective analysis, levofloxacin was as effective as tetracycline group antibiotics in effecting a cure. ²⁴ However, those treated with levofloxacin had a longer mean time to defervescence (P = 0.001) and a higher mortality with severe disease (P = 0.033). Among patients with higher APACHE II scores, mortality was significantly higher in the levofloxacin treated group. Deaths were due to septic shock, acute respiratory failure, disseminated intravascular coagulation and change of consciousness. Notably, however, four patients who did not respond to doxycycline were cured with levofloxacin, although these were all patients with mild to moderate disease. Thus, levofloxacin may be effective in instances where treatment with doxycycline fails but is not recommended for patients with severe disease.

Several issues need to be highlighted regarding the trial evidence on drugs. First, all the trials have evaluated patients with scrub typhus of mild to moderate severity and the extrapolation of findings to more severe disease remains questionable given the possible host factors (such as depressed immunity) and pathogen factors (resistant strains). The effects of antibiotic treatment on the development of the host immunity to the disease has not been studied widely, although in animal experiments it has been shown that the initiation of chloramphenicol early on in the course of illness may delay the host immune response. ²⁵ Secondly, the evidence for efficacy is mainly dependent on the locality studied. For example, even the more recent studies in South Korea demonstrate sensitivity to doxycycline whereas in resistant pockets of Thailand, rifampicin or azithromycin appear to be the better options. The situation in other regions is largely unexplored. Thirdly, whilst it may be necessary to identify new drugs to replace doxycycline in resistant areas, mass exposure of populations to such drugs should be undertaken in a controlled manner in order to avoid multi-drug resistance. Finally, diagnostic criteria for scrub typhus varied widely in these clinical trials and it is possible that some cases of suspected scrub typhus were due to other infections which give rise to a similar clinical picture.

Treatment of pregnant women and children is a special issue. The importance of treating scrub typhus in pregnancy has been highlighted by the recent report of vertical transmission ²⁶ and by evidence that scrub typhus can result in fetal loss. ²⁷ Choramphenicol is best avoided in pregnancy and in children. Tetracycline group drugs are contraindicated during pregnancy because of teratogenecity and should not be given to children because they may cause permanent staining of the teeth. Quinolones have been associated with a poor response and fetal loss when given for scrub typhus in pregnant women, ^27,28 and should clearly be avoided. Macrolides can be safely used in pregnancy and have been used safely and effectively to treat scrub typhus in pregnant women, ^21,29 and single dose azithromycin is recommended. ²⁷ Macrolides are also safe and effective for children. ³⁰

Conclusion

There have been few drug trials for scrub typhus. Azithromycin and doxycycline are equally effective drugs for the treatment of scrub typhus in areas where sensitive strains are prevalent. Doxycycline is cheaper and, despite being less tolerable, is likely to remain the mainstay of treatment. Treatment failure with a standard course of doxycycline has been observed in some studies, with a favourable response to a more prolonged course of treatment or to other drugs. Azithromycin is easier to administer, given its shorter treatment duration, and has less gastrointestinal side effects. Rifampicin has been shown to be more effective in areas with documented doxycycline resistance. In areas known to have resistant strains, such as northern Thailand, we suggest that either azithromycin or rifampicin should be used. Azithromycin is suitable for use in pregnancy and for children. There should be no hesitation in using azithromycin in patients who show a delayed response to treatment with doxycycline; rifampicin is, possibly, an effective alternative in such cases, although there is a lack of evidence available. Levofloxacin remains a second line alternative and could also be considered in patients with allergy to other drugs. Much of the evidence of efficacy of all these drugs is in cases of mild-moderate scrub typhus. In more seriously ill patients, who are unable to take drugs orally, intravenous chloramphenicol may well have a place, although trial evidence for this is unavailable. The use of levofloxacin in such patients cannot be recommended in the light of evidence of poor response in severely ill patients with scrub typhus. Further studies are clearly needed in order to evaluate the efficacy of these antibiotics in patients with severe, life-threatening scrub typhus.