Cathepsin K inhibition reduces CTXII levels and joint pain in the guinea pig model of spontaneous osteoarthritis.

Cathepsin K is a cysteine proteinase which is believed to contribute to osteoarthritis (OA) pathogenesis. This brief report evaluates the effect of the novel selective cathepsin K inhibitor AZ12606133 on cartilage metabolism in the Dunkin-Hartley guinea pig model of spontaneous OA. In parallel, electrophysiological studies were performed to determine whether acute and chronic treatment with the cathepsin K inhibitor could alter joint nociception. Acute treatment of OA knees with AZ12606133 had no effect on joint afferent nerve activity; however, prolonged (1 month) administration of the cathepsin K inhibitor delivered via a chronically implanted osmotic pump significantly reduced mechanosensitivity in response to both non-noxious and noxious joint movements. Urinal concentrations of the cartilage breakdown products cross-linked C-telopeptides of type II collagen (CTXII) were also reduced by chronic cathepsin K inhibition. These data suggest that prolonged AZ12606133 administration can reduce cartilage turnover and joint nociception in the Dunkin-Hartley guinea pig model of spontaneous OA.

When immune-neuro-endocrine interactions are disrupted: experimentally induced arthritis as an example.

We studied whether, in parallel to the activity of the hypothalamus-pituitary-adrenal axis and the sympathetic nervous system, hypothalamic cytokine expression and monoaminergic neurotransmitter concentrations are affected during the development and chronification of arthritis induced by immunization of rats with type II collagen. Corticosterone levels were increased only transiently, and were even below the normal range as the disease progressed. Increased adrenaline blood levels and hypothalamic IL-1beta and IL-6 overexpression were observed only during the induction phase of the disease. The increase in hypothalamic noradrenaline content during the symptomatic phase was paralleled by a gradual loss of sympathetic fibers in the joints. Depletion of hypothalamic noradrenergic neurons at this time did not affect the symptomatology. Contrary to observations in healthy animals, no correlation between hypothalamic IL-1beta expression and noradrenaline content was observed in rats with arthritis. The dissociation between hypothalamic cytokine gene expression and noradrenergic neuronal activity, the lack of sustained stimulation of the stress axes, and the loss of sympathetic signals in the joints indicate that the communication between afferent immune messages to the central nervous system and two main efferent anti-inflammatory pathways under control of the brain are disrupted during experimental arthritis.

Heightened flexor withdrawal response in individuals with knee osteoarthritis is modulated by joint compression and joint mobilization.

UNLABELLED: Patients with chronic pain often present with hyperalgesia, possibly due to hyperexcitability of nociceptive pathways. The aim of the present study was to investigate alterations in flexor withdrawal reflex (FWR) excitability in individuals with knee osteoarthritis (OA) and the potential effect of specific physical inputs or therapeutic interventions (ie, joint compression and mobilization) on these behaviors. Ten subjects with and 10 without knee OA (age 45-75) were recruited. The FWR was examined utilizing suprathreshold, noxious electrocutaneous stimuli applied at the medial foot. Surface electromyographic (EMG) was recorded from the tibialis anterior (TA) and biceps femoris (BF), and peak joint torques recorded at the hip, knee, and ankle. FWR threshold was ascertained and responses at 2x threshold recorded after the following conditions: a maximal, volitional, joint-compression task, a sham hands-on intervention, and a Grade III oscillatory joint-mobilization intervention. A decreased threshold-to-flexor withdrawal response was found in the OA vs control group (P < .01). EMG and joint-torque FWR responses were further augmented in the OA group following the maximal joint-compression task (P < .05), yet remained unchanged or diminished in controls. Joint mobilization, but not sham intervention, reduced reflex responses significantly, although primarily by decreasing BF activity and knee torques (P < .05). PERSPECTIVE: Application of specific physical inputs to individuals with knee OA similar to those encountered during activity of daily living or during therapeutic interventions appear to modulate involuntary, nociceptive reflex responses. Routine weight-bearing activities such as walking may potentially enhance heightened FWR responses, while joint mobilization, a commonly used clinical intervention, may diminish reflex excitability.

Disrupted brain-immune system-joint communication during experimental arthritis.

OBJECTIVE: To explore the hypothesis that, in parallel with alterations in the hypothalamus-pituitary-adrenal axis and the sympathetic nervous system, hypothalamic cytokine expression and monoaminergic neurotransmitter concentrations are affected during the course of arthritis development induced by type II collagen. This hypothesis was based on evidence that acute inflammatory processes induce cytokine expression in the brain and affect neuronal activity. We also studied whether depletion of hypothalamic noradrenaline can affect peripheral joint disease. METHODS: Hypothalamic cytokine gene expression and neurotransmitter concentration, parameters of inflammation, and joint innervation were evaluated during arthritis development in rats induced by injection of type II collagen in Freund's incomplete adjuvant. Noradrenergic neurons in the brain were depleted with 6-hydroxydopamine. RESULTS: Transiently increased corticosterone levels, followed by increased adrenaline levels and hypothalamic interleukin-1beta (IL-1beta) and IL-6 overexpression were observed only during the induction phase of the disease. Hypothalamic noradrenaline content was increased during the symptomatic phase and was paralleled by a gradual loss of noradrenergic fibers in the joints. The positive correlation between hypothalamic IL-1beta expression and noradrenaline content in control groups was not observed in rats in which arthritis developed. Depletion of hypothalamic noradrenergic neurons when arthritis was established did not affect the course of the disease. CONCLUSION: The dissociation between hypothalamic cytokine gene expression and noradrenergic neuronal activity, the lack of sustained stimulation of the stress axes, and the loss of sympathetic signals in the joints indicate a disruption in communication between afferent immune messages to the central nervous system and 2 main efferent antiinflammatory pathways under control of the brain during collagen-induced arthritis.

Therapeutic stimulation of denervated muscles: the influence of pattern.

Muscular atrophy due to denervation can be substantially reversed by direct electrical stimulation. Some muscle properties are, however, resistant to change. Using a rabbit model of established denervation atrophy, we investigated whether the extent of restoration would vary with the stimulation protocol. Five patterns, delivering 24,000-480,000 impulses/day, were applied for 6 or 10 weeks. The wet weight, cross-sectional area, tetanic tension, shortening velocity, and power of denervated muscles subjected to stimulation all increased significantly. The fibers were larger and more closely packed and there was no evidence of necrosis. There was a small increase in excitability. Isometric twitch kinetics remained slow and fatigue resistance did not improve. The actual pattern of stimulation had no influence on any of these findings. The results, interpreted in the context of ultrastructural changes and an ongoing clinical study, reaffirm the clinical value of introducing stimulation during the initial non-degenerative phase. They indicate that there would be little therapeutic benefit in adopting regimes more energetically demanding than those in current use, and that the focus should now shift to protocols that represent the least intrusion into activities of daily living.

The post-NSAID era: what to use now for the pharmacologic treatment of pain and inflammation in osteoarthritis.

Traditionally, clinicians have relied heavily on the use of NSAIDs to treat the pain of osteoarthritis, as numerous studies have proven these agents effective. However, controversy has arisen regarding their use as first-line therapy, due to increasing awareness of their cardiovascular risks. One of these agents, rofecoxib, was withdrawn from the market in 2004 due to these concerns. Since that time, numerous studies have illustrated that many of the NSAIDs, both the selective cyclooxygenase-2 inhibitors and the traditional nonselective agents, may confer similar risks of cardiovascular toxicity. Although these agents may still be useful in many patients, concerns over side effects have begun to limit their use, and patients and clinicians are reaching for alternate agents. This review highlights the evidence behind the effectiveness of other, non-NSAID pharmacologic options in the treatment of pain and inflammation in osteoarthritis.

Mechanisms of the vasorelaxant effect of Danshen (Salvia miltiorrhiza) in rat knee joints.

The aim of this study is to investigate the mechanisms of the vasorelaxant effect of the crude extract of Salvia miltiorrhiza (family: Labiatae), also known as "Danshen", in rat knee joints. Changes in blood flow of rat knee joints were measured in vivo by a laser Doppler perfusion imager. Topical administration of Danshen onto the exposed knee joint blood vessels produced dose-dependent increases in blood flow. Treatment of the rat knee joint with 2x 1 nmol of atropine, 2x 0.1 nmol of propranolol, or 2x 0.1 nmol of a mixture of pyrilamine plus cimitedine produced no change on the vasodilator response to Danshen. However, significant inhibition of the Danshen-induced vasodilator response was observed in knee joints treated with 2x 100 nmol of N(G)-nitro-L-arginine methyl ester (L-NAME), 2x 100 nmol of flurbiprofen, 2x 10 nmol of the calcitnonin gene-related peptide (CGRP) receptor antagonist CGRP(8-37), and also in knee joints but had been denervated by capsaicin treatment or by surgery. Intravenous administration of low doses of Danshen (2.5 and 6 mg/kg) did not affect the systemic blood pressure but significantly increased knee joint blood flow, whereas, high doses of Danshen (167 and 381 mg/kg) produced hypotension with concurrent decreases in knee joint blood flow. These findings indicate that the knee joint blood vessels are more sensitive to the relaxant effect of Danshen compared to blood vessels in the general circulation. The vasorelaxant effect of Danshen was found to be partly mediated by CGRP released from sensory nerves, and nitric oxide and prostaglandins also played a part. However, there is no evidence to support a role for muscarinic receptors, adrenoceptors, or histamine receptors.

Contribution of force feedback to ankle extensor activity in decerebrate walking cats.

Previous investigations have demonstrated that feedback from ankle extensor group Ib afferents, arising from force-sensitive Golgi tendon organs, contributes to ankle extensor activity during the stance phase of walking in the cat. The objective of this investigation was to gain insight into the magnitude of this contribution by determining the loop gain of the positive force feedback pathway. Loop gain is the relative contribution of force feedback to total muscle activity and force. In decerebrate cats, the isolated medial gastrocnemius muscle (MG) was held at different lengths during sequences of rhythmic contractions associated with walking in the other three legs. We found that MG muscle activity and force increased at longer muscle lengths. A number of observations indicated that this length dependence was not due to feedback from muscle spindles. In particular, activity in group Ia afferents was insensitive to changes in muscle length during the MG bursts, and electrical stimulation of group II afferents had no influence on the magnitude of burst activity in other ankle extensors. We concluded that the homonymous positive force feedback pathway was isolated from other afferent pathways, allowing the use of a simple model of the neuromuscular system to estimate the pathway loop gain. This gain ranged from 0.2 at short muscle lengths to 0.5 at longer muscle lengths, demonstrating that force feedback was of modest importance at short muscle lengths, accounting for 20% of total activity and force, and of substantial importance at long muscle lengths, accounting for 50%. This length dependence was due to the intrinsic force-length property of muscle. The gain of the pathway that converts muscle force to motoneuron depolarization was independent of length. We discuss the relevance of this conclusion to the generation of ankle extensor activity in intact walking cats. These findings emphasize the general importance of feedback in generating ankle extensor activity during walking in the cat.

A role for substance P in arthritis?

Substance P is a neuropeptide that is released from sensory nerves and which has a number of pro-inflammatory effects. In this article, we review the evidence for a role of substance P in arthritis, both in experimental animal models and rheumatoid arthritis patients. Substance P expression is altered in the joint and dorsal horn of arthritic animals, exogenous substance P and neurokinin 1 (NK(1)) receptor antagonists modulate responses in the joint, and there is some evidence for a role of substance P in human joint disease. However, the therapeutic potential of NK(1) receptor antagonists in the treatment of rheumatoid arthritis remains controversial.

Modulation of the gait deficit in arthritic rats by infusions of muscimol and bicuculline.

Gait analysis in the adjuvant-induced arthritic rat model of chronic pain was used to examine the role of GABA(A) receptors in the development of pain. Drug solutions were administered continuously at 5+/-0.75 microl/h for 14 days via Alzet osmotic pumps (2ML2) placed under the skin of the back. The GABA(A) receptor agonist, muscimol, produces a dose-dependent reversal of the gait deficits seen in arthritic rats without reducing the tibiotarsal joints inflammatory edema or the histological picture of joint erosion and inflammation. The higher infusion rate for muscimol, 20 microg/h, caused the gait for the arthritic rats to be indistinguishable from that of normal non-arthritic rats. In normal, non-arthritic rats, muscimol did not show any effect on gait. The GABA(A) receptor antagonist bicuculline showed small but significant exacerbation of stride length (P < 0.05) single and double stance time (P < 0.05) and swing time deficits (P < 0.05) in the arthritic rats, but no changes in measures of gait in the normal control rat. The results suggest that the development of arthritic pain is increased in the absence of GABA(A) receptor tone and that increasing GABA(A) receptor tone can reduce arthritic pain but does not affect the disease process.

Locations of movement-related cells in the human subthalamic nucleus in Parkinson's disease.

The subthalamic nucleus (STN) is an emerging target for deep brain stimulator (DBS) implantation for the treatment of advanced Parkinson's disease (PD). Understanding the somatotopic organization of the STN is important for surgical navigation within the nucleus. We analyzed intraoperative data obtained during 54 procedures for the implantation of STN stimulators to assess the locations of movement-related cells. Cells were considered movement-related if they exhibited modulation of the cell discharge during passive movement of the contralateral upper or lower extremity. Microelectrode track reconstructions were plotted on a human brain atlas, using the location of the DBS electrode from postoperative magnetic resonance images as a registration mark in reconstructing microelectrode track locations. Movement-related cells were predominantly located in the dorsal part of the nucleus. The majority of the cells were related to proximal joint manipulation. Arm-related cells were located laterally and at the rostral and caudal poles, whereas leg-related cells were located medially and centrally. The finding of three or more leg-related cells on a given microelectrode track was predictive of a medial localization within the motor area. Our findings are consistent with the small number of published studies on STN somatopy in the human and the nonhuman primate.

Kinaesthetic neurons in thalamus of humans with and without tremor.

Increased afferent input may alter receptive field sizes, properties and somatotopographic representation in the cortex. Changes in the motor thalamus may also occur as a result of altered afferent input. Such plasticity has been implicated in both sensory and movement disorders. Using tremor as a model of augmented afferent input to kinaesthetic/deep neurons representing the shaking limbs, we studied the representation and properties of these neurons in human thalamus in patients with resting tremor (RestTr) from Parkinson's disease, patients with action- or posture-induced tremor (ActionTr), and patients without tremor (NoTr). Data were collected during stereotactic thalamotomy or insertion of deep brain stimulators for relief of pain or movement disorder. Using microelectrode recording, 58 kinaesthetic neurons responding to wrist and/or elbow movement were studied by mapping the receptive field, carefully isolating each joint during testing. There were no significant differences in the proportions of single and multijoint responsive neurons in the different patient groups (RestTr, ActionTr and NoTr). The borders between tactile-cutaneous, deep-kinaesthetic and voluntary cell representations in the thalamus were mapped in 74 patients and compared between the different tremor groups. A significant difference in kinaesthetic representation was found: both the RestTr and ActionTr groups had a significantly greater kinaesthetic representation than the NoTr patients. There was an expansion of kinaesthetic representation in patients with chronic increased afferent drive from tremor, without alteration in RF size. No decrease in tactile representation was found, suggesting that the increase in kinaesthetic representation does not occur at the expense of tactile representation. These data suggest that plasticity can occur at the thalamic level in humans and may contribute to the pathogenesis of tremor.

Local anesthetics differentially inhibit sympathetic neuron-mediated and C fiber-mediated synovial neurogenic plasma extravasation.

UNLABELLED: Local anesthetics are used for local irrigation after many types of operations. However, recent evidence of toxic effects of local anesthetics at large concentrations during continuous administration suggests an advantage of using decreased local anesthetic concentrations for irrigation solutions. In this study, we determined whether smaller concentrations of local anesthetics may maintain an antiinflammatory and, therefore, analgesic effect without the risk of possible toxicity. Lidocaine and bupivacaine were studied for their ability to inhibit both components of neurogenic inflammation-C fiber-mediated and sympathetic postganglionic neuron (SPGN)-mediated inflammation-in the rat knee joint. Intraarticular lidocaine 0.02% reduced 5-hydroxytryptamine (5-HT)-induced (SPGN-mediated) plasma extravasation (PE) by 35%, and further decreases were obtained by perfusing larger concentrations of lidocaine. Intraarticular bupivacaine 0.025% inhibited 5-HT-induced PE by 60%, and a 95% inhibition was obtained with bupivacaine 0.05%. Larger local anesthetic concentrations were necessary to inhibit C fiber-mediated PE than those required to inhibit SPGN-mediated PE. Lidocaine 0.4% was required to reduce mustard oil-induced PE by 60%. Lidocaine 2% inhibited mustard oil-induced PE to baseline levels. Bupivacaine 0.1% was required for an 80% reduction of PE. Bupivacaine 0.25% inhibited mustard oil-induced PE to baseline levels. Our results demonstrate differential effects of local anesthetics on SPGN- and C fiber-mediated PE but confirm the concept of using smaller concentrations of local anesthetics to achieve inhibition of postoperative inflammation. IMPLICATIONS: Local anesthetic wound irrigation is often used to treat postoperative surgical pain. Large concentrations of local anesthetics are usually used, and these concentrations may have possible neurotoxic and myotoxic effects. Our results demonstrate antiinflammatory effects of lidocaine and bupivacaine at concentrations smaller than used clinically.

Innervation of the human costovertebral joint: implications for clinical back pain syndromes.

BACKGROUND: The diagnosis of pain in the upper back, shoulder, chest, and arm is often made with considerable confusion and may be accompanied by needless expense and suffering by the patient. Despite the paucity of evidence concerning the tissues and mechanisms responsible for interscapular and atypical chest pain or "pseudo-angina," practitioners of manual therapy maintain that manipulation of the costovertebral elements and associated soft tissues may be helpful in the treatment of these painful conditions. OBJECTIVE: We have examined the costovertebral complex in humans with respect to the presence of immune-like reactivity to neurofilament protein and the neuropeptide substance P and calcitonin gene-related peptide, markers that reveal the presence of axons in peripheral tissues. DESIGN: Human costovertebral complexes obtained at autopsy were processed with standard histologic examination and immunocytochemical methods to detect the presence of neurofilaments, substance P, and calcitonin gene-related peptide. MAIN OUTCOME MEASURES: Outcomes were descriptive and did not require statistical methods. RESULTS: All costovertebral joints contained innervation within the anterior capsule and synovial tissues. In 4 separate cases, the costovertebral joints contained large intraarticular synovial inclusions or "meniscoids" found to contain small bundles of axons with immune-like reactivity to substance P. Axon bundles were identified in serial section with monoclonal antibodies to neurofilaments as well as with urea-silver nitrate staining. CONCLUSIONS: The costovertebral joint has been considered a candidate for producing back pain and/or pseudo-angina that may be ameliorated by spinal manipulation. This study has demonstrated that the costovertebral joint has the requisite innervation for pain production in a similar manner to other joints of the spinal column.

Neural features of the reach and grasp.

The concept of canonical representations within the motor system has been both supported and refuted using a variety of behavioral studies. Here, based upon neurophysiological data, I discuss the relationship amongst those neuronal substrates of action and the behavioral components of a movement. A novel view of reaching and grasping has been proposed which predicts that movements with similar kinematic and dynamic properties have a similar representation within the nervous system (Smeets & Brenner, 1999). However this is broadly inconsistent with a variety of neurophysiological findings that emphasize the independence amongst representations of action.

Motor cortex modulation of exteroceptive information at bulbar and thalamic lemniscal relays in the cat.

We have studied the control of the primary motor cortex on the activity of lemniscal neurons in semi-chronic implanted cats. In each experiment, two to three foci in the primary motor cortex were identified by intracortical microstimulation at low threshold (up to 20 microA) for their capacity to evoke movements of contralateral single limb joints. Neurons belonging to the dorsal column nuclei (main cuneate nucleus and gracile nucleus), or to the ventral posterolateral nucleus, were sampled for their response to stimulation of the peripheral cutaneous fields, as well as the antidromic response to stimulation of the contralateral medial lemniscus and ipsilateral somatosensory cortex, respectively. These neurons were then tested for stimulation of the cortical foci using a current intensity equal to the threshold needed to evoke motor effects, although we reduced the duration of the stimulating trains; thus, we avoided evoking movements which could elicit afferent volleys along the somatosensory paths. It was found that the primary motor cortex was able to modulate the transmission of exteroceptive signals at the level of both dorsal column nuclei and ventral posterolateral nucleus with analogous modalities. In particular: (i) a high percentage of responses, with a prevalence of excitatory effects, was observed when the receptive field of the neurons topographically corresponded to, or was very close to, the joint controlled by a given cortical focus; (ii) in these cases, higher percentages of excitations were observed in tests which concerned the distal segments of limbs than the proximal segments; (iii) the percentage of responses became lower as the neuronal receptive field was located further from the cortical motor target, the pattern being more frequently inhibitory in nature. From a functional point of view, the motor cortex control appears to be organized in a very precise manner. Its excitatory nature might subserve integrative mechanisms by which exteroceptive information arising in a given limb segment would be enhanced by a motor command inducing movements of the same body part. Moreover, a better definition of the afferent input could be obtained by a simultaneous depression of neurons, which send towards the cortex signals from adjacent or more distant cutaneous regions. It can be hypothesized that such an organization of the cortical control could improve the discriminative somatosensory aspects during the execution of explorative movements, besides supplying a sharper cutaneous feedback to the motor cortex.

Local hemodynamics, permeability, and oxygen metabolism during acute inflammation of innervated or denervated isolated equine joints.

OBJECTIVES: To determine oxygen metabolism, permeability, and blood flow in isolated joints in response to interleukin 1beta (IL-1beta) and contribution of innervation. SAMPLE POPULATION: One metacarpophalangeal (MCP) joint of 24 adult horses. PROCEDURE: The MCP joint was isolated for 6 hours in a pump-perfused, auto-oxygenated, innervated or denervated preparation. Isolated joints were assigned to the following 4 groups: control, control-denervated, inflamed, and inflamed-denervated, and inflammation was induced by intra-articular injection of IL-1beta. Circuit arterial and venous pressures, flows, and blood gas tensions, synovial fluid production, and intra-articular pressure were measured. Total vascular resistance; oxygen delivery, consumption, and extraction ratio (ER); and permeability surface area product were calculated. Synovial membrane blood flow was determined at 0, 60, and 330 minutes. Synovial membrane wet-to-dry ratio was obtained, and permeability to macromolecules was determined by intra-articular injection of Evans blue albumin and fluorescein isothiocyanate-conjugated dextran. RESULTS: Oxygen delivery and synovial membrane blood flow progressively increased but were not different among groups. Oxygen consumption and ER significantly increased in inflamed joints, as did intraarticular pressure and synovial fluid production. Inflamed joints had greater wet-to-dry ratio. Albumin permeability significantly increased in the villous synovial membrane of the inflamed groups, and dextran permeability was increased in the innervated groups, with a trend toward increased permeability in inflamed groups. CONCLUSION: Inflammation significantly increased oxygen demand, which was initially met by increased ER. Permeability to small molecules was increased with inflammation; innervation increased permeability to large molecules. Use of an isolated joint model enabled documentation of the physiologic responses of the joint to acute inflammation.

Multielectrode nerve cuff stimulation of the median nerve produces selective movements in a raccoon animal model.

In this study, an electrode system consisting of twelve small platinum dot electrodes imbedded in a spiral silicone rubber insulating cuff was used to investigate the feasibility of selective (regional) stimulation of the median nerves of the raccoon. Acute experiments in four raccoons consisted of functional responses observations, isometric force recordings from tendon attachments and postmortem fascicular mapping. Functional responses (elbow, wrist and/or digit flexion, pronation and/or thumb abduction) to selective stimulation were noted as dependent upon cuff electrode configuration (longitudinal tripole with and without field steering, as well as a transverse bipolar arrangement) and current level (threshold, 1/2 maximal, maximal). Muscle force recruitment curves (force as a function of stimulus amplitude) were plotted for flexor digitorum superficialis, flexor digitorum profundus, flexor carpi radialis, palmaris longus and pronator teres of three raccoons. Fascicular maps at the level of the nerve cuff were created indicating the approximate position of innervation to each of the aforementioned muscles, as well as other innervation such as paw fascicles, sensory fascicles, and elbow innervation (such as coracobrachialis). The greatest selectivity was observed at or near threshold current levels. In all four raccoons studied, a threshold electrode choice and stimulation strategy could be identified enabling selective production of either digit flexion, wrist flexion and/or digit and wrist flexion. It was possible to elicit a selective pronation response at threshold in three of the four animals. Selective elbow flexion at threshold could be produced in all four experiments. With stronger currents, additional movements were usually induced. The raccoon therefore appears to be a suitable, if challenging, animal model for further development of not only nerve cuff electrode approaches but perhaps other stimulation electrode technologies prior to human neuroprosthetic studies.

Parallel organization of proprioceptive inputs from joint receptors to cortical somatosensory areas I and II in the cat.

1. Studies in monkeys indicate that proprioceptive and tactile inputs are conveyed from the thalamus to the primary somatosensory cortex (SI) and thence to the secondary somatosensory area (SII) in a serial scheme. In contrast, in the cat, tactile information is conveyed in parallel from the thalamus to SI and SII. The present study, in the cat, employed reversible inactivation of SI to determine whether proprioceptive inputs to SII from joint receptors depend on an indirect serial path via SI or are conveyed over a direct path from the thalamus. 2. SI and SII foci for knee joint inputs were determined with evoked potential mapping. Reversible inactivation of the SI focus by cooling had no effect on the amplitude, latency or time course of SII potentials evoked by joint inputs. There was also no consistent effect on the response levels of individual SII neurones examined during SI inactivation. Furthermore, there was no attenuation of the later components of the responses, and therefore no evidence that these depended on an indirect path to SII via SI. 3. Results demonstrate that proprioceptive inputs project directly from thalamus to SII over a pathway organized in parallel with that to SI, in contrast to the serial scheme reported for proprioceptive processing in primates.