The Jarisch-Herxheimer Reaction in Leptospirosis: A Systematic Review

Background Leptospirosis is an endemo-epidemic zoonotic disease associated with potentially fatal renal, cardiovascular or pulmonary failure. Recommended treatment includes antibiotics, which may induce a Jarisch-Herxheimer reaction (JHR). Since little information on the importance of this adverse event is available, we performed this review to quantify frequency and impact of JHR in leptospirosis management. Methodology/Principal Findings This review systematically summarizes the literature on the JHR in leptospirosis. To approach the broader aspects of the subject, articles considering the treatment of leptospirosis, national leptospirosis guidelines and textbook and technical reports of the World Health Organisation were reviewed. Publications describing JHR in leptospirosis are very limited and consist mainly of single case reports and small case series. A single randomized control trial specifically assessed the JHR occurrence, but it has never been systematically investigated in large trials. Not all guidelines and not all literature on leptospirosis mention this reaction which can be fatal. Conclusions/Significance Although generally assumed to be a rare event, the true prevalence of JHR in leptospirosis is unknown and the awareness of this event is insufficient. All leptospirosis guidelines and local leptospirosis protocols should stress on systematic monitoring for clinical status early after antibiotic administration. Large well designed studies are required to precise the incidence and the impact of JHR as well as the severity and rates between various antibiotics.


Introduction
Leptospirosis is a widespread zoonotic disease of worldwide distribution caused by pathogenic Leptospira, a spirochetal organism that is transmitted to humans by exposure to urine of infected mammalian reservoirs such as rodents or wild and domestic animals [1]. The disease has a wide-ranging clinical spectrum from asymptomatic forms to severe presentations. The latter are estimated to occur in 5-15% of all human infections and commonly combine jaundice, renal failure, myocarditis and/or haemorrhage [2] with a significant mortality rate. While a single experimental study demonstrated the relevance of antibiotic administration even in late stages of leptopsirosis [3] the benefit of antibiotic therapy on mortality lacks of clinical evidence [4]. However, the current treatment guidelines still rely on antibiotic administration regardless of stage or severity of the disease [5]. Initiation of chemotherapy in spirochetal diseases may precipitate a febrile inflammatory reaction [6,7], known as the Jarisch-Herxheimer reaction (JHR), originally described in patients with syphilis receiving mercury treatment [8,9]. This reaction is characterised by an acute inflammatory response associated with the release of large amounts of cytokines, resulting from clearance of spirochetes from the circulation [6,7]. Prevalence, clinical manifestations and outcome of JHR have been well studied in syphilis [10], Lyme disease [11], tick-born relapsing fever [12] and louse-born relapsing fever [13][14][15]. However, a clear idea of importance JHR in leptospirosis is lacking. In order to quantify frequency and impact of JHR in leptospirosis, we systematically reviewed the published literature and put it in a broader perspective, identifying data gaps that should be addressed.

Search Strategy and Selection Criteria
The inclusion criteria were as follows: all publications on leptospirosis treated with antibiotics providing information about the occurrence of early adverse events including JHR were selected. When the original article was not obtainable but the abstract containing the requested information was, the publication was included in the analysis. To avoid multiple counting (duplication) of identical procedures and cases, follow-up publications on identical procedures and cases were traced and excluded. The electronic database PubMed was searched with the keywords ''leptospirosis'' and ''Leptospira''. A second search was performed for ''penicillin'', ''amoxicillin'', ''ampicillin'', ''cefepime'', ''ceftriaxone'' and ''antibiotics''. A final search was done with ''Jarisch-Herxheimer reaction''. Search results were assessed using different combinations. The search was not limited by study design, language or date of publication. A similar search was conducted with Embase. The Cochrane Library was also searched for leptospirosis publications. World Health Organisation (WHO) websites and international or accessible national guidelines in endemic countries were searched, and the latest editions of five standard textbooks in infectious and tropical diseases were studied [16][17][18][19][20]. Titles and abstracts were screened, and potentially relevant case reports, case series and trials were further evaluated. Reports were included if they assessed the occurrence of JHR in leptospirosis. Reference lists of included articles were screened for further relevant publications. Study identification and data extraction were conducted independently and cross-checked for accuracy by both authors.

Data Abstraction and Methodological Assessment
The primary outcome was JHR occurrence or rates measured at time points described by investigators. Secondary outcomes include renal failure rates and mortality. Serovars were also recorded. It was impossible to retrospectively grade the severity of the JHR due to the lack of a standardized definition of the events.

Results
Twenty eight publications met our inclusion criteria, with a total number of 976 leptospirosis cases treated with antibiotics. One publication was a follow-up report on identical case series and therefore excluded from analysis [21]. Abstract or full text was not found for an additional publication which was also excluded [22]. One previous review dedicated to the topic was published in 1991 without systematically assessing the literature [23]. Among the selected studies, 11 patients presenting with concomitant infections were excluded [24].

Studies Characteristics and JHR Occurrence
Eight case reports [25][26][27][28][29][30][31][32] and five case series [23,[33][34][35][36] published between 1955 and 2012 were reviewed and analyzed. All selected patients had a laboratory confirmed leptospirosis without concomitant infection and all were treated with antibiotics. A JHR was reported in 92 patients from 1 to 48 hours after administration of the first dose of antibiotics. The nature of antibiotics regimen given varied considerably: penicillin (n = 81), ampicillin (n = 8), ceftriaxone (n = 1), cefepime (n = 1) and penicillin and ceftriaxone (n = 1). The most common features of described JHR were sudden onset of shivering or rigors (n = 6), with rise in temperature (n = 9), with (n = 8) or without (n = 1) a fall in blood pressure, occurring after administration of the first dose of antibiotics ( Table 1). The largest case series accounted for 70 cases, out of 84 case-patients treated with penicillin in Malaya in 1957 [36].
Seven randomized control trials (RCT) [21,24,[37][38][39][40][41][42], six non randomized control trials (NRCT) [43][44][45][46][47][48] were reviewed and analyzed ( Table 2). With the exception of two RCTs [41,42], all studies monitoring the efficacy of antibiotics were not designed to assess adverse events linked to antibiotics including the JHR occurrence. The single study designed to monitor adverse events focusing on leptospirosis-related JHR was performed in 1986 by Watt et al. [42]. In this study, a single Herxheimer-like reaction was observed in a patient receiving saline placebo. The authors concluded that fear of a Herxheimer-like reaction should not dissuade clinicians from administering antibiotics to patients with leptospirosis. However, studied patients received up to four doses of parenteral antibiotic or had completed up to two days of an oral antibiotic regimen before inclusion.
JHR in National and International Guidelines, Textbooks and WHO/ILS Guideline Among 3 screened national or international guidelines, all described JHR as a potential adverse event, although occurring with penicillin only [5,49,50]. Three textbooks out of the five consulted mentioned the possibility of JHR early after penicillin initiation [18][19][20].

JHR Management and Outcome
JHR management was not specified in 82 cases, but reported with no details in five cases (supportive care). In the remaining cases, the management consisted in fluid infusion, vasopressors (n = 1), corticosteroids and inotropic support (n = 1) and transient dialysis (n = 1). Overall, eleven studies mentioned immediate outcome after JHR occurrence. Among them, all cases fully recovered with the exception of one JHR-related death [34].

JHR and Leptospira Serovars
Out of 13 case reports or case series reporting JHR, the pathogen serovar was stated in 18 patients. Nine different types of serovar were involved in JHR-patients, the most prevalent being Icterohaemorrhagiae (n = 6).

Discussion
To the best of our knowledge, this systematic review is the first to formally appraise the measurement of JHR occurrence in leptospirosis treated with antibiotics. Publications describing JHR consist mainly of single case reports or case series. All RCTs included in our review but one were not designed to assess JHR incidence and therefore judged to be of poor methodological quality. Moreover, methods used to monitor JHR were not described in enough detail to reach predefined criteria and to determine whether there was a link between JHR occurrence and organ failure or death. Most reports were targeted at criteria such as duration of fever, clearance of spirochetes and length of stay. The single appropriately designed and detailed RCT had major    limitations, including a small number of subjects and a lack of JHR confirmation. However, this work will help to better estimate the potentially detrimental impact of antibiotics in leptospirosis and inform future management guidelines. Since there is little doubt that renal damage can be aggravated by vascular hypotension, particular attention should be paid to monitor blood pressure after initiation of antibiotics in suspected leptospirosis, especially in particular conditions such as pregnant, hypotensive or chronic renal insufficiency presenting patients. Of note, JHR is not reported in studies describing prognostic factors of leptospirosis but clinicians are usually facing complicated management [51,52].
Interestingly, a higher proportion of JHR occurred in early stage leptospirosis, suggesting a higher probability of the event before the natural clearance of spirochetes. It is postulated that the inflammatory process results from activation of the cytokine cascade during the degeneration of spirochetes. Therefore, the apparent lower proportion of JHR in patients with leptospirosis compared with patients diagnosed with other spirochetal diseases may be explained by a lower bacteraemia. This hypothesis requires further investigation for validation. JHR unobserved or unreported by clinicians is an additional reason for this ostensible reduced frequency. Conversely, a confusion of JHR symptoms with the aggravation of the leptospirosis itself may lead to overestimate the incidence of this event.
Interestingly, international guidelines for the management of leptospirosis briefly mention the occurrence of JHR with penicillin, omitting the details of its prevention, management or outcome. Moreover, antibiotics other than penicillin (e.g. ceftriaxone) are not evoked. Similarly, infectious diseases textbooks mentioning the event overlook this reaction with non-penicillin antibiotics. Noteworthy, recommended antibiotics for the management of leptospirosis vary considerably according to settings. For example, most commonly used drugs in the Philippines for mild leptospirosis is doxycyclin while penicillin G is recommended for treating severe cases [50]; amoxicillin and cefotaxim are preferentially used in French Caribbean or in New Caledonia in mild and severe cases, respectively [53].
Some limitations in the current review should be acknowledged. First, potential publication bias is impossible to completely exclude. Moreover, unpublished reports or reports that were not referenced in databases could have been missed despite the comprehensive search. Second, patients presenting with undetected co-infections may have been included in the study. Finally and most importantly, the diagnosis of JHR was not supported by any dosage of biological markers. In addition, the definition of JHR was not uniform, including in the largest case series published by Mackay-Dick [36] which lacks of details such as the onset of reaction. Therefore, although most of the cases presented a genuine JHR, some might have had a clinical aggravation related to the spirochetal disease regardless of the antibiotherapy.

Conclusions
The prevalence of JHR in leptospirosis treated with antibiotics is unknown and the awareness of this adverse event is insufficient. Although this review suggests that antibiotic treatment in patients with leptospirosis may result in less common and less severe JHR than in patients with other spirochetal diseases, all leptospirosis guidelines and local leptospirosis protocols should stress on systematic monitoring for clinical status early after antibiotic administration. We strongly recommend that patients receiving penicillin or other antibiotics for the management of leptospirosis to be monitored early after initiation of treatment to prevent any detrimental effects of a potential JHR. Since it is still controversial whether antimicrobials produce a beneficial effect in mild human leptospirosis, large well designed studies are required integrating a specific monitoring to precise the incidence and the impact of JHR as well as the severity and rates between various antibiotics.