Erythema Migrans-Like Lesions in the South Require Treatment Given the Current State of Knowledge

Lantos et al, in their recent effectiveness modeling study propose that the preferred clinical strategy for southern patients with erythema migrans–like (EM) lesions who lack travel to Lyme disease–endemic areas is to not treat but to observe and treat only if disseminated Lyme disease develops (Lantos et al. 2013). Their model, motivated by concern with overtreatment and overuse of antibiotics, is compromised by several issues: (1) The unsubstantiated assumption that the Lyme-like so-called southern tick-associated rash illness (STARI) transmitted by Amblyomma americanum ticks does not need treatment, (2) their assuming that only the northeastern type of Lyme disease (classic caused by Borrelia burdoferi sensu stricto) can cause disease and warrants treatment, and (3) observation of the untreated patient for possible disease symptoms of disseminated borreliosis as adequate in lieu of treatment since development, as with its cousin syphilis, can occasionally take months to years.

STARI, also known as Masters disease, by definition is the Lyme-like condition associated with the bite of the lone star tick, an aggressive tick common in the South but now found as far north as Maine. Despite extensive efforts, the causative agent of STARI remains unknown; therefore, no diagnostic laboratory methodologies exist. Neither is there a diagnostic code in the International Classification of Diseases or the Current Procedural Terminology code set maintained by the American Medical Association. The only way a definitive diagnosis of STARI can be made is for the tick-exposed patient with Lyme-like signs and symptoms to present with a history of an actual tick bite temporally associated with the illness and with the biting tick itself. Finally, the tick must be accurately identified as an A. americanum. Because both Ixodes scapularis and A. americanum are sympatric over much of the eastern United States (Leydet and Liang 2014), determining whether a newly symptomatic patient has Lyme disease or STARI may be problematic without having the offending tick. Using an elegant probability modeling strategy and geographical studies to guide decision-making, the authors propose that in “southern and southeastern states, treatment of EM as if it is Lyme disease may result in an unacceptably high risk of complications with little expected benefit.” Details on what comprises unacceptable complications are not provided. The ”little expected benefit” is apparently based on the authors' unreferenced assumption “that antibiotic treatment is not justified for…the treatment of STARI.”

Southern patients deserve rigorous evidence to justify nontreatment of Lyme-like disease. Lantos et al. do not present any. Indeed, they could not because no evidence to support nontreatment exists (Herman-Giddens 2014). There are no long-term studies of untreated STARI patients to inform as to sequelae (Kirkland et al. 1997, Masters et al. 2008, Feder et al. 2011, Herman-Giddens 2012a). That treatment is useful is suggested by several studies describing ill patients whose treatment with doxycycline resolved their symptoms and EM-like lesions (Kirkland et al. 1997, Felz et al. 1999, Wormser et al. 2005). EM rashes are known to resolve over time without treatment in both classic Lyme disease and STARI; such resolution does not indicate that no long-term sequelae will occur (Wormser 2006).

Masters and colleagues published evidence that sequelae from STARI do occur (Masters and Donnell 1995, Masters et al. 2008). They withdrew authorship of the paper resulting from a study they were conducting with the Centers for Disease Control and Prevention (CDC) (Campbell et al. 1995) because available data showing longer-lasting sequelae were excluded (Masters and Donnell 1995, Masters 2006). In their series of 45 Missouri EM patients, even with treatment, three had arthritis or carditis from 1 to 4.5 months after their rash (Masters and Donnell 1995). (Interestingly, the CDC recently reported three sudden cardiac deaths from Lyme carditis [Centers for Disease Control and Prevention 2013]).

The cause of STARI is still not known. Because the illness resembles Lyme disease, and there is more genetic variation in southern strains of B. burgdorferi sensu lato (Bbsl) and sensu stricto (Lin et al. 2001) than in the north, and Bbsl are well-established in certain ticks in many areas in the south (Lin et al. 2001, Clark 2004, Harrison et al. 2010, Smith et al. 2010, Leydet and Liang 2014), it is premature to assume that A. americanum ticks absolutely cannot transmit any strain of Borrelia. Studies of the Lyme-like illness transmitted by the lone star tick (STARI) find spirochetal like forms and/or serologic and PCR evidence in some patients, suggesting a borreliosis (Kirkland et al. 1997, Felz et al. 1999, Masters et al. 2008, Clark et al. 2013). Infections by Bbsl strains not reactive to the B31 strain used for the CDC-advised two-tiered testing could be occurring. Absence of positive cultures from STARI patients to date may not be sufficient evidence to rule out borrelial infections as a cause. Spirochetes are notoriously difficult to culture. BSK medium may not support isolation of all genotypes of B. burdorferi (Norris et al. 1997, Felz et al. 1999, Leydet and Liang 2014.)

In addition to the possibility that other pathogenic Borrelia species may be present in the South, consider the problems caused by this paper's recommendation for treating only when classic Lyme disease is observed. The authors' declaration that Lyme disease is “rare” and “highly improbable” in the South and Southeast is based on 6- to 9-year-old data from the Yale geographical studies that used questionable methods for the South and do not agree with more recent data from other sources (Diuk-Wasser et al. 2012, Herman-Giddens 2012b). Tick populations and infection rates are not static. To date, North Carolina has four widely separated counties endemic for Lyme disease by stringent CDC criteria; eight more are pending. Other southern examples include Tennessee with four endemic counties (J. Dunn, personal communication), Alabama with four (Alabama Department of Health 2013), and Florida where the entire state is considered endemic (Florida Department of Health 2014). Additionally challenging, if one were to follow the Lantos et al. recommendation that Lyme-like illness in a southern state such as North Carolina be observed rather than treated, is how to determine whether an ill patient deserves immediate treatment because endemic counties are often not contiguous. For example, what if an untraveled patient lives in a county between two endemic counties or in a next-door county or just south of the border where five Virginia counties are endemic?

We know disseminated sequelae from untreated Lyme disease and probably from STARI can produce harm; therefore, choosing to not treat but observe southern patients with Lyme-like signs and symptoms needs to be based on informed and convincing proof that there is little to no chance for harm. Right now, that does not exist. Evidence-based medicine is a popular paradigm in medical practice. Classic Lyme-endemic counties occur in certain areas of the South and Southeast, and STARI causes illness and may have long-term sequelae. No evidence exists showing STARI to be harmless. Thus, prudence, ethical standards, liability issues, and good medical practice demand a very high bar to not treat a symptomatic patient. Such level of evidence does not yet exist.