Content uploaded by Timothy J. J. Inglis
Author content
All content in this area was uploaded by Timothy J. J. Inglis on May 19, 2016
Content may be subject to copyright.
Content uploaded by Timothy J. J. Inglis
Author content
All content in this area was uploaded by Timothy J. J. Inglis on Dec 21, 2015
Content may be subject to copyright.
Review
Principia ætiologica: taking causality beyond
Koch’s postulates
Timothy J. J. Inglis
Correspondence
Timothy J. J. Inglis
tim.inglis@health.wa.gov.au
Division of Microbiology and Infectious Diseases, PathWest Laboratory Medicine WA, QEII Medical
Centre, and School of Biological and Chemical Sciences, University of Western Australia,
Nedlands, WA 6909, Australia
There is no single accepted method to establish a causal relationship between an infective agent
and its corresponding infectious disease. Different biomedical disciplines use a patchwork of
distinct but overlapping approaches. To a greater or lesser extent these are based on criteria
known as the Koch–Henle postulates, or ‘Koch’s postulates’ for short. Deficiencies in Koch’s
postulates were recognized by their principal author shortly after their formulation. Now, over a
century later, a more rigorous method to test causality has still to be finalized. One contender is a
method that uses molecular methods to establish a causal relationship (‘molecular Koch’s
postulates’). Recognizing the wider range of contemporary approaches used to build an argument
for a causal relationship, the use of a more inclusive approach to establish proof of causality is
proposed. This method uses an argument built from a series of assertions. Assertion 1:
congruence or reproducible correlation of a taxonomically defined life form with the
clinico-pathological and epidemiological features of infection. Assertion 2: consistency of the
demonstrable biological response in the subject to an encounter with the prospective infective
agent. Assertion 3: progressive or cumulative dissonance as an explanation for pathophysiological
processes at every known level of biological organization in the subject. Assertion 4: curtailment of
that pathophysiological process on the deliberate introduction of a specified biomedical
intervention. Evidence to implicate the candidate biological entity as an initiator of or primer for
cumulative dissonance places it in a subcategory of micro-organisms to be known as ‘priobes’. A
priobe is the sufficient and necessary antecedent cause of a pathophysiological process
evident as an infectious disease.
Introduction
Robert Koch and Jacob Henle are generally credited with
the first method used to establish the aetiology of a
specified infectious disease (Koch, 1884). The Koch–Henle
postulates served the emerging science of microbiology well
during its early years, and gave experi mental consistency to
the investigation of causal relationships. However, signifi-
cant limits to the postulates were soon recognized and
restricted their wider scientific application. In recent times,
the original Koch–Henle concept of causation has been
used by experimentalists with decreasing frequency. The
expanding investigative repertoire of the life sciences
dictates a need for a contemporary method for establishing
the cause of infection. Falkow recognized the need for
more rigorous genetic criteri a for the determinants of a
disease, and modelled his proposals for the molecular
biology of pathogenesis on the Koch–Henle postulates
(Falkow, 1988). While Falkow’s molecular reiteration of
the original postulates established a standard for molecular
and cell biology developments in the clinical sciences, it did
not anticipate a direct contribution to concepts of
pathogenesis from the emerging fields of molecular ecology
and microbial population genetics. The possibility that all
levels of biological organi zation might have some bearing
on the occurrence, course and outcome of infecti on
was recognized by Colwell and named ‘biocomplexity’
(Colwell, 1999). Lederberg alluded to the evolutionary
context of infection when he posed the question why do
micro-organisms not cause disease more often (Lederberg,
1999). Salyers and Witt placed even greater emphasis on
ecological determinants of bacterial infection (Salyers &
Witt, 2005). In their critique of the Koch–Henle postulates,
they highlight the need to recognize that the effect of
specific biomedical interventions might also provide proof
of cause.
These considerations are undergirded by developments in
scientific method and the philosophy of science. There is
growing ethical objection to the use of laboratory animal
models for incremental scientific gain. This places a restraint
on the use of animal models for pathogenesis research or
clinical diagnostic work. Molecular and cell biology
methods cannot fully substitute for tissue pathology, and
Journal of Medical Microbiology (2007), 56, 1419–1422 DOI 10.1099/jmm.0.47179-0
47179
G
2007 SGM Printed in Great Britain 1419
whole-organism outcome indices, such as failure to thrive,
weight loss and death. There are also significant difficulties
establishing an aetiological role for micro-organisms that
cannot be cultivated under laboratory conditions. At a
philosophical level, contemporary scientific method does
not assume absolute objectivity in the experimentalist and
places greater reliance on verification by complementary
experimental methods. In the specific case of establishing
cause, supportive evidence for sufficient cause (evidence that
the proposed causal agent was capable of a causal role) is
inadequate. Evidence for necessary cause must also be
advanced. To establish necessary cause, there must be
evidence that an outcome cannot occur without the
proposed agent. In studies on the aetiology of infection,
necessary cause alludes to Popper’s concept of falsifiability,
and to the recognition by Salyers and Witt of medical
intervention as evidence for a causal relationship (Popper,
1963; Salyers & Witt, 2005).
Central to our understanding of the origin or pathogenesis
of infectious disease is the causal microbial agent or
‘pathogen’. This term has acquired a range of overlapping
and sometimes contradictory meanings as our knowledge of
pathogenic processes has grown. Given these limitations, I
will use ‘priobe’ as a new term to denote a specific, biological
initiating or priming agent implicated as the sufficient and
necessary antecedent cause of an infective process.
The principles by which co nfirmation of an aetiological
role can be established for the inferred infective agent are
laid out in the following paragraphs.
Definition of terms
Priobes
(i) A priobe is a specific, biological initiating or prim-
ing agent implicated as the sufficient and necessary
antecedent cause of a pathological process.
(ii) All priobes are capable of initiating or priming a
biological response in a dissimilar life form.
(iii) Not all micro-organisms are priobes.
(iv) Not all priobes are micro-organisms.
(v) Most but not all priobes are unicellular.
(vi) Most but not all priobes are capable of indepen-
dent biological organization at molecular or
higher level.
(vii) Failure of the responding life form to resolve
negotiations with a priobe at any level of
biological organization leads to accumulation of
dissonant effects at that level and contributes to
the pathological state.
(viii) Most but not all priobes must maintain an
encounter with the responding life form to
sustain cumulative dissonance.
Cumulative dissonance
(i) The progressive or accumulated series of changes
initiated by a priobe at every level of biolog ical
organization in which the pathological state has
been described.
(ii) Thus, an accumulation of unresolved negotiations
in successive biolog ical dimensions arising from
encounters between distinct life forms determines
the occurrence, course and outcome of pathology.
Principles of aetiology
(i) The method recognizes that the argument for a
causal relationship will be assembled from a series
of assertions about the interactive relationship
between the proposed priobe and its corresponding
biological partner. The sequence begins with
investigation of conjecture, continues through an
accretion of experimental evidence and concludes
with demonstrable falsifiability of the core relation-
ship through deliberate, controlled intervention.
(ii) The method aims to satisfy the requirements of the
Koch–Henle postulates (Koch, 1884), Falkow’s
molecular corollary (Falkow, 1988), Colwell’s
recognition of scales of biocomplexity (Colwell,
1999), observations by Salyers and Witt on the
admissibility of intervention (Salyers & Witt,
2005) and Popper’s insistence on falsifiability
(Popper, 1963).
(iii) The method recognizes the wider biological
context of an interactive process considered to be
infection, and avoids attribution of causality from
isolated experimental, pathological or time–place-
limited epidemiological phenomena.
(iv) The method matches the level of certainty attached
to a causal relationship with the number of
assertions that can be made to support the argument.
(v) The method accepts that preliminary conclusions
as to the causation of individual clinical episodes,
public health incidents and the natural history of
emerging infections are all provisional and subject
to revision in the light of new observations.
Argument
Assertion 1: congruence
The consequences of an encounter with a taxonomically
defined, distinct life form must be congruent with the
clinical, pathological and epidemiological features of disease.
Assertion 2: consistency
The introduction of a taxonomically defined life form must
initiate a consistent biological reaction without antecedent
external cause.
Assertion 3: cumulative dissonance
The taxonomically defined life form must initiate a
progressive series of changes at every level of biological
organization in which the infection has been described.
T. J. J. Inglis
1420 Journal of Medical Microbiology 56
Assertion 4: curtailment
Restriction, alteration, destruction or removal of the
candidate priobe at any level of biological organization
must reduce the occurrence, course or outcome of the
corresponding infection.
Process
The process of aetiological certainty is shown in Table 1.
Practical application
Before Koch’s postulates there was no rigorous method to
establish causality. However, the system he introdu ced in
the postulates had methodological and epistemological
limitations (Mazzarello, 2004). The immediate quandary
Koch encountered was how his postulates could accom-
modate the recurrence of cholera epidemics in the absence
of an asymptomatic carrier state or environm ental
reservoir. Until toxigenic Vibrio cholerae was discovered
in the mucilaginous coating of blue-green algae, it was
widely believed that cholera lacked an environmental
reservoir (Islam et al., 1990; Alam et al., 2006). Further
work on the ecology of V. cholerae, and its relationship to
oceanic temperature fluctuations and climatic cycles, has
added important detail to a biocomplexity model of
cholera risk that spans the molecular and cellular scale at
one end to human social behaviour and climatic change at
the other (Lipp et al., 2002). Cholera has therefore become
the test case for biocomplexity, over a century after Koch’s
original observations on the disease. The biocomplexity
concept has been successfully employed as a model in
reverse by beginning with a point-source outbreak, and
pursuing the biocomplexity paradigm through individual
patient management and cellular interactions in the
environment, down to specific molecular markers for a
disease of environmental exposure: melioidosis (Inglis et al.,
1998, 2000, 2001; Merritt et al., 2006).
Another aspect of infectious di sease that eluded Koch and
his contemporaries was the basis of diseases that lacked a
single precipitating microbe or microbe-derived toxin. The
role of microbial consortia and polymicrobial complexes in
specific infectious disease is now better understood
through the contribution of biofilms, particularly in
medical device-associated infections (Costerton et al. ,
2005). In the current model of causality, these biofilms
can be considered as proto-tissues, entering the patho-
physiological process at the tissue level of organization.
These and other biofilm-related infections necessarily defy
analysis in a one species/one organ model of infection,
particularly when the immedi ate environment is dynam-
ically changing, such as the respiratory support devices
used for intensive care patients (Safdar et al., 2005).
It follows that the inter face between infected subject and
priobe requires analysis on both sides of the border.
Microbe/innate immunity interactions do not al ways
follow the simple course originally predicted. The con-
sequences of an encounter with a microbe or its products
can be a spectrum of evoked responses, which may in turn
be modulated by the variable condition of the infected
subject. A decade ago the variation in outcomes of severe
sepsis was attributed to immunological dissonance: a
breakdown in the balance between pro-inflammatory and
suppressive effectors (Bone, 1996). This process was
subsequently alluded to as an example of predictably
complex biological phenomena in a dysfunctional up-
scaling of host response during severe infection (Gullo,
1999). The current synthesis argues that immunological
dissonance is an important component of cumulative
dissonance because it describes how the evoked host
response can become dysfunctional and ultimately
destructive. An example of damage incurred in human
subjects as a result of cumulative dissonance is the cytokine
storm proposed for some of the more severe pathophysio-
logical aspects of epidemic influenza (Yokota et al., 2000).
Although the earliest descriptions of the process describe
cumulative dissonance in influenza, the other three criteria
are still required to establish a causal relationship between a
specific strain of influenza virus and specific disease events.
Congruence, consistency and curtailment also provide
yardsticks for more rigorous analysis of emerging infec-
tious agents. They are and have been used extensively
under different names by different biomedical disciplines,
but need to be assembled in a single cohesive argument if
truly new disease phenomena, such as severe acute
respiratory syndrome, are to be better understood.
Possibly the greatest challenge to any novel approach to
causal analysis of infectious disease is the group of
proposed infective agents that have been linked with the
transmissible spongiform encephalopathies (Aguzzi &
Heikenwalder, 2006). In some circles, viruses still provoke
discussion about whether or not they constitute a
Table 1. The process of aetiological certainty
Aetiological
certainty
Assertion Short form Explanation
Conjectural 1 Congruence Congruence with clinico-pathological and epidemiological features
Potential 2 Consistency Consistency of biological response to encounter with prospective infective agent
Candidate 3 Cumulative dissonance Progressive or cumulative dissonance at every known level of biological organization
Confirmed 4 Curtailment Curtailment of infective process on controlled introduction of biomedical intervention
Principia ætiologica
http://jmm.sgmjournals.org 1421
legitimate life form. The proposed agents of spongiform
encephalopathy are therefore right at the edge of
contemporary understanding of microbial life. Yet they
appear to behave as transmissible agents of disease. The
prion hypothesis is thus a test of the present model for
analysis of causation. From what is already known, there is
a degree of congruence between the molecular biology,
epidemiology and clinico-pathological features. This has
reproducible consistency and elements of cumulative
dissonance as the disease progresses during a period of
latency from molecular and cellular damage to the onset of
gross neurological signs. It could also be a rgued that the
effects of the offal ban in the UK in response to the bovine
spongiform encephalopathy outbreak represent evidence
for curtailment (Stevenson et al., 2000). If the remaining
details in the pathophysiology of Creuzfeldt–Jacob Disease
continue to provide evidence for the proposed four-po int
argument for causality, the putative prion agent will
deserve recognition as a provoked reaction initiator: a
priobe. However, proponents of the prion hypothesis will
then need to establish a consensus on the taxonomic status
of the putative infective agent.
Conclusion
The principles described in this account take note of
methods currently used in a variety of disciplines to assemble
a robust argument for causality. While the terms used and
emphasis placed on specific assertions may vary from
discipline to discipline, all who work in this field understand
the need to unify the evidence they present in a single,
cohesive argument. Whether that argument is used to
develop a hypothesis or to refine an experimental process,
the process described here is itself provisional and subject to
further refinement in the light of future observations.
Acknowledgements
I am grateful to my colleagues, and in particular to Dr Manfred
Beilharz, Professor Thomas Riley and Professor Ian Poxton for their
advice during the preparation of the original manuscript.
References
Aguzzi, A. & Heikenwalder, M. (2006). Pathogenesis of prion diseases:
current status and future outlook. Nat Rev Microbiol 4, 765–775.
Alam, M., Sultana, M., Nair, G. B., Sack, R. B., Sack, D. A., Siddique,
A. K., Ali, A., Huq, A. & Colwell, R. R. (2006). Toxigenic Vibrio cholerae
in the aquatic environment of Mathbaria, Bangladesh. Appl Environ
Microbiol 72, 2849–2855.
Bone, R. C. (1996). Immunologic dissonance: a continuing evolution
in our understanding of the systemic inflammatory response
syndrome (SIRS) and the multiple organ dysfunction syndrome
(MODS). Ann Intern Med 125, 680–687.
Colwell, R. W. (1999). Biocomplexity. The 99th Annual General
Scientific Meeting of the American Society for Microbiology, McCormick
Place, Chicago, IL.
Costerton, J. W., Montanaro, L. & Arciola, C. R. (2005). Biofilm in
implant infections: its production and regulation. Int J Artif Organs
28, 1062–1068.
Falkow, S. (1988). Molecular Koch’s postulates applied to microbial
pathogenicity. Rev Infect Dis 10 (Suppl. 2), S274–S276.
Gullo, A. (1999). Sepsis and organ dysfunction/failure. An overview.
Minerva Anestesiol 65, 529–540.
Inglis, T. J., Garrow, S. C., Adams, C., Henderson, M. & Mayo, M.
(1998).
Dry-season outbreak of melioidosis in Western Australia.
Lancet 352, 1600.
Inglis, T. J., Rigby, P., Robertson, T. A., Dutton, N. S., Henderson, M.
& Chang, B. J. (2000).
Interaction between Burkholderia pseudomallei
and Acanthamoeba species results in coiling phagocytosis, endamebic
bacterial survival, and escape. Infect Immun 68, 1681–1686.
Inglis, T. J., Golledge, C. L., Clair, A. & Harvey, J. (2001). Case report:
recovery from persistent septicemic melioidosis. Am J Trop Med Hyg
65, 76–82.
Islam, M. S., Drasar, B. S. & Bradley, D. J. (1990). Long-term
persistence of toxigenic Vibrio cholerae 01 in the mucilaginous
sheath of a blue-green alga, Anabaena variabilis. J Trop Med Hyg 93,
133–139.
Koch, R. (1884). Die Aetiologie der Tuberkulose Mitt Kaiser Gesundh
2, 1–88.
Lederberg, J. (1999). Pathogens at the crowded summit. The 99th
Annual General Scientific Meeting of the American Society for
Microbiology, McCormick Place, Chicago, IL.
Lipp, E. K., Huq, A. & Colwell, R. R. (2002). Effects of global climate on
infectious disease: the cholera model. Clin Microbiol Rev 15, 757–770.
Mazzarello, P. (2004). The explicative power of the rules: the case of
the Koch’s postulates. Med Secoli 16, 293–316 in Italian
Merritt, A., Inglis, T. J., Chidlow, G. & Harnett, G. (2006). PCR-based
identification of Burkholderia pseudomallei. Rev Inst Med Trop Sao
Paulo 48, 239–244.
Popper, K. R. (1963). Conjectures and Refutations. London: Routledge
and Kegan Paul.
Safdar, N., Crnich, C. J. & Maki, D. G. (2005). The pathogenesis of
ventilator-associated pneumonia: its relevance to developing effective
strategies for prevention. Respir Care 50, 725–739.
Salyers, A. A. & Witt, D. D. (2005). Bacterial Pathogenesis. Washington,
DC: American Society for Microbiology.
Stevenson, M. A., Wilesmith, J. W., Ryan, J. B., Morris, R. S., Lockhart,
J. W., Lin, D. & Jackson, R. (2000).
Temporal aspects of the epidemic
of bovine spongiform encephalopathy in Great Britain: individual
animal-associated risk factors for the disease. Vet Rec 147, 349–354.
Yokota, S., Imagawa, T., Miyamae, T., Ito, S., Nakajima, S., Nezu, A. &
Mori, M. (2000).
Hypothetical pathophysiology of acute encephal-
opathy and encephalitis related to influenza virus infection and
hypothermia therapy. Pediatr Int 42, 197–203.
T. J. J. Inglis
1422 Journal of Medical Microbiology 56
