ProSavin and OXB-102

The research record

This is the project's file for this landmark, as written and checked. Sources are listed at the end; anything not yet confirmed is marked "Not yet verified".

Who

Stéphane Palfi (Henri Mondor Hospital, Paris), Roger Barker (Cambridge) and colleagues; Nicholas Mazarakis designed the approach (per Imperial College). Sponsors: Oxford BioMedica, then Axovant/Sio Gene Therapies.

What happened

ProSavin is a lentiviral vector (a modified virus that integrates into cell DNA) carrying three genes: AADC, tyrosine hydroxylase and GTP-cyclohydrolase 1, together the machinery for making dopamine. It was injected into both putamens of 15 people with advanced Parkinson's in France and the UK, in three escalating doses (1.9×10⁷, 4.0×10⁷ and 1×10⁸ transducing units; 3, 6 and 6 people), with no control group (NCT00627588; Lancet 383:1138-46, 2014). Authors reported no serious adverse events attributed to the drug or surgery, and mean off-medication motor score (UPDRS part III) fell from 38 to 26 at 6 months and 27 at 12 months (p=0.0001 for both). The company said in 2015 that benefit persisted in most patients to three years (company statement).

The successor, OXB-102 (AXO-Lenti-PD), entered a phase 1/2 trial, SUNRISE-PD (NCT03720418). Sio reported a 21-point mean motor improvement in two evaluable patients in October 2020, then returned the rights in January 2022; the registry shows the trial terminated on 12 April 2022 with 6 people enrolled. A 2026 Movement Disorders report (Rowe, Evans, Kayhanian, Barker et al.) described the six patients (two low dose, four higher dose): eight serious adverse events, one definitely related to the treatment, all transient. It found limited efficacy data, with some stabilisation of motor function in most patients.

Why it moved us toward a cure

Like AAV-AADC, this is symptom-directed: it replaces dopamine production rather than slowing neuron loss, so it is not a cure route. It showed a different vector class could be given safely to the human brain, and it is a cautionary tale on how fragile small-company programmes are.

Outcome

Dead end as a programme, not a scientific refutation: the trial ended early on insolvency, the highest-dose cohort was never enrolled, and planned imaging analyses were not finished. Oxford BioMedica said in 2022 it would seek a new partner; later activity was not found.

Sources

• Palfi et al., Lancet 2014;383:1138-46, DOI 10.1016/S0140-6736(13)61939-X
• Imperial College news, 10 Jan 2014 · Inserm release
• Oxford BioMedica, 5 May 2015 (three-year data)
• Oxford BioMedica, 1 Feb 2022 (rights returned)
• Rowe et al., Mov Disord 2026 (Sep), PMID 42206679, abstract via Europe PMC · Trial record NCT03720418 · Cambridge repository record of the 2026 report · NeurologyLive summary
• Not yet verified: the 2026 report's DOI and the claim that the highest-dose cohort and planned imaging analyses were not completed; the 21-point figure beyond Sio's own statement; current rights holder for OXB-102.