Signal transduction blockade and cancer: combination therapy or multi-targeted inhibitors?

D. J. Kerr1,* and N. B. La Thangue2

1 Department of Clinical Pharmacology, Radcliffe Infirmary, Oxford 2 Division of Biochemistry and Molecular Biology, University of Glasgow, Glasgow, UK

* Email: david.kerr{at}clinpharm.ox.ac.uk

The huge amount of research effort that has gone into understanding the molecular and cell biological processes that give rise to cancer has provided a wealth of information on the critical mechanisms that allow a normal cell to acquire the traits of a tumour cell. This information has defined some of the key regulatory proteins and pathways that come under aberrant control in cancer, and which in turn are responsible for the abnormal growth. From the clinical perspective, therapeutic approaches that correct this aberrant activity could reinstate normal growth or delay the excessive proliferation of tumour cells.

The combined approaches of molecular biology, genomics and structural biology have yielded an increasingly large number of druggable protein targets that continue to be exploited in the drive to develop novel mechanism-based cancer therapeutics, which are expected to exhibit improved efficacy over many of the existing cytotoxic cancer drugs. Currently, it is estimated that the 10% to 15% of the 28 000 protein-encoding genes in the human genome that are likely to be druggable fall into a relatively small number of protein families. Protein phosphokinases account for ~22% of the druggable genome, of which ~50% are currently under investigation as therapeutic targets [1Go].

Phosphokinases are of particular interest in cancer because of their frequent deregulation, either through increased expression or mutation in the gene, leading to heightened activity of the target kinase [2Go]. Notable examples include the consistent chromosomal abnormality that occurs in chronic myelogenous leukaemia (CML), where a translocation event between chromosome 9 and 22, the Philadelphia chromosome, results in the bcr–abl fusion protein derived from the juxtaposition of the c-abl oncogene and the breakpoint cluster region (bcr). The fusion protein attains a constitutively active tyrosine phosphokinase activity that is essential for the growth of the leukaemic cells. Similarly, the epidermal growth factor family of tyrosine kinase receptors (ErbB) play a crucial role in regulating proliferation, and a variety of mutagenic events can give rise to increased ErbB activity, such as gene amplification and mutation, that alter ErbB protein stability and result in increased kinase activity.

Mechanism-based drugs that specifically target these cancer-specific mutations have undergone clinical exemplification. Gleevec (STI S71) inhibits the aberrant Abl tyrosine kinase activity in CML, and has demonstrated remarkable single-agent activity during CML blast crisis. However, the appearance of clinical resistance to Gleevec has begun to be a significant clinical problem. The majority of CML patients that relapse after an initial response have reactivated bcr–abl kinase activity, which frequently results from mutations rendering the kinase less sensitive to pharmacological inhibition by Gleevec [3Go].

Iressa (ZD1839) is an ATP competitive antagonist of ErbB1 that exhibits activity against different solid tumours, including non-small-cell lung cancer. In a recent randomised phase III trial it was found not to provide significant improvement in survival when combined with standard platinum-based chemotherapy. In contrast to Gleevec, the limited efficacy seen with Iressa may have resulted from recruitment of a non-ideal patient population, which, in turn, emphasises the additional need for patient stratification when dealing with single target mechanism-based therapeutics [4Go].

Consequently, deriving optimal benefit from mechanism-based drugs that inhibit discrete targets requires careful consideration and clinical planning. The traditional pharmacological paradigm that there is a dominant signal transduction pathway that is responsible for driving the malignant phenotype is becoming increasingly compromised by the appearance of resistance to target- and mechanism-based drugs, reflecting the genetic flexibility of the cancer cell genome as well as the inevitable redundancy in the pathways that govern kinase signal transduction networks.

The question therefore arises as to how best to optimise the clinical benefit of these drugs. Relevant combination therapies seem like a possible option where different mechanism-based agents are combined in a fashion that reflects the multiplicity of aberrant control pathways seen in the tumour. On the other hand, the development of single therapies that inhibit multiple targets also appears to be an increasingly feasible and attractive strategy. Below, we propose and discuss how clinical benefit from combination therapies and single-agent/multiple target drugs could be achieved.

Combination therapy with kinase inhibitors

There is a sound pharmacological basis for using combinations of potent inhibitors of distinct kinases, hoping for at least additive and at best synergic effects. Although it may be possible through genome-wide screening technologies to predict which combinations are likely to produce theoretical benefits, the likelihood is that most single target inhibitors will be tested in a variety of dose schedules, and in combination studies both with proven conventional cytotoxic agents and other newer agents. So we might see, for instance, receptor tyrosine kinase, Src kinase and MAP kinase inhibitors combined with cell cycle- and phase-specific drugs like 5-fluorouracil for colorectal cancer. There are phase I trials planned or in progress for inhibitors of each of these kinase targets, and therefore the clinical community will be challenged to develop new trial designs to find ways of incorporating and evaluating multiple inhibitors. This view reflects the plethora of novel anticancer agents in development and entering the clinic, and the fact that conventionally powered phase III clinical trials of A versus B (assuming that the novel agent B would likely be added to existing standard chemotherapy) would take too long to systematically evaluate the large number of new agents, and fail to consider the scientific basis for new agents to complement each other's activity.

A new trial design paradigm
It is possible to envisage a rolling phase I trial in which, after determining the maximum tolerated dose (MTD) or biologically effective dose with combination A (the conventional gold standard chemotherapy) + B (novel inhibitor), the dose of A and B are reduced to 75% of the recommended schedule, allowing the introduction and subsequent dose escalation of novel compound C and so on and so on. The ideal situation would be one in which there was no overlapping toxicity, implying that we could combine individual agents close to their full dose. We could decide the schedule and order in which the new agents were administered based on the scientific information underpinning the molecular targets or, if there were single-agent data showing activity (clinical or biological), we could create a hierarchy of ‘entry’ into the combination regimen.

A randomised phase II, leading to phase III trial
Assuming that the pattern of toxicity and MTD for each of the various combinations was defined in the above type of phase I trial design, we could then consider taking each combination into a factorial trial design. For example, if we assume that we build on a base of conventional, active chemotherapy and plan to introduce three novel compounds, we could undertake a factorial trial design, which would have 23 (eight) theoretical combinations.

This means that 50% of patients are exposed to each drug, which would permit calculation of the contribution of individual agents to toxicity and response. This will allow us to power the factorially randomised phase II studies to detect those combinations that had a significant impact on the proportion of patients responding (e.g. increase of ≥15% compared with chemotherapy alone), typically by randomising ~40–50 patients per arm. The ‘best’ arms would then go forward into a more formally powered phase III randomised comparison to look for reasonable improvements in progression-free and overall survival.

Multi-target kinase inhibitors

An alternative strategy to devising sophisticated clinical trial regimens for combination therapies would be to step back from the selective, single target inhibitors and consider multi-target kinase inhibitors. As an example, let us consider the flavonoid aglycones, and more specifically quercetin. There is an enormous scientific and epidemiological literature on this compound, which is generally regarded as a broad-spectrum kinase inhibitor that has a wide range of additional biochemical properties including inhibition of Na+-K+ ATPase, protein kinase C, mutant p53 and phosphatidyl-3 kinase, as well as sensitising cancer cells to several conventional cytotoxics (cisplatin, cytosine arabinoside and taxol). The first formal phase I assessment of quercetin showed that it could be administered intravenously, by brief infusion, every 3 weeks to an MTD of 1700 mg/m2, at which level, reversible nephrotoxicity was seen [5Go]. There was no obvious effect on rapidly proliferating stem-cell compartments (bone marrow, hair follicles or gastrointestinal mucosa), and hints of tumour responses were apparent in ovarian and hepatocellular cancer and refractory acute myeloid leukaemia (D. R. Ferry, personal communication). Pharmacodynamic end point measurements correlated plasma concentration of drug with inhibition of tyrosine phosphorylation in lymphocytes at clinically recommended drug doses in the phase I trial [5Go]. Since it was considered wholly impracticable to develop quercetin, we undertook to rationally design and synthesise a stable and water-soluble prodrug derivative. This compound, QC12, entered a pharmacokinetic study in which it was shown that the prodrug QC12 is hydrolysed rapidly in plasma to yield quercetin, with an estimated bioavailability of 40% to 50%. QC12 therefore has the hallmarks of a potentially effective signal transduction inhibitor. Clinically, QC12 may modulate and synergise with conventional cytotoxic agents like taxol and cisplatin, and offer significant clinical benefit [6Go].

Conclusions and perspectives

Which of these two paradigms will prove to be correct, combinations of ‘selective’ single target inhibitors or use of single agents with the capacity to block multiple targets? At the moment we have no discriminant clinical data to answer this question, but the coming years should see both concepts tested in the clinical arena.

References

1. Hopkins AL, Groom CR. The druggable genome. Nat Rev Drug Discov 2002; 1: 727–730.[ISI][Medline]

2. McLaughlin F, Finn P, La Thangue N. The cell cycle, chromatin and cancer: mechanism-based therapeutics come of age. Drug Discov Today 2003; 8: 793–802.[CrossRef][ISI][Medline]

3. Druker BJ, David A. Karnofsky Award Lecture. Imatinib as a paradigm of targeted therapies. J Clin Oncol 2003; 21 (23 Suppl): 239s–245s.[Free Full Text]

4. Dancy JF, Freidlin B. Targeting epidermal growth factor receptor – are we missing the mark? Lancet 2003; 362: 62–64.[CrossRef][ISI][Medline]

5. Ferry DR, Smith A, Malkhandi J et al. Phase I trial of the flavonoid quercetin: pharmacodynamics and evidence for in vivo tyrosine kinase inhibition. Clin Cancer Res 1996; 2: 659–668.[Abstract]

6. Mulholland P, Ferry DR, Anderson et al. Pre-clinical and clinical study of QC12, a water-soluble, prodrug of quercetin. Ann Oncol 2001; 12: 245–248.[Abstract]





This Article
Full Text (PDF)
E-letters: Submit a response
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Disclaimer
Request Permissions
Google Scholar
Articles by Kerr, D. J.
Articles by La Thangue, N. B.
PubMed
PubMed Citation
Articles by Kerr, D. J.
Articles by La Thangue, N. B.