Diseases are increasingly researched on the molecular level as well as redescribed and explained by way of molecular mechanisms. One result is that what before was considered one or a few diseases are split into a multitude of diseases or disease subtypes in a fine-grained fashion according to molecular variations, mutations, or gene expressions.
Cancer is traditionally classified according to actual tumor site, or to the site of cancer origin. In contrast, there is a trend in contemporary cancer research of categorizing cancer types and subtypes in terms of genetic mutations, receptor types or other molecular level features. There is an ambition to understand and explain cancer development in a more fine-grained manner, and to develop new and better cancer treatments.
We investigate this change in description and understanding of cancer with an attention to the separation between treatment focus (the goal to treat cancer) and explanatory focus (the goal to find and understand cancer causing mechanisms). We shed light on how these different angles, although they often go hand in hand, may give rise to different takes on cancer categorization where the latter is more substantial and the former more instrumental than the other. We discuss how categorizations of cancer subtypes often are based on the possibility of diagnosis and intervention which may or may not overlap with the mechanism of disease development.
Targeted cancer therapies may intervene directly in disease mechanisms, they may intervene via a disease cause without intervening directly in the mechanism in which this cause plays a role, or they may intervene via a factor not directly related to the disease process at all.
In research on nano-carriers, the more relevant situation seems to be the second case, where a causal factor in cancer development is used as a handle (to attach the carrier to the cell) but where the drug works on a relatively coarse level by killing the cell when it divides. It will vary to what degree the causal factor used as a handle is implied in cancer development (and this does not really matter), and although differences between cells when it comes to such handles represent actual differences between cancer cells, these differences may or may not correspond to differences in developmental history and prognostic relevance of the different cells.
In principle, variation in biomarkers can be disconnected from the cancer causing mechanisms in the various cells and clones due to effects of passenger mutations. On this background, we argue that molecular categorization of new cancers or cancer subtypes based on genetic variation and thereby variation in targetable ligands and biomarkers, is mainly instrumental in the sense of helping diagnosis and treatment, and is not necessarily based on features explaining or reflecting the origin and development of the disease.
The current classification system in psychiatry (as exemplified by the DSM) exhibits severe problems, and its recent revision, culminating in the DSM-5, has left many disappointed. On the one hand, there are controversial debates on the criteria for individual diagnoses and the question, whether they pathologize normal feelings and behavior, for example in the cases of ADHD or depression. On the other hand, there are also numerous critics who question the overall system. It is often argued that the heterogeneity of groups picked out by the DSM’s polythetic criteria, the excessive rates of comorbidity, and the lack of predictive success of the DSM diagnoses indicate a severe lack of validity. This lack is commonly attributed to the DSM’s phenomenological approach to classification, the current system being based on co-occuring, observable symptoms. The main proposal for improving the situation is to change the classification from a phenomenology-based one to an etiology-based one that groups symptoms according to our best scientific theories about their underlying causes. Proponents of such an etiological revolution often present it as a move forward towards a more scientific, evidence-based nosology. Even more cautious criticisms often seem to assume that the change towards a more valid etiological classification is only a matter of time, awaiting further research results.
What I want to show is, first, that the question of the classificatory basis is not one that can be answered by empirical evidence alone. Instead, it requires judgments on what level of evidence is needed to justify changes as well as judgments on what kind of evidence is most important. Second, in making these judgments we need to weigh the needs of clinical practice and scientific research.
Regarding the question of how much evidence is enough to legitimate a more radical revision, it is important to note the DSM’s multiple purposes. While it aims to be a suitable basis for research, it also thoroughly shapes psychiatric practice. Changes can be very consequential in that they affect patient’s diagnoses and possibly treatment, might impact questions of reimbursement, and even change public views of mental disorder and normality. Therefore, before one starts a revolution, there should be solid evidence that this will improve the situation in terms of science as well as health care. What exactly that means is moreover not a purely scientific question but calls for value-judgments on the weighing of inductive risks and consequences of possible errors.
In consequence, the needs of clinical practice and of scientific research do at present stand in conflict with each other. While the former calls for a conservative approach and high standards on evidence before every radical change, the DSM’s problems as a basis of scientific research are indeed severe and call for pluralistic explorations of possible alternatives and causal explanations. The central difficulty in psychiatric classification is, accordingly, not just a lack of validity or a lack of evidence, but lies in trade-offs between the different demands of research and practice.
Philosophers interested in the role of values in science have focused much attention on the argument from inductive risk. In the 1950s and 1960s, a number of authors argued that value judgments play an ineliminable role in the acceptance or rejection of hypotheses (Hempel 1965; Rudner 1953). No hypothesis is ever verified with certainty, and so a decision to accept or reject a hypothesis depends upon whether the evidence is sufficiently strong. But whether the evidence is sufficiently strong depends upon the consequences (including ethical consequences) of making a mistake in accepting or rejecting the hypothesis. Recent philosophers of science have not only revived this argument; they have also extended it. While Rudner and Hempel focused on one point in the appraisal process where there is inductive risk – namely, the decision of how much evidence is enough to accept or reject a hypothesis – more recent philosophers of science have argued that there is inductive risk at multiple points in the research process. Douglas (2000) argues that inductive risk is present in the choice of methodology (e.g., the setting of a level of statistical significance), the characterization of evidence, and the interpretation of data. Wilholt (2009) argues that there is inductive risk in the choice of model organism. The upshot of these and other extensions of the Hempelian/Rudnerian argument from inductive risk – which I will call the classical argument from inductive risk – is, again, that the research process is shot through with inductive risk. Indeed, it can seem that, as a result of these extensions, there is inductive risk at any point in the research process at which a decision must be made.
While I applaud the extensions of the classical argument from inductive risk, and while I think that they provide valuable insights into the ways in which value judgments operate in the appraisal of research, I will argue that some of the purported extensions of the classical argument do not fit cleanly within the schema of the original argument and that, for the sake of conceptual clarity, they should simply be treated as different arguments. I will discuss the growing problem of overdiagnosis of disease due to expanded disease definitions in order to show that there are some risks in the research process that are important – and that should be taken seriously by philosophers of science – that very clearly fall outside of the domain of inductive risk. Finally, I will introduce the notion of epistemic risk as a means of characterizing such risks. This more fine-grained taxonomy of risks in the research process will help to clarify the different roles that values can play in science.