Georg Ivanovas From Autism to Humanism - systems theory in medicine
Sometimes it is quite easy to understand when a therapy goes well. But often, especially in chronic diseases it seems to be difficult to judge all the changes occurring during the course of a therapy. This is even more difficult when the aim is not only to make symptoms go away (first order therapy), but when a fundamental change and an improvement of the robustness is intended. There might be an initial deterioration and inner rhythms might provoke sudden crises. But how can this be distinguished from a worsening of the situation? Even when the condition of the patient improves there might be some doubt whether the change is only a suppression which reduces robustness.
Most practitioners have a certain feeling for such processes. They possess a private theory, perceive patterns without a theoretical concept (chap. 4.9). This constitutes to a large extend what is called experience and intuition. The problem arises, however, when one tries to formulate a formal model of such developments, that is, when one tries to formulate the laws of an individual prognosis. According to my understanding this is only possible with a concept of ‘biological hierarchies’ and we are far from having such a concept (Redner 2008).
In technical systems the hierarchy is mostly modular. “The term modular is used to denote linearly ordered scalar hierarchies in which the characteristic scale of each successive level of the hierarchy is around one order of magnitude larger than the characteristic scale of the preceding level. The most basic examples are provided by number systems such as the decimal, with its successive units, tens, hundreds, thousands, etc. places….Modular hierarchies arise from the instance of an approximately linear relationship between the entropy and the number of equiprobable states of a system” (Smith J 2002).
The line of command in a linear and modular system is rather simple. The higher instance commands the lower one, just as in the army. Its hierarchy can be depicted in the form of a dendrogram. In contrast, a biological hierarchy is nonlinear and has recursive loops (Clauset et. al. 2008). Physiology processes dissolve into cybernetic functions with no centre or decision maker left. Although hierarchical structures might be found in certain traits the of sub-systems (TSH commands the thyroid gland), it makes no sense to say that TSH is higher in the hierarchy than thyreoglobuline as the thyroid gland is cyclic organised. Moreover, the heart is not higher in the hierarchy than the lungs, and so on. Despite the fact that a lot of control mechanisms (by far not all) are triggered by centres in the brain, it is inappropriate to say that the brain is the instance of control, either. By that the terms ‘brain’ or ‘centre’ would become explanatory principles (chap. 3.3). The inner clock, for example, is generated on the level of the clock genes with different centres in the brain involved processing daylight, meals etc. (chap. 4.7). But the brain does not ‘control’ the inner clock in the modular meaning. That is, the master clock of the hypothalamic suprachiasmatic nucleus is not a master in a classical sense, it is more a moderator or facilitator.
Also cognition possesses a kind of hierarchical structure involving different parts of the brain. More general traits are processed by many neurons, more specific traits by fewer (Blech 2008a). In the recognition of the actress Halle Berry, many neurons are connected to the trait ‘human, woman’, fewer to the colours of her body and her hair and even fewer to characteristic movements, etc. Finally, there is one neuron always active in the perception of Halle Berry independent from the form or context she is perceived. This neuron represents the so-called grandmother neuron of neuroscience. But this single cell, although at the top of the cognition - pyramid of specificity (Blech 2008a) does not do the work of cognition, nor does it command. The whole pattern is necessary (Quiroga et al 2008).
Moreover, the usual ‘command model’ of the nervous system saying that nerves interact through the release of neurotransmitters in synapses and thus somehow ‘command’ the following nerve seems not to be accurate, either. There exists a major extrasynaptic interaction between nerves. Parallel nerves influence each other biochemically without synapses. Thus, the function of the brain seems to be much more chaotic than previously thought (Kukley et al 2007). Although this recent finding has not been verified by other groups, the spatial biochemical coupling of nerves makes a lot of sense in a non-modular understanding of organisation.
The appearance of such nonlinear processes can be modelled by self-organization. But there remains an ontological problem, or more accurate, an ontological unease. Is there a higher pattern, a principle or force behind this self-organization? This question is the root of many philosophical speculations. For some time the idea prevailed that the genome might play the role of a ‚master plan’. But the correlation between a certain trait and a gene is often as low as 8 % for most of today’s ‘genetically caused diseases’ (PaísanRuiz et al 2004). As the RNA is able to restore a gene deleted in the former generation (chap. 4.3) there is clearly a wider organisation ‘commanding’ in a certain sense, even the DNA.
To model such and other principles of physiological organisation an own class of ‘biological hierarchies’ is necessary (Smith J 2002), a model based on the principles of cybernetics and general systems theory. It has to be teleological. That is, it has to assume that the organisation has a purpose and that the observed processes are not accidental. The main purpose is, of course, the autopoiesis of the living. Autopoiesis defines the reconstruction of the own structure and organisation as the central characteristic for living beings and implies a ‘downward causation’ as the necessary means to do so (chap. 4.8). This downward causalisation cannot have the form of a modular command. It is the equifinal ability of a system.
The teleological view implies that it is the ‘aim’ of the cybernetic circles to regulate the variables of the system against perturbations within a viability zone (Gershenson 2007: 2).
Thus, symptoms reveal the best possible reaction of the human under the given circumstances. Sometimes this ‘best reaction’ is harming like the lethal overreaction of the immune system in avian flu (chap 2.8.c), the occurrence of sepsis or an anaphylactic shock. In these cases an external stimulus is capable to induce the destruction of the whole organism as the metabolism cannot control anymore the always occurring tendency of cybernetic runaways (chap. 6.7).
But not every contact with a new germ and not every stimulus in allergy leads to death. This implies that the inner regulation is normally able to control such a process at a certain point. A simple swelling of the face after a wasp’s sting reflects a better internal control than an anaphylactic shock. Flushing or sweating under stress is less a threat than a stomach ulcer. If someone after a painful experience develops diarrhoea the person is in a better equilibrium than someone falling into depression.
This simple reflection enables already to formulate a basic theory of biological hierarchies. On the premise that the organism tries to maintain its integrity (in controlling overshoots) an organism is as healthier as it is able to control runaways early. Nevertheless, runaways have to occur to a certain extend otherwise the organism is rigid, has no rhythms and is not able to compensate inner and outer fluctuations. In contrast, a crisis which remains in a manageable range is an indication of a functioning adaptation, and in chronic disease it is a sign of recovery demonstrating the reappearance of inner rhythms (chap. 5.3.b).
With such a blueprint the physician is able to judge certain developments in the course of a therapy. When a patient with gastritis and depression (with suicidal thoughts) does better with his stomach, but remains unchanged in his depressive state the situation is no improvement. But an amelioration on the emotional level accompanied by a worsening of the gastritis would be an improvement of the whole situation. As long as the stomach pain remains in a manageable range and is of limited duration it might be regarded as an initial deterioration. That is, fluctuations and long term consequences have to be taken into account, in order to judge a present situation. For example, is a viral disease with the occurrence of high fiver for 2 or 3 days a good sign? Or is it better to have no fever at all, or to show only minor reactions (no ‘specific symptoms’) but to remain in a state of unease for weeks? From the point of view of adaptation the first choice is preferable.
That is, the physician has to understand the organism in its interconnectedness if he wants to judge the outcome of his intervention. He has to estimate in how far the intervention is beneficial for the patient in his integrity. The simple vanishing of a sign or symptom does not inform about the general condition of the human (chap. 2.5.e, 6.4, 6.6, 6.8).
Models of biological hierarchies taking the complexity of inner regulation into account exist already, at least in a rudimentary form. They are found in different methods of empirical medicine. Best known is the Hippocratic individual prognosis. Unfortunately there are only general guidelines which are not applicable to current medicine. More explicit are the models of acupuncture and homeopathy, which are presented in detail (appendix VI).
Such contributions of the empirical medicine might give an idea of how an individual prognosis and biological hierarchies could be modelled, how fluctuations and inner rhythms occur in the process of an individual therapy.
What would be the benefit of such a model? It would help to improve the general situation, the adaptability and robustness of the patient. It would enable the physician to understand which patient really profits from a therapy. That is, treatment could be much more individual. The current approach allows often only the statement that a patient has a 10% probability to benefit of a given treatment, the so-called NNT (the ‘number needed to treat’, chap. 2.1.f). Such a model might even be able to detect impending side effects of a therapy quite early. Experience and theoretical concepts like synergetics (chap. 4.11) suggest that the deterioration of the whole system has early first signs which could be detected with an appropriate model.