Cell Replacement Therapy for Parkinson's Disease

How Close Are We to the Clinic?

Javier Ganz; Nirit Lev; Eldad Melamed; Daniel Offen


Expert Rev Neurother. 2011;11(9):1325-1339. 

In This Article

Clinical CRT Trials in PD

Initial experiments of DA cell transplantation in human PD patients began three decades ago. The first clinical trials used adrenal medullary or fetal VM allografts (transplants between genetically dissimilar individuals within the same species) transplanted into the striatum as a source for DA neurons.[48,112,113] These first studies did not show significant improvements; however, several open-label clinical studies that followed demonstrated encouraging results.[114–121] The first human studies were carried out on a small number of patients and, despite cautious optimism, the improvements found were modest.[23,24,48] Several studies followed demonstrating a wide variability of patients' outcomes. Some exhibited marked improvement, in some cases even permitting withdrawal of antiparkinsonian medication; others showing modest improvements, and others did not display any benefit. Motor aspects as quantified by the motor part of the UPDRS were the major outcome measured, while non-motor features of PD were not investigated. Motor improvements, when they occurred, had a positive impact on activities of daily living and quality of life.[114–119] Tremor and postural instability improved least, similar to the poor response of these symptoms to L-DOPA therapy. The clinical improvements reported by these open-label neural grafting trials have been long-lasting. Post-mortem studies undertaken in grafted patients that have died demonstrated survival of grafted fetal VM DA neurons with local reinnervation of the striatum by these cells for long periods postoperatively.[120,122] Functional imaging by F-DOPA PET also demonstrated that the grafted patients had increased fluorodopa uptake.[123,124] Such studies have suggested that approximately 100,000 DA neurons need to be present within the grafted striatum to achieve significant clinical benefit, and that those of nigral origin are most competent in innervating the striatum.[125]

Two double-blind placebo-controlled trials of fetal VM transplantation in PD followed these promising initial studies. The first of these controlled trials was published in 2001 by Freed et al. and included 40 PD patients aged 34–75 years, suffering from advanced disease (mean disease duration: 14 years).[29] Patients were randomly assigned to receive a VM transplant or sham surgery – a burr hole with no penetration of the dura – thereby achieving blindness of the patient and the assessing neurologist. Transplants included VM of two embryos (7–8 weeks old) and were transplanted into the putamen on both sides of the patient's brain. VM tissue was cultured 1–4 weeks prior to transplantation and unusually, prepared into strands of tissue.[126] Immunosuppression was not given, unlike the open-label studies that had all used standard immunotherapy.[29] The primary outcome, which was a change in a subjective global rating of clinical improvement at 1 year post-transplantation, did not reveal significant improvement.[29] Significant improvements were seen on UPDRS motor scores in 'off' times in grafted patients who were younger than 60 years of age. However, subsequent analysis suggested that the main determinant of this improvement was the preoperative L-DOPA responsiveness rather than the patients' age. Ma et al. reported the long-term outcome in 33 of the original trial participants who were followed for 2 years after transplantation and 15 of these subjects who were followed for 2 additional years.[127] They claimed that a high residual preoperative level of dopamine in the anterior putamen, as determined by F-DOPA PET, was associated with better clinical outcome.[127] PET scanning showed significant increases in F-DOPA uptake in the putamen of transplanted patients and post-mortem examinations showed DA neuronal survival and fiber outgrowth in the grafts. These results suggest that clinical benefit and graft viability are sustained up to 4 years after transplantation. Moreover, the dependence of clinical (but not imaging) outcomes on subject age and sex at 1 year may not persist over the long term. However, these results must be cautiously interpreted since the patient number was not large enough, longer follow-up duration is needed and the symptomatic benefit was smaller than expected by PET results. Moreover, the number of surviving grafted dopamine neurons was low compared with those previously reported in open-label studies.[29] The main motor improvements were similar to those seen with DA medications and included rigidity and bradykinesia, while tremor did not change.

The second double-blind sham surgery controlled trial[128] involved 34 patients with advanced PD, aged 30–75 years. Patients were randomized to receive either bilateral transplants (from one or four donors in each side) or sham surgery. The surgical technique differed from the previous study, solid pieces of VM tissue were obtained from 6–9-week fetuses, stored in a hibernation medium for 2 days and surgery was performed by two-stage procedure separated by a week. All patients received immunosuppression with cyclosporine for 6 months postoperatively.[128] Primary outcome was motor component of the UPDRS in 'off' state, between the baseline and the final 24-month visit. Although there was a clear trend for benefit, it did not reach statistical significance. Post hoc analysis demonstrated significant motor benefits in patients suffering from milder disease.[128] Importantly, patients in both the one- and four-donor transplant groups showed significant motor improvement compared with placebo at 6 and 9 months post-transplant, but not thereafter. It was speculated that discontinuation of the immunosuppressive therapy triggered an immune response to the graft, resulting in loss of transplant function. Indeed, the magnitude and time course of the initial improvement (up to 6–9 months) was similar to that reported in earlier open-label studies.[112,121,129] Despite a possible immune system-mediated impairment in graft function, PET scanning showed significant bilateral increases in striatal fluorodopa uptake in transplanted groups. Post-mortem data from this study also showed that the DA neurons survived in large numbers (~100,000 per putamen in the four-donor group and 30,000 in the one-donor group), with marked reinnervation of the striatum.[128] These findings are surprising and the explanation that the clinical benefits were transient owing to immune mediated graft impairment does not settle well with the increased striatal uptake demonstrated by PET and the post-mortem findings. It seems contradictory to hypothesize that an immune response reduced the function of the graft after 9 months and on the other hand to observe significant increases in F-DOPA signal at the end of the study and significant reinnervation of the striatum at autopsy, therefore, the immune involvement should only be viewed as speculation.

These studies proved the concept that transplanted DA neurons could survive in the brains of PD patients, become functionally integrated and suggest persistent clinical improvement. Nevertheless, the studies showed that fetal VM transplants produced very variable responses. The reason for this variability was not clear, and possible explanations include technical issues, such as tissue preparation, implantation procedures, or patient selection variables, such as age, levodopa response or disease stage.[125]

One of the big questions about CRT for PD is why all of the encouraging results obtained in animal models have failed to be translated in human clinical trials? Several possible explanations could explain these phenomena. First, experimental models always only partially recapitulate the disease. The animal models in which experimental CRT were tested, such as 6-OHDA in rats or MPTP in monkeys, consist of the administration of a neurotoxic substance that, through an acute injury, produces massive DA neuron death. Yet, PD pathology in humans is not an acute pathology, but a progressive chronic degenerative process occurring in the CNS. Second, the environment that the transplanted cells are exposed to differs in the diseased brain and in animal models brains. Chronic diseases, in particular neurodegenerative diseases, such as PD, present a highly deranged environment, which includes, among others, heavy oxidative stress, protein aggregation and trophic support deficiencies induced by malfunction of neuron support cells. These mechanisms are less prominent in animal models. Third, most transplanted patients suffered from long-lasting severe disease, therefore a selection of better suited patients could ameliorate the results. These differences are not trivial and might be of major importance when trying to translate a therapy developed in animal models to humans. Therefore, encouraging results from animal models prompt experiments in humans patients; however, expectations should be limited and defining the best suited patient for the therapeutic intervention is imperative.