Patient selection may be the explanation for the negative results seen in the neurotransplantation study for Parkinson's disease (PD) reported by Olanow and colleagues. 1 Their primary outcome variable, the United Parkinson's Disease Rating Scale (UPDRS) motor "off " score, failed to improve in transplant patients despite positron emission tomography (PET) scan and autopsy evidence of graft survival. Their results are in contrast with our earlier double-blind study in which grafted subjects showed significant improvement in motor "off " scores for the group as a whole ( p Ο 0.04) as well as for the younger subjects ( p Ο½ 0.01). 2 Our subsequent analysis has shown that the preoperative response to L-dopa is the best predictor of transplant outcome. 3 Before transplant, our younger group (Υ
60) improved 79% with L-dopa, whereas older subjects (61-75 years) improved only 54%. 2 Regression analysis using all subjects showed that motor "off " scores improved by 33% of the L-dopa response at 1 year and 50% at 2 years after transplant. There was no independent contribution of age or disease severity. Transplant subjects in the study by Olanow and colleagues 1 had an average improvement in "off " scores of only 58% with L-dopa preoperatively. We would suggest that they check the relation between preoperative L-dopa response and subsequent transplant outcome.
There were other differences in the two studies. We used no immunosuppression and yet observed survival of transplants in 85% of patients by PET scan or autopsy. Our findings indicate that immunosuppression is probably unnecessary for human fetal neurotransplantation.
Both double-blind studies included subjects who had failed drug therapy, with most having drug-induced dyskinesias. Since 1990, we have noted dyskinesias after transplant that usually improve with reductions in L-dopa doses. 2 In 3 of our 34 transplant patients 4 and in 3 of their 21 patients, deep brain stimulating electrodes into pallidum or subthalamic nucleus (STN) have been needed to control dyskinesias. Olanow and colleagues reported mild "off " dyskinesias in 53% of transplant patients, demonstrating the partial dopamine effect of transplants in this dyskinesia-prone group.
We agree that neither transplants nor deep brain stimulation into STN can improve the "best on " state produced by L-dopa and both are associated with some risk of persistent dyskinesias. 5 Clinical signs that do not respond to L-dopa probably represent more extensive brain pathology than the loss of dopamine neurons. Who is the "ideal " transplant subject? From our data, it is a patient disabled by the "off " state who has an excellent response to L-dopa (ΟΎ70% improvement in UPDRS motor "off " scores) and who does not have drug-induced dyskinesias. An open question is whether transplants can prevent the progression of PD and the development of dyskinesias in patients who do not already have that drug complication.