Examination of distal involvement in cisplatin-induced neuropathy in man. An electrophysiological and histological...

Cisplatin is a widely used anti-neoplastic agent with dose-dependent sensory neuropathy as a major side-effect. The mechanism for the neuropathy is poorly understood; it may be caused by a lesion of the dorsal root ganglion cells or by a distal axonopathy. This distinction is important since regeneration in a neuronopathy is impossible, whereas recovery may occur if the axon is affected only distally. The most distal part of the sensory nerve fibre is, however, not accessible for conventional electrophysiological examination. To ascertain whether the distal receptor-associated part of the fibre is involved, we have used a method previously untested in patients with neuropathy. In 26 males treated with cisplatin for testicular cancer 3-6.5 years previously, and in 22 normal males, the compound sensory action potentials evoked by a tactile probe were recorded through needle electrodes placed close to the sural and median nerves. The responses were compared with action potentials evoked by electrical stimulation of the same nerves. Biopsies from the distal sural nerve at the dorsolateral aspect of the foot were obtained in three patients and in four subjects not treated with cisplatin. Sixteen patients had received a conventional dose (307-435 mg/m2) of cisplatin and 10 patients had received a high dose (553-1197 mg/m2). Two-thirds of the conventional dose patients and all the high dose patients had mild to severe sensory loss and reduced or absent tendon reflexes. The amplitude of the electrically evoked sensory action potential decreased with increasing dose of cisplatin and was correlated with the reduction of vibration sense. Tactile responses, probably originating mainly from Pacinian corpuscles, were, with the exception of two high dose patients, recorded from all sural and median nerves. The two high dose patients without a tactile response had a severely reduced or no electrically evoked response at the sural nerve. The sural nerve biopsies from high dose patients showed loss of large fibres; Pacinian corpuscles were obtained in two of these patients and contained normal axons. Our findings do not suggest that cisplatin causes a primarily distal lesion with sparing of more proximal parts of the peripheral nerve. We interpret the results as being consistent with a neuronopathy affecting primarily large sensory neurons. Brainstem and somatosensory evoked potentials and H-reflexes suggested that the spinal cord and brainstem were affected as well.