Monday, February 8, 2010
Descartes & Captain Pike
NEJM has a paper which pushes the last several years' worth of fMRI studies in patients in the PVS or MCS (persistent vegetative state; minimally conscious state) to the next level. (See here for other blog postings about this.) There have been a handful of reports in the last few years of patients clinically diagnosed as being in a PVS, who have some fMRI evidence of cortical activity (e.g. have 'normal' responses to hearing their name spoken, demonstrate normal cortical responses to instructions such as imagining they are playing tennis - by normal I mean show responses similar to healthy controls in that the appropriate areas of the motor cortex show increased activity, etc.). This study is builds on that, and the authors describe identifying one subject who could answer yes/no questions via fMRI.
Let me explain. First, the proof of principle. They had 16 healthy subjects imagine either playing tennis or walking around a familiar environment imaging what they'd see: reliably in these subjects these tasks would cause predictable, and predictably different, activity seen on fMRI (e.g. in the supplemental motor cortex for the motor task and in the parahippocampal gyrus for the spatial task). Then they told the subjects that they were to answer yes/no questions by, for instance, imagining the motor task for Yes and the spatial task for No. The healthy subjects could do this, and essentially they demonstrated you could use this technique to have people answer Y/N questions simply by having them imagine different things.
Next they tried the imaging tasks on 54 patients who were diagnosed as being in the PVS or MCS. 5 of them could do the imagining task (get results similar to the healthy subjects). Then they chose one patient to test the Y/N answering protocol (it is unclear to me why this specific patient was chosen or why they didn't investigate or didn't report their findings on the other four), and found that this patient could in fact give accurate Y/N answers to biographical questions (name of his father, etc.). He did this for 5/6 questions, repeatedly. This patient, a young man over 3 years out from a traumatic brain injury, was considered to be in a PVS (although the authors indicate some ambivalence about this - there seems to be some question if he was in a MCS: regardless, he was profoundly brain injured and would likely be considered by most to be in a PVS).
My thoughts on this: I think there is no doubt that the previously-established clinical diagnoses of PVS and MCS are going to be modified by this sort of information, and it seems apparent that a small percentage of patients diagnosed as being in a PVS actually have some cortical activity and awareness (on some level, in some way - see below). The sort of idea of the PVS being a 'clean,' homogeneous diagnostic and phenotypic category (no 'higher' cortical activity, no awareness of surroundings) clearly is wrong (at least for a few patients - this study would suggest 2-9%). However, until any of this translates into improved prognostic data (the ability to predict who, if anyone, has a chance of some recovery) or effective therapies are developed (perhaps aided by this technique to identify those patients like to benefit), the clinical situation hasn't really changed. And, as we've discussed on the blog before, the idea that there might be people misdiagnosed as being in the PVS doesn't change in any fundamental way the ethics of giving, withholding, or withdrawing life-prolonging medical treatments in these patients. The fundamental question remains would the patient her/himself want to be prolonged, with medical treatments, in such a state with such a prognosis, and allowing that that patient may have some awareness doesn't necessarily change one way or the other the decision-making. It might for some families/individuals, but not predictably. If you believe in the 'clean' PVS then you can argue that at least they're not suffering so why not continue; if you don't, then you have the added dilemma of a patient truly locked-in, unable to communicate thirst, pain, hunger, love, etc. and that's not necessarily 'better' than being 'vegetative.'
As perplexing and challenging as this research is, I think we have to keep this perspective at the forefront as these sorts of responses are inevitable. This research is important, and it might change things, but not in the way some automatically assume.
Of course the logical extension of this technology is The Captain Pike scenario (for those few 'PVS' patients who can respond); those patients, if felt to be sufficiently decisional, could theoritically communicate their own wishes.
How one would determine that - and the gulf between answering 5 biographical Y/N questions and establishing that a patient is 'conscious,' is discussed in an associated editorial, which has a more detailed exploration of how we really don't know what any of this means yet.
The mind is an emergent property of the brain and cannot be "seen" in images. The article by Monti et al is provocative; however, physicians and society are not ready for "I have brain activation, therefore I am." That would seriously put Descartes before the horse.(Image via Memory-Alpha.org, which uses a Creative Commons share-alike license.)
(Eric Widera at Geripal has also posted about this article; he has a nice summary of some of the media reporting, such as it is, about this paper.)