Combination therapies are increasingly becoming an important part of modern anticancer therapy. By exploiting targeted, mechanism-based treatments through the use of rational combinations, the multifactorial complexity of cancer may be addressed.
Advances in the comprehension of biological systems, driven by genomics and proteomics, have provided a framework within which preclinical scientists can predict the biological responses resulting from the modulation of multiple independent targets in combination. These advances will require a change in drug discovery, with the emphasis moving from single to multitargeted approaches using cellular, phenotypic analyses. Ultimately, combination therapies will enable more efficacious treatments to be tailored to individual patient needs.
The potential impact of combining targeted anticancer agents has yet to be fully realized, although a range of novel therapeutics are being tested in the clinic. We can clearly learn, however, from other disease areas, in particular, HIV infection.
Following the discovery of HIV in the early 80s, initial monotherapy treatments, primarily with nucleoside reverse transcriptase inhibitors (NRTIs), showed response rates that were both limited and transient. Similar results are being seen now with anticancer monotherapies. These disappointing results significantly enhanced the development and subsequent combination of NRTIs with HIV protease inhibitors and nonnucleoside reverse transcriptase inhibitors. This highly active antiretroviral therapy, improved disease control, resulting in a decrease in HIV-related deaths.
We can draw analogies between novel anticancer approaches and the development of anti-HIV therapies because cancer is a multifactorial disease that is often associated with the perturbation of complex signaling pathways with built-in redundancy and biological buffering capacity. With the exception of those targeting defined genetic alterations in cancer, treatments that only affect a single target and pathway are likely to be stymied by this redundancy.
Consequently, innate and acquired resistances are important issues for targeted agents when dosed as monotherapy. In contrast, rationally designed therapies targeting multiple oncogenic drivers have the potential to overcome this biological redundancy while increasing response rates and overcoming acquired resistance.
The increasing number of anticancer agents targeting specific pathways, in contrast to cytotoxic agents, enables a rationally designed approach to combining compounds based on disease etiology. Through utilization of these agents in combination, using either concurrent or sequenced regimens, tumor biology can be exploited using a number of strategies.
These include exploiting multiple targets on parallel pathways, inhibiting targets at multiple levels on a single pathway, and targeting both the host and tumor components, as well as, utilizing agents with different mechanisms of actions against the same target. In order to leverage maximum benefit, alignment of these strategies to the etiology of the disease will be critical.
With the shift toward combination strategies, new tools and techniques will be required. Discovery campaigns aimed at monotherapies are typically built around single targets, using biochemical and/or cellular assays, whereas combination interactions can only be elucidated using cell lines with intact signaling networks.
The scientific rationale for determining synergy with in vitro and in vivo combination experiments is outlined in a 2006 Pharmacology Review article by T.C. Chou, which suggests that an ideal isobologram of in vitro affects should be created using a full rank matrix of combinations to be tested, similar to the adjacent schematic, wherein drug A (novel therapeutic) and drug B (novel therapeutic or current standard of care) are combined in ratios bracketing the IC50 value for each compound independently. The experimental results, often derived from inhibition of cell proliferation experiments, are then analyzed and plotted as an isobologram to determine the combination interaction.
Typically, combinations that show promising synergy in cell-based proliferation/survival experiments consisting of a single cell line are assessed further using cellular phenotypic and mechanistic assays across a panel of cancer cell lines, providing a view of the molecular basis of the interaction. The use of isogenic cell lines can also play a role in confirming the mechanistic basis of unusual interactions.
Confirmation of in vitro observations using in vivo assays in order to evaluate the hypothesis in a more clinically relevant model is also important, not only to support clinical hypothesis testing, but also to to enable a better understanding of the link between tumor-derived cell line models and the tumor grown in vivo.
One method for testing combinations in vivo is the use of bioluminescent cell lines in a subcutaneous or orthotopic tumor model with biophotonic imaging as a measurement of the efficacy of the combination in vivo. Given the cost constraints associated with in vivo studies, an alternative constant-ratio combination design experiment can be conducted instead of a full-rank matrix approach. The constant-ratio determination uses a fixed drug combination with a dose-response curve generated for the fixed combination.
Although not unique to combination therapies, there is a clear need to better understand and enhance the capabilities of preclinical anticancer models to predict efficacy in the clinical setting. The theoretical bases of in vitro experimental design and combination data analysis have evolved significantly since the 1980s. In recent years, major advances have been made in both academia and industry in assembling panels of tumor cell lines to model, at least in vitro, the diverse etiologies of particular disease types.
Alignment of these tumor cell line panels to their respective clinical tumors at the genetic, epigenetic, genomic, and proteomic level will be key in their utility to predict clinical responses.
In addition to enhancing the predictivity of combination efficacy, these models will be key in the identification and validation of proof of mechanism and proof of principle biomarkers that can be used in preclinical and clinical settings as advanced surrogate endpoints. A key success factor, not just for combinations and individualized medicine but for oncology treatments in general, is the development of biomarkers that track efficacy throughout treatment regimens.
The need for better biomarkers is obvious and part of the FDA’s critical path initiative, but the specific implication for combination therapies and targeted treatments is that the clinician will be able to gauge efficacy and monitor treatments earlier in the regime, resulting in early readouts of efficacy.
To compliment this, there is also a need for in vivo models that confirm in vitro observations in conjunction with drug safety endpoints. Although xenograft models can be useful, the development and use of orthotopic, genetic, and autochthonous models will be important in the assessment of novel agents in combination. Ultimately, feedback from the clinic will be required to reshape and optimize preclinical models to generate combination hypotheses to test in the clinic.
Tailored for Patients
The application of personalized medicine and disease stratification to meet an individual’s particular genetic background and disease etiology could be realized by a drug combination approach. Preclinical models could be used to characterize specific biomarkers associated with increased sensitivity to particular combination therapies.
Biopsies of patient tumors could then be analyzed ex vivo, leading to the selection of the most effective combination for an individual patient. Technological developments that enable simple and robust identification and routine analysis of predictive biomarkers, in both preclinical and clinical samples, will be key.
Development of a number of technologies, in particular within the proteomics field (e.g., reverse protein arrays), promises to facilitate the further understanding, not only of expression levels, but also of intracellular pathways, enabling the rational choice of novel agents for treatment in combination from both preclinical and clinical samples.
Fundamental advances in the state of science, coupled with the realization of the difficulties associated with single-molecule approaches, have necessitated alternative therapeutic approaches. Genomics, proteomics, and other systems biology research efforts have provided a better understanding, and capability, to measure the underlying pathways that must be modulated to control and remediate disease.
This comprehension coupled with an explosion in the development of small molecule and biological agents against multiple cancer targets across a range of oncogenic signaling pathways provides us with a toolbox to enable the rational design and development of combination treatments. By addressing biological complexity in a more holistic way, combination therapy will result in more effective treatments that ultimately will enable personalized medicine.