Objectives: To explore whether interpersonal and in- Results: Patients with brachial plexus blocks experi-
termanual sensory referral occurs following anesthetic enced touch sensations in the anesthetized arm when
block of a limb and to test theories of disinhibition of mir- watching another persons arm being touched or when
ror neuron activity and transcallosal referral. the contralateral intact hand was touched.
N
EURONS IN THE VENTRAL limb of another person being touched? In
premotor area of primates the absence of inhibition by the sensory sig-
ordinarily fire when the nal, will MNS activity reach the threshold
primate performs a spe- for conscious perception, leading you to feel
cific action such as reach- touch quale on your limb? To explore this
ing for an object, putting it in his or her possibility our laboratory had previously ex-
mouth, or pushing it. These motor com- amined 3 patients with phantom limbs.6 Be-
mand neurons orchestrate the sequence cause these patients lacked tactile input from
of muscle twitches required for the action.1 the phantom limb, we predicted that they
Asubsetoftheseneuronsmirrorneurons would experience tactile quale in their phan-
fire even when the primate watches tom limb while observing touch to an-
someone else move his or her hand, as if the other person, and this is indeed what we
neuron was simulating the other primates found. This finding suggests that ones own
movements and intentions.2 There are also skin is the only barrier to experiencing
mirror neurons for touch; neurons in S2 fire someone elses touch sensation. We call this
when you are touched and also when you clinical sign hyperempathy.
watch someone else being touched.3,4 Inthatstudy,patientswithphantomlimbs
If sensory mirror neurons fire when you were seen several months after amputation,
observe touch, why do you not actually feel allowing time for long-term cortical reorga-
touch quale when another person is nization.6 Stronger support for the MNS
touched? You may empathize with the sen- disinhibition hypothesis would be obtained
sation but will not actually feel touch quale if hyperempathy occurred rapidly after de-
on your skin (although this does happen afferentation, such as under anesthesia.
in a rare form of congenital synesthesia Another prediction about sensory refer-
called mirror-touch synesthesia5). One pos- ral under anesthesia can be made based on
sibility is that the absence of tactile recep- transcallosal connections between bilat-
tor activity in your skin sends a null signal eral somatosensory cortices. Bilateral hand
Author Affiliations: Center for to your somatosensory cortex indicating representation is found in the macaque so-
Brain and Cognition, University that you are not being touched and pre- matosensory cortex,7 and in humans the
of California, San Diego
(Ms Case and
venting the mirror neuron system (MNS) same arm muscles used in executing an ac-
Dr Ramachandran); and activation from reaching the threshold of tion are activated while observing it in both
University of California conscious sensation. the ipsilateral and contralateral forearms8 of
San Diego Medical Center, What if you are deprived of sensory in- the observer. We therefore predicted inter-
Hillcrest (Dr Abrams). formation about a limb and observe the same manual referral of sensation from the in-
Characteristic A B C D E F
Sex/Age, y M/21 F/25 M/70 M/29 M/44 F/57
Handedness R R R R R R
Blocked side R R L R L R
Referral to blocked arm Yes Yes Yes No No Yes
Referral to intact arm Yes; less No Yes; less No No No
Intermanual referral Yes Yes Yes No No Yes
very subtle sensitivity in a small part of the fourth and fifth dig-
A C its of the anesthetized arm but was included in the analysis be-
cause all other parts of his arm lacked sensation.
Next, a volunteers arm was positioned close to the intact arm
Model Intact hand Blocked hand Model or blocked arm (2 trials on each side in alternating order), and
the patient was asked to watch while the volunteer was stroked
on the surface of the arm, dorsum of the hand, fingers, thumb,
B D and palmar surface by the experimenter for approximately 20 sec-
onds in each location. The patient was asked to report any sen-
sations felt in either arm. Patient F was tested using a double-
S2 S2
blind procedure by a research assistant who was unaware of the
purpose of the experiment. The patient was asked by the assis-
tant to report sensation anywhere in the body. Sometimes when
the patients indicated that they felt their blocked arm to be po-
Figure. Interpersonal sensory referral. A, Patient observes touch to models
hand proximate to own intact hand. Patients own hand is not touched. B,
sitioned in a different location from where it really was, the mod-
Right somatosensory cortex receives null signal indicating lack of touch to els hand was placed near the felt location of the arm so that it
intact hand. Mirror neurons fire, representing observed touch to intact hand. was clearly visible. For example, patient F spontaneously re-
Null signal inhibits mirror neurons. C, Patients blocked hand is not touched. ported that her anesthetized arm felt like it was lying across her
Patient observes touch to models hand proximate to blocked hand. D, Left chest, even though she knew it was lying at her side. She re-
somatosensory cortex receives no signal from the blocked hand. Mirror
neurons fire, representing observed touch to blocked hand.
ferred to this perceived limb as her third arm.
During the blocked arm trials, 4 of the 6 patients reported that
they could feel, in their anesthetized arm, the touch adminis-
tact hand to the anesthetized hand due to commissural in- tered to the volunteer. Sensory referral tended to emerge quickly
put from the intact hand and lack of inhibition from the an- but not immediately. Patients expressed a variety of reactions in-
esthesizedhand.Indeed,ourlaboratoryhaspreviouslyshown cluding amusement, surprise, and disbelief. Patient A and C also
that touching the intact hand of a patient whose arm had reported feeling some referred sensation in the intact arm dur-
been amputated (with the patients eyes closed during the ing the intact arm trials. This sensation was less pronounced than
touch) resulted in sensations perceived simultaneously in the sensation during the blocked arm trials (patients A and C es-
the phantom arm in a systematic, topographically organized timated 50% and 5% more sensation, respectively, on the blocked
manner (although there were minor deviations from topog- side). Patient B reported no referral to the intact arm, and pa-
raphy presumably due to noise during reorganization).9,10 tient F reported 1 or 2 instances of referral to the blocked arm;
Intermanual referral was also subsequently reported in a this referral was negligible (infrequent and very weak) com-
patient with deafferentation caused by stroke.11 pared with the referral from the blocked side to the blocked arm.
We also asked 2 patients (C and F) to observe an ice cube
placed on the volunteers hand. Patient C felt cold on the blocked
METHODS side but not the intact side. Patient F felt heaviness and numb-
ness in response to ice on the blocked side but not on the in-
We tested 6 randomly selected patients undergoing brachial plexus tact side. This implies that there might be mirror neurons for
blocks in preparation for orthopedic surgery to determine whether cold; such neurons might have been constructed through Heb-
interpersonal referral (hyperempathy) and intermanual referral bian links formed between observing ice visually and touch-
would occur immediately following deafferentation (Table). Three ing it for a lifetime.
additional patients were enrolled in the study and then with- Why did any sensory referral to the intact side occur? We
drawn because they no longer wished to participate after sur- believe this referral can be explained by transcallosal connec-
gery was completed (n=1) or because of clear sensation in the tions. In a healthy individual, sensory input from the right arm
anesthesized arm due to the diminishing effect of the anesthetic is sent not only to the contralateral (left) hemisphere but also
(n=2). Clinical examination that was conducted before surgery to the ipsilateral (right) hemisphere via commissural transcal-
revealed no neurologic abnormalities or preexisting sensory im- losal fibers. This may contribute to partial tonic inhibition of
pairment recorded on the patients medical charts. The patients MNS activity. When input from the right arm is blocked, its
were tested in the postoperative recovery room within 30 min- transcallosal contribution is also removed (Figure). This would
utes after their surgical procedure; patients were awake and able lower the threshold for MNS activity on the right hemisphere
to understand and respond to our questions, but the blocked arm (left arm) due to lower overall sensory input. This argument is
remained anesthetized. supported by elegant work in monkeys and flying foxes show-
We first tested sensation in the blocked arm to ensure that ing an immediate expansion of receptive fields when transcal-
the block remained complete. One patient (patient C) showed losal input is blocked, suggesting that callosal input provides