|
Home
|
|
Symptoms
|
|
Live Discussion
|
|
Diagnosis
|
|
Treatment
|
|
Area Support
|
|
Library
|
|
Research
|
|
Lymelinks
|
|
Contact
|
|
Pets & Lyme
|
|
DONATIONS
|
|
Drug Info
|
|
Medical Dictionary
|
|
Board of Directors
|
|
|
No Warranties or Representations
The data and information presented in this web site are presented in good faith and believed to be accurate. Any and all liability for the content or any omissions including any inaccuracies, errors, or misstatements in such data or information is expressly disclaimed. The web site is compiled for the sole purpose of informing community members of resources and information pertaining to Lyme Borreliosis Disease and its coinfections.
The Canadian Lyme Disease Foundation, Directors and members are not liable for any direct or indirect damages or any damages whatsoever resulting from loss of use, data or profits, whether in an action of contract, negligence or other tortious action arising out of or in connection with the use or performance of information available from this website.
Consult a qualified Lyme ( Borreliosis ) Disease literate doctor for medical advice if Lyme Disease is suspect.
|
William T. Harvey, MD ILADS.org
'Lyme disease': ancient engine of an unrecognized borreliosis pandemic?
W. T. Harvey, P. Salvato
Diversified Medical Practices, Houston, Texas, USA
Summary
Unexpectedly we have found large numbers of chronically ill Borrelia
burgdorferi PCR- and seropositive patients in Houston, Texas, a zoonotically
'non-endemic' area. In order to understand this finding prior to sufficient
data availability, we chose to examine critically currently accepted but
troublesome 'Lyme disease' concepts. Our method was to analyze each
foundation 'Lyme disease' premise within the context of available medical
and veterinary literature, then to reconstruct the disease model consistent
with the preponderance of that data. We find the present conceptualization
of the illness seriously truncated, with a high likelihood of two distinct
but connected forms of human B. burgdorferi infection. The yet-unrecognized
form appears to have a broader clinical presentation, wider geographic
distribution, and vastly greater prevalence. We conclude that 'Lyme disease'
currently acknowledges only its zoonosis arm and is a limited
conceptualization of a far more pervasive and unrecognized infection state
that must be considered a global epidemic. ª
2003 Elsevier Science Ltd. All rights reserved.
INTRODUCTION
'No part of the aim of normal science is to call forth new
sorts of phenomena; indeed those that will not fit the box are not seen at
all'. Thomas Kuhn: The Structure of Scientific Revolutions (1).
In 2000, Paul Ewald advanced the hypothesis that evolutionary biology principles, if
followed faithfully, predict that infection likely underpins many current
illnesses without known etiology, perhaps via a single agent (2).
Here we
present evidence that may identify one such agent.
Twenty-six months ago,
our practice began to test chronically ill patients with multi-system
presentation for Borrelia burgdorferi (Bb) infection. Our criteria further
included suspicion of protracted infection and inability to otherwise find a
diagnosis. About a third of initial tests were positive via CDC Western blot
criteria or serum/urine PCR; however, repeat testing eventually revealed
that most are positive. We had not expected these results, as we are in
Southeastern Texas, a 'nonendemic' region. (The prevalence of
Borrelia-infected ticks in Texas is about 1-2%.) (3-5). Concurrently, we
recognized a striking similarity in symptoms and signs of test-positive
individuals to other untested patients of ours. Most fit within the
presentation criteria of 'late, disseminated Lyme disease' but with some
prominent differences. We began antibiotic treatment of all test-positive
patients, and most, regardless of presentation, began noticeable improvement
within 3 to 6 months. Since no history of erythema migrans (EM) rash or
illness following tick bite was reported by these patients, and most had
been ill for many years with similarly ill family members, we set out to
understand what we were confronting. Our experience did not match the CDC
case definition or the epidemiological evidence for late. To resolve this
conundrum, we concluded that our best initial strategy was to derive our own
conclusions by careful assessment of all available relevant data.
METHOD
We began by examining the present conceptualization of 'Lyme disease' for
comprehensiveness and accuracy. We chose the US Centers for Disease Control
and Prevention(CDC) 'surveillance case definition' of 'Lyme disease' as the
current standard model of human Bb infection. We next isolated what we
believed to be the linchpin premises supporting that model, examining each
against available pertinent National Library of Medicine (NLM) medical and
veterinary Bb data. Each premise was addressed independently with relevant
data categorized and summed. We then derived a modified illness model of
human Bb infection by combining the reassessed premises. Finally, we
examined this newly-derived illness model within the context of current
medical nosology and the realities of contemporary clinical medicine for
plausible fit and superior predictability potential.
FINDINGS
Semantics: 'Lyme disease' examined.
The origin and conceptualization of
'Lyme disease' 'Lyme disease' is the label given to a human illness first
recognized in Old Lyme, CN in 1975 (6). The initial cases resulted from a
zoonosis present in local vertebrate reservoirs/ hosts and transferred
incidentally to humans via an arthropod vector, Ixodes scapularis (6-10).
The responsible agent was later identified as B. burgdorferi, a spirochetal
bacterium, with human disease considered to result exclusively from the
genospecies cluster B. burgdorferi sensu lato (Bbsl ). (7,11). Globally, the
species presently within this cluster include B. burgdorferi sensu stricto
(Bbss), B. afzelii, B. garinii, B. lonestari, and possibly B. valaisiana
(12,13). As the knowledge base expanded internationally in the subsequent 26
years, the zoonosis-only conceptualization of 'Lyme disease' persisted with
emphasis on acute and early stages. The exclusive focus on zoonosis resulted
in a major portion of resources being concentrated on reservoir and vector
prevalence (14). The 1975 conceptual model reached 'standard-of-care' status
little changed with publication of the CDC 'case classification' in 1997
(15). 'Lyme disease' as defined by the CDC: the prevailing view The CDC
initially published case definitions for Public Health Surveillance in
October, 1990.
For the first time, uniform criteria were available to be
used in case reporting that included 'Lyme disease' (16). The full CDC
criteria for 'Lyme disease' were published in the May, 1997 MMWR (CDC
Morbidity and Mortality Weekly Report) for surveillance purposes and
included clinical description, laboratory diagnostic criteria, confirmed
case classification, and relevant definition of terms (15).
We refer the
reader to the referenced MMWR for the complete surveillance criteria, all
contained under the title: 'Lyme Disease (Revised 9/96) clinical
description'. More extensive descriptions of the illness presentation,
course, epidemiology, and diagnosis may be found in subsequent CDC sources
(17,18). We included these in our synthesis of the most comprehensive
current conceptualization of 'Lyme disease', but considered the 1997
'surveillance case definition' as authoritative. The foundation premises of
the 'Lyme disease' model following are what we consider to be the foundation
premises used to create and support the present internationally endorsed
conceptualization (model) of 'Lyme disease' believed to represent all human
Bb-induced illness. (All illness models incorporate a set of premises
derived from numerous observations, assumptions and definitions by inductive
reasoning. Collectively, the premises deductively create a conceptual
framework that constitutes the model. The model is then used deductively to
derive diagnosis and treatment.)
We state each premise separately, then
compare to the available published data. (We reviewed 951 peer-reviewed
papers and 13 books to cover what we deemed the relevant literature.) A
re-summarization completes the process. Assessment of these 13 premises
constitutes the body of our argument. Before examining the premises which we
believe support the concept of 'Lyme disease', we stress that our use of
this label will be limited to its presently defined conceptual boundaries.
The CDC defines 'Lyme disease', exclusively as a zoonotic illness.
Congenital and gestational transfer cases have been disregarded for reasons
not evident to us. This limited perspective is the first important
illness-model error that will further compound as we examine its premises.
The 13 premises are grouped into five general areas: Initial Clinical
Presentation (1-3), Testing/Confirmation(4-5), Pathogen Transfer, (6-9),
Course and Outcome(10-11), and Distribution and History (12-13).
Premise 1: The erythema migrans rash consistently heralds initial B.
burgdorferi infection In most early and many recent studies, the presence of
an EM rash is presumed to consistently verify initial 'Lyme disease' human
inoculation of Bbsl via tick (17). This assumption is, in fact, the
foundation of many important conclusions reached about 'Lyme disease'
despite numerous papers that acknowledge secondary EM lesions occur (15,17).
Available data do not support the EM rash to be dependable for diagnosis.
The EM rash may follow initial tick inoculation (primary), or it may occur
months to years later (secondary) (19-28). The rash also frequently fails to
appear following tick inoculation (26). It is not known to be a marker in
gestational or congenital Bbsl transfer (13). Baranton's recent data show
that some Lyme-derived Bbsl variants do not cause EM lesions (29).
Premise 2: Borrelia burgdorferi infection is sub-clinical in some infected
humans with assumed benign outcome Copious data support that B. burgdorferi
infection can be 'sub-clinical' and thus unnoticed in infected individuals
(20,30-54) and animals (55,56). The average symptomatic/asymptomatic ratio
(S/A) in these studies from endemic areas, primarily using serology for
diagnosis, is close to 1:1, much as Steere first noted in 1986 (57). The S/A
ratio in 'non-endemic' areas may be similar or higher (58). Notably, no
recent 'Lyme disease' study referenced by the CDC considers the asymptomatic
state to be significant (17,18). The question of whether sub-clinical cases
activate or fail to activate at some future date is intriguing. Lack of
sufficiently prolonged longitudinal follow-up in available studies fails to
provide an answer. Baranton's recent data, however, show differing
pathogenicity among Bbsl species, lending support to the reality of some
prolonged innocuous sub-clinical infections although study design limits
insight into ultimate outcome (59). When prolonged longitudinal outcome has
been considered, several studies do support reactivation among patients
(31,33,36,37,41,43,49,53,57,60).
Premise 3: Arthritis is a primary late musculoskeletal 'Lyme' infection
sequela, where 'late' in CDC parlance is vague Objective joint swelling is a
'late Lyme disease' symptom criterion (15). This frequently-documented
standard, however, was derived primarily from data limited to patients
meeting strict 'Lyme disease' requirements, necessitating vector inoculation
for inclusion and recent enough for high antibody levels (6,61-69).
False-seronegative individuals would have necessarily been rejected,
possibly including those with low antibody levels and those infected by
sexual, congenital or gestational routes. Further, we find no concise or
consistent definition of 'late' in published 'Lyme disease' definition
criteria, obfuscating the time when arthritis might be expected to appear
(17,18). In our 'non-endemic' region, we have rarely seen rheumatic joint
presentation in our 455 seropositive or PCR positive patients. Many have
migratory and intermittent arthralgias, however, and most have been ill far
longer than one year. We find no data to support our clinical experience,
however, likely because our patients fall outside the 'Lyme disease'
inclusion criteria.
Premise 4: Humans with late stage 'Lyme disease' show high antibody levels
and high numbers of Western blot test bands The data support this position
when applied strictly to presently defined 'Lyme disease' (17,70). These
data do not address non-zoonotic transfer cases. An exhaustive review
addressing 'Lyme disease' antibody data may be found in a recent work by
Gardner (13). IgG and IgM response curves are reproducible within reasonably
consistent ranges (8,71-74). Of note, however, referenced patients were
studied only a limited number of months following initial vector
inoculation, most less than a year (25,72,75-78). We found no study
characterizing immune reactivity to Bbsl in untreated patients from
non-endemic regions and where symptoms have been present for one year to
decades. Consistently, most serious studies have examined and tested only
patients from limited geographic areas where high tick infection rate and
acute human disease coincide. The immune reaction of infected patients not
meeting 'Lyme disease' criteria have fallen outside rigorous scrutiny.
Gardner and others have shown conclusively that a group of Bbsl-infected
humans was not inoculated transdermally but rather acquired their disease
congenitally or gestationally (13,79-81). How might their antibody picture
appear? Gardner's exhaustive review of antibody production following
gestational Bbsl transfer is instructive. In her Table 11-8, 72% of neonates
with tissue-verified borreliosis did not produce antibodies in sufficient
quantity to be seropositive (13). Review of normal human fetal and neonatal
antibody production in general reveals as well lagging IgG and IgM antibody
levels to age one year (graph data, P46)
(82). Beyond one to three years, we find no clarifying data.
Premise 5: Serologic testing to verify spirochete viability in late 'Lyme
disease' cases is reliable Numerous potential problems confound conclusions
from the available in-vivo serology data (37,83-86). Terms such as
'symptomatic' are typically defined within the case definition of early (We
assume much less than one year after inoculation) 'Lyme disease'. The
presumed presence of B. burgdorferi in the human host is dependent on a
strong history of vector inoculation and the subsequent (early) pattern of
antibody response. Other requirements imposed for 'proof of infection'
include endemic area residence, length of tick attachment, or recent memory
of EM-like rash. High background seropositivity in 'non-endemic' areas is
dismissed as 'false' without adequate proof. Studies are also insufficiently
longitudinal. The extensive data linking serologic outcome with spirochete
presence in the host limit this connection to relatively early 'Lyme
disease' and thus exclude late and all non-zoonotic patients. Once Bbsl
disseminates in the host, other immunerelated factors apply that have not
been sufficiently addressed in the 'Dearborn criteria' defining
seropositivity (87). Pleomorphism, variable antigen presentation, immune
avoidance, individual immune variance, host-derived enzyme cloaking, immune
complex sequestration, and antibody inaccessibility to spirochete-privileged
sites argue against sustained or consistent immune response (88-92). Recent
findings by Wang and Hilton suggest the presence or absence of Bb antibody
production is associated with unique individual HLA specificities of the
Class II (93). Eighteen of 44 (41%) variously symptomatic patients were
found seronegative where infection was verified by PCR to Osp A in
cerebrospinal fluid or mononuclear cells. Other data support serology test
uncertainty: (1) Seronegativity does not prove absence of a viable Bb
infection (45,90,93), which is consistent with the principle that negative
findings cannot be used to prove lack of positivity. (2) Antibodies may
exist minimally or rarely in very late Bbsl-infected humans (90). (3)
Culture and histological methods have been used extensively by
veterinarians, and provide substantial data supporting the inaccuracy and
insensitivity of serology in identifying living B. burgdorferi in non-human
subjects (51,52,60,94-98). (4) By ''similarity'', the presence of antibodies
to T. pallidum generally means the presence of spirochetes (99). An
evolutionary biology perspective is further to the point. Any persistent
pathogen (relevant if Bbsl survives into late illness) must effectively
escape the immune system. An example is Chlamydia pneumoniae infection where
antibodies appear only when the agent is causing active pneumonia, yet the
organism persists primarily unnoticed and undetected (2). Bbsl prevalence
data are rife with a mixture of asymptomatic seropositive as well as
symptomatic seronegative findings (43-45,49,53,65,90,100-107). Many
presumptions have been used to rationalize this data. Curiously, none have
considered the possibility that the subject pools may include a high number
of intra-human transfer cases. In summary, the preponderance of available
data cast serious doubt on the validity of current serology criteria for
diagnosing viable human Bb infection.
Premise 6: The presumed US human 'Lyme disease' agent is limited to one
species of Bbsl: B. burgdorferi sensu stricto (Bbss) Until very recently,
the presumed sole United States (US) human 'Lyme disease' agent is the
species Bbss. The preponderance of available data, based on the assumption
that all B. burgdorferi human infection is zoonotic, supports this
assumption (29,59). James has now published evidence, however, that Borrelia
lonestari infects humans in the US via the vector Amblyomma americanum (12).
Borrelia valaisiana has also been found to infect humans in two US cases
(13). Bbss was the first species to be identified shortly after discovery of
the disease in the Northeastern US. Because the illness was immediately
assumed solely a zoonosis, this assumption resulted in the tendency to look
for other possible species and strains less among ill humans than in vectors
and animal reservoirs. Such a 'self-fulfilling' assumption built into the
'Lyme disease' model may have helped assure that the only species identified
until recently would be the prevailing regional endemic zoonotic species.
Our Houston clinical experience of numerous patients with Acrodermatits
chronicum atrophicans (ACA), typically found in Borrelia afzelii, support
the likelihood that other Borrelia genospecies cause human disease within
the US. Premise 7: 'Lyme disease' is exclusively a vector-borne (primarily
arthropod) illness To date, the vector considered primary for transmission
of Bbsl to humans is the arthropod (17) likely related to its role in the
initial 1975 recognition of 'Lyme disease' in humans (6). Many arthropod
species have been found infected with Bbsl and causal transfer established
(13). The tick has been studied in North America exhaustively, having the
characteristics of a highly effective vector: long life, vertebrate blood
meal feeding, and bacterial transovarial passage (13). Its role in 'Lyme
disease' is assured, because it is the vector in what is considered
exclusively a zoonosis. Data are available, however, that expand the
possible diversity of Borrelia vectors worldwide beyond the arthropod. Other
possible carriers include the flea (108,109), mosquito (110-112), fly (111),
and mite (113). Related enzootic cycles have been only rarely examined,
although some data link non-arthropod vectors with animal hosts
(110,112-114). We suggest that early, 'Lyme disease' sustained myopic focus
on the arthropod as sole vector in the spread of 'Lyme disease' within the
zoonosis context likely delayed early consideration of other enzootic cycles
as well as non-zoonotic Bbsl transfer directly between humans. We propose
the human may well be the most likely 'vector' for Bbsl transfer to other
humans. The label 'Lyme disease' has become, by convention, a semantic
boundary that excludes consideration that an infectious agent responsible
for a zoonosis may also exist independently as a non-zoonosis. CDC-defining
criteria do not address human congenital transfer and in at least one
reference deny without proof that sexual transfer occurs (17). This mindset
assures that Bbsl cases falling outside 'Lyme disease' criteria have not
been considered in most research, nor reported to local health agencies.
Premise 8: Congenital (vertical) transmission between humans does not occur
The CDC position on intra-human Bbsl transmission is that 'Lyme disease
bacteria are not transmitted from person-to-person' (17). Current human and
veterinary data make this position indefensible (79,80). Schlesinger and
MacDonald reported the first human congenital transfer cases of Bbsl.
Gardner provided the initial and now most recent exhaustive review of
available human gestational transfer cases (13,81). Her credible supporting
studies utilized histological, PCR, or culture identification of Bb in both
mother and newborn or aborted fetus. She reviewed 263 Bbsl-infected cases
and summarized the birth outcomes. If mothers are untreated, Gardner notes
the high percentage of negative pregnancy outcomes along with symptomatic,
as well as seemingly asymptomatic, neonates. Indirect data supports the
possibility of human congenital B. burgdorferi transfer (95,96,115),
including similarity to other spirochetal diseases such as Treponema
pallidum (116-118). Contrary data suggest that congenital human (24,119-122)
and congenital animal (123-126) transfer does not occur. Use of the 'Lyme
disease' model for these studies (with inclusion criteria of EM rash, tick
attachment history, or endemic region residence) necessarily excludes
congenital transfer, which obviates their conclusions (15). Most of the
human data were based on simple surveys of birth outcome, without
satisfactory proof of spirochetal absence (likely a current impossibility).
The contrary veterinary data appear credible and employ a search for
spirochetes by culture or histological methods. These data support animal
species that exhibit congenital transference and those that may not, which
suggested species-specific transfer differences. Not unexpectedly, we find
no serious or credible epidemiological studies that have attempted to
identify the true rate of human congenital Bbsl transfer. The only method we
have of estimating congenital human Bb transfer is by other intra-human
illnesses. Transfer rates of Cytomegalovirus and Toxoplasmosis range from
14% to 59% (127). The congenital transfer rate of Treponema pallidum has
been reported as high as 68% in one cohort of treated infected mothers
(116). There is evidence to support the possibility that Bb may present
clinically differently in congenitally infected versus vector-inoculated
humans, and a review of similar chronic trans-placental diseases in humans
is instructive (82,127). Common in congenital infection are 'silent'
transfer, differential neonate illness presentation, and a negative effect
on later immune competence. The general principles of neonate immune
function, adult immune function, and transplacental transfer of pathogens
provide further insight into the relationship between trans-placental agents
and a new and developing immune system (13). This information collectively
suggests that silent or atypical birth presentation may be common, possibly
resulting in delayed or complete lack of recognition of the transfer.
Premise 9: Sexual (horizontal) transfer between humans does not occur The
CDC position on sexual intra-human Bbsl transmission is that it does not
occur (17). We find no study that addresses sexual transmission of Bb among
humans; conversely, we find no study supporting that it does not occur.
Inferential data, however, suggest the possibility of human sexual transfer.
The data come from sound veterinary studies (96,98,115), the finding of Bb
in human semen and breast milk (128,129), and by similarity to Treponema
pallidum where sexual transfer is abundantly documented (117,130,131). Our
clinical experience strongly suggests that predictable, possibly inevitable
Bbsl transfer between sexually active couples occurs. The preponderance of
infected spouses we have tested to date also exhibit positive serology or
PCR for Bbsl presence.
Premise 10: 'Lyme disease' is not considered a persistent infection,
implying self-limited outcome Most CDC-referenced studies support this
assumption. There is insufficient or no follow up after initial diagnosis or
treatment in these studies, however, to support this position. Extensive use
of unsupported presumptions is troubling as well. The latter include
labeling patients with persisting or recurring disease characteristics as
'reinfected' without serology or tissue evidence, or pre-746 Harvey and
Salvato Medical Hypotheses (2003) 60(5), 742-759 ª 2003 Elsevier Science
Ltd. All rights reserved. suming lack of infection because subsequent
positive antibody tests do not meet 'Lyme disease' inclusion criteria
(patients were not from 'endemic' areas, etc.)(17,18). Substantial data
support the probability that human Bbsl infection can persist indefinitely.
This state may obtain even when treatment is provided according to 'standard
guidelines' (52,90,132-135).
The supporting rationale for persistence is summarized as follows:
(1) Latency and relapse are widely observed Bb phenomena
(43,89,94,100,136,137).
(2) Symptoms frequently reemerge following therapy (89,94,100,137).
(3) Many mechanisms of potential survivability have been found in the highly
complex and adaptable Bb organism (138-141).
(4) An inert survival state is implied by the lengthy time to grow viable
spirochetes from EM incubated cultures (43,142).
(5) Animal models support extensive survival of Bb in tissue despite lack of
detectable presence in body fluids (51,52).
(6) Cyst forms have been found in-vivo to transfer infection directly
without reversion to spirochete form, suggesting a possible alternative
mechanism for silent transfer (143).
(7) Recent T. pallidum data unexpectedly support prolonged human spirochetal
infection despite use of standard treatment protocols (144).
Premise 11: Long-term 'Lyme disease' sequelae are autoimmune-induced or the
result of past infection damage This premise is a corollary of premise 10,
where longterm infection sequelae are used to rationalize lack of infection
persistence. Very few 'Lyme disease' CDC-referenced studies conclude that
long-term sequelae are a result of chronic infection (17,18). Several
hypotheses, nevertheless, have been advanced to address the nature of 'late'
sequelae. Autoimmune effect is one proposed mechanism derived from indirect
evidence (145-147). Another is anatomic damage assumed induced by Bbsl in
earlier infection (148). Both positions are hypothetical and use unsupported
assumptions. On the other hand, substantial data suggest that late sequelae
are the result of persistent infection (see Premise 3). We believe this
large number of published studies supporting that a high probability of
persistent Bbsl infection casts doubt on the above two mechanisms as primary
determinants of pathology. They may, we believe, be included within the
context of persistence as potential contributory mechanisms of ongoing
pathology. A search for other clinical outcomes of prolonged Bb infection in
published data yields no clear answer. The CDC position in 2001 is limited
to a few sentences: 'Infrequently, Lyme disease morbidity may be severe,
chronic, and disabling. An ill-defined post-Lyme disease syndrome occurs in
some persons following treatment for Lyme disease. Lyme disease is rarely,
if ever, fatal.' (17). Most published research avoids comment on longterm
sequelae (149). Because of the present dearth of relevant data, we propose
use of another perspective to address the question of sequelae from late
active Borrelia infection. Late effects differing from early effects is used
as a rationale that, because of this difference, support that active
infection no longer exists. Examination of other persistent infections
contradicts this argument. Many infections often present with dissimilar
acute and late effects. Examples are Chicken Pox later appearing as
'Shingles', and 'strep throat' manifesting eventually as Rheumatic Fever.
Some chronic infections have no acute phase. An example is the virus HHV-8
later manifesting as Kaposi's Sarcoma. Thus, there exists the possibility in
late Bbsl infection of not recognizing the presentation. Asymptomatic
patients with late infection may also be easily overlooked, and assumed
non-infected (20,30-54). Further, ill patients presenting with disseminated
symptoms without meeting defined 'Lyme disease' endemicity criteria are also
at serious risk of not being considered Bbsl infected.
Premise 12: Lyme disease is geographically constrained to areas of high
zoonosis prevalence, mostly in North America and Eurasia Gardner has
comprehensively summarized international 'Lyme disease' distribution data
(13). The resulting map concentrates illness primarily into a Northern
Hemisphere temperate zone belt covering most of Europe and the United
States. Expectedly, maps of zoonotic endemicity overlie the illness maps
faithfully. We conclude a high likelihood that 'Lyme disease' is constrained
to areas of high zoonotic endemicity simply because endemic area occurrence
is an inclusion criterion. This illustrates the circularity of creating a
predicted disease outcome by limiting its definition. An extensive search of
published literature reveals that distribution of human borreliosis may be
much broader than described, practically is essentially globally
disseminated. Bbsl presence in humans, other vertebrate reservoirs or both,
have been reported from over thirty countries on six continents and several
islands(5,22,54,55,58,110-112,114,150-179). Failure to document the full
geographic extent of the organism may stem from simple lack of public health
resources in most countries or lack of recognition of the disease in humans.
We find no credible studies of human Bbsl infection prevalence conducted
outside 'endemic' zoonotic regions.
Premise 13: 'Lyme disease' is a contemporary human illness Bbsl was first
acknowledged as a human pathogen in the US medical literature from 1982 to
1983 (7-9,11). The limited historical data that address earlier human
infection do so indirectly by examining reservoir (Peromyscus,
Massachusetts, 1894) or vector (Ixodes ricinus, Germany,1884) infection
using museum DNA evidence(163,180), or disease categorizations based on skin
manifestations of unknown etiology (26,30,172,181-185). Data from outside
the 'Lyme disease' zoonosis model vaguely suggest the possibility that Bbsl
is not a recent pathogen in nature, including human infection. Isolated
papers examining Bb dissemination address such possibilities as: (1) The
birth of the pathogen as a transkingdom mutation from African Swine Fever
virus (186). (2) European 'Lyme disease' gradient rising from West to East
(22). (3) Extensive presence of Borrelia garinii and afzelii in Eurasia
(13). (4) Extensive presence of the spirochete (Garinii and Afzelii only) in
Northeast Asia (Vladivostok) in a common Ixodid vector providing opportunity
for Siberian-Alaska land bridge transfer 10,000-30,000 BC (187). (5) A
recent hypothesis that Bbsl may be the protective agent of juvenile and
adult arthritis in Louisiana Tchefuncte Indians between 500 BC and 300 AD
(187,188). (6) Documented human presence in central and southern South
America. (7) Bb sensu stricto main genospecies in North America (18). (8)
Evidence for spread of Bb sensu stricto from the Western hemisphere to
Europe after 1492 (189). Together, we suggest these data hint at a possible
but unexamined circum-global dissemination of the pathogen over many human
generations.
We believe the global occurrence of B. burgdorferi and its many
strains provides the strongest evidence to support the likelihood that Bbsl
has been present in nature and in humans for centuries to millennia.
Protracted existence of the spirochete, if validated, would provide strong
support for broad intra-human spread that began some indefinite time
following early vector-to-human transfer. Re-synthesis of premises by
'preponderance of data' weighting 'Lyme disease' was the label initially
given to the illness conceptualization (disease model) of human Bb
infection. The model congealed about 20 years ago as a zoonosis principally
from locally available information. Subsequent worldwide data appear to have
been gathered within the contextual boundaries of this initially conceived
model. We find that the preponderance of this data support the conclusion
that the zoonotic model was, and remains, incomplete, and includes only a
portion of all B. burgdorferi infected humans. The data suggest there may
exist a much larger unrecognized pool of Bbsl-infected individuals sustained
by persistent intra-human transfer that we provisionally call 'Epidemic
Borreliosis'. A summary comparing these two populations is shown in Table 1.
Clinical diagnosis of long-infected patients has been inconsistent and
puzzling, specifically regarding the signs of EM rash and arthritis. The
erythema migrans rash, initially considered the herald lesion for infection,
actually occurs both sporadically in initial inoculation and later as
secondary lesions. Its absence alone is thus of no value in rejecting a
diagnosis, although its presence alerts to the probability of infection. A
symptomatic state may be present or absent in the initial presentation,
where absence of symptoms can mask the presence of a non-pathogenic strain.
Arthritic joints are considered common in disseminated zoonotic Bbsl, but
paradoxically, only intermittent and migratory joint pain is described in
very late borreliosis. Laboratory tests are presently reliable for
supporting a diagnosis of recent vector-transferred 'Lyme disease' but seem
highly unreliable if the transfer was zoonotic more than a year earlier, or
was congenital. In these cases where antibodies are likely sparse, serology
is valid only when positive. Negative results are necessarily inconclusive
and may be seriously misleading, regardless of symptoms. The argument
supporting 'false positive' serology, when based on zoonosis criteria, is
invalid if Bbsl infection is widespread from prolonged intra-human transfer.
The pathogen responsible for 'Lyme disease' is a limited subset of the
genospecies B. burgdorferi.
Until recently, only one human pathogen had been
identified in the continental United States: B. burgdorferi sensu stricto
(Bbss).
The exclusive position of this species likely arose from limiting
the early search for vectors to the geographic region where 'Lyme disease'
was initially discovered. Given the global diversity of species such as
Borrelia afzelii and garinii, the reality of intra-human transfer, and the
probability of prolonged infection, we expect extensive regional diversity
of Bbsl species in humans both in endemic and non-endemic regions of the
earth. We propose that anticipating other species will improve
identification by broadening the carriers tested to humans as well as
zoonotic vectors and reservoirs. Transfer of Bbsl to humans occurs via both
zoonotic vectors ('Lyme disease') and other humans. Congenital transfer is
fact. Animal data support that sexual transfer can occur, and other data
suggest its possibility.
'Lyme disease' reservoirs and vectors may be even
more globally widespread than currently modeled, increasing the probability
of broader and historically longer inoculation of Bbsl into the human
population.
The finding of Bbsl virtually all countries where it is sought
also implies more uniform global distribution of the infection in humans.
We
expect regionally endemic 'Lyme disease' cases now to be far fewer than
intra-human disease cases (Epidemic Borreliosis), and that many of the
latter are unrecognized principally due to mislabeling. Regardless of
initial transfer route, human infection with pathogenic Bbsl may persist for
life. Symptoms and signs may vary from sub-clinical to extensive and severe,
including cycling between these states (33,77,190,191). Clinical sequelae
from prolonged infection appear likely and may be cumulative with various
mechanisms operant. An unknown number of sub-clinical cases may remain
undetected for life, regardless of whether latency persists or unrecognized
activation occurs. If lifetime persistence is the rule, then all living,
untreated patients infected at any time during their lives remain infected.
The preponderance of all reviewed data suggest that B. burgdorferi may have
been present in both natural reservoirs as well as in humans via intra-human
transfer for centuries or millennia.
We propose that this concept, if
verified, predicts a much larger current population of congenitally and
possibly sexually infected individuals worldwide than infected via zoonotic
vectors.
In summary, we propose a significantly modified human Bbsl
infection illness model that incorporates 'Lyme disease' only as one engine
feeding a larger reservoir of chronic, Borrelia-infected humans (Fig. 1). We
believe zoonosis was the likely source of initial human disease, and
continues to contribute newly infected cases. We further propose that
vertical and horizontal intra-human transmission over generations has likely
had a non-linear amplifying effect on human prevalence. If true, this
transmission mechanism now significantly exceeds the contribution of new
cases from zoonotic vectors, and has reached pandemic proportion on all
continents where humans reside. These conclusions strongly support our
clinical experience.
Table 1
Comparison of proposed illness characteristics
'Lyme disease'a Epidemic Borreliosisb What is the initial disease
presentation? Erythema Migrans (EM) rash Frequent but inconsistent
Secondary; occasional Symptoms present if 'early' None to flu-like In
neonates: none to fatal Symptoms present if 'late' None to multi-organ None
to multi-organ Joint symptoms if 'late' Arthritis Arthralgiasc Cardiac signs
if 'late' High-degree block Arrhythmias, T-waves unstablec Typical illness
length at initial presentation <1 year <2 years.
How is the infection
diagnosed? Number of Bb species One (region-specific) Many (non
region-specific) Serum antibody levels High and constant Low or occasional
Accuracy of serology Accurate High number of false negatives Relationship of
serology to region of diagnosis Direct None Usefulness of EM rash Alerts to
recent inoculation Announces infection presence Usefulness of arthritis
Suggests 'late' infection stage Suggests zoonosis transfer How is the
disease conferred to humans? By Zoonotic Vector? Yes No Congenitally? No Yes
(proven) Sexually No Yes (not studied) Regional? Yes: endemic areas No:
anywhere What is the disease course & outcome? Self-limited? Defined as
likely No; likely lifelong infection Latent? Yes Yes Activate or reactivate?
No by assumption Yes Considered long-term sequelae Autoimmunity/residual
damage Infection persistence/(mechanisms not elucidated) Asymptomatic
seropositive patient infection status Considered infected only if from
endemic region Infected regardless of region of residence What other
epidemiological factors pertain? Primary worldwide vector Arthropod Human
Bbsl presence in human population (time) Not addressed Millennia
Distribution in humans Confined to endemic areas Worldwide, diffuse a 'Lyme
disease' - All human cases of human Borrelia burgdorferi infection within
the defined limits of the CDC case definition (15). b Epidemic borreliosis -
all human Borrelia burgdorferi infection cases outside the CDC case
definition of 'Lyme disease'. Includes Zoonosis vector-transferred
infections more than one year old and all congenital, gestational and sexual
intra-human transferred infections whether symptomatic or asymptomatic. c
Author's clinical experience.
CONCLUSIONS: We propose there are at least two similar and unified, but
distinct forms of human B. burgdorferi infection: 'Lyme disease', and
'Epidemic Borreliosis' (disease spread directly between humans). Late (more
than one year old) zoonotic disease may overlap both forms. 'Lyme disease'
is the only presently acknowledged Bb illness form, conceptualized as a
zoonotic disease where intra-human transfer is considered rare. As defined,
'Lyme disease' is primarily located in limited geographic areas, is
clinically recognized relatively early after inoculation, and the reported
case numbers are small. Human infection in this model is considered
accidental and 'self-limited'. We propose the existence of a much larger
'non- Lyme' pool of B. burgdorferi-infected humans with a clinical
presentation of extraordinary variability, global geographic distribution,
and far greater prevalence. Transfer is intra-human (congenital and almost
certainly sexual) and is initially silent or unrecognized. If not
successfully treated, infection is life-long, and latency, late activation,
and reactivation are common. Zoonotic cases more than one year old may
present similarly. We label this larger pool 'Epidemic Borreliosis'.
Combining both the 'Lyme' and 'non-Lyme' concepts results in a significantly
altered model of human B. burgdorferi infection. Zoonotic borreliosis is
fact and is the milieu within which the complete human disease has existed,
perhaps for millennia. Zoonotic transfer was likely the initial route of
human inoculation and continues with regularity into the larger pool of
infected humans in zoonotically endemic regions. We believe that human
endemicity is virtually ubiquitous wherever humans live worldwide and has
now reached pandemic proportion. Overlap of these two groups occurs where
competent infected vectors exist, but we believe the numbers of 'non-Lyme'
cases predominate significantly even here. Infection prevalence has not
likely reached numerical stability, since the amplifying effect of
congenital transfer, coupled with the current global population expansion,
suggests the probability of continuing prevalence rise. We propose that
'Lyme disease' is a limited conceptualization of a far more pervasive
Borrelia infection state that is now an unrecognized global epidemic.
DISCUSSION: Our proposed model further challenges many aspects of medical
science now believed to be true. Not only does it consign 'Lyme disease' to
a minor role in B. burgdorferi infection prevalence but supports the idea
that a zoonosis can initiate what can later become a vastly more extensive
intra-human infectious disease. As a medical model, we find this revised
concept works with exceptional success. Until now, the current model has
seriously limited our capacity to diagnose and treat many patients. In our
patient population, the revised model provides a rational mechanism that far
better explains our experience. It effectively resolves the conflicting
viewpoints held by clinicians and academicians about what has been labeled
'chronic Lyme disease' and is now allowing us to resolve or minimize illness
in most of these patients. Unexpectedly, the revised model has provided us
much more. Most of our patients arrive with a diagnosis from diverse
specialty areas but unsuccessfully treated. Use of this reframed model
provides rational insight into many of these cases. It provides a successful
diagnosis and treatment strategy that, when applied, resolved many patients'
symptoms, thus suggesting to us that B. burgdorferi may underpin these
illnesses as cofactor or origin. We wondered if all of these 'atypical'
Borrelia-infected patients might give us a clue to the true magnitude of the
infection prevalence. A large number of clinically-similar (to late Bbsl
infection) medical conditions with unknown etiology exist within the
inclusive medical framework. Aaron, reviewing evidence from unexplained
medical conditions (chronic fatigue syndrome, fibromyalgia, the irritable
bowel syndrome, multiple chemical sensitivities, temporomandibular disorder,
tension headache, interstitial cystitis, and the post-concussion syndrome),
found substantial clinical overlap (192). Clauw likewise found similar
clinical overlap among several of these illness categories, as have others,
that include Gulf War Syndrome variants, overtraining syndromes, and
numerous 'functional somatic syndromes' (193-197). A single paper by Pachner
and Steere written in 1985 provides a credible rationale for most of the
neurological symptoms and signs described in these illnesses as well as in
human Bbsl infection (198). Later papers describe a persistent, infection-
based inflammation that may provide the fundamental pathology mechanism
(106,145,199-215). These 'orphan' illnesses that constitute most of our
(now) Borrelia-positive Houston patients sum to at least a double-digit
prevalence in the United States even if we consider only four of these
'chronic syndromes'. Estimates of case-definition fibromyalgia include 2-4%
(193,216,217), chronic fatigue syndrome, 0.42% (218), Gulf War syndrome, 4%
(219) and multiple chemical sensitivities, 2-5% (220). Vague model boundary
limits in these similar 'syndromes' coupled to illness labels where no
prevalence data is available make this information unquestionably imprecise.
However, when combined with the unknown but finite prevalences of the many
other illness categories mentioned above including unknown overlap, however
they hint that the combined number is not small. We wondered whether our
proposed model could generate such numbers as the infection rates reported
by the CDC support a much lower prevalence (18). When the CDC data are
examined using our derived assumptions, however, and a 'zoonosis-only'
prevalence is generated, the outcome is 0.6% (Appendix A). We next generated
a crude estimate of the expected background prevalence of 'Lyme disease' in
non-endemic regions using the assumptions of Masters (221). The resulting
point prevalence is 2% once system stability has been reached. If congenital
transfer is added and assumed ongoing for 1000 years, which we think not
unreasonable, the point prevalence in 2000 AD becomes 6.5%. If sexual
transfer is further added with the same assumptions at a 50% transfer rate,
the combined point prevalence becomes 15.5%. Details of this exercise are
found in Appendix B. Another combined prevalence estimate based only on
symptoms was generated by one of us in 1993 from an annual medical history
form. 2683 employees of a Department of Energy plant were queried regarding
30 common symptoms and signs of late Bb infection. Endemicity and EM or tick
bite criteria were excluded. 12.8% of the employees met similar symptom
criteria. (Appendix C: unpublished data). These unexpected numbers of
possible Bbsl-infected patients hidden for decades by mislabeling, fit
comfortably within our proposed model and are then not difficult to explain.
We propose that where data were not initially available, temporary
hypothetical bridges, although dissimilar, were necessarily created by early
investigators to fill their model framework gaps. Our proposed model now
fills in most such gaps for all these illnesses. It also revises framework
elements of other illnesses we had considered unassailable parts of the
standard medical paradigm. We believe failure to recognize the breadth of
this infection is readily explainable by inadvertent research errors: (1)
most data have been derived only from zoonotically endemic areas, (2)
'validation' rested on inadequate serologic diagnostic methodology, and (3)
controls, when used, were useless since half those infected are
'sub-clinical'. Clearly, discovery of the illness in an area of high
zoonotic endemicity contributed to early and continuing clinical myopia but
was the necessary first step in its recognition. Other factors contribute to
clinical recognition failure. Silent transfer, latency, late activation, and
recurrent activation likely combine to create a setting resistant to
standard epidemiological detection methods. The pathogen's extreme
complexity is another probable contributor. Its adaptability, pleomorphism,
genetic diversity, and differential tissue tropism create extraordinary
symptom variability. Likewise, activation of numerous latent viruses and
opportunistic bacteria from immune depression in late disease may further
expand illness complexity. Such varied presentation is not likely to have
previously been considered to have a single infectious etiology, thus which
excludes Bbsl from most differential diagnoses. We propose that zoonotic
transfer combined with human transfer on a global scale for centuries can
indeed result in double-digit prevalence. These numbers applied to our model
hint at the highly improbable: that the prevalence of all humans infected
with Bbsl could constitute an even larger percentage of the population by
including the sub-clinical cases mostly excluded from epidemiological
surveys to date. The technological solution that can validate our proposed
model is a more sensitive and specific laboratory test likely not based on
serum antibody presence. We do not offer a specific solution but propose
that detection of the organism itself or unique biochemical markers altered
by the infection are required. Because of unpredictable latency and
inability to use controls, use of Koch's postulate or sophisticated
epidemiological methods are no longer adequate in cases such as this and
have likely reached the historical limits of their usefulness here (2).
Despite the need for dramatically improved detection methods, there exists
even now an instrument capable of recognizing the scope of this illness: the
astute clinician willing to carry what Carl Sagan called the 'burden of
skepticism' (222). This perspective is an essential medical tool, as the
'system' within which the clinician works requires reframing prevailing
cognitive paradigms before unfamiliar ideas can be 'seen' at all (222,223).
(The history of medicine, in fact, is built on examples of mindset that
delayed recognition or evolution of most illness concepts (224).) We believe
that reliance on familiar models is ultimately the principal reason that
what we here term 'Epidemic Borreliosis' remains hidden from the view of
science. Our purpose in publishing this newly proposed model is to encourage
skepticism by investigators as well as clinicians: to consider the
possibility that 'Lyme disease' is an inadequate conceptualization of all
human Bb infection. We are confident that once considered, others will 'see'
what we are finding in clinical practice (223). We thus propose that 'Lyme
disease' is only the herald encounter with a human infectious disease of
currently inconceivable proportion. We anticipate that if our model is
validated and the proposed high prevalence of B. burgdorferi in humans is
verified, the conceptual framework of this and many other human diseases
will be radically altered.
RECOMMENDATIONS: Our recommendations are based on verifying or disproving the
disease model we present here and are derived from the gaps remaining in the
data we have reviewed. The initial task must be to identify all humans
infected with B. burgdorferi. This likely requires first understanding
antibody status in late (beyond 18 months) infections and the
pathophysiological mechanisms linking Bbsl presence and human disease.
Commercial tests to reliably detect living Bbsl in humans as well as
reservoirs and vectors must then follow to reveal the agent's true worldwide
prevalence. Extensive effort will be required to prove or disprove
persistence, and to determine all disease entities associated with Bbsl
infection: whether cause-and-effect, co-factor, or unrelated. Finally, the
full extent of epidemiological science must be applied to determine the
scope and efficiency of human congenital transfer and to investigate sexual
Bbsl transfer. Answers will guide development of preventive strategies.
Concurrently, treatment modalities and schedules to eradicate B. burgdorferi
from all patients regardless of infection route or duration, must be
created. If our experience holds, this will be a difficult task, and will
require serious and rapid commitment from all nations.
To the Top
|
| | |
