Skip Navigation LinksHome > May 2014 - Volume 26 - Issue 3 > Anticancer drug development: moving away from the old habits
Current Opinion in Oncology:
doi: 10.1097/CCO.0000000000000076

Anticancer drug development: moving away from the old habits

Ades, Felipea; Zardavas, Dimitriosa,b; Aftimos, Philippea; Awada, Ahmada

Free Access
Article Outline
Collapse Box

Author Information

aMedical Oncology Clinic

bBreast International Group Headquarters (BIG-aisbl), Institut Jules Bordet, Université Libre de Bruxelles, Brussels, Belgium

Correspondence to Ahmad Awada, Jules Bordet Institute, 121 Boulevard de Waterloo, 1000 Brussels, Belgium. Tel: +32 2 541 36 42; fax: +32 2 541 33 39; e-mail:

In the 1902 masterpiece silent movie by Georges Méliès, Le Voyage dans la Lune (a trip to the moon), a group of bright scientists were launched to the moon (Fig. 1). While exploring, they met a moon inhabitant and, in fear, their first reaction was to destroy him with an umbrella attack. This innovative movie from the beginning of cinema is an amusing story made to entertain people. However, interesting is how it portrays the reaction of human beings to something they do not understand. When confronted with fear, the instinctive survival mechanism is to attack. This could well be a selected behavior according to Darwin's theory of evolution. When it comes to research about cancer drugs, it is very likely that we act like George Méliès’ astronauts.

Image Tools

Oncological drug research started in the 1940s during the Second World War (WWII). After a German air raid sunk a North American ship carrying chemical weapons in the Italian port of Bari, thousands of civilians and allied soldiers were exposed to deadly mustard gas [1]. It was noticed that among the many injuries caused by mustard gas, myeloid and lymphoid suppression were among them. Shortly thereafter, mustard agents started to be used to treat hematologic malignancies, and several other alkylating agents were developed to treat other cancers [2]. A later important political landmark was the signing of the National Cancer Act of 1971 by United States (US) President Richard Nixon, starting the so-called War on Cancer. The aim of this effort was to ‘strengthen the [US] National Cancer Institute in order to more effectively carry out the national effort against cancer’ [3]. Since then, oncological drug development has expanded quickly, and today we have an extensive portfolio of cytotoxic drugs able to kill cells, cancerous or not.

Ironically, the modern era of drug development started during WWII and was boosted by the War on Cancer. So, despite being so aggressive in reacting to this unknown disease, why are we still so far from curing most cancers? And why are our treatments so toxic until today? The answer to these questions can be related to our perception of our ‘enemies’ and to our ways of developing strategies to ‘kill’ them.

The early focus of standard oncological drug research was the cancer cell itself. The objective was to eliminate it by whatever means. Briefly, drug research routinely starts in cell cultures, and if the drug is effective enough to produce cell death, it moves to animal models. If in animal models the drug kills the cancer grafts (and not the animals), the drug moves to clinical research. If successful in the clinical setting, it then becomes available for clinical practice. The advantage of such a research model is that we do not need to understand cancer biology completely, we just need to kill the cancer.

This approach relies on trial and error until we find a useful drug. Thousands of drugs have been screened this way. Drugs have been found in different, sometimes unusual, settings: in bacteria and other microorganisms (anthracyclines from Streptomyces peucetius[4]), in vegetation (taxanes from the Pacific yew tree [5], Vinca alkaloids from Catharanthus roseus[6] and irinotecan from the Camptotheca tree [7]), and in the toxins of marine organisms (trabectedin from the sea squirt Ecteinascidia turbinata[8]). Synthetic drugs have also been developed specifically for cancer-killing purposes (antimetabolites, alkylating agents from chemical weapons [2]).

One classical concept to measure the capacity of a drug to kill cells is fractional cell kill, described as the concentration of a drug delivered during a certain time period able to kill a fraction of the cells [9]. Because cytotoxic drugs act by causing cell damage, the higher their concentration and the longer cells are exposed to them, the more effective they are in producing (normal and cancer) cell death. Consequently, the more damage the drug produces, the more likely it will be further developed. It is a model made to select the most ‘toxic’, but potentially manageable drugs.

Once drug development moves into the clinical phase, there is another important setting for increasing toxicity. Phase I studies are small studies that aim to determine the safety of new drugs in a small number of patients. The traditional study design calls for administering increasing doses of a drug to sets of three patients. The doses are escalated until two out of three (or six) patients receiving the same dose experience a limiting toxicity. When this happens, the dose is reduced to the previous level. In North America, the maximal tolerated dose (MTD) is the recommended phase II dose (RP2D), whereas in the rest of the world, the MTD is considered the dose level above the RP2D.

Phase II studies test efficacy, namely drug activity. Typically, the identified MTD, or RP2D, is delivered to a larger and more homogeneous cohort of patients, and the drug's capacity to kill the tumor is measured. A frequently used endpoint for these studies is the response rate. In other words, a visual quantification of cell killing is used, endorsing the selection of cytotoxic drugs.

Why do we have to push the drugs to the limit of toxicity in phase I? The classic answer would be that we are trying to replicate in vivo the fractional cell kill of the preclinical models, and that the higher the dose and greater the time exposure, the more cancer cells we will kill. We want ‘toxic’ drugs.

So why have we not been able to cure the majority of cancers with this approach? The answer is that toxicity is not necessarily correlated with efficacy, and unfortunately the human body is much more complex than cancer cell cultures and animal models. The simple empirical (-reactive) approach is not enough to produce highly effective drugs to overcome the biological complexity of cancer.

The old research model used to develop cytotoxics is no longer adapted to our current understanding of cancer biology. This model does not take into account several variables in the processes of carcinogenesis and progression that could determine drug efficacy and tumor sensitivity/resistance. A cell culture cannot capture all the complexity of tumor heterogeneity, its interaction with the immune system, the blood flow, and other components of a tumor's microenvironment. For example, if a drug acts by modulating the biological processes of the cancer cell, or if the microenvironment is tested using the old model, most probably the drug will be rejected because at best it will stabilize cell growth and not produce cell death. The old model is probably also not suitable enough to test drug combinations based on biological agents.

Today we have a clear view that cancer is not a single disease biologically and clinically, but a broad group of heterogeneous malignancies. Even for tumors arising in one specific organ, such as in the case of lung [10] and breast cancers [11], the complexity of the carcinogenic processes can be distinct to the point of producing malignancies with completely different prognosis and response to treatment. Tailoring therapy to match not only the complexity of cancer cell biology but also its relationship to a tumor's surrounding tissues is a main concern. Fortunately, this type of approach is becoming available in the clinical setting for some tumor types. However, to achieve such a level of tailoring, the ‘trial-and-error’ approach is no longer suitable. Understanding tumor biology is a prerequisite for improving drug development.

The old model of drug development basically answers two questions: ‘Which drugs kill cancer cells? What is the maximum dose of a drug that I can administer to humans?’ In the era of precision medicine and targeted drugs, many more questions need to be asked: ‘Why did this cell become a cancer cell? Can it be targeted? Can I selectively eliminate it? Is it possible just to stop the cancer from growing without killing it? Can I tag cancer cells to help the immune system get rid of them? Can I modulate the microenvironment to prevent cancer from spreading? Can I combine drugs with complementary mechanisms of action? Is it possible to act in different ‘compartments at the same time (cancer, microenvironment, blood vessels, and immune system)?’ ‘If so, how much of the drug(s) is (are) necessary to produce the effect? How do I combine drugs to increase treatment effect without increasing toxicity? If I combine drugs upfront, will it be possible to prevent or delay resistance?’

With our current knowledge, it is not yet possible to cure most tumor types, even at early stages. The hope for the future is to turn those incurable cancers into ‘chronic diseases’ by administering sequential effective and tolerable treatments. Targeted and sequential therapies have recently improved survival rates for some cancers. We are now able to administer several lines of treatment able to produce cancer cell death or cytostatic effects; however, at the end of the day, metastatic cancer remains a progressive condition, and the only thing that has become chronic is the administration of treatments (and their toxicities).

Our will to improve must be greater than the strength of an old habit. It is clear that the old models of drug discovery do not fit the needs of modern drug development. We could be discarding promising cancer drugs just because we do not have or we do not use the correct tools and methods to test or develop them. The driving force behind future drug development must change from ‘cancer-cell killing’ only to ‘understanding-targeting-modulating-correcting cancer’ in its microenvironment, in addition to killing cells.

In the preclinical setting, the focus must change from the drug to the disease. We have reached the limits of the efficacy of the ‘trial-and-error’ approach. Only by understanding the disease biology will we be able to develop effective drugs. Although it is not yet time to retire the ‘fractional-cell-kill’ concept, it should not be our only evaluation tool. New rational designed combinations of drugs at possibly ‘lower’ doses could have a much higher effect than a single drug administered at high concentration, if they present additive, or better, synergistic mechanisms. Again, single cell cultures do not capture the complexity of the interactions of cancer with the surrounding cells, the immune system, and blood vessels. In this regard, animal models can continue to play a role in developing selective, multitargeted drugs or combinations.

In the clinical setting, translational research, not only restricted to the so-called biomarkers but also extended to other techniques such as molecular imaging, plays a key role. Every clinical trial that does not incorporate translational research loses an opportunity to better understand cancer biology and drug efficacy and to improve future treatments. In modern clinical trials, translational research is not an extra, it is a must. Early phase clinical trials must focus on finding the ‘minimal effective dose’ and not the ‘maximum tolerated dose’ (Fig. 2 suggests a rational path for clinical drug development). Research efforts must, therefore, be made to determine surrogate markers to test drug effectiveness instead of purely toxicity. Pushing drugs to the limit of toxicity is no longer an acceptable goal, as it does not necessarily translate into better outcome. The aim of early phase trials should extend beyond the classical methodology and endpoints and become increasingly therapeutic. Early metastatic setting, neoadjuvant and window trials provide a great opportunity to evaluate tumors, their surroundings, and immune response before and after therapy, allowing us to assess the effect of our treatments not only in vivo, but also ‘in human’.

Image Tools

Initiatives to improve clinical research efficacy are already being carried out. The BATTLE (Biomarker-integrated Approaches of Targeted Therapy for Lung Cancer Elimination) trial was one of the pivotal studies to incorporate molecular markers in treatment selection. Pretreated lung cancer patients were assigned to four different treatment groups according to the presence of biomarkers analyzed in fresh tissue biopsy. A Bayesian adaptive algorithm was used for patient allocation to different treatment arms. This trial represents an important landmark for personalized treatment research for proving the feasibility of this method of clinical research exemplified by showing the efficacy of sorafenib in mutant-KRAS lung cancer patients [12]. In breast cancer, a similar strategy is also under evaluation; the I-SPY2 (NCT01042379) is a neoadjuvant trial using a similar adaptive method to assign patients to conventional chemotherapy with one of 12 investigational products [13]. Table 1 shows a sample of genotype-driven phase 1 clinical trials in oncology. A personalized medicine program initiated in 2007 performed molecular testing of 1283 patients with advanced cancers referred to a phase 1 program. Patients were treated with matched therapy, when a targetable molecular aberration was identified, or with standard phase 1 treatment, if no aberration was identified or no matched therapy was available. Patients treated with matched therapy had a higher response rate (27 versus 5%, P < 0.0001) and experienced longer time to treatment failure (5.2 versus 3.1 months, P < 0.0001) and overall survival (13.4 versus 9 months, P = 0.017) [14].

Table 1
Table 1
Image Tools

Functional imaging guidance is a promising instrument to improve clinical research. It is a valuable tool to evaluate early treatment response, with the potential to help patient selection, quick treatment dose escalation, and selecting biopsy sites. In a correlated study of the neoadjuvant Neo ALTTO trial, HER2-positive breast cancer patients were evaluated with 18F-FDG PET/computed tomography (CT) at baseline and after 2 and 6 weeks of treatment. This study showed that metabolic responders had higher pathologic complete response rates than nonresponders [15]. Similarly, in another analysis of two phase I trials determining the MTD for two different MEK inhibitors, 18F-FDG PET/CT was able to predict 97% of the nonresponders according to the Response Evaluation Criteria In Solid Tumors (RECIST) criteria [16].

An example of the use of functional imaging to select patients for chemotherapy is the PEPITA (Preoperative chemosensitivity testing as Predictor of Treatment benefit in Adjuvant stage III colon cancer) study in colorectal cancer [17]. In this trial, patients will receive one cycle of neoadjuvant FOLFOX and will be evaluated by 18F-FDG PET/CT at baseline and after chemotherapy. Patients will then be treated postoperatively at the physicians’ discretion. The objective is to evaluate whether 18F-FDG PET/CT responders derive a higher benefit of adjuvant chemotherapy, and in that way select patients for this therapeutic approach. The ENCHANT-1 (EvaluatiNg CHaperone inhibition by gANetespib in breasT cancer) is an additional trial that incorporates PET as a marker of early response. This is a phase II metastatic breast cancer study in which patients will be treated with first-line ganetespib and evaluated with PET/CT prior to and after its administration. Early metabolic nonresponders will be excluded from the trial and receive standard of care, whereas responders will continue to receive the study drug until disease progression [18].

Functional imaging can also be used for the identification of drug targets. This strategy has been tested with success in different disease settings. A radiolabeled trastuzumab molecule was synthesized to evaluate the effectiveness of Hsp90 inhibition in animal models. A Fab fragment of trastuzumab was labeled with 68Ga, a positron emitter, which allows the sequential PET imaging of HER2 expression. It was used to quantify the loss and recovery of HER2 induced by an Hsp90 inhibitor [19]. In another innovative approach, a labeled antibody targeting L858R mutant epidermal growth factor receptor (EGFR) receptor was created. This mutation confers resistance to EGFR treatment strategies. The radiolabeled antibody was successful in identifying the mutant EGFR receptor in mouse models that can lead to better patient selection for anti-EGFR treatment in the near future [20]. In prostate cancer xenograft models, the activity of an antiandrogen treatment was measured through PET imaging with a fully humanized, radiolabeled antibody against the prostate-specific membrane antigen (PSMA). Variations in the expression of this marker were used to quantitatively measure the treatment response [21].

Sequential tumor biopsy is a feasible [22,23] and useful tool to evaluate tumor biology before and after drug exposure, being able to assess drug effects ‘in human’. In a breast cancer neoadjuvant study, patients not achieving pathologic complete response had their residual tumors sequenced. A great diversity of molecular alterations was observed in the residual disease, providing insights into the disease biology and resistance mechanisms that will hopefully help to better tailor future adjuvant studies [24].

An approach to tackle the emergence of treatment resistance is the sequencing of circulating cell-free tumor DNA released from the cancer cells into the plasma. The increase in the quantification of mutant alleles was correlated with resistance to treatments. The emergence of PIK3CA mutation was associated with resistance to paclitaxel. Similarly, the detection of plasmatic mutated alleles of RB1, MED1, and EGFR was related to resistance to cisplatin, tamoxifen and trastuzumab, and gefitinib, respectively [25]. In another proof-of-concept study measurement of circulating tumor DNA was more effective in evaluation of treatment response of metastatic breast cancer patients than CA15-3 and circulating tumor cell quantification [26].

In conclusion, ‘better trials for better drugs’ are already a strategy that can be seen in the horizon. Adaptive clinical trials using biomarkers to assign patient selection were proven to be feasible and are a reality in oncology clinical research. New translational tools are likely to optimize the ability to evaluate tumor response, to identify drug targets, and to detect the emergence of drug resistance. It is, however, unlikely that one of these approaches will be able to, independently, answer all the questions that modern oncological clinical research poses. They make themselves part of the major effort of the scientific community to overcome the challenges in cancer drug development and the hope is that, combined, they will shorten drug development and fulfill the goal of personalized cancer medicine.

Back to Top | Article Outline



Back to Top | Article Outline
Conflicts of interest

The authors have no conflicts of interest.

Back to Top | Article Outline


1. Reminick G. Nightmare in Bari: the World War II Liberty ship poison gas disaster and coverup. Palo Alto, California:Glencannon Press; 2001 .

2. Goodman LS, Gilman A, Brunton LL. Goodman & Gilman's manual of pharmacology and therapeutics. 12th ed. New York:McGraw-Hill Medical; 2008 .

3. National Cancer Act of 1971 – OGCR.[Accessed 13 November 2012].

4. Grein A. Antitumor anthracyclines produced by Streptomyces peucetius. Adv Appl Microbiol. 1987; 32:203–214.

5. Adams JD, Flora KP, Goldspiel BR, et al. Taxol: a history of pharmaceutical development and current pharmaceutical concerns. J Natl Cancer Inst Monographs. 1993. 141–147.

6. Himes RH. Interactions of the catharanthus (Vinca) alkaloids with tubulin and microtubules. Pharmacol Ther. 1991; 51:257–267.

7. Torck M, Pinkas M. Camptothecin and derivatives: a new class of antitumor agents. J Pharm Belg. 1996; 51:200–207.

8. Vincenzi B, Napolitano A, Frezza AM, et al. Wide-spectrum characterization of trabectedin: biology, clinical activity and future perspectives. Pharmacogenomics. 2010; 11:865–878. doi: 10.2217/pgs.10.69

9. Wilcox WS, Schabel FM, Skipper HE. Experimental Evaluation of Potential Anticancer Agents XV. On the relative rates of growth and host kill of ‘single’ leukemia cells that survive in vivo cytoxan therapy. Cancer Res. 1966; 26:1009–1014.

10. Pao W, Girard N. New driver mutations in nonsmall-cell lung cancer. Lancet Oncol. 2011; 12:175–180. doi:10.1016/S1470-2045(10)70087-5

11. Perou CM, Sørlie T, Eisen MB, et al. Molecular portraits of human breast tumours. Nature. 2000; 406:747–752. doi:10.1038/35021093

12. Kim ES, Herbst RS, Wistuba II, et al. The BATTLE trial: personalizing therapy for lung cancer. Cancer Discov. 2011; 1:44–53. doi:10.1158/2159-8274.CD-10-0010

13. Barker AD, Sigman CC, Kelloff GJ, et al. I-SPY 2: an adaptive breast cancer trial design in the setting of neoadjuvant chemotherapy. Clin Pharmacol Ther. 2009; 86:97–100. doi:10.1038/clpt.2009.68

14. Tsimberidou AM, Iskander NG, Hong DS, et al. Personalized medicine in a phase I clinical trials program: the MD Anderson Cancer Center initiative. Clin Cancer Res. 2012; 18:6373–6383. doi:10.1158/1078-0432.CCR-12-1627

15. Gebhart G, Gámez C, Holmes E, et al. 18F-FDG PET/CT for early prediction of response to neoadjuvant lapatinib, trastuzumab, and their combination in HER2-positive breast cancer: results from Neo-ALTTO. J Nucl Med. 2013; 54:1862–1868.

16. Kraeber-Bodéré F, Carlier T, Naegelen VM, et al. Differences in the biologic activity of 2 novel MEK inhibitors revealed by 18F-FDG PET: analysis of imaging data from 2 phase I trials. J Nucl Med. 2012; 53:1836–1846. doi:10.2967/jnumed.112.109421

17. Hendlisz A, Golfinopoulos V, Deleporte A, et al. Preoperative chemosensitivity testing as Predictor of Treatment benefit in Adjuvant stage III colon cancer (PePiTA): protocol of a prospective BGDO (Belgian Group for Digestive Oncology) multicentric study. BMC Cancer. 2013; 13:190doi:10.1186/1471-2407-13-190

18. Awada A, Rodriguez AA, Kong A, et al. The ENCHANT-1 trial (NCT01677455): an open label multicenter phase II proof of concept study evaluating first-line ganetespib monotherapy in women with metastatic HER2-positive or triple-negative breast cancer (TNBC). J Clin Oncol. 2013; 31:(suppl):abstr TPS1136[Accessed 22 November 2013].

19. Smith-Jones PM, Solit DB, Akhurst T, et al. Imaging the pharmacodynamics of HER2 degradation in response to Hsp90 inhibitors. Nat Biotechnol. 2004; 22:701–706. doi:10.1038/nbt968

20. Yeh HH, Ogawa K, Balatoni J, et al. Molecular imaging of active mutant L858R EGF receptor (EGFR) kinase-expressing nonsmall cell lung carcinomas using PET/CT. Proc Natl Acad Sci U S A. 2011; 108:1603–1608. doi:10.1073/pnas.1010744108

21. Evans MJ, Smith-Jones PM, Wongvipat J, et al. Noninvasive measurement of androgen receptor signaling with a positron-emitting radiopharmaceutical that targets prostate-specific membrane antigen. Proc Natl Acad Sci U S A. 2011; 108:9578–9582. doi:10.1073/pnas.1106383108

22. Gomez-Roca CA, Lacroix L, Massard C, et al. Sequential research-related biopsies in phase I trials: acceptance, feasibility and safety. Ann Oncol. 2012; 23:1301–1306. doi:10.1093/annonc/mdr383

23. Lee J, Hays JL, Noonan AM, et al. Feasibility and safety of sequential research-related tumor core biopsies in clinical trials. Cancer. 2013; 119:1357–1364. doi:10.1002/cncr.27916

24. Balko JM, Cook RS, Vaught DB, et al. Profiling of residual breast cancers after neoadjuvant chemotherapy identifies DUSP4 deficiency as a mechanism of drug resistance. Nat Med. 2012; 18:1052–1059. doi:10.1038/nm.2795

25. Murtaza M, Dawson S-J, Tsui DWY, et al. Noninvasive analysis of acquired resistance to cancer therapy by sequencing of plasma DNA. Nature. 2013; 497:108–112. doi:10.1038/nature12065

26. Dawson SJ, Tsui DWY, Murtaza M, et al. Analysis of circulating tumor DNA to monitor metastatic breast cancer. N Engl J Med. 2013; 368:1199–1209. doi:10.1056/NEJMoa1213261

© 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins