A Mountain out of a Molehill: Do We Treat Acute Leptospirosis, and If So, with What?

The leptospirosis field tends to lag behind other areas of infectious diseases, not least because diagnosing this disease is difficult and requires specialized, not readily available laboratory methods. A lingering issue is the optimal treatment of acute leptospirosis, particularly with regard to whether antimicrobial treatment should be used. In the 1980s, 2 articles were published that dealt with the efficacy of treating definitively diagnosed cases of severe leptospirosis in the Philippines [1] and Barbados [2]. These 2 articles reached opposite conclusions regarding the potential benefit of treatment for leptospirosis. In the San Lazaro Hospital in the Philippines, Watt et al. [1] found that intravenous penicillin was beneficial in reducing the duration of fever and end-organ complications. In contrast, the Barbados study [2] found that penicillin seemed to have no clinically significant benefit. This debate seems to have continued to the present. For example, at the Third Triennial Meeting of the International Leptospirosis Society (October 2002), data were presented to support the efficacy of β-lactam antibiotics (penicillin or ampicillin) for the treatment of severe leptospirosis in urban slum dwellers in Salvador, Brazil. This study was conducted because it was unclear to attending physicians whether the institution of such therapy would be useful on the basis of the debate noted above. In most other parts of the world, including Barbados, where previous studies of treatment of severe leptospirosis with penicillin had not revealed significant benefit, the consensus was that no physician would withhold antimicrobial treatment for laboratory-diagnosed, or even clinically suspected, leptospirosis, whether mild or severe. Two major issues remain with regard to the treatment of leptospirosis. How do we identify patients who will benefit from treatment? And how can we optimize timely delivery of appropriate and cost-effective therapy? In this issue of Clinical Infectious Diseases, Panaphut et al. [3] report their study of a patient population in northeastern Thailand at high risk for occupationally acquired leptospirosis. The objective of their study was to compare the efficacies of ceftriaxone and penicillin for the treatment of acute, severe leptospirosis. The authors apparently believed that comparison of ceftriaxone with placebo was unethical. The investigators enrolled patients with a high likelihood of having severe leptospirosis on the basis of a clinical presentation of fever accompanied by jaundice, an elevated serum creatinine level, and/or refractory hypotension. Laboratory diagnosis of leptospirosis was based on the results of serological tests with a dipstick assay that detects genus-specific antileptospiral IgM antibodies and in which no laboratory confirmation of other common “tropical infection” was found—in particular, melioidosis, malaria, and scrub typhus. Of 372 patients with clinically suspected leptospirosis, 173 with positive dipstick assay results were enrolled in the study and randomized; 162 of these patients completed the trial. Of the 173 patients with positive dipstick assay results, 110 had definitive evidence of leptospiral infection, as demonstrated by the “gold standard” microagglutination test (MAT): 100 had a 4-fold increase in titer, and 10 had seroconversion. On the basis of the MAT results, the authors concluded that the most frequent infecting Leptospira interrogans serovars were Sejroe, Saigon, and Bratislava, and the authors suggested that there is a changing spectrum of infecting serovars in the study region, a theory based on comparisons with MAT reactivities in other previous studies. However, the only definitive way to identify the infecting serovar (and, thus, the presumptive zoonotic source) is by culture isolation. Thus, the authors could have been more circumspect about the possibility of the changing epidemiology of leptospirosis transmission [4, 5]. Almost one-quarter of dipstick-positive patients did not have confirmation of infection by MAT, either because of the lack of paired serum samples or because there were negative results of MAT. However, it is well known that not all patients with even culture-confirmed leptospirosis develop agglutinating antibodies, so the lack of MAT confirmation in cases does not invalidate the study results. It would have been nice, however, to determine whether the proportion of dipstick-positive and MAT-unconfirmed cases differed between the penicillin and ceftriaxone treatment groups, which were otherwise well matched with regard to clinical parameters. The results of the study were straightforward. During a follow-up period of ∼1 week, the median duration of fever in the ceftriaxone and penicillin groups was 3 days. There were no clinically significant differences in complications between the 2 groups, and the mortality rates were identical. Five patients died in each group (overall case fatality rate, 5.8%), with cause of death attributed to pulmonary hemorrhage (n = 3), multiorgan failure (n = 2), hyperkalemia (n = 1), uremic encephalopathy (n = 1), and acute respiratory distress syndrome (n = 1); the cause of death was not reported for 2 patients. Previously reported data report mortality rates ranging from 0.1% to >20%, depending on location [6]; however, the relationship of host and leptospiral virulence factors to morbidity and mortality remains obscure. The authors concluded that ceftriaxone is as effective as penicillin for the treatment of acute, severe leptospirosis. Why should ceftriaxone rather than penicillin be used for the treatment of leptospirosis, when it is more expensive? There are at least 2 major advantages, including frequency and route of administration (ceftriaxone once a day given either intravenously or intramuscularly vs. penicillin 4 times per day given only intravenously), and the fact that, at least in the United States, there has been a serious lack of availability of aqueous penicillin. The cost-benefit ratio of ceftriaxone versus the personnel cost of administering the drug, although advantageous in industrialized countries, may be less so in developing regions, where nursing support costs less. Also, in many cases of penicillin allergy, it is possible to safely administer a cephalosporin. Therefore, it is highly reassuring to know that there is an alternative to penicillin for the treatment of severe leptospirosis. Clinical trials comparing doxycycline (and perhaps other non–β-lactam antibiotics) with penicillin or ceftriaxone would be useful as well, to have more diversified armamentarium for treating leptospirosis. Clinical investigators in the field of leptospirosis have made a mountain out of a molehill regarding the question of whether acute leptospirosis should be treated. Even in Barbados, where a previous study of penicillin treatment for acute leptospirosis suggested limited benefit, clinicians continue to use antimicrobial treatment to treat confirmed and clinically suspected cases. With the landmark study of Panaphut et al. [3], it is clear that severe leptospirosis should be treated with antimicrobial therapy, and that clinical trials of new antibiotics for the treatment of severe leptospirosis in the future must use either penicillin or ceftriaxone (and not placebo) as the comparator. Whether mild leptospirosis should be treated with antimicrobial therapy remains unknown, although by the time a laboratory diagnostic result returns for patients who present with an undifferentiated febrile illness, clinical symptoms have usually resolved, and treatment is probably unnecessary. Yet the most astute clinician cannot predict whether a nonspecific febrile presentation will differentiate into severe leptospirosis. We await the deployment of more accurate, specific, and accessible methods to detect the initial presentation of acute leptospirosis (either through antigen detection or nucleic acid amplification techniques), as well as the development of clinical prediction capabilities, to allow us to make rational treatment decisions for all forms of leptospirosis. I thank Adam S. B. Richards for critical review of the article in manuscript.