Intermittent convection-enhanced delivery of GDNF into rhesus monkey putamen: absence of local or cerebellar toxicity

Glial cell line-derived neurotrophic factor (GDNF) has demonstrated neurorestorative and neuroprotective effects in rodent and nonhuman primate models of Parkinson’s disease. However, continuous intraputamenal infusion of GDNF (100 µg/day) resulted in multifocal cerebellar Purkinje cell loss in a 6-month toxicity study in rhesus monkeys. It was hypothesized that continuous leakage of GDNF into the cerebrospinal fluid compartment during the infusions led to down-regulation of GDNF receptors on Purkinje cells, and that subsequent acute withdrawal of GDNF then mediated the observed cerebellar lesions. Here we present the results of a 9-month toxicity study in which rhesus monkeys received intermittent intraputamenal infusions via convection-enhanced delivery. Animals were treated with GDNF (87.1 µg; N = 14) or vehicle (N = 6) once every 4 weeks for a total of 40 weeks (11 treatments). Four of the GDNF-treated animals were utilized in a satellite study assessing the impact of concomitant catheter repositioning prior to treatment. In the main study, eight animals (5 GDNF, 3 control) were euthanized at the end of the treatment period, along with the four satellite study animals, while the remaining eight animals (5 GDNF, 3 control) were euthanized at the end of a 12-week recovery period. There were no GDNF-related adverse effects and in particular, no GDNF-related microscopic findings in the brain, spinal cord, dorsal root ganglia, or trigeminal ganglia. Therefore, 87.1 µg/4 weeks is considered the no observed adverse effect level for GDNF in rhesus monkeys receiving intermittent, convection-enhanced delivery of GDNF for 9 months. Electronic supplementary material The online version of this article (10.1007/s00204-018-2222-z) contains supplementary material, which is available to authorized users.


GENERAL INFORMATION
Histologic tissue processing was conducted by Seventh Wave Laboratories LLC, Chesterfield, MO, USA under the supervision of Kim Shevlin. The pathology evaluation was conducted by Dr. Kristen J. Nikula of Seventh Wave. Results of microscopic evaluations were recorded electronically in Provantis (version 9.1).
The death details (i.e., Day of Death, Day of Necropsy, and Mode of Death), necropsy information, and day on study (i.e., Study Day No.) for each animal as reflected in the appendices were transcribed into the Provantis pathology system by Seventh Wave to facilitate reporting.

Study type as listed in table headers of the Individual Animal Data refers only to Seventh
Wave's portion of the study.
Tissues from 20 male Rhesus monkeys were received at Seventh Wave Laboratories for sampling, processing, and slide production. The monkeys were treated via intermittent bilateral intraputamenal convection-enhanced delivery catheters (two per animal, one in each putamen) at a volume of approximately 130 μL/catheter, including 65 µL test or control article and 65 µL diluent (artificial cerebrospinal fluid; aCSF) to clear the dead space of the system.
Study design information received from the Sponsor was as follows:

METHODS
Macroscopic observations at necropsy were provided as gross necropsy forms for reference in this pathology report. As these data will be presented elsewhere in the study report, they are not re-presented here.
The following tissues were received after each necropsy in 10% neutral buffered formalin or 4% paraformaldehyde from the testing facility as collected during necropsy: brain including cerebellum (21 cassettes), dorsal root ganglia (cervical, lumbar, thoracic), spinal cord *To determine the incidence, main findings were considered present in an animal if they occurred in one or more of the sections examined from that animal. If a finding was present, the severity shown is the highest severity for the finding in the animal. CT = catheter track. Sporadic, incidental, "spontaneous" findings are not listed in this table.  *To determine the incidence, main findings were considered present in an animal if they occurred in one or more of the sections examined from that animal. If a finding was present, the severity shown is the highest severity for the finding in the animal. CT = catheter track. Sporadic, incidental, "spontaneous" findings are not listed in this table.

Control -Terminal GDNF -Terminal GDNF-Satellite
Other than GDNF immunostaining, which was observed only in animals administered r-metHuGDNF, microscopic findings were generally consistent with expected responses to an implanted device of this type (Butt, 2011;Polikov, et al., 2005). These findings were confined to the catheter tracks and adjacent neuropil and exhibited the same characteristics and similar incidence and severity across the control group and those administered test article. The only exception to this was moderate neutrophilic inflammation (infiltration, neutrophil) and an associated slightly increased tissue reaction (considered as the combination of mixed cell perivascular cuffs, gliosis/astrocytosis, and gray matter vacuolation) around the right catheter track of one control animal (V000909). One animal administered r-metHuGDNF (V002611) exhibited an accumulation of pigmented macrophages (correlate to the necropsy observation of discoloration) and a tissue reaction (gliosis/astrogliosis) around the left catheter track that was slightly higher in severity but similar in character to that noted in controls. These findings in animal V002611 may have been related to the clinical observations (Veterinary Care Summary signed by Dr. Diane Stockinger, 10/23/15) that included post-surgical issues associated with the initial implantation of the catheters such as swelling and intermittent fluid accumulation caudal to the incision site. The histopathological findings noted around the left catheter track in this animal were considered consistent with these clinical events. The pigmented macrophages contained pigment consistent with hemosiderin, which is an indicator of previous local hemorrhage around the catheter. Some degree of local hemorrhage is an anticipated event with implantation of the catheter. Accordingly, minimal to mild numbers of pigmented macrophages were noted around the catheter track in all other animals in the study and although they occurred with slightly higher severity (moderate) in V002611 the severity of prior hemorrhage suggested by these pigmented macrophages was still within the range of what might be expected with implantation of the catheter. Despite the slight increase in severity of the pigmented macrophages and surrounding glial reaction around the catheter in this animal, these changes were not of high enough severity to be considered adverse findings particularly given the limited distribution and absence of any associated degenerative findings. r-metHuGDNF was not thought to play a role in the development of these findings given the comparable characteristics of these findings to those noted in many other animals including controls.
None of the findings in the terminal and satellite sacrifice groups were considered adverse based on the limited distribution and severity with the one exception being the moderate neutrophilic inflammation in control animal V000909. The inflammatory process in this animal was considered adverse given its severity level as well as its character, which was highly suggestive of a response to local bacterial infection. This inflammatory process was not considered directly related to the catheter itself but instead specific to this particular animal's clinical history of picking at the sutures and wound associated with the incision site overlying the implanted system. These factors, along with repeated manual manipulation of the external port and surrounding tissues by the animal, likely predisposed the local soft tissue environment around the port to bacterial growth as suggested by the clinical history of a purulent discharge from this area. Details of the clinical observations in this animal are documented in the clinical history (Veterinary Care Summary signed by Dr. Diane Stockinger, 10/23/15). Bacterial growth in this area could serve as a nidus of infection that subsequently tracked along the catheter and into the brain as suggested by the acute inflammatory process observed in this animal.
Overall, the findings observed in the terminal and satellite sacrifice groups were comparable in incidence and severity suggesting that the repositioning of the catheters in the satellite group animals did not exacerbate the anticipated catheter-related findings. *To determine the incidence, main findings were considered present in an animal if they occurred in one or more of the sections examined from that animal. If a finding was present, the severity shown is the highest severity for the finding in the animal. CT = catheter track. Sporadic, incidental, "spontaneous" findings are not listed in this table. Microscopic findings were generally consistent with expected responses to an implanted device of this type (Butt, 2011;Polikov, et al, 2005) with the same characteristics and similar incidence and severity in control animals and those given r-metHuGDNF. All findings were confined to the catheter tracks and adjacent neuropil and were not considered adverse based on the limited distribution and severity and/or the lack of associated degenerative changes with the one exception being the moderate neutrophilic inflammation (infiltration, neutrophil) and associated slightly increased tissue reaction (white matter vacuolation and gliosis/astrogliosis) around the catheter track of a single monkey administered r-metHuGDNF (V002608). The findings in this animal were similar to that described with the vehicle control animal (V000909) with an inflammatory process consistent with a response to local bacterial infection. V002608 had a history of issues with the external port requiring repair during the study. V002608 also had a clinical observation of purulent discharge around the port which is likely an indication of a local bacterial infection in adjacent subcutaneous tissues providing a likely source for subsequent tracking of bacteria along the exterior of the catheter into the brain as suggested by the acute inflammatory process observed microscopically in this animal. The findings in V002608 were considered adverse given their severity as well as their character that was suggestive of local bacterial involvement. The findings in V002608 were not considered r-metHuGDNFrelated because of the similarities to findings noted in a control animal (V00909). These findings were also not considered a direct result of the catheter system itself but instead a secondary event related to issues with the port which required re-stabilisation surgery. Details of the clinical observations are documented in the clinical history (Veterinary Care Summary signed by Dr. Diane Stockinger, 10/23/15).

Control -Recovery GDNF -Recovery
Other than the acute inflammatory process described above for a single r-metHuGDNF recovery animal (V002608), findings from the recovery groups had the same characteristics and a similar incidence and severity to those noted in the terminal sacrifice and satellite sacrifice animals. This is consistent with the conclusion that these effects are related to the implantation of, and tissue reaction to, the catheter only and not influenced by r-metHuGDNF administration.
There were no r-metHuGDNF-related or catheter-related microscopic findings in the cerebellum, dorsal root ganglia, trigeminal ganglia, or spinal cord in the recovery sacrifice group.

Non-r-metHuGDNF-Related, Non-Catheter Track-Related Findings
All other microscopic findings in monkeys in the terminal, satellite, or recovery groups were considered spontaneous and/or incidental because they occurred at a sporadic, low incidence and/or occurred at a similar incidence and severity in the control group and those administered r-metHuGDNF, and/or their incidence and severity were as expected for rhesus monkeys. This included a few animals (V001954, V001935, and V001633) with a small number of axon spheroids noted in brain sections away from the catheter track. These axon spheroids were considered incidental, age-related findings as has been previously

CONCLUSION
Other than GDNF immunostaining, intermittent administration of r-metHuGDNF to Rhesus monkeys via bilateral intraputamenal convection-enhanced delivery catheters at a volume of 65 μL/catheter (0.67 µg/µL), followed by 65 µL aCSF, every four weeks for a total of 11 doses did not result in any r-metHuGDNF related microscopic findings in the brain (including cerebellum), spinal cord, dorsal root ganglia or trigeminal ganglia. GDNF immunostaining demonstrated immunohistochemically was noted in the brain of all r-metHuGDNF-treated animals including recovery animals where the staining intensity was mildly reduced relative to terminal sacrifice animals.
Microscopic findings related to the implantation of, and tissue reaction to, the intraputamenal catheters were noted in all animals but were generally limited in distribution to the catheter tract and immediately surrounding tissues and were largely consistent with the anticipated response to the implantation of this type of device. These findings were not considered adverse due to their limited distribution and severity and there were no discernible differences in character, incidence and severity across the vehicle control and r-metHuGDNF treatment groups with only a few notable exceptions, each of which was associated with the clinical history in these particular animals.
Two of these animals, including a control (V000909) and an animal administered r-metHuGDNF (V002608), exhibited moderate neutrophilic inflammation in and around the catheter track in the brain with a mild to moderate tissue reaction in the surrounding neuropil (such as white matter vacuolation and/or gliosis/astrogliosis). Given the severity, and that the character of this inflammation was suggestive of local bacterial infection, these findings were considered adverse but they were not considered r-metHuGDNF-related or directly associated with the catheter itself. These findings were unique to these two animals and they were associated with a common clinical history involving chronic issues with the external port and surrounding soft tissues including purulent discharge. One of these animals (V000909) demonstrated repeated manual manipulation of the tissue overlying the implanted system which perpetuated and exacerbated the issue. As such, these findings were not considered directly the result of the catheter itself but instead related to the chronic clinical issues associated with the tissue overlying the extracranial parts of the device.
Another exception was an r-metHuGDNF administered terminal sacrifice animal (V002611) that had findings consistent in character with all the other animals in the study, but with a slightly higher severity of gliosis/astrocytosis and pigmented macrophages around the catheter track. The pigment was consistent with hemosiderin, which is an indicator of previous hemorrhage, and some degree of local hemorrhage is an anticipated event with catheter implantation. These findings were associated with clinical observations including post surgical swelling and intermittent fluid accumulation caudal to the incision site. These clinical observations may have been indicative of similar underlying issues along the catheter track as was suggested by the slightly enhanced microscopic findings noted in this animal. As such, the slight enhancement of these findings was not considered r-metHuGDNF related, particularly considering the occurrence in just a single animal and the overall consistency in character with findings noted in vehicle control and other r-metHuGDNF treated animals. The findings in this animal were also not considered adverse given the limited distribution, the lack of associated degenerative findings, and the limited severity, even though it was slightly increased over other animals In conclusion, there were no r-metHuGDNF-related histopathology findings noted in this study, with the exception of GDNF immunostaining present in the brain of r-metHuGDNFtreated animals as anticipated. All other notable findings were localized to the catheter track and generally of the type, incidence and severity anticipated with the implantation of this type of device. The only adverse findings noted in the study were not r-metHuGDNFrelated as they occurred in a vehicle control terminal sacrifice animal and in an r-metHuGDNF-treated recovery sacrifice animal and they were also not considered directly related to the catheter itself but instead secondary to chronic clinical issues associated with the external port or soft tissues in these two particular animals.  (1)  (  (  (  (  (  (  (  (                   (                    The area that appears to be either an artefact hole due to a section that is not flat/complete or a potential catheter track does not appear on the recut slilde and therefore is considered an artefact. The GDNF slide shows some brown staining in the meninges and along the surface. This stain is interpretted as non-specific "edge effect" because it includes nuclear staining and appears non-specific. CT MINERALIZED MATERIAL; Focal; Minimal: The mineralized material is morphologically similar to the material observed in catheter tracks. The material is localized in to the mininges and a sulcus and the location and morphology suggest that is is associated with the start of catheter track that is just tangential to the section. Although the catheter track is not observed, the finding is entered as CT mineralized material because of the morphologic characteristics and the material may have been displaced during removal of the catheters at necropsy.

BS/PONS O-LT;
There is a focus in the brainstem where cells exhibit a punctate or granular brown staining in the GDNF-stained slide. However, the same brown, granular pigment is present in the H&E-stained slide. Therefore, this stain is interpretted as an endogenous pigment and not as specific GDNF immunostaining.

BS/PONS O-RT;
There is a focus in the brainstem where cells exhibit a punctate or granular brown staining in the GDNF-stained slide. However, the same brown, granular pigment is present in the H&E-stained slide. Therefore, this stain is interpretted as an endogenous pigment and not as specific GDNF immunostaining.

SN N;
There is a focus in the brainstem where cells exhibit a punctate or granular brown staining in the GDNF-stained slide. However, the same brown, granular pigment is present in the H&E-stained slide. Therefore, this stain is interpretted as an endogenous pigment and not as specific GDNF immunostaining.

Nature of Incident:
The tissue as listed below was not examined.

Animal Number
Missing Tissue Pathologist Comments V002614 Trigeminal Ganglia Insufficient tissue to evaluate following recut or reharvest

Corrective Action:
The following is a description of actions taken for specific non-examination reasons: • For tissues that were insufficient to evaluate, a recut or reharvest (as appropriate) was performed.

Impact of Deviation:
The tissue that was suboptimal (insufficient) or missing was not target tissue and was present in a sufficient number of animals to allow the pathologist to evaluate the test article effect within the dose group. Therefore, this deviation from the protocol has no adverse impact on the pathologist's ability to evaluate test article effect.