IS6 tion the neighbor asked, “But why, then. are we searching here?“. Replied the Mullah. “Because there is more light here”. We would like to conclude our response to Dr. Kruger’s letter by reiterating in agreement and endorsement of his own ‘final words‘. “If we want public support and understanding of our activities. are we not responsible for selecting the most reliable (and we would add. reler~unrl experimental models”.
Bennett Blumenkopf Jonathan J. Lipman Department of Neurological Surgery Vunderhilt LbCersity Medical Center NushGlle. TN 37232, USA
PAIN 02065
Autotomy Kruger
sense
and nonsense.
Reply to L.
Since it was first described. well over 100 studies have been published on autotomy, and the cumulative evidence linking it to abnormal sensory experience is strong, albeit circumstantial. To mention just one example, resection of the neuroma, even before autotomy actually begins, delays its onset and development by the amount of time required for a new neuroma to form (Seltzer et al. 1991). Rather than attempt an overall summary, I would like to point out a few misconceptions concerning the autotomy model that are aired from time to time. (1) Phantoms and phantom pains are very common in amputees and patients with total limb denervation (e.g., Sherman et al. 1984). However, for obvious reasons. humans do not often, and are nor expected to intentionally chew their anaesthetic limb (autotomy). There are exceptions, however: very young children and the severely retarded (e.g., Lesch-Nyhan syndrome). Rats are less aware of consequences than are people. (2) Likewise, even in the presence of intense dysaesthesias, autotomy is expected to be minimal when there is hyperalgesia. Scratching and biting of the limb would hurt! It should only be prominent when dysaesthesia is referred to an anaesthetic limb (as in anaesthesia dolorosal. What Blumenkopf and Lipman (19911 showed is that anaesthesia alone, without dysaesthesia, is not sufficient. Dr. Kruger seems not to be aware that according to Bennett and Xie (1988). localized autotomy occurs in about 70% of rats with nerve constriction hyperalgesia. Although it was not checked, I venture that in these animals it was largely restricted to patches of skin rendered anaesthetic by the lesion. (3) Dorsal rhizotomy triggers autotomy, probably because of dysaesthesias originating in, the deafferentated CNS (as in plexus avulsion injuries, for example). Peripheral nerve section also triggers autotomy. However. it does riot follow that after peripheral injury the underlying dysaesthesias must originate in the CNS. Ectopic afferent discharge is known to develop in the PNS after nerve injury, and there is a good deal of evidence linking it to the sort of sensations that probably trigger autotomy in nerve-injured rats. Certainly in man, the intensity of ectopic PNS activity correlates closely with the intensity of neuropathic dysaesthesias (e.g., Nordin et al. 1984). On the other hand, I have little doubt that in both man and animal there is also a CNS contribution. Most autotomy re-
search is aimed at understanding the causes of neuropathic pain, and these are probably peripheral and central. Depending on the dose used. ricin applied to major nerves kills associated dorsal root ganglia, producing the equivalent of dorsal rhizotomy. Thus, ricin deafferentation is expected to trigger autotomy, and it does. This observation certainly does not prove that ectopic PNS activity is irrelevant. (4) In autotomy studies, the foot is totally denervated, and consequently the self-inflicted toe lesions are necessarily painless. Only if autotomy indeed reflects ongoing neuropathic pain is there an ethical issue. and this is no different from the partial nerve injury models in which hyperalgesia occurs in conjunction with ongoing pain. The models are different and complementary. In both models the justification is the clear promise that this research holds out for pain patients. Research on rodents with complete transection of hind limb nerves has produced a wealth of new and relevant knowledge on topics such as ectopic neural discharge, CNS changes, heritability and othera. While we should certainly be sensitive to, and attempt to minimize, animal suffering, our prime concern must remain humans. If the anti-vivisection movement should force us into a PR battle on autotomy, we halv the wherewithal1 to respond.
References Bennett, G.J. and Xie, Y.K.. A peripheral mononeuropathy in rats that produces disorders of pain sensation like those seen in man, Pain, 33 (1988) 87-107. Blumenkopf. & and Lipman, J.J.. Studies in autotomy: its pathophysiology and usefulness as a model of chronic pain. Pain, 45 (1991) 203-210. Nordin, M., Nystrom, B., Wallin, U. and Hagbarth, K.-E., Ectopic sensory discharges and paresthesiae in patients with disorders of peripheral nerves, dorsal roots and dorsal columns, Pain, 20 (1984) 231-245. Seltzer, Z., Paran, Y., Eisen. A. and Ginzburg, R., Neuropathic pain behavior in rats depends on the afferent input from nerve-end neuroma including histamine-sensitive C-fibers, Neurosci. Lett., 128 (1991) 203-206. Sherman. R., Sherman, C. and Parker, L., Chronic phantom and stump pain among American veterans: results of a survey, Pain 18 (1984) 83-95.
Marshall
Devor
Life Sciences Institute Hebrew lJnic,ersity of Jerusalem Jerusalem 91904. Israel
PAIN 02066
Comments on Marchand (1991) 249-257
et al.,
PAIN,
Dr. Marchand and co-workers have made an important tion to the study of the psychophysical effects of spinal cord tion (SCS) for the relief of pain in man. By applying contemporary sensory testing methods, they have been able previously unappreciated phenomena.
45
contribustimulasensitive to detect
157 Using sensitive methods for testing temperature discrimination, the authors have shown significant decrements in temperature discrimination, heat pain thresholds and heat pain ratings within the projected area of spinal cord stimulation paresthesias. The magnitude of the effect on experimental pain ratings is calculated to be 20-28% and is reported to be comparable to the effects of stimulation upon patients’ spontaneous clinical pain ratings. This comparison is awkward, however, as is the inference that these effects have a common mechanism: “ since we observed comparable DCS-related modulation of the cutaneous heat pain and the varied clinical pain, such changes probably affect cutaneous and deep nociceptors similarly”. The goal of clinical application of spinal cord stimulation is the relief of chronic pain, which may relate to neural injury or deafferentation rather than nociception. Incremental temperature sensibility and acute heat pain threshold measurements with 2-5 set stimuli, as employed in this protocol - and even sustained presentation of painful thermal stimuli - may not be a relevant model (Gracely 1991) for therapeutic application of SCS. In fact, effects of SCS or any other pain relieving measure upon biologically useful pain or sensibility are not desirable in a chronic, clinical application, When we examined the effects of SCS upon temperature discrimination, 2-point discrimination, and light touch in a quantitative fashion (North et al. 19771, albeit using less sensitive measures than in the present study, we took a different perspective; we concluded that there was no clinically significant effect upon these important, normal sensory functions. Patient reports of impairment of normal sensory function by SCS are unusual. Clinical pain ratings in this study understandably differed from patients’ global ratings of pain and its relief by stimulation during everyday activities, as these ratings were obtained only with the subjects at rest, in an experimental setting. As the authors point out, patient’s overall ratings of everyday pain relief by SCS may reflect relief from pain at its highest intensity (beyond the scale displayed on the ordinate in Fig. 1). In accord with the authors’ premise, as we have reported (North et al. 1977, 1991a), the average SCS patient reports marked reduction in the percentage of time at the maximum pain intensities on the McGill Pain Questionnaire. Other phenomena may contribute further to the observed discrepancy (North et al. 1977, 1991a): patients commonly report a latency to pain relief, from the time stimulation begins, in excess of the 30 min allowed in this protocol, and they sometimes report persistence of relief, after stimulation ends, in excess of the 8 h required by the protocol. Details of the stimulation parameters used in this study, and of the relationship between the effects observed and the extent of overlap of pain by paresthesias, were not specified and would be of interest. The authors implicitly recognize the importance of overlap of each patient’s topography of pain by stimulation paresthesias as a necessary condition for obtaining pain relief. Patients with the most successful stimulator implants, from a technical point of view, typically would have the smallest areas of extraneous stimulation paresthesias, perceived outside the area of pain, within which the effects described in this paper could be studied. The study population includes an indeterminate number of patients with estimated pain relief as low as 30% - a failure by the clinical criteria used in all reported series of SCS patients. The authors conclude that “even with patients selected for their successful use of DCS, the clinical and experimental pain perception is reduced only a small amount. Thus, these relatively small reductions must be weighed against the invasive nature of electrode implantation”. The terms ‘relatively small’ imply some basis for comparison, but none is given. As the patients studied are not all ‘successful’ by standard criteria and as the experimental measures chosen are of questionable relevance to routine clinical application of spinal cord stimulation, this conclusion is not supported. It has been our experience that the benefits as well as the risks of spinal cord stimulation compare quite favorably with other invasive proce-
dures for chronic, intractable lumbar pain syndromes (North et al. 1991al. Over the past 20 years, we have observed no major morbidity (threatening life or neurologic function) in our SCS patients at Johns Hopkins (North et al. 1977, 1991a, b; Long et al. 1981); with contemporary percutaneous methods of electrode insertion (which may or may not have been used in this series) the risks are lower than ever. ‘Invasive’ treatments are not uniformly or inherently more morbid than non-invasive, medical therapy, viz., the well-known, occasionally lethal complications of treatment with opiates, tricyclics, or anticonvulsants. The techniques reported by Dr. Marchand et al. are an important addition to our methods for studying acute pain. As they are refined and expanded and as their relevance to chronic pain is clarified, their application to other pain relieving techniques will be of considerable interest.
References Gracely, R.H., Experimental pain models. In: M.B. Max, R.K. Portenoy and E.M. Laska (Eds.), Advances in Pain Research and Therapy. Vol. 18. The Design of Analgesic Clinical Trials, Raven Press, New York, 1991. Long, D.M., Erickson, D., Campbell, J. and North, R., Electrical stimulation of the spinal cord and peripheral nerves for pain control, Appl. Neurophysiol., 44 (1981) 207-217. North, R.B., Fischell, T.A. and Long, D.M., Chronic stimulation via percutaneously inserted epidural electrodes, Neurosurgery, 1 (1977) 215-218. North, R.B., Ewend, M.G., Lawton, M.T. and Piantadosi, S., Spinal cord stimulation for chronic, intractable pain: superiority of ‘multichannel’ devices, Pain, 44 (1991a) 119-130. North. R.B., Ewend, M.G., Lawton, M.T., Kidd, D.H. and Piantadosi, S., Failed back surgery syndrome: 5-year follow-up after spinal cord stimulator implantation, Neurosurgery, 28 (1991b) 692-699.
Richard B. North Department of Neurosurgery Johns Hopkins 600 North Wolfe Street Baltimore, MD 21205, USA
PAIN 02067
Reply to Dr. R.B. North Dr. North makes several important comments about our recent article (Marchand et al. 1991). We agree with Dr. North that experimental measures of cutaneous nociception are not the best models of chronic pain syndromes that may involve neural injury or deafferentation. However, our study addressed an intriguing puzzle about spinal cord stimulation, i.e., although physiological data in humans and animals indicate that dorsal column stimulation alters activity in normal nociceptive pathways (Hillman and Wall 1969; Brown and Martin 1973; Feldman 1975; Handwerker et al. 1975; Foreman et al. 1976; Lindblom et al. 19771, clinical studies have not confirmed that normal nociception is altered by such stimulation (Lindblom and
Bennett Blumenkopf Jonathan J. Lipman Department of Neurological Surgery Vunderhilt LbCersity Medical Center NushGlle. TN 37232, USA
PAIN 02065
Autotomy Kruger
sense
and nonsense.
Reply to L.
Since it was first described. well over 100 studies have been published on autotomy, and the cumulative evidence linking it to abnormal sensory experience is strong, albeit circumstantial. To mention just one example, resection of the neuroma, even before autotomy actually begins, delays its onset and development by the amount of time required for a new neuroma to form (Seltzer et al. 1991). Rather than attempt an overall summary, I would like to point out a few misconceptions concerning the autotomy model that are aired from time to time. (1) Phantoms and phantom pains are very common in amputees and patients with total limb denervation (e.g., Sherman et al. 1984). However, for obvious reasons. humans do not often, and are nor expected to intentionally chew their anaesthetic limb (autotomy). There are exceptions, however: very young children and the severely retarded (e.g., Lesch-Nyhan syndrome). Rats are less aware of consequences than are people. (2) Likewise, even in the presence of intense dysaesthesias, autotomy is expected to be minimal when there is hyperalgesia. Scratching and biting of the limb would hurt! It should only be prominent when dysaesthesia is referred to an anaesthetic limb (as in anaesthesia dolorosal. What Blumenkopf and Lipman (19911 showed is that anaesthesia alone, without dysaesthesia, is not sufficient. Dr. Kruger seems not to be aware that according to Bennett and Xie (1988). localized autotomy occurs in about 70% of rats with nerve constriction hyperalgesia. Although it was not checked, I venture that in these animals it was largely restricted to patches of skin rendered anaesthetic by the lesion. (3) Dorsal rhizotomy triggers autotomy, probably because of dysaesthesias originating in, the deafferentated CNS (as in plexus avulsion injuries, for example). Peripheral nerve section also triggers autotomy. However. it does riot follow that after peripheral injury the underlying dysaesthesias must originate in the CNS. Ectopic afferent discharge is known to develop in the PNS after nerve injury, and there is a good deal of evidence linking it to the sort of sensations that probably trigger autotomy in nerve-injured rats. Certainly in man, the intensity of ectopic PNS activity correlates closely with the intensity of neuropathic dysaesthesias (e.g., Nordin et al. 1984). On the other hand, I have little doubt that in both man and animal there is also a CNS contribution. Most autotomy re-
search is aimed at understanding the causes of neuropathic pain, and these are probably peripheral and central. Depending on the dose used. ricin applied to major nerves kills associated dorsal root ganglia, producing the equivalent of dorsal rhizotomy. Thus, ricin deafferentation is expected to trigger autotomy, and it does. This observation certainly does not prove that ectopic PNS activity is irrelevant. (4) In autotomy studies, the foot is totally denervated, and consequently the self-inflicted toe lesions are necessarily painless. Only if autotomy indeed reflects ongoing neuropathic pain is there an ethical issue. and this is no different from the partial nerve injury models in which hyperalgesia occurs in conjunction with ongoing pain. The models are different and complementary. In both models the justification is the clear promise that this research holds out for pain patients. Research on rodents with complete transection of hind limb nerves has produced a wealth of new and relevant knowledge on topics such as ectopic neural discharge, CNS changes, heritability and othera. While we should certainly be sensitive to, and attempt to minimize, animal suffering, our prime concern must remain humans. If the anti-vivisection movement should force us into a PR battle on autotomy, we halv the wherewithal1 to respond.
References Bennett, G.J. and Xie, Y.K.. A peripheral mononeuropathy in rats that produces disorders of pain sensation like those seen in man, Pain, 33 (1988) 87-107. Blumenkopf. & and Lipman, J.J.. Studies in autotomy: its pathophysiology and usefulness as a model of chronic pain. Pain, 45 (1991) 203-210. Nordin, M., Nystrom, B., Wallin, U. and Hagbarth, K.-E., Ectopic sensory discharges and paresthesiae in patients with disorders of peripheral nerves, dorsal roots and dorsal columns, Pain, 20 (1984) 231-245. Seltzer, Z., Paran, Y., Eisen. A. and Ginzburg, R., Neuropathic pain behavior in rats depends on the afferent input from nerve-end neuroma including histamine-sensitive C-fibers, Neurosci. Lett., 128 (1991) 203-206. Sherman. R., Sherman, C. and Parker, L., Chronic phantom and stump pain among American veterans: results of a survey, Pain 18 (1984) 83-95.
Marshall
Devor
Life Sciences Institute Hebrew lJnic,ersity of Jerusalem Jerusalem 91904. Israel
PAIN 02066
Comments on Marchand (1991) 249-257
et al.,
PAIN,
Dr. Marchand and co-workers have made an important tion to the study of the psychophysical effects of spinal cord tion (SCS) for the relief of pain in man. By applying contemporary sensory testing methods, they have been able previously unappreciated phenomena.
45
contribustimulasensitive to detect
157 Using sensitive methods for testing temperature discrimination, the authors have shown significant decrements in temperature discrimination, heat pain thresholds and heat pain ratings within the projected area of spinal cord stimulation paresthesias. The magnitude of the effect on experimental pain ratings is calculated to be 20-28% and is reported to be comparable to the effects of stimulation upon patients’ spontaneous clinical pain ratings. This comparison is awkward, however, as is the inference that these effects have a common mechanism: “ since we observed comparable DCS-related modulation of the cutaneous heat pain and the varied clinical pain, such changes probably affect cutaneous and deep nociceptors similarly”. The goal of clinical application of spinal cord stimulation is the relief of chronic pain, which may relate to neural injury or deafferentation rather than nociception. Incremental temperature sensibility and acute heat pain threshold measurements with 2-5 set stimuli, as employed in this protocol - and even sustained presentation of painful thermal stimuli - may not be a relevant model (Gracely 1991) for therapeutic application of SCS. In fact, effects of SCS or any other pain relieving measure upon biologically useful pain or sensibility are not desirable in a chronic, clinical application, When we examined the effects of SCS upon temperature discrimination, 2-point discrimination, and light touch in a quantitative fashion (North et al. 19771, albeit using less sensitive measures than in the present study, we took a different perspective; we concluded that there was no clinically significant effect upon these important, normal sensory functions. Patient reports of impairment of normal sensory function by SCS are unusual. Clinical pain ratings in this study understandably differed from patients’ global ratings of pain and its relief by stimulation during everyday activities, as these ratings were obtained only with the subjects at rest, in an experimental setting. As the authors point out, patient’s overall ratings of everyday pain relief by SCS may reflect relief from pain at its highest intensity (beyond the scale displayed on the ordinate in Fig. 1). In accord with the authors’ premise, as we have reported (North et al. 1977, 1991a), the average SCS patient reports marked reduction in the percentage of time at the maximum pain intensities on the McGill Pain Questionnaire. Other phenomena may contribute further to the observed discrepancy (North et al. 1977, 1991a): patients commonly report a latency to pain relief, from the time stimulation begins, in excess of the 30 min allowed in this protocol, and they sometimes report persistence of relief, after stimulation ends, in excess of the 8 h required by the protocol. Details of the stimulation parameters used in this study, and of the relationship between the effects observed and the extent of overlap of pain by paresthesias, were not specified and would be of interest. The authors implicitly recognize the importance of overlap of each patient’s topography of pain by stimulation paresthesias as a necessary condition for obtaining pain relief. Patients with the most successful stimulator implants, from a technical point of view, typically would have the smallest areas of extraneous stimulation paresthesias, perceived outside the area of pain, within which the effects described in this paper could be studied. The study population includes an indeterminate number of patients with estimated pain relief as low as 30% - a failure by the clinical criteria used in all reported series of SCS patients. The authors conclude that “even with patients selected for their successful use of DCS, the clinical and experimental pain perception is reduced only a small amount. Thus, these relatively small reductions must be weighed against the invasive nature of electrode implantation”. The terms ‘relatively small’ imply some basis for comparison, but none is given. As the patients studied are not all ‘successful’ by standard criteria and as the experimental measures chosen are of questionable relevance to routine clinical application of spinal cord stimulation, this conclusion is not supported. It has been our experience that the benefits as well as the risks of spinal cord stimulation compare quite favorably with other invasive proce-
dures for chronic, intractable lumbar pain syndromes (North et al. 1991al. Over the past 20 years, we have observed no major morbidity (threatening life or neurologic function) in our SCS patients at Johns Hopkins (North et al. 1977, 1991a, b; Long et al. 1981); with contemporary percutaneous methods of electrode insertion (which may or may not have been used in this series) the risks are lower than ever. ‘Invasive’ treatments are not uniformly or inherently more morbid than non-invasive, medical therapy, viz., the well-known, occasionally lethal complications of treatment with opiates, tricyclics, or anticonvulsants. The techniques reported by Dr. Marchand et al. are an important addition to our methods for studying acute pain. As they are refined and expanded and as their relevance to chronic pain is clarified, their application to other pain relieving techniques will be of considerable interest.
References Gracely, R.H., Experimental pain models. In: M.B. Max, R.K. Portenoy and E.M. Laska (Eds.), Advances in Pain Research and Therapy. Vol. 18. The Design of Analgesic Clinical Trials, Raven Press, New York, 1991. Long, D.M., Erickson, D., Campbell, J. and North, R., Electrical stimulation of the spinal cord and peripheral nerves for pain control, Appl. Neurophysiol., 44 (1981) 207-217. North, R.B., Fischell, T.A. and Long, D.M., Chronic stimulation via percutaneously inserted epidural electrodes, Neurosurgery, 1 (1977) 215-218. North, R.B., Ewend, M.G., Lawton, M.T. and Piantadosi, S., Spinal cord stimulation for chronic, intractable pain: superiority of ‘multichannel’ devices, Pain, 44 (1991a) 119-130. North. R.B., Ewend, M.G., Lawton, M.T., Kidd, D.H. and Piantadosi, S., Failed back surgery syndrome: 5-year follow-up after spinal cord stimulator implantation, Neurosurgery, 28 (1991b) 692-699.
Richard B. North Department of Neurosurgery Johns Hopkins 600 North Wolfe Street Baltimore, MD 21205, USA
PAIN 02067
Reply to Dr. R.B. North Dr. North makes several important comments about our recent article (Marchand et al. 1991). We agree with Dr. North that experimental measures of cutaneous nociception are not the best models of chronic pain syndromes that may involve neural injury or deafferentation. However, our study addressed an intriguing puzzle about spinal cord stimulation, i.e., although physiological data in humans and animals indicate that dorsal column stimulation alters activity in normal nociceptive pathways (Hillman and Wall 1969; Brown and Martin 1973; Feldman 1975; Handwerker et al. 1975; Foreman et al. 1976; Lindblom et al. 19771, clinical studies have not confirmed that normal nociception is altered by such stimulation (Lindblom and
