Medical Hypotheses
Volume 67, Issue 4 , 2006, Pages 860-864

Is Lyme disease always poly microbial? The jigsaw hypothesis

David C. Owen

University of Wales, College of Medicine, Accident and Emergency, Heath
Park, Cardiff, United Kingdom

Received 22 March 2006; accepted 30 March 2006. Available online 30 June
2006.

Summary

Lyme disease is considered to be caused by Borrelia species of bacteria
but slowly evidence is accumulating which suggests that Lyme disease is
a far more complex condition than Borreliosis alone. This hypothesis
suggests that it may be more appropriate to regard Lyme disease as a
tick borne disease complex. Over recent years numerous different
microbes have been found in ticks which are known to be zoonotic and can
coinfect the human host. The hypothesis suggests that multiple
coinfections are invariably present in the clinical syndromes associated
with Lyme disease and it is suggested that these act synergistically in
complex ways. It may be that patterns of coinfection and host factors
are the main determinants of the variable clinical features of Lyme
disease rather than Borrelia types. An analogy with a jigsaw puzzle is
presented with pieces representing Borreliae, coinfections and host
factors. It is suggested that many pieces of the puzzle are missing and
our knowledge of how the pieces fit together is rudimentary. It is hoped
that the hypothesis will help our understanding of this complex,
enigmatic condition.

Introduction

Lyme disease is a condition associated with spirochete bacteria of the
Borrelia genus transmitted by ticks of the Ixodes species. This
association was made in 1982 [1]. An expanding clinical entity of Lyme
disease was already recognised [2]. Since its recognition Lyme disease
has caused controversy within the medical profession with frequent lack
of agreement about diagnosis and treatment. This paper suggests that a
possible source of the confusion has been the preoccupation with
Borrelia as the cause of Lyme disease. A hypothesis is presented that
Borrelia alone does not result in the clinical syndromes we associate
with Lyme disease.

The prevalence of Lyme disease in the USA is reported to be increasing
[3]. This may partly reflect a true increase in incidence but is also
likely to reflect increasing recognition of the disease. The CDC
recognises that the reported cases of Lyme disease underestimate the
true prevalence of Lyme disease. The true prevalence remains unknown [3].

The CDC criteria for reporting Lyme disease includes two tier
serological testing [3] but it is accepted that this may not be
appropriate for diagnosis in a clinical setting [3]. Both false positive
and false negative results can occur using serologically based tests and
laboratories have been found to yield inconsistent results when
reporting results of serological testing [4] and [5].

Culturing of spirochetes is difficult and Borrelia PCR testing has been
suggested to hold promise for diagnosis [6] but the sensitivity of the
test is low especially if blood is used as the material for testing [6].

Where the main clinical hallmark of Lyme disease, the Erythema migrans
rash, is present diagnosis of Lyme disease in an endemic area is
relatively straightforward and the diagnosis should be purely clinical
[7]. Lyme disease may be present in the absence of EM however [8] and
while some fear this may result in over-diagnosis of Lyme disease [8] it
should be recalled that the true prevalence of Lyme disease is not known.

In Lyme disease classical inflammatory indices may be normal and
objective features traditionally associated with infection such as
temperature are usually absent although subjective symptoms such as
chills are common [9]. As we have noted diverse presentations of Lyme
disease occur and the encephalopathy may be the presenting feature [9]
and [10].

The optimum treatment for Lyme disease has not been established [9].
Courses of treatment recommended in many medical textbooks may be
adequate for early infection but will frequently fail in chronic Lyme
disease. Retrieval of Borrelia by culture and PCR has been reported from
Lyme disease patients even after repeated antibiotic courses [11] and [12].

Coinfections are being increasingly recognised in Lyme disease but their
precise role is proving difficult to elucidate. There are many different
micro-organisms found in ticks which are capable of causing a Zoonosis.
They include species of Anaplasma (Erhlichia) [13], Babesia [14],
Bartonella [15] Mycoplasma [16] and Rickettsiae [17] as well as others.
All may be transmitted by other vectors as well as ticks and all have a
very wide geographic distribution.

The hypothesis

This hypothesis suggests that Lyme disease only results when certain
synergistic combinations of micro-organisms occur in the host.
Borreliosis alone may be an insufficient condition for the development
of Lyme disease.

It is postulated that only in the presence of other micro-organisms in
addition to Borrelia can Lyme disease develop. These other
micro-organisms may be tick borne, other vector borne or non vector
borne microbes and may be bacterial, protozoan, viral or even fungal.

Pathogenic micro-organisms which may be chronic and are not associated
with traditional markers of infection have been termed stealth organisms
but this term has not been widely used in the medical literature to date.

An analogy with a jigsaw puzzle is suggested whereby symptomatic disease
only occurs when certain pieces of the jigsaw puzzle are in place, the
pieces representing coinfections, known and unknown, and host factors.

Evidence for the hypothesis

(1) Coinfections. Coinfections are recognised to be important in Lyme
disease [18] and it is not possible to consider the topic of Lyme
disease without also considering coinfections although some authors have
attempted to do this [7].

In one study of 240 patients diagnosed with Lyme disease 11% were found
to have evidence of concurrent Babesia infection [19]. In another study
of 27 Lyme disease patients 100% were found to have evidence of
persistent Mycoplasma fermentans infection by PCR analysis despite
previous lengthy antibiotic treatment [20].

The frequent finding of coinfections in Lyme disease patients is one
piece of evidence suggesting that Borrelia alone may be insufficient for
the development of Lyme disease. Just as it can be difficult to confirm
the presence of Borrelia in Lyme disease it can also be difficult to
confirm the presence of coinfections. This implies that we are likely to
underestimate the frequency with which coinfections occur in Lyme disease.

Attempts to distinguish which symptoms are due to Borrelia and which are
due to coinfections in a patient can be difficult. For example it has
been reported that the symptoms of Babesia and Borrelia may be difficult
to distinguish [19] and [21]. Presentation of Anaplasma infection is
usually non specific with flu like symptoms similar to those which may
occur in early Borrelia infection [22].

Since it is likely that there are coinfections which remain to be
discovered a picture emerges that in Lyme disease multiple coinfections
may be present in most if not all cases and attempting to ascribe
different symptoms to different organisms will prove difficult.

Detection of the presence of the organisms associated with Lyme disease
can be difficult for the following reasons:

(a) The organisms are few in number.

(b) The organisms are difficult to culture.

(c) There may be failure to elicit an antibody response due to
insufficient antigen numbers, immunomodulation or immune evasion by the
organisms.

(2) Known carrier state for Borrelia. There is evidence for a carrier
state of Borreliosis [23] whereby an individual may harbour Borrelia but
be asymptomatic. The frequency with which this occurs is not known since
patients who are asymptomatic rarely have tests which would indicate the
presence of viable Borreliae. In the context of this hypothesis latent
Borreliosis may exist because of the absence of other necessary
micro-organisms although many alternative mechanisms could be postulated
to account for the phenomenon.

(3) Variable clinical manifestations of Lyme disease. We have already
noted that the presentation of Lyme disease can be highly variable and
there could be many possible reasons for this. One possible reason is
that the variable manifestations correlate with variations in type and
load of coinfections. As we have noted attempts to distinguish which
symptoms are due to which coinfection in Lyme disease are difficult as
would be expected if a complex, variable, poly-microbial eco-system is
present.

One aspect of the variable clinical features of Lyme disease is the
apparent difference in presentation between different geographic areas.
For example differences have been noted between the American and
European forms of the disease [7]. While this may in part be due to
differences between strains of Borreliae it may also be that the
differences reflect a differing coinfecting load in the two groups.

(4) Multi-system nature of Lyme disease. Lyme disease is a multi-system
illness and this is recognised in the CDC criteria for diagnosis.
Objective criteria for multi-system involvement may be absent and
patients with confirmed Borreliosis on occasion have symptoms of
multi-system involvement without signs [24]. Lyme disease patients
frequently have differing symptoms [3], [9] and [25] and this might lead
one to question how it is that one condition can result in such a vast
array of symptoms. This hypothesis suggests that the answer is that Lyme
disease is not one condition but a disease complex caused by a variable
system of interdependent micro-organisms.

(5) Difficulty in treating Lyme disease. It is known that Lyme disease
can be difficult to treat [9] and [12] and there is evidence that Lyme
disease treatment is more likely to be successful when undertaken early
on [26]. Various reasons have been suggested for the difficulties in
treating the condition and there is evidence for a cystic, dormant form
of Borrelia which is unresponsive to antibiotics [27]. In the context of
the present hypothesis another possible reason for the difficulties in
treatment could be the complex, variable microbial load.

(6) Lyme like illness with Erythema migrans in the apparent absence of
Borrelia burgdorferi ? Masters or STARI disease. The condition often
referred to as Masters disease is a condition which may follow the bite
of a Lonestar tick, Amblyomma americanum [28]. Masters Disease resembles
Lyme disease but despite intensive efforts Borreliae have so far not
been recovered from patients with this condition and the putative
organism Borrelia lonestari has not been cultured [29] and [30]. Within
lonestar ticks Borreliae have been found but their role in the
pathogenesis of masters disease remains unproven [31]. Erythema migrans
does occur following lone star tick bites [28] and Erythema migrans is
normally considered to be pathognomonic of Borreliosis. However it may
be the case that other types of micro-organisms found in the lonestar
tick can produce Erythema migrans in the absence of Borrelia and this is
consistent with the present hypothesis. Using a jigsaw puzzle analogy
there may be sufficient pieces in place for Erythema migrans to develop
in the absence of the jigsaw piece which represents Borrelia. The
suggestion that Borrelia may sometimes not be necessary for the
development of a Lyme disease like illness goes beyond the current
hypothesis but it is plausible. In a recent study [32] it was found that
the clinical features associated with Erythema migrans in Missouri where
Lone star ticks are common were quite different to patients presenting
with Erythema migrans from New York State where B. burgdorferi is
prevalent in Ixodes ticks. While different Borrelia types may explain
the differences in clinical features an alternative explanation is that
the differences reflect the presence of different complements of
coinfection.

Evidence for synergism between Borrelia and coinfections

Implicit in this hypothesis is the concept that in Lyme disease Borrelia
acts in concert with other micro-organisms for their mutual benefit but
to the detriment of the host. It should be appreciated that if this
hypothesis is true it will be virtually impossible to study synergism
within the human host since by the time people become symptomatic they
will invariably be multiply coinfected.

In one study in humans [19] it was found that clinical symptoms appeared
to be more severe in patients who had both Borrelia and Babesia
infection than in either condition alone.

Using a mouse model it has been found that coinfection with Babesia
microti and B. burgdorferi results in greater severity of carditis and
arthritis than by either infection alone [33].

A similar finding of increased severity of Lyme arthritis has been found
in mice coinfected with Anaplasma phagocytophilum, the agent of HGE, and
B. burgdorferi [34]. In another study in mice coinfected with Anaplasma
and Borrelia it was shown that an increase in spirochaete numbers
occurred in the presence of Anaplasma [35]. It was postulated that the
ability of A. phagocytophilum to functionally impair neutrophils was a
possible mechanism for the observed synergy [35].

Beyond modification of the host immune response there are many possible
mechanisms in which synergism between tick borne pathogens may occur but
such discussion is beyond the scope of this article. Borrelia is a
highly complex microbe [36] and certainly has the genetic capacity for
complex interactions with the host, other tick borne and possibly non
tick borne microbes.

Implications of the hypothesis

If it is true that Lyme disease is invariably due to the presence of a
complex of different micro-organisms then the implications are profound.
The role played by Borrelia in the condition is diminished and it may
become necessary to redefine Lyme disease in terms of a tick borne
disease complex rather than a condition caused by Borrelia.

The hypothesis implies that Lyme disease as a tick borne disease complex
will remain an illness which is diagnosed clinically. Difficulties will
arise when tick borne micro-organisms are found in a patient who
exhibits features of another illness and it may be that dual diagnosis
will become more common.

An interesting implication of the hypothesis is that Lyme disease may
develop in stages analogous to adding pieces to a jigsaw puzzle. Latent
Borreliosis from an earlier tick bite may not become symptomatic until
the necessary coinfections have been introduced.

An important implication of the hypothesis relates to treatment. The
hypothesis suggests that treatment may need to be tailored to patients
according to the complement of coinfecting organisms. This tailoring
will be difficult to achieve until methods of detection of Borrelia and
its coinfections are improved. In the meantime the hypothesis suggests
that it is unlikely a definitive treatment protocol for Lyme disease
will be found. If synergism is occurring between pathogens the
hypothesis lends support to either a monotherapy or combination therapy
approach as being reasonable.

The hypothesis has implications for clinical research into Lyme disease.
It will be difficult to produce matching controls of Lyme disease
patients for clinical trials as it is unlikely the full complement of
the infecting load will be the same for any two cases. At the scientific
level the hypothesis suggests that research into Lyme disease should
focus on mechanisms of microbial interaction within the host. Such an
ecological approach will be difficult technically but modern molecular
biological techniques have enabled this work to begin [37].

Using the jigsaw puzzle analogy it is likely that numerous pieces of the
puzzle are yet to be found: It is highly likely new coinfections and
host factors will be discovered in this truly complex and highly evolved
disease.


References

[1] W. Burgdorfer, A.G. Barbour, S.F. Hayes, J.L. Benach, E. Grunwaldt
and J.P. Davis, Lyme disease ? a tick-borne spirochetosis?, Science 216
(1982), pp. 1317?1319. Abstract-ScienceDirect Navigator |
Abstract-ScienceDirect Navigator | Abstract-EMBASE | Abstract-MEDLINE

[2] A.C. Steere, S.E. Malawista, J.A. Hardin, S. Ruddy, P.W. Askenase
and W.A. Andiman, Erythema chronicum migrans and Lyme arthritis, the
enlarging clinical spectrum, Ann Intern Med 86 (1977), pp. 685?698.
Abstract-EMBASE | Abstract-MEDLINE | Abstract-ScienceDirect Navigator

[3] CDC. Lyme disease ? United States, 2001?2002. MMWR 2004;53(17)365?9.

[4] L.L. Bakken, K.L. Case, S.M. Callister, N.J. Bourdeau and R.F.
Scell, Performance of 45 laboratories participating in a proficiency
testing program for lyme disease serology, JAMA 268 (1992), pp. 891?895.
Abstract-MEDLINE | Abstract-EMBASE | Abstract-ScienceDirect Navigator

[5] S.W. Luger and E. Krause, Serologic tests for lyme disease ?
interlaboratory variability, Arch Intern Med 150 (1990), pp. 761?763.
Abstract-ScienceDirect Navigator | Abstract-EMBASE | Abstract-MEDLINE

[6] M.E. Aguero-Rosenfeld, G. Wang, I. Schwartz and G.P. Wormser,
Diagnosis of Lyme borreliosis, Clin Microbiol Rev 18 (2005), pp.
484?509. Abstract-ScienceDirect Navigator | Abstract-EMBASE |
Abstract-Elsevier BIOBASE | Abstract-MEDLINE

[7] U.R. Hengge, A. Tannapfel, R. Erbel, G. Arendt and T. Ruzicka, Lyme
borreliosis, Lancet Infect Dis 3 (2003), pp. 489?500. SummaryPlus | Full
Text + Links | PDF (2295 K)

[8] L.H. Sigal, Toward a more complete appreciation of the clinical
spectrum of Borrelia burgdorferi infection: early lyme disease without
Erythema migrans, Am J Med 114 (2003), pp. 74?75. SummaryPlus | Full
Text + Links | PDF (51 K)

[9] ILADS Working Group, ILADS guidelines for Lyme disease, Expert Rev
Anti-infect Ther 2 (2004) (Suppl. 1).

[10] B.A. Fallon and J.A. Nields, Lyme disease: a neuropsychiatric
illness, Am J Psychiatr 151 (1994), pp. 1571?1583. Abstract-MEDLINE |
Abstract-EMBASE | Abstract-ScienceDirect Navigator

[11] V. Preacc-Mursic, K. Weber, H.W. Pfister, B. Gross and A. Baumann,
Survival of Borrelia burgdorferi in antibiotically treated patients with
Lyme borreliosis, Infection 17 (1989), pp. 355?359.

[12] J. Oksi, M. Marjamaki, J. Nikoskelainen and M.K. Viljanen, Borrelia
burgdorferi detected by culture and PCR in clinical relapse of
disseminated Lyme borreliosis, Ann Med 31 (1999), pp. 225?232.
Abstract-MEDLINE | Abstract-EMBASE | Abstract-Elsevier BIOBASE |
Abstract-ScienceDirect Navigator

[13] J.H. McQuiston, C.D. Paddock, R.C. Holman and J.E. Childs, The
human ehrlichioses in the United States, Emerg Infect Dis 5 (1999), pp.
635?640.

[14] P.J. Krause, Babesiosis, Med Clin N Am 86 (2002), pp. 361?375.

[15] E. Eskow, R.S. Rao and E. Mordechai, Concurrent infection of the
central nervous system by Borrelia burgdorferi and Bartonella henselae.
Evidence for a novel tick-borne disease complex, Arch Neurol 58 (2001),
pp. 1357?1363. Abstract-EMBASE | Abstract-Elsevier BIOBASE |
Abstract-MEDLINE | Abstract-ScienceDirect Navigator | Full Text via CrossRef

[16] J.G. Tully, D.L. Rose, C.E. Yunker, J. Cory, R.F. Whitcome and D.
Willamson, Helical mycoplasmas (spiroplasmas) from ixodes ticks, Science
212 (1981), pp. 1043?1045. Abstract-MEDLINE | Abstract-EMBASE |
Abstract-ScienceDirect Navigator

[17] D. Raoult and V. Roux, Rickettsioses as paradigms of new or
emerging infectious diseases, Clin Microbiol Rev 10 (1997), pp. 694?719.
Abstract-ScienceDirect Navigator | Abstract-MEDLINE | Abstract-EMBASE

[18] C. Thompson, A. Spielman and P.J. Krause, Co-infecting
deer-associated zoonoses: Lyme disease, babesiosis and ehrlichiosis,
Clin Infect Dis 33 (2001), pp. 676?685. Abstract-EMBASE |
Abstract-Elsevier BIOBASE | Abstract-MEDLINE | Abstract-ScienceDirect
Navigator | Full Text via CrossRef

[19] P.J. Krause, S.R. Telford, A. Spielman, V. Sikand, R. Ryan and D.
Christianson et al., Concurrent lyme disease and babesiosis. Evidence
for increased severity and duration of illness, JAMA 275 (1996), pp.
1657?1660. Abstract-MEDLINE | Abstract-EMBASE | Abstract-ScienceDirect
Navigator

[20] Horowitz R. International Lyme and associated diseases conference.
Philadelphia, October 2005.

[21] R.K. Pruthi, A. Spielman and S.R. Telford, Human babesiosis, Mayo
Clin Proc 70 (1998), pp. 853?862.

[22] J.P. Olano and D.H. Walker, Human ehrlichioses, Med Clin N Am 86
(2002), pp. 375?391.

[23] H.W. Pfister, V. Preac-Mursic, B. Wilske, K.M. Einhauple and K.
Weinberger, Latent Lyme neuroborreliosis: presence of Borrelia
burgdorferi in the cerebrospinal fluid without concurrent inflammatory
signs, Neurology 39 (1989), pp. 1118?1120. Abstract-ScienceDirect
Navigator | Abstract-EMBASE | Abstract-MEDLINE

[24] E.L. Logigian, R.F. Kaplan and A.C. Steere, Chronic neurological
manifestations of Lyme disease, N Eng J Med 323 (1990), pp. 1438?1444.
Abstract-MEDLINE

[25] K.B. Liegner, P. Duray, M. Agricola, C. Rosenkilde, L.A. Yannuzzi
and M. Ziska et al., Lyme disease and the clinical spectrum of
antibiotic responsive chronic meningocephalomyelitides, J Spirochaetal
Tick-borne Dis 4 (1997), pp. 61?73.

[26] R.P. Smith, R.T. Schoen, D.W. Rahn, V.K. Sikand, J. Nowakowski and
D.L. Parenti et al., Clinical characteristics and treatment outcome of
early lyme disease in patients with microbiologically confirmed Erythema
migrans, Ann Int Med 136 (2002), pp. 421?428. Abstract-EMBASE |
Abstract-ScienceDirect Navigator | Abstract-MEDLINE

[27] S.H. Brorson, In vitro conversion of Borrelia burgdorferi to cystic
forms in spinal fluid and transformation to mobile spirochetes by
incubation in BSK-H medium, Infection 26 (1998), pp. 144?150.

[28] E.J. Masters and H.D. Donnell, Lyme and/or Lyme like disease in
Missouri, Missouri Med 92 (1995), pp. 346?353. Abstract-MEDLINE

[29] A.G. Barbour, G.O. Maupin, G.J. Teltow, C.J. Carter and J. Piesman,
Identification of an uncultivable Borrelia species in the hard tick
Ambylomma americanum: possible agent of a Lyme disease-like illness, J
Infect Dis 173 (1996), pp. 403?409. Abstract-MEDLINE | Abstract-EMBASE |
Abstract-Elsevier BIOBASE | Abstract-ScienceDirect Navigator

[30] A.M. James, D. Liveris, G.P. Wormser, I. Schwartz, M.A. Montecalvo
and B.J.B. Johnson, Borrelia lonestari infection after a bite by an
Amblomma americanum tick, J Infect Dis 183 (2001), pp. 1810?1814.
Abstract-MEDLINE | Abstract-ScienceDirect Navigator | Abstract-Elsevier
BIOBASE | Abstract-EMBASE | Full Text via CrossRef

[31] P.M. Armstrong, S.M. Rich, R.D. Smith, D.L. Hartl, A. Spielman and
S.R. Telford, A new Borrelia infecting lone star ticks, Lancet 347
(1996), pp. 67?68. SummaryPlus | Full Text + Links | PDF (330 K)

[32] G.P. Wormser, E. Masters, J. Nowakowski, D. McKenna, D. Holmgren
and K. Ma et al., Prospective clinical evaluation of patients from
Missouri and New York with Erythema migrans-like skin lesions, Clin
Infect Dis 41 (2005), pp. 958?965. Abstract-MEDLINE |
Abstract-ScienceDirect Navigator | Abstract-Elsevier BIOBASE | Full Text
via CrossRef

[33] M.H. Moro, O.L. Zegarra-Moro, J. Bjornsson, E.K. Hofmeister, E.
Bruinsma and J.J. Germer et al., Increased arthritis severity in mice
coinfected with Borrelia burgdorferi and Babesia microti, J Infect Dis
186 (2002), pp. 428?431. Abstract-MEDLINE | Abstract-ScienceDirect
Navigator | Abstract-EMBASE | Abstract-Elsevier BIOBASE | Full Text via
CrossRef

[34] V. Thomas, J. Anguita, S.W. Barthold and E. Fikrig, Coinfection
with Borrelia burgdorferi and the agent of human granulocytic
ehrlichiosis alters murine immune response pathogen burden and severity
of Lyme arthritis, Infect Immun (2001), pp. 3359?3371. Abstract-Elsevier
BIOBASE | Abstract-MEDLINE | Abstract-ScienceDirect Navigator |
Abstract-EMBASE | Full Text via CrossRef

[35] K. Holden, E. Hodzic, S. Feng, K.J. Freet, R.B. Lefebvre and S.W.
Barthold, Coinfection with Anaplasma phagocytophilum alters Borrelia
burgdorferi population distribution in C3H/HeN mice, Infect Immun
(2005), pp. 2440?3444.

[36] C. Fraser, S. Casjens, W.M. Huang, G. Sutton and R. Clayton et al.,
Sequence analysis of the multi-element genome of Borrelia burgdorferi,
the Lyme disease spirochete, Nature 309 (1997), pp. 580?586.
Abstract-ScienceDirect Navigator | Abstract-MEDLINE | Abstract-EMBASE |
Full Text via CrossRef

[37] P.A. Rosa, K. Tilly and P.E. Stewart, The burgeoning molecular
genetics of the Lyme disease spirochaete, Nat Rev Microbiol 3 (2005),
pp. 129?143. Abstract-EMBASE | Abstract-Elsevier BIOBASE |
Abstract-MEDLINE | Full Text via CrossRef