ESSAY Glioblastoma Adjuvant Chemotherapy www.advhealthmat.de Designing Next-Generation Local Drug Delivery Vehicles for Glioblastoma Adjuvant Chemotherapy: Lessons from the Clinic Anthony Tabet, Melanie P. Jensen, Christopher C. Parkins, Parag G. Patil, Colin Watts, and Oren A. Scherman* 1. Glioblastoma and Clinical To date, the clinical outcomes and survival rates for patients with glioblas- Translation toma (GB) remain poor. A promising approach to disease-modification The clinical outcomes and 5-year survival involves local delivery of adjuvant chemotherapy into the resection cavity, rate for patients with glioblastoma (GB) thus circumventing the restrictions imposed by the blood–brain barrier. The make it among the most pernicious and clinical performance of the only FDA-approved local therapy for GB [carmus- challenging diseases to treat. Despite all tine (BCNU)-loaded polyanhydride wafers], however, has been disappointing. the resources, time, and talent focused There is an unmet medical need in the local treatment of GB for drug delivery on developing targeted and/or local vehicles that provide sustained local release of small molecules and combina- delivery technologies by the biomaterials community for GB, the clinical perfor- tion drugs over several months. Herein, key quantitative lessons from the use mance of the FDA-approved therapy car- of local and systemic adjuvant chemotherapy for GB in the clinic are outlined, mustine ((BCNU)-loaded polyanhydride and it is discussed how these can inform the development of next-generation wafers) and clinical trials of other mate- therapies. Several recent approaches are highlighted, and it is proposed that rial approaches have been discouraging. long-lasting soft materials can capture the value of stiff BCNU-loaded wafers As disappointing is the remarkably stag- nant clinical translation of next-generation while addressing a number of unmet medical needs. Finally, it is suggested material approaches for GB. Despite that improved communication between materials scientists, biomedical encouraging preclinical results from scientists, and clinicians may facilitate translation of these materials into the hydrogels and modified wafer formula- clinic and ultimately lead to improved clinical outcomes. tions loaded with more efficacious chemo- therapies, a total of zero have completed even a phase I clinical trial. Other strate- gies, including convection-enhanced delivery, microsphere A. Tabet, C. C. Parkins, Prof. O. A. Scherman formulations, or drug-loaded nanoparticles have seen limited, Melville Laboratory for Polymer Synthesis albeit some, translation into the clinic with mixed results. This Department of Chemistry University of Cambridge lackluster progress can be attributed, in part, to the paucity of Lensfield Road, Cambridge CB2 1EW, UK communication between material scientists, biomedical scien- E-mail: [email protected] tists, and clinicians. When examining the purported clinical Dr. M. P. Jensen, Prof. C. Watts relevance of embedding certain material properties into for- Division of Neurosurgery mulations, it is clear that some widely known truths about the Department of Clinical Neurosciences Addenbrooke’s Hospital nature of GB progression among clinicians have not reached University of Cambridge the biomaterials community. Hills Road, Cambridge CB2 0QQ, UK Furthermore, a closer examination of the lessons from the Prof. P. G. Patil BCNU wafers and other clinical trials of GB drug delivery Department of Neurosurgery materials may enrich and inspire materials scientists to create University of Michigan Medical School new systems that satisfy unmet medical needs identified by Ann Arbor, MI 48109, USA the clinical community. In tandem, clinicians and biomedical Prof. C. Watts Department of Neurosurgery scientists may benefit from a short review highlighting the bio- Birmingham Brain Cancer Program compatibility, safety, longevity, kinetics, tunability, and efficacy Institute of Cancer and Genomic Sciences of promising new drug delivery materials without inundation University of Birmingham by chemical and physical characterizations or discussions. Birmingham B15 2TT, UK Another key challenge in treating GB is an incomplete The ORCID identification number(s) for the author(s) of this article understanding of disease pathophysiology, such as mecha- can be found under https://doi.org/10.1002/adhm.201801391. nisms driving intrinsic and adaptive GB cell chemoresistance. DOI: 10.1002/adhm.201801391 A combined approach where biomedical scientists and material Adv. Healthcare Mater. 2019, 8, 1801391 1801391 (1 of 6) © 2019 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.advancedsciencenews.com www.advhealthmat.de scientists work in parallel and in close communication with resistance driven by suboptimal drug exposure.[12] This key clini cians will be key for the timely development of optimal lesson has been lost with the shift in focus from systemic to therapeutic options. locally administered chemotherapy. The antitumor effect of any chemotherapeutic agent, regardless of route of administration, is time-critical: contingent on duration of exposure for time- 2. Principles of Cancer Treatment dependent agents, and repeated high-dose exposure over time for concentration-dependent agents. Following a landmark phase III trial in 2005, the Stupp pro- Optimizing chemotherapy dosing regimens also depends tocol was adopted as the standard of care for newly diagnosed on the trade-off between increasing drug dose/exposure to GB: maximal safe surgical excision of the tumor mass, fol- enhance antitumor activity and minimizing concomitant tox- lowed by radiotherapy and concomitant temozolamide (TMZ), icity. The extended 12-cycle TMZ regimen may improve sur- followed by adjuvant TMZ.[1] Despite the added ≈2.5 months vival relative to 6 cycles, but at the expense of increased hema- survival afforded by the addition of TMZ, the 5-year survival tological toxicity.[11] Indeed, the minimum survival threshold rate[2] remains poor at 5%. A number of factors make GB dif- at which patients accept chemotherapy closely relates to the ficult to treat: 1) the high proliferative and infiltrative capacity, severity of its toxic side effects.[13] Reasonably, this principle heterogeneity, and intrinsic and acquired chemoresistance of was not prioritized in the development of the BCNU wafers, neoplastic cells; 2) the tumor microenvironment; for example, given that local delivery of BCNU circumvents the problem the ability to induce anergic states in surrounding lymphocytes of high toxicity associated with systemic administration.[14,15] and glial cells, restricting the antitumor immune response; BCNU wafers seem to be associated with a number of local and 3) the brain macroenvironment; namely the surrounding adverse events (namely cerebral oedema, intracranial infec- blood–brain barrier and blood–tumor barrier restrict the ability tion, and pericavity necrosis), more than those expected of drugs to reach the brain parenchyma.[3] Despite these unique from resection alone.[16–19] Side effects peak in the weeks/ challenges, the goal of adjuvant oncological therapy (whether months following implantation, and can persists for up to local or systemic) remains constant: maximizing cytotoxicity 6 months.[17,19,20] Interestingly, this more closely corresponds and reducing the risk of recurrence, while minimizing associ- to the time-scale of polymer degradation than the release ated toxicity and the emergence of resistance. Indeed, the FDA- kinetics of BCNU, suggesting that these side effects are more approved BCNU wafer technology was developed with these closely related to persistence of the wafer within the resection needs in mind. By delivering chemotherapy directly into the cavity than to early BCNU release.[7–10] In light of this, it is per- tumor cavity, the restrictions imposed by the blood–brain bar- haps unsurprising that phase III trials comparing BCNU to rier are circumvented, and higher local drug concentrations can placebo wafer have reported similar rates of adverse effects.[21] be achieved with limited systemic toxicity. However, drawing on After FDA-approval of the BCNU wafers, and with treatment general pharmacodynamics principles, are BCNU wafers opti- extended to patients who were not eligible in the initial clinical mally designed to meet these targets? trials, mounting concerns were reported through case reports/ Based on in vitro colony-forming inhibition studies, the cyto- series, and these may more closely reflect the comparison of toxic actions of antitumor agents are, somewhat imprecisely, wafer to standard resection.[22] classified as time-dependent (cell cycle phase-specific agents) or Two chemotherapeutic principles, the importance of sus- concentration-dependent (cell cycle nonphase-specific agents).[4] tained/frequent drug exposure and the efficacy versus toxicity This categorization has proved useful when designing clinical trade-off, were perhaps too readily dismissed in the move from dosing regimens for systemically administered agents. In systemic to local chemotherapy delivery. Immediately postim- the case of TMZ, peak concentration, rather than prolonged plant, BCNU is rapidly released from the wafer, in a “1-cycle” exposure is thought to be more important for treatment effi- fashion: the wafer delivers the majority of its chemo-payload cacy, consistent with its cell cycle nonphase-specific mode of in under 1 week.[9,10,22] This explains, in part, why the effi- action.[5,6] Accordingly,