I am relatively new to the field of pain neurophysiology. Before joining the NocionsLab in January 2014, I had been working on brain-computer interfaces – a topic that involved the use of similar neuroimaging techniques (EEG, fMRI, TMS), for fairly different purposes.
Something that struck me almost right away about the study of pain neurophysiology was how lively – to use a euphemism – were the ongoing debates within the field. Just to mention a few: Are there pain-specific regions in the brain? Can we still talk about a pain matrix? And what about the overlap between social rejection and brain regions activated by pain? Researchers in the field all appeared to have very strong and incontrovertible – yet conflicting – responses to each of these questions. For the first time, my colleagues introduced me to a world of heated journal correspondences, quarrelsome conference symposia, and particularly stinging peer reviews.
This was certainly very different from what I was used to in the BCI field, which (at least to me) always appeared as a rather cozy environment, in which ‘whatever works’ (i.e. whatever brain state classification method reaches a ‘good’ accuracy) was okay.
My first reaction was undoubtedly of excitement. I always thought that discussion and confrontation are the best way to progress within a field. And yes, also disagreement can trigger interesting considerations and ideas.
What took me a while to grasp was that the new direction of my own research – using depth electrodes implanted in the human insula to investigate the responses to different kinds of nociceptive and non-nociceptive stimuli – would have catapulted me right into the eye of the hurricane.
My first pain paper, “Nociceptive Local Field Potentials Recorded from the Human Insula Are Not Specific for Nociception” was published in PLoS Biology last January (it was actually accepted in November). In brief, using depth intracerebral recordings, we showed that nociceptive (CO2 laser) and non-nociceptive (vibrotactile, auditory, and visual) stimuli elicit very similar responses (local field potentials, LFPs), appearing as large biphasic waves, in both the anterior and posterior human insula.
Nociceptive and non-nociceptive responses also appear to originate from the same sources, and can be mainly explained by multimodal activity. In previous literature (e.g. https://www.ncbi.nlm.nih.gov/pubmed/12538410, https://www.ncbi.nlm.nih.gov/pubmed/24916602), these insular responses were often considered as a signature of pain perception and its modulation, and were therefore used to draw strong conclusions about pain processing. The main results of our study confute this widespread assumption. Moreover, they cast doubt on the possibility of finding a so-called “primary cortex for pain” within the insula. Indeed, it is true that we can observe patterns of activity for each sense in a given primary cortex, but the activity relative to the sense for which a primary cortex si specific is always predominant. It is also true that the results of our study do not exclude the possibility of finding nociceptive-specific (or pain-specific) neurons within the insula. However, what we could not observe is a specific “spot” mainly dedicated to nociception and/or pain.
I was aware that my main results would have been greeted with mixed feelings. The typical commenters of my conference presentations (including OHBM, NeuPsig and EFIC) belonged to three main categories:
1) The bored ones:
- «So what? We have known that the insula is multimodal since decades!»
- «Again with this specificity issue? I thought we were over it!»
2) The skeptics:
- «This is simply not possible; you must be doing something wrong with your analysis!»
- «You are ignoring that so many researchers much more experienced than you have demonstrated the opposite»
- «You simply have a negative result, i.e. not finding a difference between nociceptive and non-nociceptive responses».
3) The enthusiasts (these were actually quite a large number):
- «Oh, finally some results that will help toning down all that pain-specificity delusion!»
Regardless our excitement, the editors of the journals to which we first submitted our manuscript all seemed to belong to the bored category. I got a number of desk rejections which all sounded more or less like “your data is not really striking enough to fit into our high level journal”. In fact, the hard part was managing to get our paper out for review. The nicer part is that, as soon as it did go out for review, it was accepted after the first round of revision.
As I am all for transparency, here you can read the reviewers’ comments, together with our point-by-point reply. Again, you can see how the comments included very mixed opinions, from “very interesting work”, to “this data is not convincing even to a believer”, or “we knew this already”. I think that the reviewers’ comments, even the more critical ones, really helped us to strengthen the paper and make our message clearer. This was probably one of my most satisfying rounds of review.
When the paper finally came out, I was impressed by the fact that people were sharing it on Twitter and Facebook within hours after its publication. The reactions appeared to be more positive than negative, at least on Twitter. I often joked about how there must be some sort of bias: aren’t nasty reviewers ever on Twitter? In general, I could see huge interest in our results. Neuroskeptic wrote a very to the point post about the article, linking it to the “myth” of the pain matrix. (More recently, even Gizmodo featured a post describing, together with other studies, my work). Several people contacted me to congratulate me for the new publication, as well as to ask questions. I had some very interesting message exchanges with researchers who wanted to perform related studies, and with researchers who were discussing the articles in their labs (e.g. during journal clubs).
Ok, I cannot say that there weren’t also some criticisms…
@70Hertz it might help if you stick your electrode in the dpIns, bit of a misrepresentation otherwise…..
— Jonathan Brooks (@LexAtom) January 7, 2016
Here is a list of FAQ, which I thought could be interesting to answer more extensively in a post.
Why didn’t you record from the dorsal posterior insula (dpIns), as in the recent study by Segerdahl et al. published on Nature Neuroscience? That is supposed to be the real “ouch zone” within the insula!
It is true that at the time the paper was submitted, none of the recruited patients had electrodes implanted in the dpIns. As researchers, we obviously cannot decide which regions of the insula will be sampled using depth electrodes. Indeed, this was a reasonable criticism of our study, in which otherwise wide portions of the anterior and posterior insula were investigated.
What is important to take into account, however, is that the LFPs we recorded in the other portions of the insula are in no way different (in terms of latency and morphology) from the ones recorded in the dpIns in other studies (see this paper by Frot et al.). It is therefore quite reasonable to believe (although at this point we still could not prove it), that these LFPs represent the same kind of activity.
Since our recordings are ongoing, more recently, we were actually able to record stimulus-evoked responses also in a patient who had an electrode implanted in the dpIns (work under review). Not surprisingly, we observed both nociceptive and non-nociceptive LFPs in the dpIns, exactly like in the other portions of the insula.
(Regarding the paper by Segerdahl et al. about a specific involvement of the dpIns in the perception of pain, André Mouraux and I already expressed our criticisms here, as a comment to a related letter written by Karen Davis et al. I find this debate on f1000 particularly interesting and really recommend reading it.)
How can you exclude the presence of pain-specific neurons within the insula?
I cannot. In no way, the recording of LFPs can allow drawing conclusions on the presence/absence of pain-specific or nociceptive-specific neurons. All we can say is that, at the macroscopic level, we do not observe any compact “spot” in the insula which is exclusively dedicated to pain or nociception.
LFPs elicited by stimuli from different modalities have different latencies. How can this be, given that you write that they reflect the same type of activity?
The differences in latencies can be explained by the difference in time required by the sensory afferent volleys to reach the cortex. For instance, the greater latency of the nociceptive responses compared to the vibrotactile responses can be explained by the fact that small diameter A∂ fibers conveying nociceptive input have a slower conduction velocity than large-diameter aβ fibers conveying vibrotactile input.
How does the PICA work exactly, and why doesn’t it differentiate LFPs with different latencies?
Indeed, the time course of multimodal neural activity contributing to the different sensory modalities is not identical because of physiological reasons. PICA isolates spatially independent sources of activity contributing to the LFPs, regardless of their time course. I think that this is explained particularly well in this paper by Mouraux & Iannetti (2009).
As I said, overall, I received very positive feedback about my work. I have also received harsher comments, mostly by researchers who disagree with my main conclusions (i.e., there is no primary cortex for pain in the insula). I was a bit upset when one researcher told me “I could not finish reading your paper because I found it too irritating, with no space for further interpretation”.
Well, I might be sometimes quite determined about my ideas, but I do love discussion.