Multiple Sclerosis Research Repository

Per “Chemistry Review”: 

1. From “Chemistry Review”: “A. Recommendation on Approvability[.]  a. Recommendation[.]  The Office of Pharmaceutical Quality (CDER, does not recommend approval of STN 761053 for Ocrevus (ocrelizumab) manufactured by Genentech, Inc.  The data submitted in this application are not adequate to support the conclusion that the manufacture of ocrelizumab is well controlled and consistently produces a product that is pure and has appropriate potency.” (page not numbered)

2. From “Chemistry Review”: “C. Benefit/Risk Considerations[.] 

The indications proposed for ocrelizumab include the treatment of RMS and PPMS; a treatment for PPMS is considered an unmet medical need.

There are substantial concerns regarding the state of control of the ocrelizumab drug substance (DS) manufacturing process, particularly as it relates to potency.  The complement-dependent cytotoxicity (CDC) activities of the initial DS batches, including process validation batches, manufactured using the proposed commercial process were consistently lower than the CDC activities of the pivotal clinical study materials.  THis decrease in potency is similar to the consistent decrease in potency seen during a previous process change and appears to be related to differences in the glycosylation profiles of the DS batches.  The proposed CAPA for correction of the hypothesized root cause for the potency change was evaluated through a manufacturing campaign run during the review cycle and was shown to not correct the potency control issue.  Therefore, the root cause of the difference in the CDC activity between the pivotal cinical study materials and proposed commercial materials remains unknown, and it is not clear that DS with the appropriate level of potency can be consistently manufactured.  In addition, the antibody-dependent cellular cytotoxicity (ADCC) activities of the DS manufactured under the proposed CAPA were significantly higher than the ADCC activities of the pivotal clinical study materials.  The magnitude of the ADCC activity increase and the lack of understanding of the product MoA render this a potential safety issue and further indicate that the manufacturing process is not in a state of control.  Following discussions with the Agency, additional process controls were implemented.  However, verification that the latest proposed CAPA will correct the manufacturing issues and result in a process that can consistently produce DS with acceptable levels of CDC and ADCC activities has not been provided.  Additional testing, including that to assess ADCC, has been added to the DS lot release specifications to ensure that the materials released to the market will be of appropriate quality.  Additional studies to evaluate the manufacturing process controls and confirmation of the suitability of the qualified assays for use in the QC lot release setting will be performed as PMCs.  

There are also significant concerns regarding the stability of the Ocrevus drug product (DP).  Degradation of the polysorbate 20 excipient was identified in both historical DP batches and the validation batches.  This excipient is included in the formation to [greyed out], which impacts in vitro potency and may also impact PK and immunogenicity.  Visible free fatty acid particles were observed at the 18 month timepoint in the DP validation batch stability samples held under recommended storage conditions and at the 3 month time point in samples held under accelerated degradation conditions.  It is not clear when an impact to other aspects of product quality might occur.  In addition, while shipping conditions may also support the formation of aggregates, the DP shipping validation performed for ocrelizumab did not include an evaluation of the impact of shipping on product quality attributes or consider that the degradation of polysorbate may render aged ocrelizumab more susceptible to mechanical stress-induced aggregation during shipping.  WIth respect to the final point of use of the DP, while it appears possible to remove the free fatty acid particles via use of an infusion set containing a low protein binding inline filter, it is not clear whether the end user would be able to differentiate between vials containing these particles and particles that would be less benign and indicative of more substantial product quality issues.  Given the available data, the expiry period will be set at 15 months, and the expiry period will not be extended unless additional data to support control of polysorbate 20 and ocrelizumab degradation can be provided.  Additionally, as part of the mitigation strategy, polysorbate 20 testing was added to the DP stability protocol; confirmation that the current assay is appropriate for use in this setting will be performed as a PMC.  An appropriate shipping validation study evaluating DP, including DP with degraded polysorbate, will also be performed as a PMC.” (page not numbered)

3. From: “Chemistry Review”: “The BLA contains 24 months (updated during review cycle) of real time stability data for the three DS validation batches manufactured using Process v1.0 (commercial process) and 60 months of real time stability data for three DS batches manufactured using Process v0.4.  However, adequate data are not provided to support that the Process v0.4 lots are representative of Process v1.0” (p. 22)

4. From: “Chemistry Review”: “The applicant committed to performing a comprehensive leachable study and providing a toxicological assessment of all identified leachables.  However, these data were not available during the review cycle” (p 24)

5. From: “Chemistry Review”: “g. Dating period and storage conditions: The DP is intended to be stored [greyed out] degrees C.  A dating period cannot currently be assigned.  While the BLA contains 36 months of real time stability data for three DP lots manufactured using Process v0.4 DS, and these data indicate acceptable stability, these lots are not considered to be fully representative of the lots currently being manufactured due to the changes in manufacturing and product quality of DS and issues with polysorbate 20 degradation discussed above.  The 18 month real time stability data for three DP lots manufactured using Process v1.0 DS indicate that there are stability issues.  It is not clear at what point the issues would more significantly impact product quality of the batches on the stability protocol, and it is not clear whether new batches may be impacted to a greater extent and exhibit product quality issues earlier than 18 months.” (p 26)

6. From: “Chemistry Review”: “1. BLA does not reflect the procedure followed during manufacturing operations” (p 27)

7. From “Chemistry Review”: “Reviewer comment: It was stated that the minimum sample dilution was 1:20.  However, no additional data were provided to support the minimum dilution determination” (p 221)

8. From “Chemistry Review”: “Reviewer comment: The validation study results appear to be acceptable; however, the applicant did not provide any data in the submission to support the assay validation results reported in Table 5, including intra- and inter-assay precision and accuracy, relative sensitivity, and specificity.  Therefore, the adequacy of the partial validation study cannot be determined.” (p 228)

Per the “Pharmacology Review(s)”:

1. “The adequacy of the nonclinical program has been called into question by a lack of sufficient comparability between the drug material used in the nonclinical studies and the to-be-marketed drug product.  In particular, the comparability of the material used in the chronic toxicity study and the embryofetal study has not been established” (p 1)

2. “Conclusions: The pharmacology/toxicology reviewer and supervisor conducted a thorough evaluation of the nonclinical information submitted in support of this NDA.  There appears to be a lack of comparability of the drug material used in the chronic toxicity and reproductive toxicity studies and the to-be-marketed product.  In addition, there is concern that immune function in offspring of drug-treated dams has not been adequately assessed.

The reproductive toxicity studies may be acceptable if adequate comparability can be

 demonstrated between the drug material used in the study and the to-be-marketed clinical drug product.  Repeat of the chronic toxicity study was not considered necessary given the available clinical data.  Additional data addressing recovery of B-cell function in offspring of dams treated with ocrelizumab appears warranted” (p 1-2)

3. “There were numerous deviation from GLP, including (but not limited to) misdosing, deviations from scheduled sampling collection times (TK, clinical pathology), incorrect storing of samples (e.g., room temperature instead of under refrigeration).” (p 9)

4. “According to the sponsor, multiple methodological issues precluded a valid assessment of TDAR” (p 9)

5. “…the format of the study reports (.e.g, the lack of separate summary data tables) made it unnecessarily difficult to conduct a thorough review of the data” (p 14)

6. “Overall, there are two primary deficiencies in the nonclinical data conducted to support approval: -The DS material used in the embryofetal development study (v0.1) does not appear to be sufficiently comparable (based on information provided by the CMC team) to that (v 0.4) used in the pivotal clinical studies[;] -The lack of an adequate TDAR evaluation in the pre- and postnatal development (PPND) study precluded an evaluation of potential adverse effects on immune function” (p 15)

7. “At this time, it appears that there are insufficient CMC data to support approval.  If the to-be-marketed drug product is determined by the CMC team to not be comparable to the drug product(s) used in the pivotal nonclinical studies, additional nonclinical studies may need to be repeated” (p 15)

8. “These data indicate that the product derived from the 2006 process may not be comparable to the product derived from the original process” (p 28)

9. “(Injection site reactions: Fibroplasia, graded minimal to mild, was observed at the injection site in LD, MD, and HD animals (dose-related in incidence).  The sponsor related these findings to mechanical injury to tissues due to the injection.  There is no explanation as to why, if this is the case, no control animals show the reaction to injection” (p 65)

10. “Observations and Results: This study was conducted to compare toxicity of product produced by the 2006 process to product from the original manufacturing process.  Full toxicology parameters were monitored.  The results indicated that the PD andt toxicity of the two clones were similar.  However, there appeared to be a consistent reduction in exposure in the group that received the Clone [greyed out] (2006 process).  It is not clear from the information provided, if any of the general toxicity studies were conducted using the product from the 2006 (new) process.  If the product derived from the 2006 studies were conducted using product from each process.  If the CMC team determines that the products from the 2 processes are not comparable, the nonclinical studies could be invalidated (p. 76)

11. “Stillbirths: Stillbirths occurred in 2 LD animals and one HD animal.  The sponsor concluded that, due to the lack of an increase in incidence with dose, the stillbirths were not related to the test article.  However, no stillbirths occurred in the control group.  A relationship to the test article cannot be ruled out since all stillbirths occurred only in dams that were dosed with ocrelizumab” (p 98)

1. “A significant concern regarding manufacturing was raised by the CMC team that could influence the validity of the nonclinical studies.  Over the course of drug development, significant changes were made in the ocrelizumab manufacturing process.  According to the CMC, the assays used to determine potency (ADCC and CDC) have shown significant variation among the processes (v0.1-v0.4).  At the time of this review, comparability (as determined by the CMC team) between the product intended for market, the product used in the pivotal clinical studies, and the product used for reproductive toxicology is in question.  The sponsor should be asked to confirm that the product used in the pivotal clinical studies is comparable to the lots used in the reproductive toxicology studies and the product intended for market.  If comparability cannot be demonstrated, some nonclinical studies may need to be repeated (p. 111-112).