Can Central Pain be separated from peripheral activity?

Central and peripheral, somatic and autonomic, thalamic and cortical, sensory and motor. These modalities merge in a confusing pattern in CP.

Pet scans. Incredibly high technology. Absolutely revolutionary in opening up new knowledge about neuroanatomy. As to pain, however, the big problem is that PET scans for pain do not agree with each other, sometimes in the same patient. Pain, the supposedly monolithic entity, the discrete force that cannot be missed, is actually like minestrone soup. It has a lot of ingredients and we still have no good way to determine WHICH element of pain is dominant or present in any given PET scan.

At first, CP patients rejoiced at the PET. It was said to be able to SEE pain. At last, pain subjects could get the obnoxious accusers to be quiet, it seemed. They were in hopes of finally having something objective to defend against the doubters of pain, something they could even cram down their throats and force them to apologize for saying the whole thing was an exaggeration. It is bad enough to suffer terrible scathing pain without having some “expert” tell you that you are mistaken.

Unfortunately, pain is just too complex, and PET too new, for CP patients to finally become certified as bona fide pain subjects. The skepticism being a not insignificant part of suffering pain, we hope for better information, fast.

In a recent article, we reviewed Witting et al who showed NO thalamic activation in allodynic brushing. Since the thalamus traditionally has been considered the sensory relay, this was inconsistent with nearly every assumption about central pain. An axiom is that NOTHING sensory gets from cord to brain without going through thalamus. Yet, here we are, after Witting’s study, with a nonactivating thalamus and an orbitofrontal cortex which should accordingly show nothing at all.

That Witting found a quiet thalamus led us to postulate that pain signal passing through thalamus normally creates a model or template of something familiar, even if (as in CP) it shouldn’t. When there is signal according to the rules, perhaps even through quietude of the thalamus, which in that case would constitute a failure to alert pain, the orbitofrontal cortex is nevertheless alerted and is involved in central pain. This is the reverse of existing theory, so the OFC is going to get some more attention, you can be sure.

Something seems to be causing the brain to expect a pain signal to arrive from the thalamus, but when it never comes, the cortex concludes things are REALLY bad, and lets loose with the worst pain state known to mankind. Other explanations feel better, but they do not explain a quiet thalamus in the face of touch, particularly signal arriving from hypersensitized peripheral areas.

It may be comparable to an English speaking person hearing, say, Spanish. The speech is heard, but the meaning is missing. This could be disturbing. Perhaps the thalamus is speaking in a normal tone of voice, but in a different language. When something ought to be up, but isn’t, this negative message has meaning. All CP occurs ONLY when there is some sensory diminution. This fact may be our starting point for an answer to the mystery of the quiet thalamus.

There is a ton of bidirectional trafficking between thalamus and cortex and so it is not impossible that the orbitofrontal cortex activation represents determination to QUIET the thalamus, rather than being the mother lode, the tiger’s lair, of terrible pain. This is unlikely however, since there is presently no known corticothalamic tract by which the OFC may gain power over the thalamus. The OFC is a heavyweight in aversive learning, however, so perhaps it may have some switching function.

However, certain other central pain DOES show activation of the thalamus, in varying nuclei, with no two authors on PET studies in exact agreement, sometimes even in the same patient on two different occasiona. We laud their efforts and hope they refine things further and realize what all with central pain realize–that pain can be bewildering, so a muddled PET is consistent with the manifestations and expererience of central pain. Remember, doctors like to tack the term “bizarre” onto our dysesthesia, so what did they expect in the PET studies?

Traditionally, pain in the thalamus was thought to go via ventroposteromedial and ventroposterolateral nuclei (VPM and VPL). Recent work has focused more on the caudal nucleus (Vc). VMpo would be posterior to the VM. We apologize for not including an illustration but the newer, accurate maps of the thalamus are highly copyrighted. The ones that are available online, such as
do not even show the caudal nucleus, much less designate an area of grouped nuclei as ventromedial. One online drawing at does show the caudal nucleus in a cut through the thalamus, but unless you are an anatomist, this rendering may not help you get a good feel for relationships between the various nuclei. The position of the orbitofrontal cortex may be seen at ; however, its role in pain is a recent discovery, so much of the description of function is misleading.

A general atlas of the brain can be seen at McGill’s excellent site,

Another possibility is that evoked central pain is much more peripheral than we thought. The hypersensitized signal may be given a free pass through the thalamus because it has some kind of special permit, which causes no alarm until it reaches the cortex, where it wreaks havoc in sensitized cortical areas ALREADY producing spontaneous pain.

As stated, we have a problem. There ARE certain observations on PET in central pain, but what are these PET scans actually measuring? Is it central pain? Is it a failure which leads to central pain? Or what?

We have already pointed out that NORMAL pain sensation is allocated into at least three brain parts, the somatosensory cortex, which tells WHERE the stimulus is, the parietal operculum which tells WHAT kind of stimulus is present, and the insular cortex which tells the PAINFULNESS of pain.

It is well to remember that this triad of pain functions in normal pain may say NOTHING about central pain. After all, CP does not “feel” like ordinary pain.

Whether the thalamus is quiet or whether it activates may not determine in itself whether central pain will occur. It may be a marker, present some of the time and missing at others. Witting’s targetting of the OFC may well cause that area to be more carefully studied in future CP surveys.

We also have the problem that, increasingly, the existence of an evoked pain in most CP cases forces us to conclude that the peripheral nervous system is somehow involved. What other conclusion can we reach when evoked pain comes from stimulus of the peripherae. What kind of computational signal is going up in CP is not known, but there is clearly hypersensitization and neuroinflammation in the dorsal horn in CP. The magnitude is sky high, but what is the algorhythm, and what is the signal pattern which apparently can skate right on through the thalamus without creating a stir.

Evocation from light touch means that SOME peripheral signal can affect CP, but where the connection is remains a mystery.

Patients with chronic neuropathic pain (non-CRPS) and brush-evoked allodynia watched a reflected image of their corresponding but opposite skin region being brushed in a mirror. Unlike complex regional pain syndrome Type 1, this process did not evoke any sensation at the affected area (‘dysynchiria’). We conclude that central nociceptive sensitisation alone is not sufficient to cause dysynchiria in neuropathic pain. The results imply a difference in cortical pain processing between complex regional pain syndrome and other chronic neuropathic pain.

From the reverse perspective of peripheral neuropathic pain, as in CRPS (complex regional pain syndrome), we have the strange phenomeon of dysynchiria. This means that watching a mirror image of the skin on the unaffected side being brushed can evoke allodynia. This would suggest a CENTRAL component or even the primary nexus of peripheral neuropathic pain as being in the CENTRAL nervous system. CP patients do not do this, at least none have been reported. You could try to stimulate above the level of the lesion and watch in the mirror.

See eg. Acerra et al in Neurology. 2005 Sep 13;65(5):75 and also the comment on this in the same Neurology. 2005 Sep 13;65(5):666-7.

Also see Kramer et al in Eur J Pain. 2007 Apr 17; “Dysynchiria is not a common feature of neuropathic pain.”

Along the same lines, Dr. Marshall Devor has communicated to us that he now suspects there may be an element of peripheral dysfunction in the evoked pain of central pain. (They don’t come any smarter than Marshall Devor).

Many studies are indicating that we really don’t understand the function of the thalamus in central pain.

In the surveys, not a few of the CP patients have symptoms which are identical to CRPS, including skin atrophy at distal extremities, especially the feet, loss of hair on the area etc. This is supposed to be attributable to the autonomics, but really, there is no proof. It is true that some people with CRPS have positive thermograms, but others with identical symptoms do NOT have positive thermograms (temperature differences on the affected side). Calling all the different manifestations of CRPS by a number, eg CRPS type 1, does not really tell us what is going on. Shott, for example, thinks Central Pain is carried in visceral afferents which travel with blood vessels, thereby bypassing injured cord.

Now, back to the PET literature. Kim et al in J Neurosci. 2007 May 2;27(18):4995-5004 examined by PET patients with post stroke pain (which is central pain). Remember these patients will usually have unilateral pain on the side OPPOSITE to the stroke.

Kim found that:

1) thalamic lesions must extend posterior to the ventral caudal nucleus (Vc) and include ventral medial posterior nucleus (VMpo), to result in loss of cold sensibility and PSCP.

2) All lesions associated with PSCP involved posterior Vc;

3) Two lesions also involved nuclei posterior to Vc, but not VMpo.

4) ALL patients tested had alterations of cold pain sensation and tactile sensation, as measured by von Frey hairs. Three patients had altered cool sensation, and the patient with the least involvement of Vc had normal cool thresholds, suggesting that a critical volume of Vc must be involved before cool sensation is impaired.

5) Perception of warm was impaired only in lesions involving nuclei posterior to Vc. Heat pain perception was never affected.

5) In a subject with cold allodynia, a single-subject protocol PET study measured the responses to immersion of either hand in a 20 degrees C waterbath. The scan during stimulation of the affected hand was characterized by intense activation of contralateral sensorimotor cortex.

“Sensorimotor cortex” or S1 plus M1 means activity on BOTH sides of the central sulcus, or in other words S1 which is known as the primary somatosensory cortex and is in back of the groove, and M1 or the primary motor cortex which sits just in front of the groove. These areas meet at the base of the groove or central sulcus, with sensory touching motor at the center of the base. The central sulcus is a groove which runs transversely across the brain at about the top of the head. It is a landmark in neuroanatomy. To give you an idea why MRI cannot see the ST tracts, neuroradiologists are not always accurate even in identifying the comparatively huge central sulcus.

Kim concluded that “Therefore, there are modality-specific subnuclear structures in the posterior thalamus, but lesions of Vc not involving VMpo are sufficient to impair cold sensibility and to PRODUCE (our caps) CPSP.”

Our question is with his word “produce”. No one actually knows what produces CP. Witting found the thalamus could be QUIET during CP. KIm here found it active.

So what is the thalamus actually doing with signal. The fact that one patient responded to the cold water bath by intense activation of the contralateral sensorimotor cortex COULD mean merely that the thalamus was registering the cold and passing it along. Was this patient’s thalamus fighting an uphill battle to get through an injured thalamus and therefore fired intensely to reach S1, or was the intensity actually a reflection of the hyperpathia?

We don’t know and neither does anyone else and that is what is troubling us. witting’s study of the OFC must be repeated to see if we must now take the orbitofrontal cortex into account. The mere location of a cold sensation to the hand does not necessarily equal central pain. It only tells WHERE, as the S1 does. What thalamic firing or not firing does equal has yet to be determined.

We need more PET studies and also tensor MRI analysis to see flow in the tracts. A huge amount of brain matter is devoted to aversive learning because the noting of pain, and deriving information relative to pain, is so essential to survival.

One begins to suspect that aversive learning is far from simple, and that the thalamus, if it fails to activate when it normally should, in allodynic brushing, tells us there is much to be learned about cortical activity, as well as peripheral input in central pain.

At this time, both PET and tensor MRI would be necessary to really evaluate patients. Of course we could just take their word for it, but no one is able to do that, all pain specialists being mules from Missouri, the “show me” state. Of course scientists would like to know for academic purposes, and these people DO take pain patients seriously. They do it all the time with rats so why doubt humans?

Knowing the area of skin where the stimulus is generated (perhaps determined by SI, (as it normally should be) does NOT tell us how any given signal comes to be central pain.

In the surveys, we know that patients can be overwhelmed by the noise of the MRI magnet, the hyperpathia of a failed or difficult IV attempt etc, or may even be troubled by one central pain more than another. The starting point must be a good history, taken in the framework of knowledge about CP. Only then, can the PET be interpreted. It would be nice if the difficult background were not necessary, but when it comes to PAIN you are talking about a vast, comprehensive assemblage of brain operations. It will take time to pinpoint the actual PAIN, but thanks to the neuroradiologists, we are getting there. We are just as happy with negative PET results as positive ones, because it corrects mistaken assumptions. We are also not unhappy with PET variations in the same patient because any person with CP can tell you that sometimes it is worse than others, sometimes different, and sometimes something you cannot explain.

We are very glad to see Kim’s attempt to do dynamic PET, and hope that others will follow his example, and also be aware of time dynamics. There are complicated potentiations going on and disappearing, so we hope to see PET carefully quantified as to time elapsed from stimulus. The decay period is just as interesting as the incremental increase period, when it comes to evoked central pain.