Enormous effort, by many people, has helped us understand how important good evidence of correlations is in choosing medical treatments, and has characterised how to go about carefully and responsibly performing randomised controlled trials (RCTs) and observational studies on various populations of people. We are trying to do a similar job for understanding – and also explicitly characterising – how evidence of mechanisms helps choose medical treatments.
An obvious example of evidence of mechanism is technological breakthrough that allows us to see aspects of what is happening in cells for the first time. A really famous case is Rosalind Franklin’s X-ray crystallography photograph of DNA, which helped us understand its structure: the famous Photograph 51, now the name of a London play starring Nicole Kidman. Understanding DNA is now crucial to understanding many mechanisms of disease, and cure.
But it is important to realise that many cases are not at all obvious. Our use even of the so-called ‘gold standard’ RCT is astonishingly integrated with discovering mechanisms of disease.
A simple way of seeing the connection is to see that they share a similar problem. First, when discovering a mechanism, we don’t know what the components are, and we have to find them. Philosopher Carl Craver (2007) gives a theoretical ‘mutual manipulability’ account of what the relevant components of mechanisms are, based on experimental practice for discovering mechanisms. Roughly, we affect (inhibit or excite) the functioning of the whole mechanism, and try to detect changes in possible components; and we affect the possible component, and try to detect changes in the functioning of the whole. Second, when running RCTs, we need to decide what we are going to measure – what will count as having the disease, and what will count as cure or improvement.
Actually, these problems are often addressed simultaneously, in work that narrows down possible components of mechanisms, suggesting drugs that are trialled in RCTs, moving back to the lab, and so on. This is very clear in published reviews and results of RCTs in breast cancer, which are always discussing lab work on cancerous cells.
Triple-negative breast cancer is really important, although it amounts to 15-20% of all cases of breast cancer, because it’s much more common in younger women, and it’s harder to treat successfully. If it weren’t for our understanding for the possible mechanisms of cure, we would never think to classify this particular subset of breast cancers in this way, and make a special effort to run RCTs on them: ‘Triple-negative breast cancers are defined as tumors that lack expression of estrogen receptor (ER), progesterone receptor (PR), and HER2.’ (Foulkes et al. p1938.) Why do we? Well, we can establish whether cells express these things from biopsies. And, while we have extra treatments which target these receptors, and really improve prognosis, those treatments are not effective for triple-negative cases.
So it’s particularly urgent to continue to continue to seek extra treatments for triple-negative cases. But we don’t really know what we are looking for, beyond a way to target those particular cancerous cells, without harming healthy cells. So we continue to use a combination of studying the cells for new components of mechanisms we can target, and looking for significant relationships in RCTs and observational studies studying this particular subgroup.
For example, hope was offered by lab work suggesting that inhibiting DNA repair could be effective, and a small phase 2 trial of iniparib, which inhibits usual DNA repair activity without serious toxic effects, seemed promising (O’Shaugnessy et al. 2011, p206). Sadly, a large phase 3 trial was very much less promising (O’Shaugnessy et al. 2014), illustrating how lab work, or even phase 2 trials, often fail to transfer to human populations.
The chances are we’re going to make progress by identifying further subtypes of triple-negative breast cancer. For example, there is a subgroup of women with triple-negative breast cancer whose tumours seem to be particularly sensitive to chemotherapy, so that it works very well for them, but we don’t really know why (Foulkes et al. p1945). Understanding this and developing new treatments will require laboratory work to discover drugs, which target particular mechanism components, but then have to be trialled in the far more complex context of real human populations, which are then interpreted to inform ongoing laboratory work, and so on.
Cancer scientists will keep at it. After all, this integrated, dynamic methodology is how we got effective drugs to successfully target breast cancer cells expressing the receptors above – significantly improving treatment of 80-85% of cases.
Craver: Explaining the Brain, 2007, Clarendon Press: Oxford.
Foulkes, Smith, and Reis-Filho: ‘Triple-Negative Breast Cancer’ in N Engl J Med 2010; 363:1938-48.
O’Shaughnessy, Osborne, Pippen, Yoffe, Patt, Rocha, Koo, Sherman, and Bradley: ‘Iniparib plus Chemotherapy in Metastatic Triple-Negative Breast Cancer’ in N Engl J Med 2011; 364:205-14.
O’Shaughnessy, Schwartzberg, Danso, Miller, Rugo, Neubauer, Robert, Hellerstedt, Saleh, Richards, Specht, Yardley, Carlson, Finn, Charpentier, Garcia-Ribas, and Winer: ‘Phase III Study of Iniparib Plus Gemcitabine and Carboplatin Versus Gemcitabine and Carboplatin in Patients With Metastatic Triple-Negative Breast Cancer’ in J Clin Oncol 2014, 32: 3840-3847.