An FDA oncology analysis of toxicities associated with PBD-containing antibody-drug conjugates

• Published in: Regulatory toxicology and pharmacology 2019-10, Vol.107, p.104429, Article 104429
• Main Authors: Saber, Haleh, Simpson, Natalie, Ricks, Tiffany K., Leighton, John K.
• Format: Article
• Language: English
• Subjects: Animals; Antibody-drug conjugate; Benzodiazepines - toxicity; First-in-human dose selection; Haplorhini; Humans; Immunoconjugates - toxicity; Maximum Tolerated Dose; Mice; Oncology drug development; Pyrroles - toxicity; Pyrrolobenzodiazepine; Rats
• ISSN: 0273-2300; 1096-0295
• DOI: 10.1016/j.yrtph.2019.104429

With a new generation of antibody-drug conjugates (ADCs) that contain a drug-to-antibody ratio (DAR) of 2, the question remains whether advances in technology have resulted in more stable and tumor-specific ADCs. These ADCs are anticipated to cause minimal systemic exposures of payloads, with toxicities being evident mainly at tumor sites. We examined 15 ADCs with PBD-dimer payloads and a DAR of 2 and concluded that dose limiting toxicities in animals and in humans are generally related to the payload. Both the payloads and the ADCs had pro-inflammatory responses causing severe toxicities that were at times of low incidence, making it difficult to assess a cause-effect relationship. Due to their low incidence, single-patient cohorts may not detect these events and such design may not be suitable in first-in-human (FIH) trials. The commonly proposed approach by the sponsors for FIH dose selection was 1/6th highest non-severely toxic dose (HNSTD) in monkeys. This approach resulted in an acceptable balance of safety and efficient dose escalation in phase 1 trials, when using data from repeat-dose toxicology studies and body surface area for scaling. No sponsor used the data generated in rodents or proposed novel approaches for FIH dose selection. •Toxicities of PBD-ADCs are related to the payload.•Toxicities were comparable in animals and humans and included myelosuppression and pro-inflammatory responses.•Nonclinical recommendations in ICH S9 and ICH S9 Questions and Answers are applicable to PBD-ADCs.•Single-patient cohorts may not be a suitable design for a first-in-human (FIH) study.•1/6th HNSTD (body surface area scaling) in monkeys was a common approach for FIH dose selection.