Nat Rev Drug Dis_2005

May 2005 Vol 4 No 5 S5 | Editorial

doi:10.1038/nrd1748

S6 | Jean-Pierre Issa, Hagop Kantarjian & Peter Kirkpatrick doi:10.1038/nrd1726

S8 | Lee M. Ellis & Peter Kirkpatrick doi:10.1038/nrd1727

S10 | Richard Goldberg & Peter Kirkpatrick doi:10.1038/nrd1728

S12 | Ching-Hon Pui, Sima Jeha & Peter Kirkpatrick doi:10.1038/nrd1729

S14 | Jonathan Dowell, John D. Minna & Peter Kirkpatrick doi:10.1038/nrd1730

S16 | David Gibbs, Ann Jackman & Peter Kirkpatrick doi:10.1038/nrd1731

S19 | doi:10.1038/nrd1737

2005 Nature Publishing Group

Welcome to Hot Drugs Cancer 2005, a collection of articles on the new oncology drugs approved by the US FDA last year. The articles provide a two-page snapshot of key information

on the featured drugs. The first page discusses the science behind the drug and how it was developed, including background on the disease area, the rationale for the development of the drug, its mechanism of action and major clinical results. And on the second page, experts in the field consider questions related to advances in the treatment of the disease in question and the clinical impact of the new drug.

The drugs featured represent an intriguing mix of old and new approaches to cancer therapy. Azacitidine, now approved for the treatment of myelodysplastic syndromes, was originally turned down for approval as a cytotoxic agent more than twenty years ago. However, since then, its ability to exert antineoplastic effects by reversing aberrant DNA methylation has been recognized. And clofarabine and pemetrexed are both examples of how optimization of traditional anticancer drug classes nucleoside analogues and antifolates, respectively can still result in important treatment advances.

The other three drugs highlighted all come from the new breed of molecularly targeted agents. Bevacizumab, an antibody against vascular endothelial growth factor, could be viewed as a culmination of more than thirty years of research aimed at targeting tumour angiogenesis as a therapeutic strategy. Completing the collection, the antibody cetuximab and the small-molecule kinase inhibitor erlotinib represent different approaches to targeting another key signalling pathway in cancer, the epidermal growth factor receptor pathway. Finally, complementary market data for oncology drug development is also provided.

We are pleased to acknowledge the financial support of Genentech, ImClone Systems and Roche in the production of Hot Drugs Cancer 2005. Thanks to this support, the content will be freely available online for six months on a dedicated web focus, which can be found at http://www.nature.com/nrd/focus/hotdrugs/2005/index.html. The editorial content is the sole responsibility of the Nature Publishing Group.

M AY 20 05 | S5

I N T R O DU C T I O N

EDITORIAL OFFICENature Publishing Group The Macmillan Building4 Crinan StreetLondon N1 9XW, UKTel: +44 (0)20 7833 4000Fax: +44 (0)20 7843 3629

Supplement Editor: Peter KirkpatrickEditorial Production: Susanne HarrisCopy Editor: Daniel JonesManaging Production Editor: Judith ShadwellElectronic Production: Alexander ThurrellSponsorship: Claire HinesDesign and Layout (Sponsor Profiles): Suzanne ColemanProduction: Kelly HopkinsPublisher: Adam SmithEditor-in-Chief, Nature Publications: Philip CampbellPublishing Director: Peter Collins

World Wide Web: www.nature.com/nrd/focus/hotdrugs/2005

NATURE REVIEWS JOURNALSNature Reviews CancerChief Editor: Ezzie Hutchinsonwww.nature.com/reviews/cancer

Nature Reviews Drug DiscoveryChief Editor: Peter Kirkpatrickwww.nature.com/reviews/drugdisc

Nature Reviews GeneticsChief Editor: Magdalena Skipperwww.nature.com/reviews/genetics

Nature Reviews ImmunologyChief Editor: Elaine Bellwww.nature.com/reviews/immunol

Nature Reviews MicrobiologyChief Editor: David OConnellwww.nature.com/reviews/micro

Nature Reviews Molecular Cell BiologyChief Editor: Arianne Heinrichswww.nature.com/reviews/molcellbio

Nature Reviews NeuroscienceChief Editor: Rachel Joneswww.nature.com/reviews/neuro


Azacitidine

H O T D R U G S | C A N C E R

S6 | MAY 2005 www.nature.com/reviews/drugdisc

Azacitidine

Azacitidine (Vidaza; Pharmion), an inhibitorof DNA methylation, was approved by theUS FDA for the treatment of myelodysplasticsyndromes in May 2004. It is the first drug tobe approved by the FDA for treating this rarefamily of bone-marrow disorders, and hasbeen given orphan-drug status. It is also apioneering example of an agent that targetsepigenetic gene silencing, a mechanismthat is exploited by cancer cells to inhibit theexpression of genes that counteract themalignant phenotype.

Myelodysplastic syndromes (MDS) are a het-erogeneous group of bone-marrow disordersthat are characterized initially by ineffectivehaematopoesis (the formation and develop-ment of blood cells) and subsequently by thefrequent development of acute myeloid leuk-aemia (AML)1,2. Typical symptoms of haem-atopoietic failure include infection, bleeding,bruising and fatigue1,2.

Standard treatment for patients with MDSconsists of supportive measures such as bloodtransfusions and administration of haemato-poetic factors to correct cytopaenias, andantibiotics to treat opportunistic infections1,2.At present, the primary potentially curativetreatment is allogeneic stem-cell transplanta-tion, but such therapy is often not appropriatefor MDS patients owing to their advanced ageor accompanying diseases1,2. There is thereforea great need for novel agents for the manage-ment of MDS. Advances in the understandingof the pathogenesis of MDS have led to thetesting of a number of such agents, includingDNA-methylation inhibitors such as azacitidine(FIG. 1).

Basis of discoveryDNA methylation is a key epigenetic mecha-nism that results in the heritable silencing ofgenes without a change in their codingsequence3,4. Epigenetic processes are requiredfor the normal development of mammaliancells, but are not used for the dynamic regu-lation of gene expression1. However, it is nowknown that malignant cells can exploit theprocess of DNA methylation to silence theexpression of genes that counteract the malig-nant phenotype, such as tumour-suppressorgenes3. Leukaemias and MDS are character-ized by the hypermethylation and conse-quent silencing of multiple genes1,3, whichhas led to interest in inhibiting DNA methyl-ation as a therapeutic strategy for treatingthese diseases.

Drug propertiesAzacitidine is a nucleoside analogue of cyti-dine that specifically inhibits DNA methyl-ation by trapping DNA methyltransferases4

(FIG. 1). It was originally developed as a cyto-toxic agent, and an application to the FDArequesting its approval as such was turneddown more than 25 years ago. The discoveryin the early 1980s that it was a hypomethyl-ating agent4, and the elucidation of the role ofDNA hypermethylation in cancer, haveprompted its re-evaluation and eventuallyled to its recent approval. Azacitidine isthought to exert its antineoplastic effects inpart by causing hypomethylation of DNAand consequent reactivation of previouslysilenced genes, including tumour-suppressorgenes4,5. It might also have activity throughincorporation into RNA, as well as directcytotoxicity.

Clinical dataA randomized, open-label, controlled trialcompared the safety and efficacy of subcuta-neous azacitidine plus supportive care withsupportive care alone (observation group) in~200 patients with any of the five subtypes ofMDS5. Azacitidine was administered at asubcutaneous dose of 75 mg per m2 daily forseven days every 4 weeks; the dose wasincreased to 100 mg per m2 if no beneficialeffect was seen after two treatment cycles5.Patients in the observation group were allowedto cross over to the azacitidine group on thebasis of several criteria of worsening disease5.

The primary efficacy endpoint wasresponse rate. After excluding several patientswho were found to have AML at baseline, theoverall response rate (complete response andpartial response) in patients who weretreated with azacitidine was 15.7%, whichwas statistically significantly higher than theresponse rate of 0% in the observation group(before crossing over) 5. Response occurred inall MDS subtypes, as well as in patients withbaseline diagnosis of AML, and the medianduration of clinical response (complete orpartial response) was estimated as 330 days;75% of the responding patients were still inpartial response or better at completion oftreatment5. Patients who crossed over toreceive azacitidine had a response rate of12.8%5.

Azacitidine was also studied in a similarmanner in two single-arm, open-label studiesin patients with MDS and AML. Similar pro-portions of patients had a complete or partialresponse to azacitidine as in the randomizedtrial5.

IndicationsAzacitidine is approved by the FDA for thetreatment of patients with the following MDSsubtypes: refractory anaemia or refractoryanaemia with ringed sideroblasts (if accompa-nied by neutropaenia or thrombocytopaeniaor requiring transfusions), refractory anaemiawith excess blasts, refractory anaemia withexcess blasts in transformation, and chronicmyelomonocytic leukaemia5.

OO

OH OH

HO N

NN

NH2

a b

Z

Z

Z

DNAreplication

Strandseparation

DNMTs

CAS number: 320-67-2

Azacitidine4-amino-1--D-ribofuranosyl-s-triazin-2(1H)-one;C8H12N4O5; Mr = 244;

Figure 1 | Azacitidine and DNA methylation. a | Azacitidine. b | A family of DNA methyltransferases(DNMTs) catalyse the methylation of the 5 position of the cytosine ring. After intracellular conversion to 5-aza-2-deoxycytidine (decitabine), azacitidine (Z) is incorporated in place of cytidine into DNA, where it actsas a direct and irreversible inhibitor of DNMTs. Cells then divide in the absence of DNMTs, which results inprogressive DNA hypomethylation and reactivation of previously silenced genes1,3,4. Azacitidine alsoincorporates into RNA, but very little is known about the effects of this. Pink circles, methylated CpG;yellow circles, unmethylated CpG. Adapted from REF. 4.


NATURE REVIEWS | DRUG DISCOVERY MAY 2005 | S7


Analysing clinical issues for myelodysplasticsyndrome (MDS) and epigenetic therapies ingeneral are Jean-Pierre Issa, M.D., Director ofthe Translational Research Laboratories in theDepartment of Leukemia, and HagopKantarjian, M.D., Chairman of the Departmentof Leukemia at the University of Texas MDAnderson Cancer Center, USA. Dr Issas interestsare the roles of epigenetic changes in ageing andcancer, and the rapid translation of basic researchin epigenetics to the treatment of patients withleukaemias and solid tumours. Dr Kantarjiansinterests are novel therapies for leukaemias.

What are the current key issues in the treatmentof MDS?MDS is a very heterogeneous group of diseaseswith vastly different outcomes depending onclinico-pathological features1: it can rangefrom a chronic, indolent disease to an aggressivemalignancy with a median survival similar tothat of metastatic lung cancer. A key issue inthe management of MDS is to reduce this het-erogeneity using molecular profiles, which willthen be essential to select patients for therapy.The International Prognostic Scoring System(IPSS) uses a combination of clinical datawith cytogenetics to achieve a certain degree ofselection6, but it remains imperfect. Moreimportantly, the IPSS score does not predictwho is going to respond to azacitidine or othertreatment modalities in MDS. A concertedeffort to supplement the IPSS using gene-expression profiles, DNA-methylation profilesor higher-resolution chromosomal analysis isessential to making progress in this disease.Given that the reported major response rate toazacitidine is less than 20%7, this effort shouldthen be applied to identifying those patientsmost likely to respond to this drug.

How would you expect MDS treatment toevolve in the next 5 years? Azacitidine should be the cornerstone of effortsto move MDS treatment towards real life-prolonging options. Given that MDS is pri-marily a disease of older individuals, aggressivetherapies such as combination chemotherapyand stem-cell transplantation are simply notrealistic for most patients. There is much inter-est therefore in exploring less toxic agents inthis disease, and learning how to integrate theminto a multi-agent therapeutic approach.

Among the classical therapies, the cytotoxiccytosine analogue cytarabine has a response rateof about 15% in MDS when used at low doses,and the topoisomerase inhibitor topotecan alsohas reported activity in MDS8. There is also

reported synergy between hypomethylatingagents and topoisomerase inhibitors9, whichshould be explored in MDS. A drawback forthese approaches is toxicity, which has to bekept at a minimum in an elderly population.Given that all patients who initially respondto azacitidine eventually relapse, it might beattractive to investigate whether agents suchas cytarabine and topotecan can prolongremissions in this disease.

Newer agents in MDS that are attractingattention include the nucleoside analogueclofarabine10, the farnesyl-transferase inhibitortipifarnib8 and the thalidomide analoguelenalidomide11. Although all these agents areclinically promising, there is no compellingrationale for combining them with azacitidine.The deoxyribose form of azacitidine, decitabine,also has significant clinical activity in MDS. Thiscompound is a more potent hypomethylatingagent, but does not incorporate into RNA. Itsrelative clinical efficacy compared to azacitidinedeserves investigation. It is also not knownwhether there is complete overlap between themechanism of clinical activity of the two com-pounds. From a molecular standpoint, the mostrational combinations involving azacitidine ordecitabine are combination epigenetic therapy,and combinations that exploit gene reactiva-tion. DNA methylation-associated gene silen-cing in cancer involves a cascade of orchestratedmolecular events12 that include methyl-bindingproteins, histone deacetylases, histone methy-lases and polycomb-group proteins. Each ofthese pathways is attractive as a target for epi-genetic therapy, and histone-deacetylaseinhibitors are already in clinical trials.Valproicacid, a widely used medication for epilepsy, isalso a histone-deacetylase inhibitor. Remark-ably, it has single-agent activity (thoughmodest) in MDS13, and trials of combinationsof valproic acid with hypomethylating agentsare ongoing. Finally, given that azacitidine acti-vates the expression of readily targetable path-ways4, combinations that exploit this are alsoattractive. For example, hypomethylating agentssensitize many cells to the action of retinoicacid, and trials of combinations of azacitidinewith all-trans retinoic acid are ongoing.

How applicable might DNA-methylationinhibitors be to other leukaemias and cancers? One of the most important areas of researchis understanding the azacitidineMDS con-nection. There is nothing extraordinary aboutDNA methylation in MDS15, and why thatdisease should be particularly responsive isunknown. There is evidence that azacitidine has

activity in AML16, but it could be argued thatAML and MDS form a continuum of diseasesto explain this fact. Possible explanations for thecurrent clinical state of affairs can be found inthe facts that hypomethylating agents workbetter and with less toxicity at low doses17, andthat epigenetic therapy works slowly7. MDS is adisease of elderly patients with few therapeuticoptions and a more indolent course than acuteleukaemias. By necessity, investigators interestedin MDS had to test agents at low, non-toxicdoses, and could wait several months beforedetermining response or failure. The discoveryof MDS as a disease responsive to hypomethyl-ating agents might owe more to these issuesthan to real biology. If this hypothesis is correct,epigenetic therapy at low doses extended over aperiod of several months of treatment mightwell prove useful in other malignancies.

Jean-Pierre Issa and Hagop Kantarjian are at theDepartment of Leukemia, University of Texas MDAnderson Cancer Center, Houston, Texas 77030,USA. Peter Kirkpatrick is at Nature Reviews DrugDiscovery. Correspondence to J.-P. I. & P. K.e-mails: [email protected];[email protected]

doi:10.1038/nrd1726

1. List, A. F., Vardiman, J., Issa, J.-P. & DeWitte, T. M.Myelodysplastic syndromes. Hematology (Am. Soc.Hematol. Educ. Program) 297317 (2004).

2. Hofmann, W.-K. et al. Myelodysplastic syndromes.Hematol. J. 5, 18 (2004).

3. Herman, J. G. & Baylin, S. B. Gene silencing in cancer inassociation with promoter hypermethylation. N. Engl. J.Med. 349, 20422054 (2003).

4. Egger, G., Liang, G., Aparicio, A. & Jones, P. A. Epigeneticsin human disease and prospects for epigenetic therapy.Nature 429, 457463 (2004).

5. FDA labelling information [online], (2004).

6. Greenberg, P. et al. International scoring system forevaluating prognosis in myelodysplastic syndromes. Blood89, 20792088 (1997).

7. Silverman, L. R. et al. Randomized controlled trial ofazacitidine in patients with the myelodysplastic syndrome: a study of the cancer and leukemia group B. J. Clin. Oncol.20, 24292440 (2002).

8. Faderl, S. & Kantarjian, H. M. Novel therapies formyelodysplastic syndromes. Cancer 101, 226241 (2004).

9. Anzai, H., Frost, P., & Abbruzzese, J. L. Synergisticcytotoxicity with 2-deoxy-5-azacytidine and topotecan invitro and in vivo. Cancer Res. 52, 21802185 (1992).

10. Kantarjian, H. et al. Phase 2 clinical and pharmacologicstudy of clofarabine in patients with refractory or relapsedacute leukemia. Blood 102, 23792386 (2003).

11. List, A. et al. Efficacy of lenalidomide in myelodysplasticsyndromes. N. Engl. J. Med. 352, 549557 (2005).

12. Bird, A. DNA methylation patterns and epigenetic memory.Genes Dev. 16, 621 (2002).

13. Kuendgen, A. et al. Treatment of myelodysplasticsyndromes with valproic acid alone or in combination withall-trans retinoic acid. Blood 104, 12661269 (2004).

14. Wijermans, P. et al. Low-dose 5-aza-2-deoxycytidine, aDNA hypomethylating agent, for the treatment of high-riskmyelodysplastic syndrome: a multicenter phase II study inelderly patients. J. Clin. Oncol. 18, 956962 (2000).

15. Claus, R. & Lubbert, M. Epigenetic targets in hematopoieticmalignancies. Oncogene 22, 64896496 (2003).

16. Saiki, J. H. et al. 5-azacytidine in acute leukemia. Cancer42, 21112114 (1978).

17. Issa, J. P. et al. Phase I study of low-dose prolongedexposure schedules of the hypomethylating agent 5-aza-2-deoxycytidine (Decitabine) in hematopoieticmalignancies. Blood 103, 16351640 (2004).

EPIGENETIC THERAPIES AND MDS | VIEW FROM THE CLINIC


Bevacizumab



Bevacizumab

Bevacizumab (Avastin; Genentech/Roche),an antibody against vascular endothelialgrowth factor, was approved by the US FDAin February 2004 for the first-line treatmentof metastatic colorectal cancer incombination with 5-fluorouracil-basedchemotherapy. It is the first approved agentto target tumour angiogenesis.

Angiogenesis the development of new bloodvessels from pre-existing vasculature has akey role in normal development, but also inseveral diseases, such as cancer. For a tumour togrow beyond a certain size, it needs a network ofblood vessels that supply nutrients and oxygen,and remove waste products1. So, understandinghow angiogenesis is regulated in cancer has beena major area of cancer research, particularly inthe past decade.

Basis of discoveryThe hypothesis that inhibiting angiogenesismight be an effective anticancer strategy wasput forward more than 30 years ago2. Earlierexperiments had provided evidence that

tumour angiogenesis was mediated by dif-fusible factors produced by tumour cells, andthis hypothesis stimulated efforts to identifythese factors1.

In the 1980s, such efforts led to the isola-tion of vascular endothelial growth factor(VEGF), a potent stimulator of the growthof endothelial cells, the main type of cell inthe inside lining of blood vessels1. It is nowknown that VEGF, which activates receptortyrosine kinases on the surface of endothelialcells (FIG. 1), is a key regulator of normal andpathological blood vessel growth1. In theearly 1990s, the demonstration that inhibi-tion of VEGF-induced angiogenesis using amonoclonal antibody against VEGF markedlysuppressed tumour growth in vivo3 led to thedevelopment of bevacizumab.

Drug propertiesBevacizumab is a humanized version of themonoclonal antibody against VEGF that wasused in early proof-of-principle experiments3,4.It binds to VEGF, which prevents its interac-tion with the VEGF receptor tyrosine kinases

VEGFR1 and VEGFR2 (FIG. 1), and inhibits thegrowth of human tumour cell lines in mice1.Following early clinical trials showing thatbevacizumab as a single agent was relativelynon-toxic, and that it could be added to stan-dard cytotoxic chemotherapy regimes, largeclinical trials were initiated in several cancertypes, including colorectal cancer1.

Clinical dataIt has been estimated that colorectal canceris the fourth largest cause of cancer deaths5.For many years, chemotherapy of the diseasehas been based on traditional cytotoxicdrugs, in particular 5-fluorouracil (5-FU) incombination with leucovorin (LV). Morerecently, the addition of the cytotoxic drugsirinotecan (Camptosar; Pfizer) or oxaliplatin(Eloxatin; Sanofi-Aventis) to 5-FU/LVregimes has been shown to prolong survivalin advanced colorectal cancer.

Bevacizumab, in combination with 5-FU-based chemotherapy, was studied in random-ized, controlled trials as a first-line treatmentfor patients with metastatic carcinoma of thecolon or rectum6,7. In the pivotal Phase IIIstudy, which involved more than 800 patients,bevacizumab (5 mg per kg every 2 weeks) orplacebo was given to the patients in additionto bolus-IFL (irinotecan 125 mg per m2

intravenously, 5-FU 500 mg per m2 intra-venously and LV 20 mg per m2 intravenouslyonce weekly for 4 weeks every 6 weeks).Median overall survival was significantlyincreased from 15.6 months in the bolus-IFL+ placebo arm to 20.3 months in the bolus-IFL + bevacizumab arm6,7. Similar increaseswere also seen in progression-free survival(6.4 versus 10.6 months), overall responserate (35 % versus 45%) and duration ofresponse (7.1 months versus 10.4 months).

Bevacizumab was generally well toleratedin these clinical trials; however, some seriousand unusual toxicities were noted. In particular,bevacizumab was associated with gastro-intestinal perforations and wound healingcomplications in ~2% of patients7. Otheradverse events associated with bevacizumabuse include thromboembolic complications,hypertension, bleeding and proteinuria7.

IndicationsBevacizumab, used in combination with intra-venous 5-FU-based chemotherapy, is approvedby the FDA for the first-line treatment ofpatients with metastatic carcinoma of the colonor rectum7.

VEGFR1 VEGFR2

Decoy effect on VEGF signalling Induction of uPA, tPA, MMP9 Vascular-bed-specific release of growth factors

Proliferation Migration Survival Angiogenesis Permeability

VEGF

Bevacizumab

Figure 1 | Simplified view of VEGF signalling and tumour angiogenesis. The receptors for vascularendothelial growth factor (VEGF, also known as VEGF-A) VEGFR1 (also known as Flt1) and VEGFR2(also known as Flk1 or KDR) are expressed on the surface of blood endothelial cells. VEGFR2 isthought to be the major mediator of endothelial cell mitogenesis, survival and microvascular permeability,whereas VEGFR1 does not seem to mediate an effective mitogenic signal in endothelial cells, but doeshave other activities that can be important in tumour growth and metastasis, including the induction ofmatrix metalloproteinases (MMPs). tPA, tissue plasminogen activator; uPA, urokinase-type plasminogenactivator. Adapted from REF. 1.




Analysing clinical issues for anti-VEGFtherapies is Lee M. Ellis, M.D., Professor ofSurgical Oncology and Professor of CancerBiology at the University of Texas MDAnderson Cancer Center, Texas, USA. Hisresearch interests are defining the role ofangiogenesis in primary and metastatic humancolorectal, pancreatic and gastric cancers, andevaluating mechanisms of anti-angiogenictherapies, such as those targeted against VEGF.

What impact have anti-VEGF therapies such asbevacizumab had on the treatment of cancer? The results from the original Phase III trialscomparing IFL to IFL and bevacizumab6, andthe subsequent FDA approval of bevacizumabin combination with intravenous 5-FU-basedregimens, have changed the standard of carefor patients with metastatic colorectal cancer(CRC). More recently, it has been announcedthat in Phase III trials, bevacizumab providedbenefit to patients with advanced non-small-celllung cancer (NSCLC). In addition, Phase I/IIstudies have suggested that the addition ofbevacizumab to other therapeutic modalities(chemotherapy and radiation therapy) mightbe efficacious in patients with other malignan-cies, such as pancreatic carcinoma, renal-cellcarcinoma and locally advanced rectal cancer.

Another Phase III trial in metastatic CRCevaluated patients randomized to FOLFOXwith or without the VEGF receptor tyrosinekinase inhibitor PTK787/ZD222584. It has beenannounced that after ~1,200 patients had beenaccrued , an investigator analysis demonstratedan improvement in progression-free survival,but a central radiology review did not note astatistically significant improvement in progres-sion-free survival. Data on overall survivalshould be available in 2006.

How well are the mechanisms of benefit of anti-VEGF therapies understood?It is important to differentiate anti-VEGFtherapy from other anti-angiogenic therapies,many of which are specifically intended todisrupt endothelial cell signalling or survivalpathways. VEGF is a pluripotent factor thathas multiple effects on the vasculature, includ-ing induction of angiogenesis and enhance-ment of endothelial cell survival, the ability toinduce vascular permeability, and productionof a dilated and disorganized vascular net-work. This altered tumour vascular networkleads to inefficient blood flow, which can the-oretically be normalized with anti-VEGFtherapy, augmenting delivery of chemotherapyand oxygen8.

It was initially believed that the VEGF recep-tors (VEGFRs) are present only on endothelialcells, but recent studies have demonstrated thatVEGFRs are present on tumour cells9,10. So,another potential mechanism of action of anti-VEGF therapy is a direct effect on tumour cells,where it might inhibit processes involved intumour progression and metastasis.

In addition to the above novel mechanismsfor anti-VEGF therapy, it is still believed thatanti-VEGF therapy can inhibit tumour angio-genesis. A detailed analysis has shown that theaddition of bevacizumab to chemotherapyleads to a remarkable improvement in progres-sion-free survival relative to the incrementalimprovement in response rate as seen withchemotherapy alone (A. Grothey, personalcommunication). This observation suggeststhat anti-VEGF therapy might indeed be anti-angiogenic, although this is indirect evidence.

How do you see anti-VEGF therapy evolving inthe next few years?Although most studies with bevacizumab havebeen carried out in patients with advanced-stagedisease, future studies will include its use incombination with other therapies in the neo-adjuvant and adjuvant settings. Interestingresults have already been obtained with the useof bevacizumab in addition to chemoradiationtherapy for locally advanced rectal and pan-creatic cancers11,12, and ongoing trials in CRCshould provide some insight into the appropri-ate use of this agent in the adjuvant setting.However, one must consider the long-termeffects of bevacizumab. In a meta-analysis ofrandomized Phase III trials, a 2.3-fold increasein arterial thrombotic events, including stroke,myocardial infarction, angina and transientischaemic attacks, was noted. This is particularlyimportant in patients who will be receivingbevacizumab in the adjuvant setting, and long-term follow up of cardiovascular events is cru-cial in such studies, especially as many patientswould remain disease-free without therapy orwith chemotherapy alone as adjuvant therapy.

Although anti-angiogenic therapy was origi-nally intended for use as single-agent therapy,and was then combined with chemotherapy,there is also great potential for use in combina-tion with other biological agents. One veryinteresting observation during the past year hasbeen the results from the BOND2 study (a fol-low-up of the BOND1 study of cetuximab withor without irinotecan in irinotecan-refractorypatients). In this study, patients with metastaticCRC who had progressed on irinotecan-basedtherapy were randomized to receive cetuximab

plus bevacizumab or cetuximab plus beva-cizumab plus irinotecan13. The combination ofthe two biological agents alone led to a remark-able 23% response rate in the refractory setting,and the addition of irinotecan to the two bio-logical agents led to a 38% response rate. Theseresults suggest that biological agents can becombined without undue toxicity and that dualtargeting of various mediators of tumour pro-gression is efficacious in and of itself. However,the addition of chemotherapy to the biologicalagents is likely to achieve an even better result.

In summary, it is clear that bevacizumabimproves the effects of chemotherapy inpatients with metastatic CRC and NSCLC, andpossibly other cancers. The efficacy of single-agent anti-VEGF therapy requires further study,in both the advanced-disease setting as well asthe adjuvant setting. It is imperative that wedetermine the underlying mechanism of actionof anti-VEGF drugs in order to develop predic-tive markers and better define the optimal use ofthese agents, as well as their potential short- andlong-term toxicities.

Lee M. Ellis is at the University of Texas MDAnderson Cancer Center, PO Box 301402,Houston, Texas 77230-1402, USA. PeterKirkpatrick is at Nature Reviews Drug Discovery.Correspondence to L.M.E. & P. K.e-mails: [email protected];[email protected]

doi:10.1038/nrd1727

1. Ferrara, N. et al. Discovery and development of bevacizumab,an anti-VEGF antibody for treating cancer. Nature Rev. DrugDiscov. 3, 391400 (2004).

2. Folkman, J. Tumor angiogenesis: therapeutic implications.N. Engl. J. Med. 285, 11821186 (1971).

3. Kim, K. J. et al. Inhibition of vascular endothelial growthfactor-induced angiogenesis suppresses tumour growth invivo. Nature 362, 841844 (1993).

4. Presta, L. G. et al. Humanization of an anti-vascularendothelial growth factor monoclonal antibody for thetherapy of solid tumors and other disorders. Cancer Res.57, 45934599 (1997).

5. Parkin, D. M. et al. Cancer burden in the year 2000. Theglobal picture. Eur. J. Cancer 37, S4S66 (2001).

6. Hurwitz, H. et al. Bevacizumab plus irinotecan, fluorouracil,and leucovorin for metastatic colorectal cancer. N. Engl. J.Med. 350, 23352342 (2004).

7. FDA labelling information [online] (2004).

8. Jain, R. K. Normalization of tumor vasculature: an emergingconcept in antiangiogenic therapy. Science 307, 5862 (2005).

9. Dias, S. et al. Autocrine stimulation of VEGFR-2 activateshuman leukemic cell growth and migration. J. Clin. Invest.106, 511521 (2000).

10. Fan, F. et al. Expression and function of vascular endothelialgrowth factor receptor-1 on human colorectal cancer cells.Oncogene 24, 26472653 (2005).

11. Willett, C. G. et al. Direct evidence that the VEGF-specificantibody bevacizumab has antivascular effects in humanrectal cancer. Nature Med. 10, 145147 (2004).

12. Crane, C. H. et al. Preliminary results of a Phase I study ofrhuMab VEGF (bevacizumab) with concurrent radiotherapy(XRT) and capecitabine (CAP). Proc. ASCO GastrointestinalCancers Symp. A85 (2004).

13. Saltz, L. B. et al. Interim report of randomized Phase IItrial of cetuximab/bevacizumab/irinotecan (CBI) versuscetuximab/bevacizumab (CB) in irinotecan-refractorycolorectal cancer. Proc. ASCO Gastrointestinal CancersSymp. A169b (2005).

ANTI-VEGF THERAPIES | VIEW FROM THE CLINIC


Cetuximab



Cetuximab

Cetuximab (Erbitux; ImClone Systems/Bristol-Myers Squibb) is a monoclonalantibody that binds to the epidermal growthfactor receptor, which is important in thegrowth of many cancers. In February 2004,it was granted accelerated approval by theUS FDA for the treatment of metastaticcolorectal cancer on the basis of tumourresponse rates in Phase II trials.

Colorectal cancer is one of the most com-monly diagnosed cancers, and has been esti-mated to be the fourth largest cause of cancerdeaths worldwide1. Recently, there have beenimportant advances in the therapy of this dis-ease, such as the introduction of the cytotoxicdrugs irinotecan (Camptosar; Pfizer) andoxaliplatin (Eloxatin; Sanofi-Aventis), and theantibody bevacizumab (Avastin; Genentech/Roche), which targets tumour angiogenesis.Nevertheless, a major need remains for novelagents for treating colorectal cancer, especiallyfor those patients who fail to respond to current treatments or who develop resistanceto them.

Basis of discoveryAdvances in the understanding of the aber-rant signalling pathways involved in cancerhave led to considerable efforts to developtherapies that target specific elements ofthese pathways, in the hope that such agentswill have greater activity and higher speci-ficity for tumour cells than traditional cyto-toxic drugs. Receptor tyrosine kinases which have key roles in signalling pathwaysthat regulate processes such as cell prolifera-tion, apoptosis and angiogenesis haveattracted much attention, as several of themhave been strongly implicated in tumourgrowth and progression, including the epi-dermal growth factor receptor (EGFR)2,3

(FIG. 1). The EGFR is expressed in a widerange of solid tumours, including colorectalcancers4, which has prompted the develop-ment of a number of agents that inhibit itsactivity24.

Drug propertiesCetuximab is a recombinant, human/mousechimeric monoclonal antibody that bindsspecifically to the extracellular domain of thehuman EGFR2,5,6. Bound antibody competi-tively inhibits the binding of epidermalgrowth factor (EGF) and other ligands toEGFR, blocking receptor phosphorylationand activation of receptor-associated kinases,

thereby inhibiting downstream signal trans-duction2,5,6 (FIG. 1). In preclinical studies,cetuximab showed promising anticanceractivity both alone and in combination withtraditional cytotoxic drugs2,5,7, prompting itsclinical evaluation in a range of cancers inwhich EGFR is thought to have an importantrole, including colorectal cancer.

Clinical dataCetuximab was studied in trials involvingpatients with EGFR-expressing metastaticcolorectal cancer, whose disease had pro-gressed after receiving an irinotecan-containingregimen6. In a randomized controlled trialconducted in 329 such patients, patientsreceived either cetuximab (400 mg per m2

initial dose, followed by 250 mg per m2

weekly until disease progression or unaccept-able toxicity) and irinotecan, or cetuximabalone6. Patients receiving cetuximab and

irinotecan had an objective response rate of22.9%, and those receiving cetuximab alonehad objective response rate of 10.8%. Themedian duration of response was 5.7 monthsin the combination arm and 4.2 months inthe monotherapy arm, and the median timeto disease progression was 4.1 months in thecombination arm and 1.5 months in themonotherapy arm6.

IndicationsCetuximab, used in combination with irino-tecan, is approved by the FDA for the treatmentof EGFR-expressing, metastatic colorectalcarcinoma in patients who are refractory toirinotecan-based chemotherapy6.

Cetuximab administered as a single agentis approved by the FDA for the treatment ofEGFR-expressing, metastatic colorectal car-cinoma in patients who are intolerant toirinotecan-based chemotherapy6.

TK TK TK TK

Autophosphorylation

Gefitinib, erlotinib

Activation of signal-transductioncascades (for example, MAPK)

Apoptosis Invasion andmetastasis AngiogenesisCellproliferation

P P

EGFR

Ligand (EGF, TGF)

Ligand binding

Cetuximab

Figure 1 | EGFR and the mode of action of cetuximab. The epidermal growth factor receptor (EGFR) is one of four members of the erbB family of receptor tyrosine kinases, which consist of an extracellulardomain that can bind ligands, a transmembrane domain and an intracellular tyrosine kinase domain8. A simplified illustration of the EGFR signal transduction pathway is shown. Binding of a ligand to EGFRcauses receptor dimerization (either with another EGFR monomer or with another member of the erbBfamily), leading to tyrosine kinase activation8. The resultant receptor autophosphorylation initiates signal-transduction cascades involved in cell proliferation and survival8. Cetuximab blocks binding ofligands to EGFR, thereby inhibiting receptor phosphorylation and downstream events. Agents thatinhibit the tyrosine kinase activity of EGFR have also been developed and, of these, two have beenapproved by the FDA so far: gefitinib (Iressa; AstraZeneca) and erlotinib (Tarceva; Genentech/Roche), for the treatment of advanced non-small-cell lung cancer. MAPK, mitogen-activated protein kinase; TGF-, transforming growth factor-; TK, tyrosine kinase domain.




Analysing clinical issues for colorectal canceris Richard M. Goldberg, M.D., Professor ofMedicine, Division Chief, Director of OncologyServices, and Associate Director for ClinicalResearch at the Lineberger ComprehensiveCancer Center, University of North Carolina,USA. His research interests include clinicaland translational research in gastrointestinalcancers, particularly colorectal cancer (CRC),and the development and integration of noveltherapies into treatment programmes. He hasrun a number of Phase I, II and III clinicaltrials testing new chemotherapy regimens inpatients with advanced malignancies andgastrointestinal cancers.

What are the current key areas in the pharma-cotherapy of CRC? With neo-adjuvant therapy, one area that showspromise is investigations into the potential rolesof irinotecan, oxaliplatin, bevacizumab andcetuximab with radiation and 5-FU in the neo-adjuvant management of locally advancedrectal cancer. Combinations of these therapieswith 5-FU and radiation seem to be pushing thepathological complete remission rate up fromless than 10% with 5-FU plus radiation to above25% in some preliminary reports9.

There is likely to be evolution in the opti-mal adjuvant therapy for stage III colon cancer.Bevacizumab, cetuximab and other agentsunder investigation might well improve curerates when combined with chemotherapy.Current trials, such as NSABP CO-8 and theIntergroup N-0147, and future studies indevelopment, will help to address these issues.

For treatment of patients with advanceddisease, a key issue is how to optimize the useof targeted agents, including those approvedfor use (bevacizumab and cetuximab) andthose under investigation (PTK787, IMC18F1,AZ2171, VEGF-Trap and others). Recently,results were presented from the BOND2 study,a randomized Phase II trial which assignedpatients with advanced CRC that wasrefractory to irinotecan to bevacizumab incombination with cetuximab plus or minusirinotecan10. This study suggests that therapywith an antibody to VEGF and an antibodyto EGFR can result in tumour shrinkage andmeaningful benefit in many patients, and thatthis effect is augmented by the addition ofirinotecan. These data raise the possibility thatcombined biological therapies, such as beva-cizumab plus cetuximab as used in theBOND2 trial, might become a viable strategyfor the management of advanced diseasewithout chemotherapy.

Prevention strategies in CRC have been setback considerably by the recent toxicity findingsassociated with treatment using rofecoxib as ameans of polyp prevention11. The identificationof new prevention strategies has the potential tomake the biggest impact on deaths from CRC.Early detection methods such as CT colonogra-phy and faecal screening for tumour-associ-ated DNA when further refined also have thepotential to reduce the mortality from CRC.

What have been the most significant recentadvances in disease pharmacotherapy?We now have proof that neo-adjuvant radia-tion and chemotherapy improves local controland quality of life in locally advanced rectalcancer12. The introduction of monoclonalantibodies targeting EGFR and VEGF into thetherapeutic armamentarium has opened thedoor to a new paradigm of therapy. The test-ing and approval of irinotecan- and oxaliplatin-based therapies for treatment of CRC have ledto median survivals exceeding 20 monthswith combinations of chemotherapy with or without biologicals compared with the 12 months seen with 5-FU or capecitabinealone13. In addition, various small-moleculeinhibitors of tyrosine kinases are now beingevaluated in clinical trials. Finally, there is agrowing appreciation of the importance ofpharmacogenetic factors that affect drugresponse and toxicity, such as the finding thatpatients with the UGT1A1 7/7 polymorphismare more susceptible to toxicity. These datacan be used to influence physician choice ofdrug regimens14.

How would you expect the treatment of CRC toevolve in the next 5 years? In the next 5 years, I expect 5-year overallsurvival for stage II colon cancer to approach85%, 5-year overall survival for stage IIIcolon cancer to approach 75%, 5-year overallsurvival for stage IV CRC to approach 15%,and management of advanced disease toresult in median survivals of 3036 months.These advances should come from continuedimprovements in surgical techniques andstaging, more effective chemotherapy andbiological therapy combinations, and theintroduction of new drugs. I would anti-cipate the approval of 23 new agents for themanagement of CRC, including new anti-angiogenic agents and possibly small-mole-cule inhibitors. In general, there will be ashift in drug development to biologicals fromchemotherapy agents because of the proofthat the cetuximab and bevacizumab trials

have provided that biologicals are usefultools for treating CRC.

Owing to the availability of novel agentssuch as cetuximab and bevacizumab, andadvances in the characterization of tumourand patient characteristics, in 5 years weshould be on the verge of individualizedchemotherapy/biological therapy prescrip-tions, an approach that will minimize the toxic-ity and maximize the activity of treatmentprogrammes. Nevertheless, it is clear that thereis much to learn before we can optimize theuse of novel therapies. The recent disappoint-ing announcement on the CONFIRM1 trial, inwhich the addition of PTK787 to FOLFOXfailed to reach statistical significance in thepredetermined prolongation of time to pro-gression, highlights the uncertainty of newdrug development in this era.

Richard Goldberg is at the University of NorthCarolina at Chapel Hill, 27599-7305, USA. PeterKirkpatrick is at Nature Reviews Drug Discovery.Correspondence to R.M.G. & P.K.e-mails: [email protected];[email protected]

doi:10.1038/nrd1728

1. Parkin, D. M. et al. Cancer burden in the year 2000. Theglobal picture. Eur. J. Cancer 37, S4S66 (2001).

2. Baselga, J. The EGFR as a target for anticancer therapy focus on cetuximab. Eur. J. Cancer 37, S16S22 (2001).

3. Dancey, J. & Sausville, E. A. Issues and progress withprotein kinase inhibitors for the treatment of cancer. NatureRev. Drug Discov. 2, 296313 (2003)

4. Salomon, D. S. et al. Epidermal growth factor-relatedpeptides and their receptors in human malignancies. Crit. Rev. Oncol. Hematol. 19, 183232 (1995).

5. Goldstein, N. I. et al. Biological efficacy of a chimericantibody to the epidermal growth factor receptor in ahuman tumor xenograft model. Clin. Cancer. Res. 1,13111318 (1995).

6. FDA labelling information [online] (cited 6 May 2004)

(2004).

7. Prewett, M. C. et al. Enhanced antitumor activity of anti-epidermal growth factor receptor monoclonal antibodyIMC-C225 in combination with irinotecan (CPT-11) againsthuman colorectal tumor xenografts. Clin. Cancer Res. 8,9941003 (2002).

8. Yarden, Y. & Sliwkowski, M. X. Untangling the ErbBsignalling network. Nature Rev. Mol. Cell Biol. 2, 127137(2001).

9. Hofheinz, R.-D. et al. Phase I trial of capecitabine andweekly irinotecan in combination with radiotherapy forneoadjuvant therapy of rectal cancer. J. Clin. Oncol. 23,13501357 (2005).

10. Saltz, L. B. et al. Interim report of randomized Phase II trialof cetuximab/bevacizumab/irinotecan (CBI) versuscetuximab/bevacizumab (CB) in irinotecan-refractorycolorectal cancer. Proc. ASCO Gastrointestinal CancersSymp. A169b (2005).

11. Solomon, S. D. et al. Cardiovascular risk associated withcelecoxib in a clinical trial for colorectal adenoma prevention.N. Engl. J. Med. 352, 10711080 (2005).

12. Sauer, R. et al. Preoperative versus postoperativechemoradiotherapy for rectal cancer. N. Engl. J. Med. 352,17311740 (2004).

13. Grothey, A. et al. Overall survival in advanced colorectalcancer correlates with the proportion of patients treatedwith 5-fluorouracil/leucovorin, irinotecan, and oxaliplatin. J. Clin. Oncol. 22, 12091214 (2004).

14. Iyer, L. et al. UGT1A1*28 polymorphism as a determinant ofirinotecan disposition and toxicity. Pharmacogenomics J.2, 4347 (2002).

COLORECTAL CANCER | VIEW FROM THE CLINIC


Clofarabine



Clofarabine

Clofarabine (Clolar; Genzyme), a purinenucleoside antimetabolite, was grantedaccelerated approval by the US FDA for thetreatment of paediatric patients with relapsedor refractory acute lymphoblastic leukaemia inDecember 2004. It is the first new drug forpaediatric leukaemia to be approved in morethan a decade, and the only one to receiveapproval for paediatric use before adult use.

Leukaemias, such as acute lymphoblasticleukaemia (ALL), are the most commonmalignancies in children and adolescents. Inrecent decades, major progress has been madein understanding the pathobiology of acuteleukaemias and the host pharmacogeneticfactors that affect drug metabolism and dis-position in these diseases1. In parallel withthese advances, the cure rate of childhoodALL reached ~80% in some clinical trials inthe 1990s1. Although part of this success canbe credited to the modification of therapyaccording to stringently defined risk groups,the greatest contribution has come from theoptimal use of drugs developed between the1950s and 1970s. Nevertheless, the prognosisfor children with ALL who do not achieve ormaintain complete remission is very poor,and so new drugs and therapeutic approachesare urgently needed.

Basis of discoveryNucleoside analogues, such as cytarabine, arean important class of highly effective cytotoxicdrugs for the treatment of leukaemias andother haematological cancers2. The purinenucleoside analogues cladribine and fludara-bine (FIG. 1a) have shown activity in acuteleukaemias, but at dose levels associated withprohibitive neurotoxicities2. Both drugsinhibit DNA synthesis after intracellular con-version to their active triphosphate metabo-lites, but their primary mode of action is dif-ferent. Cladribine triphosphate particularlyinhibits ribonucleotide reductase, therebydecreasing cellular deoxynucleotide pools,whereas fludarabine triphosphate primarilyinhibits DNA polymerases2. Clofarabine (FIG.1b) was rationally developed on the basis ofexperience with cladribine and fludarabine,with the aim of improving on their efficacyand minimizing toxicity2,3.

Drug propertiesClofarabine is converted intracellularly bydeoxycytidine kinase to the 5-monophos-phate metabolite, and then by mono- and

diphosphokinases to the active 5-triphos-phate form. Clofarabine triphosphate inhibitsDNA synthesis through its inhibitory actionon both ribonucleotide reductase and DNApolymerases, and is cytotoxic to rapidly pro-liferating and quiescent cancer cell types in vitro2,46. Clofarabine was well toleratedand showed significant activity in a Phase Itrial in paediatric patients with refractoryand relapsed leukaemia7, prompting its eval-uation in Phase II trials in paediatric andadult patients with ALL and acute myeloidleukaemia (AML)6,8.

Clinical dataThe efficacy and safety of clofarabine (52 mgper m2 per day intravenously for 5 days,repeated every 26 weeks following recoveryor return to baseline organ function) wereevaluated in a single-arm study involving 49paediatric patients (median age, 12 years) withrelapsed or refractory ALL6. Most patients(46/49) had received 24 prior regimens. Thestudy endpoints were the rate of completeremission (CR), defined as no evidence ofcirculating blasts or extramedullary disease, anM1 bone marrow (5% and




Analysing clinical issues for acute leukaemiasare Ching-Hon Pui, M.D., Director of theLeukemia/Lymphoma Division of St JudeChildrens Research Hospital, AmericanCancer Society F. M. Kirby Clinical ResearchProfessor and Professor of Pediatrics at theUniversity of Tennessee Health Science Center,USA, and Sima Jeha, M.D., Director ofDevelopmental Therapeutics in the Leukemia/Lymphoma Division of St Jude ChildrensResearch Hospital and Associate Professor ofPediatrics at the University of Tennessee HealthScience Center. Dr Puis research interest isstudying the biology and improving theoutcome of childhood leukaemias. Dr Jehasresearch interest is developing new agents forthe treatment of leukaemias and lymphomas.

What are the unmet needs in acute leukaemia?Although cure rates for paediatric ALL in thepast decade reached ~80%, the prognosis forpatients who do not achieve or maintain com-plete remission remained dismal. Adults withALL are more likely than children to have drug-resistant disease at presentation and less likely totolerate chemotherapy or to comply with treat-ment plans. Their 5-year survival rates are onlyin the range 3040%, despite the frequent use ofhaematopoietic stem-cell transplantation1.

Treatment results also remain disappoint-ing for both children and adults with AML:only about half of the childhood cases, andonly about one-third of cases in patients 18 to60 years old can be cured with contemporarytreatments9,10. The full extent of the problemcan be better appreciated when one considersthat more than half of all cases of AML affectpatients 60 years of age or older, the vastmajority of whom received only palliativetherapy until recently. This elderly popula-tion of AML patients is growing rapidly andcould double by 203010. There is therefore anurgent need for more effective and well-toler-ated treatments for children with AML andfor all adults with acute leukaemia, especiallythe elderly.

What progress has been made in recent years innew drug development for acute leukaemias?The past decade has seen considerableprogress in defining the molecular pathwaysthat drive the pathogenesis of acuteleukaemia. These analyses have contributed tomore precise leukaemia classification systems,allowing more selective conventional therapy,and have also identified several pivotal genesor fusion genes whose protein products seemsuitable for targeted therapy, including FLT3,

PTEN, NOTCH1 and RAS1. The paradigm ofsuch therapy is imatinib mesylate (Gleevec;Novartis), a potent inhibitor of the BCRABLtyrosine kinase that is used to treat Philadelphiachromosome-positive leukaemias. It is welltolerated, even in the elderly, and can inducenot only clinical but also molecular remissionsin most patients treated11,12.

Although not directed to molecular targets,several new formulations of existing agents,such as L-asparaginase, vincristine and cytara-bine, have been produced in an effort toimprove the efficacy and reduce the toxicity ofthe parent compounds. Epigenetic therapywith decitabine and 5-azacitidine to reverseaberrant methylation, or with histone deacetyl-ase inhibitors (SAHA and valproate) to reacti-vate silenced tumour-suppressor genes, has alsoshown promise13. Novel nucleoside analoguesseem to be preferentially effective againstparticular subtypes of acute leukaemia: gem-citabine for ALL, 5-azacitidine and decitabinefor myelodysplasia preceding AML, nelarabinefor T-cell ALL and cladribine for AML2,14.Fludarabine is effective for both lymphoid andmyeloid leukaemia. Finally, clofarabine, whichwas rationally designed to incorporate the mostfavourable properties of its congeners fludara-bine and cladribine2, has now been approvedfor use in paediatric patients with relapsed orrefractory ALL6.

What impact will clofarabine have on thetreatment of acute leukaemias?In addition to providing a valuable treatmentoption for paediatric patients with ALL, clofara-bine also has substantial activity againstAML7,8,14. In Phase I trials, 1 of 16 adults and 1of 8 children with AML achieved completeremission7,14. More encouraging are the resultsof Phase II trials in adults with AML. In onestudy of 31 adults with relapsed AML, com-plete remissions were achieved in 6 of 8patients in first relapse after an initial remis-sion of 12 months or more; in 2 of 11 in firstremission of less than 12 months; and in 5 of12 in second or subsequent relapse8. In anotherstudy, clofarabine induced complete remissionin 16 of 27 patients (6079 years of age) withnewly diagnosed AML who were consideredunfit for conventional chemotherapy; the drugwas associated with grade 3 toxicities in only 3patients15. On the basis of the hypothesisthat clofarabine might modulate cytarabinetriphosphate accumulation and the lack ofoverlapping toxicity, Faderl and colleaguescombined this agent with cytarabine in PhaseII adult trials16,17. The combination treatment

was active in myeloid malignancies and welltolerated. Because clofarabine inhibits therepair of DNA damage, ongoing studies aretesting this analogue in combination withDNA-damaging agents, such as topoisomeraseII inhibitors and alkylators18. Taken together,these preliminary results warrant furtherinvestigation of clofarabine to further define itsrole in the treatment of acute leukaemias.

Ching-Hon Pui and Sima Jeha are at St JudeChildren's Research Hospital, 332 North LauderdaleStreet, Memphis, Tennessee 38105-2794, USA.Peter Kirkpatrick is at Nature Reviews DrugDiscovery. Correspondence to C.H. P. and P.K. e-mails:[email protected]; [email protected]

doi:10.1038/nrd1729

1. Pui, C. H. et al. Acute lymphoblastic leukemia. N. Engl. J.Med. 350, 15351548 (2004).

2. Faderl. S. et al. The role of clofarabine in hematologic andsolid malignancies - development of a next-generationnucleoside analog. Cancer Res. 31 Mar 2005 (doi: 10.1002/cncr.21005).

3. Montgomery, J. A. et al. Synthesis and biologic activity of22-fluoro-2-halo derivatives of 9--D-arabinofuranosyladenine.J. Med. Chem. 35, 397401 (1992).

4. Parker, W. B. et al. Effects of 2-chloro-9-(2-deoxy-2-fluoro--D-arabinofuranosyl)adenine on K562 cellular metabolismand the inhibition of human ribonucleotide reductase andDNA polymerases by its 5-triphosphate. Cancer Res. 51,23862394 (1991).

5. Carson, D. A. et al. Oral antilymphocyte activity and inductionof apoptosis by 2-chloro-22-arabino-fluoro-22-deoxy-adenosine. Proc. Natl Acad. Sci. USA 89, 29702974 (1992).

6. FDA labelling information [online], (2004).

7. Jeha, S. et al. Clofarabine, a novel nucleoside analog, isactive in pediatric patients with advanced leukemia. Blood103, 784-789 (2004).

8. Kantarjian, H. et al. Phase 2 clinical and pharmacologicstudy of clofarabine in patients with refractory or relapsedacute leukemia. Blood 102, 23792386 (2003).

9. Pui, C. H. et al. Childhood and adolescent lymphoid andmyeloid leukemia. Hematology (Am. Soc Hematol. Educ.Program) 118145 (2004).

10. Stone, R. M. et al. Acute myeloid leukemia. Hematology(Am. Soc. Hematol. Educ. Program) 98117 (2004).

11. Wassmann, B. et al. A randomized multicenter open labelphase II study to determine the safety and efficacy ofinduction therapy with imatinib (Glivec, formerly STI571) incomparison with standard induction chemotherapy in elderly(>55 years) patients with Philadelphia chromosome-positive(Ph+/BCR-ABL+) acute lymphoblastic leukemia (ALL)(CST1571ADE 10). Ann. Hematol. 82, 716720 (2003).

12. Thomas, D. A. et al. Treatment of Philadelphia chromosome-positive acute lymphocytic leukemia with hyper-CVAD andimatinib mesylate. Blood 103, 43964407 (2004).

13. Egger, G., Liang, G., Aparicio, A. & Jones, P. A. Epigeneticsin human disease and prospects for epigenetic therapy.Nature 429, 457463 (2004).

14. Kantarjian, H. M. et al. Phase I clinical and pharmacologystudy of clofarabine in patients with solid and hematologiccancers. J. Clin. Oncol. 21, 11671173 (2003).

15. Burnett, A. K. et al. A phase 2 evaluation of single agentclofarabine as first line treatment for older patients with AMLwho are not considered fit for intensive chemotherapy.Blood 104, 248a (2004).

16. Faderl, S. et al. Results of a phase 1-2 study of clofarabinein combination with cytarabine (ara-C) in relapsed andrefractory acute leukemias. Blood 105, 940947 (2005).

17. Faderl, S. et al. Clofarabine plus cytarabine (ARA-C)combination is active in newly diagnosed patients (pts) =age 50 with acute myeloid leukemia (AML) andmyelodysplastic syndrome (MDS). Blood 104, 250a (2004).

18. Faderl, S. et al. Phase I study of clofarabine plus idarubicinand clofarabine plus idarubicin plus cytarabine (ARA-C) inpatients (PTS) with relapsed and primary refractory acutemyeloid leukemia (AML), myelodysplastic syndrome (MDS),and myeloid blast phase of chronic myeloid leukemia (CML).Blood 104, 501a (2004).

ACUTE LEUKAEMIAS | VIEW FROM THE CLINIC


Erlotinib



Erlotinib hydrochloride

Erlotinib hydrochloride (Tarceva; OSI Pharma-ceuticals/Genentech/Roche), a member of a class of targeted anticancer drugs thatinhibit the activity of the epidermal growthfactor receptor, was approved by the USFDA in November 2004 for the treatment ofadvanced non-small-cell lung cancer afterfailure of at least one prior chemotherapyregimen. It is the first such drug to demon-strate an increase in survival in Phase III trials in patients with advanced non-small-celllung cancer.

Lung cancer has been estimated to be the lead-ing cause of cancer mortality worldwide1. Themost common form non-small-cell lungcancer (NSCLC), which accounts for ~75% ofcases is too advanced to be operable in>50% of patients, and standard first-linechemotherapy based on platinum agents onlyimproves survival modestly2. Second-linetreatment options in patients with advancedNSCLC are limited; docetaxel is the only estab-lished choice to be approved by the FDA.

The limited efficacy and lack of specificityof cytotoxic chemotherapy in solid tumourssuch as NSCLC has provided an impetus todevelop therapies that aim to specifically targetcancer cells by modulating the aberrant molec-ular pathways underlying tumour growth andprogression, in the hope of achieving greaterefficacy with fewer side effects. In particular,

protein kinases have emerged as key regulatorsof all aspects of cancer, and many kinaseinhibitors are now being developed.Agents thatinhibit the activity of cell membrane receptortyrosine kinases, such as the epidermal growthfactor receptor (EGFR)3, are among the mostadvanced in clinical development.

Basis of discoveryEGFR is part of the ERBB family of receptortyrosine kinases, which has four closely relatedmembers: EGFR (ERBB1), HER2 (ERBB2),HER3 (ERBB3) and HER4 (ERBB4). Eachmember consists of an extracellular ligand-binding domain, a transmembrane domain andan intracellular tyrosine kinase domain4 (FIG. 1a).Ligand binding to EGFR causes receptor dimer-ization, either with another EGFR monomer orwith another member of the ERBB family.Dimerization activates the tyrosine kinaseactivity in the intracellular domain, which leadsto receptor autophosphorylation and the initia-tion of signal-transduction cascades involved incell proliferation and survival4.

Activation of EGFR has been implicated inprocesses involved in tumour growth and pro-gression, including cell proliferation, inhibi-tion of apoptosis, metastasis and angiogenesis3

(FIG. 1a). EGFR is expressed in various solidtumours, including 4080% of NSCLCs, pro-viding a strong rationale for testing EGFRinhibitors in patients with such tumours5,6.

Several possible approaches to targetingEGFR have been investigated, includingmonoclonal antibodies directed against theextracellular ligand-binding domain, such ascetuximab (Erbitux; Imclone Systems/Bristol-Myers Squibb), and small-molecule inhibitorsof the intracellular tyrosine kinase domain3. Asmall-molecule EGFR inhibitor, gefitinib(Iressa; AstraZeneca), was approved by theFDA for the third-line treatment of patientswith advanced NSCLC in May 2003, and hasnow been joined by erlotinib.

Drug propertiesErlotinib (previously known as OSI-774 andCP-358774; FIG. 1b) is a small molecule thatcompetes with the binding of ATP to theintracellular tyrosine kinase domain of EGFR,thereby inhibiting receptor autophosphory-lation and blocking downstream signal trans-duction79. It showed promising anticancereffects in various preclinical cancer models7,8,prompting its clinical evaluation in a range ofcancers, including NSCLC.

Clinical dataErlotinib hydrochloride (150 mg orally oncedaily) was evaluated in a randomized, doubleblind, placebo-controlled trial involving 731patients with locally advanced or metastaticNSCLC after failure of at least one chemo-therapy regimen9. The primary endpoint wassurvival, which was significantly longer inpatients receiving erlotinib hydrochloride:median overall survival in these patients was 6.7months compared with 4.7 months in patientsreceiving placebo9. Median progression-freesurvival and tumour response rates in patientsreceiving erlotinib hydrochloride were 9.9weeks and 8.9%, respectively, compared with7.9 weeks and 0.9%, respectively, in patientsreceiving placebo9.

Erlotinib hydrochloride has also been eval-uated in combination with platinum-basedchemotherapy (carboplatin and paclitaxel, orgemcitabine and cisplatin) in two placebo-controlled, randomized trials involving morethan 1,000 first-line patients with locallyadvanced or metastatic NSCLC. No clinicalbenefit was demonstrated from the addition oferlotinib hydrochloride in these trials9.

IndicationsErlotinib hydrochloride is approved by theFDA for the treatment of patients with locallyadvanced or metastatic NSCLC after failure ofat least one prior chemotherapy regimen9.

a

TK TK TK TK

Autophosphorylation

Erlotinib

Activation of signal-transductioncascades (for example, MAPK)

Apoptosis Invasion andmetastasis AngiogenesisCellproliferation

EGFR

Ligand (EGF, TGF)

Erlotinib hydrochloride

N-(3-ethynylphenyl)-6,7-bis(2-methoxyethoxy)-4-quinazolinamine; C22H23N3O4HCl; Mr = 429.90; CAS registry number: 183319-69-9

b

N

N

HN

OO

OO

HCl

P P

Figure 1 | EGFR signalling and erlotinib. a | Simplified illustration of signal transduction through theepidermal growth factor receptor3,4 (EGFR). Ligand binding leads to receptor dimerization. This results inreceptor autophosphorylation, which is inhibited by erlotinib. b | Structure of erlotinib hydrochloride.MAPK, mitogen-activated protein kinase; P, phosphate group; TGF, transforming growth factor; TK,tyrosine kinase domain.




What are the current key issues in the treatmentof non-small-cell lung cancer?One current key issue in the treatment ofNSCLC is that a plateau seems to have beenreached with regards to the efficacy of standardchemotherapy. Numerous trials have nowdemonstrated that platinum-based chemo-therapy combinations are the most effectivetreatment in advanced NSCLC. The drug thatis combined with platinum (a taxane, gemcit-abine or vinorelbine) seems to affect only thetoxicity profile of the regimen, as all have simi-lar efficacy. Attempts to enhance the activity ofthese regimens by adding a third chemotherapyagent have been uniformly unsuccessful.

Patients vary dramatically in their responseto these standardchemotherapies, and anothermajor issue is the development of pharmaco-genomic tests based on a molecular assessmentof the tumours from individual patients, andalso interindividual variations in drug metabo-lism, to determine which drugs would be bestfor which patients.

A further key to advancing therapy is theidentification of novel molecular targets,such as EGFR, as well as developing drugsthat effectively act on these targets, such aserlotinib and gefitinib. In addition, an impor-tant current focus is identifying those patientslikely to benefit from these therapies.

What data is available that might help under-stand patient response to EGFR inhibitors?The discovery of somatic mutations in the tyro-sine kinase domain of EGFR has identified asubset of NSCLC patients that are exquisitelysensitive to the EGFR tyrosine kinase inhibitors(TKIs)1012. However, the randomized single-agent trial of erlotinib indicates that there mustbe patients without EGFR mutations that alsoderive benefit from drug treatment. Finding

molecular correlates of these responses (such asEGFR amplification or gene-expression signa-tures) will be important for identifying theresponding patients prospectively.

William Pao and colleagues have foundidentical mutations in EGFR that correlatewith response to erlotinib13. The same clinicalcharacteristics that are predictive for responseto gefitinib also identify those patients mostlikely to benefit from erlotinib. Women,never-smokers, those with adenocarcinomaor bronchoalveolar carcinoma, and patients ofEast Asian descent are all more likely torespond to treatment with an EGFR TKI, andthese clinical features correlate with the presenceof an EGFR mutation.

Recently, it was found that in some patientswho have relapsed on erlotinib or gefitinib, theirlung cancers contain, in addition to an EGFRtyrosine kinase domain mutation, a secondmutation (T790M) that confers the drug resis-tance14,15. Of interest, tumour cells with theT790M mutation were still sensitive to anotherEGFR TKI, indicating that drugs that overcomethis resistance can be developed.

What might be the best approaches for exploit-ing targeted therapies in NSCLC?For many targeted therapies, there is a goodpreclinical rationale to suggest that they mightbe more effective in the setting of minimal bulkdisease (that is, the adjuvant setting after defini-tive surgery for early-stage disease). Both of theavailable EGFR TKIs (gefitinib and erlotinib)are currently being investigated in the adjuvantsetting in NSCLC, after the administration ofadjuvant chemotherapy. Of course, the ultimatein early treatment is use of the drugs forchemoprevention of the development of lungcancer in high-risk individuals. There is grow-ing evidence that abnormalities of EGFR sig-nalling occur in smoking-damaged lungepithelium as a precursor to cancer, and TKIsas oral agents with little toxicity could have animportant role in such prevention.

With regards to combinations of TKIs withother drugs in clinically evident disease, all ofthe published Phase III trials that have investi-gated adding a targeted agent to standardchemotherapy for advanced NSCLC havefailed to demonstrate a benefit for the combi-nation (including four trials with either gefi-tinib or erlotinib). Two such trials have yet tobe reported one with bexarotene and onewith bevacizumab, and results of these trialsare eagerly anticipated. In the case of the EGFRinhibitors, the failure of these drugs to enhancethe activity of standard chemotherapy might

have been due to the patient populationstudied. These trials were conducted beforethe discovery of EGFR mutations that are pre-dictive for response to the EGFR TKIs. Giventhe relatively low incidence of these mutations,the majority of patients included in these trialsprobably had wild-type EGFR in theirtumours and would therefore be less likely tobenefit from this therapeutic approach. Trialsin patients with tumour-acquired EGFRmutations are being planned.

With regards to combining targeted agents,some preliminary data are available. RoyHerbst and Alan Sandler conducted a Phase IItrial of erlotinib and bevacizumab in advancedNSCLC16. The combination was safe and well-tolerated, and enough activity was seen to justifya Phase III comparison with standard chemo-therapy in the second-line setting in metastaticdisease.Jonathan Dowell and John D. Minna are at theUniversity of Texas Southwestern Medical Center at Dallas, Dallas, Texas 75390-8593, USA. PeterKirkpatrick is at Nature Reviews Drug Discovery.Correspondence to J.D.M. and P.K.e-mails: [email protected];[email protected]:10.1038/nrd1730

1. Parkin, D. M. et al. Cancer burden in the year 2000. The global picture. Eur. J. Cancer 37, S4S66 (2001).

2. Cersosimo, R. J. Lung cancer: a review. Am. J. HealthSyst. Pharm. 59, 611642 (2002).

3. Dancey, J. & Sausville, E. A. Issues and progress withprotein kinase inhibitors for the treatment of cancer. NatureRev. Drug Discov. 2, 296313 (2003).

4. Yarden, Y. & Sliwkowski, M. X. Untangling the ErbB signallingnetwork. Nature Rev. Mol. Cell Biol. 2, 127137 (2001).

5. Baselga, J. Why the epidermal growth factor receptor? Therationale for cancer therapy. The Oncologist 7 (S4), 28 (2002).

6. Salomon, D. S. et al. Epidermal growth factor-relatedpeptides and their receptors in human malignancies. Crit. Rev. Oncol. Hematol. 19, 183232 (1995).

7. Moyer, J. D. et al. Induction of apoptosis and cell cycle arrestby CP-358,774, an inhibitor of epidermal growth factorreceptor tyrosine kinase. Cancer Res. 57, 48384848 (1997).

8. Pollack, V. A. et al. Inhibition of epidermal growth factorreceptor-associated tyrosine phosphorylation in humancarcinomas with CP-358,774: dynamics of receptorinhibition in situ and antitumor effects in athymic mice. J. Pharmacol. Exp. Ther. 291, 739748 (1999).

9. FDA label information [online], (2004).

10. Lynch, T. J. et al. Activating mutations in the epidermal growthfactor receptor underlying responsiveness of non-small-cell lungcancer to gefitinib. N. Engl J. Med. 350, 21292139 (2004).

11. Paez, J. G. et al. EGFR mutations in lung cancer:correlation with clinical response to gefitinib therapy.Science 304, 14971500 (2004).

12. Shigematsu, H. et al. Clinical and biological features ofepidermal growth factor receptor mutations in lung cancers.J. Natl Cancer Inst. 97, 339346 (2005).

13. Pao, W. et al. EGF receptor gene mutations are common inlung cancers from never smokers and are associated withsensitivity of tumors to gefitinib and erlotinib. Proc. NatlAcad. Sci USA 101, 1330613311 (2004).

14. Kobayashi, S. et al. EGFR mutation and resistance of non-small-cell lung cancer to gefitinib. N. Engl. J. Med. 352,786792 (2005).

15. Pao, W. et al. Acquired resistance of lung adenocarcinomasto gefitinib or erlotinib is associated with a second mutationin the EGFR kinase domain. PLoS Med. 2, e73 (2005).

16. Sandler, A. et al. Phase I/II trial evaluating the anti-VEGFMAb bevacizumab in combination with erlotinib, aHER1/EGFR-TK inhibitor, for patients with recurrent non-small cell lung cancer. J. Clin. Oncol. 22, 14S (2004).

NON-SMALL-CELL LUNG CANCER | VIEW FROM THE CLINIC

Analysing clinical issues for targetedtherapies for non-small-cell lung cancer areJohn D. Minna, M.D., Professor and Directorof the Hamon Center for TherapeuticOncology Research at the University of TexasSouthwestern Medical Center, USA, andJonathan Dowell, M.D., Assistant Professor,Department of Internal Medicine at theUniversity of Texas Southwestern MedicalCenter. Dr Minnas research interest is todetermine all of the molecular abnormalitiesleading to the pathogenesis of lung cancer andto translate this information into new methodsfor the diagnosis, prevention, and treatment ofthis disease. Dr Dowells research interest is inclinical and translational trials in lung cancer.


Pemetrexed



Pemetrexed disodium

In February 2004, pemetrexed disodium(Alimta; Eli Lilly), an anticancer drug thattargets folate-dependent reactions that areessential for cell proliferation, became the firstdrug to be approved by the US FDA for thetreatment of the rare cancer malignant pleuralmesothelioma. Its accelerated approval forthe second-line treatment of non-small-celllung cancer followed in August 2004.

Drugs that target folate-dependent reactionshave a long history in cancer therapy. Amino-pterin, which inhibits dihydrofolate reductase(DHFR), entered clinical use in the late 1940s,and its less-toxic derivative methotrexate,which was introduced soon after, is still widelyused to treat various cancers1,2.

Basis of discoveryReduced folates are essential cofactors in theenzymatic synthesis of thymidine and purinenucleotides (FIG. 1a). Methotrexate, which isclosely related to folic acid, has several effects onfolate metabolism that lead to impaired nucleo-tide production, which ultimately disruptsDNA synthesis and cell proliferation1.

First, inhibition of DHFR by methotrexateleads to depletion of the intracellular pool ofreduced folates (FIG. 1a). In addition, and as withnatural folates, the enzyme folyl-polyglutamylsynthase (FPGS) can attach several glutamatemolecules to methotrexate, which has a num-ber of important effects1. Polyglutamationdecreases cellular efflux, which prolongs drugaction; indeed, the increased propensity ofcancer cells to polyglutamate methotrexatecompared with normal cells might underliesome of the selectivity of the drug. Further-more, methotrexate polyglutamates are alsoinhibitors of DHFR, and have increased potencycompared with methotrexate as inhibitors ofthe folate-dependent enzymes thymidylate syn-thase (TS) and glycinamide ribonucleotideformyltransferase (GARFT)1,2 (FIG. 1a).

The utility of methotrexate has led to con-siderable efforts to develop antifolates that havea broader spectrum of activity and which cir-cumvent methotrexate resistance1,2. Pemetrexedis the first of this next generation of antifolatesto be approved as an anticancer drug in theUnited States.

Drug propertiesPemetrexed (FIG. 1b) was discovered throughscreening of derivatives of a known GARFTinhibitor for anticancer activity3. However,in contrast to the parent compound and

methotrexate, its primary site of action wasfound to be TS. After being transported intothe cell via the reduced folate carrier, which isinvolved in the uptake of physiological folates,pemetrexed is polyglutamated by FPGS2.These polyglutamate derivatives (the pen-taglutamate is the predominant intracellularform) are potent inhibitors of TS, and are alsoweaker inhibitors of GARFT24. Pemetrexedand its polyglutamate derivatives also inhibitDHFR, but with ~1,000-fold less potency thanmethotrexate2. As with methotrexate, peme-trexed polyglutamates have an increased intra-cellular half-life, resulting in prolonged drugaction in malignant cells2,5.

Pemetrexed showed promising therapeuticeffects in preclinical and early clinical studiesin a range of cancers2, including malignantpleural mesothelioma (MPM) and non-small-cell lung cancer (NSCLC), which have sincebecome the first indications for which peme-trexed has been studied in Phase III trials6,7.

Clinical dataMPM is a rare cancer that is usually associatedwith asbestos exposure. Pemetrexed disodium(as an intravenous infusion) in combinationwith the platinum-based cytotoxic drug cisplatinwas compared with cisplatin alone in a random-ized study involving 448 chemotherapy-naivepatients with MPM5,6. To decrease adverseevents, patients were given folic acid and vita-min B

12supplements. Patients receiving both

pemetrexed disodium and cisplatin had a

significantly increased median overall sur-vival time of 12.1 months, compared with 9.3months for those receiving cisplatin alone5,6.

NSCLC is the most common form of lungcancer, which is the leading cause of cancermortality worldwide. First-line therapy ofNSCLC is based on platinum agents, and doce-taxel is the standard second-line treatmentoption. Pemetrexed disodium was comparedwith docetaxel in a randomized trial involving571 patients with locally advanced metastaticNSCLC after prior chemotherapy; patientstreated with pemetrexed disodium also receivedfolic acid and vitamin B

12supplements. Peme-

trexed disodium did not show superiority overdocetaxel on the primary endpoint of survival,and there were no statistically significant differ-ences between pemetrexed disodium and doc-etaxel with respect to the secondary endpointssuch as objective response rate5. However,pemetrexed disodium had a more favourablesafety profile than docetaxel; for example, itcaused significantly less neutropenia and febrileneutropenia5,7.

IndicationsPemetrexed disodium in combination with cis-platin is approved by the FDA for the treatmentof patients with MPM whose disease is unre-sectable or who are otherwise not candidatesfor curative surgery5. Pemetrexed disodium as asingle agent is also approved by the FDA for thetreatment of patients with locally advanced ormetastatic NSCLC after prior chemotherapy5.

L-Glutamic acid, N-[4-[2-(2-amino-4,7-dihydro-4-oxo-1H-pyrrolo[2,3-d]pyrimidin-5-yl)ethyl]benzoyl]-,disodium salt, heptahydrate; C20H19N5Na2O67H2O; Mr = 597.49; CAS number: 150399-23-8

Pemetrexed disodium heptahydrate

O

NH

CO2

HN

HN

N

O

H2N

CO2 Na

Na

7H2O

a b

DHF

10-CHO-THF

GAR

Purinenucleotides

THF5,10-CH2-THF

TS

DHFR

GARFT AlCARFT

dTMP

dUMP

DNA

DNA, RNA

Figure 1 | Folate metabolism and pemetrexed. a | Simplified illustration of some key enzymaticreactions of folate metabolism, showing enzymes affected by pemetrexed, or its polyglutamates. Thesemultiple drug effects impair nucleotide synthesis and thereby inhibit cell proliferation. b | Structure ofpemetrexed disodium. AICARFT, aminoimidazole carboxamide ribonucleotide formyltransferase; DHFR,dihydrofolate reductase; GARFT, glycinamide ribonucleotide formyltransferase; THF, tetrahydrofolate, TS,thymidylate synthase.




Analysing issues that affect the clinicalapplication of antifolates are David Gibbs, PhD,FRACP, Medical Oncologist at the ChristchurchHospital, New Zealand, and Ann Jackman, PhD,Professor of Biochemical Pharmacology at theInstitute of Cancer Research, UK. Dr Gibbs'interests are the development of targeted anti-folates, and the management of lung cancerand ovarian cancer. Dr Jackman's interests areantifolate drug development and overcomingdrug resistance in ovarian cancer by combiningtargeted agents with cytotoxic drugs.

What are the key issues that need to be addressedto maximise the benefits of antifolates?The key issues in developing the clinical utilityof antifolates are common to the developmentof other anticancer drugs: namely, strategies toreduce normal tissue toxicity, strategies toabrogate the development of drug resistanceand strategies to predict the patients that aremost likely to benefit from treatment.

There is evidence that clinical response to TSinhibitors is predicted by levels of TS express-ion8,9, but this is not yet used as a predictor inclinical practice. There is also evidence that celllines acquire resistance to antifolates, such aspemetrexed, by downregulation of the reducedfolate carrier (RFC) and FPGS10, which suggeststhat measurement of levels of a few key proteinsmight allow prediction of antifolate activity.

Normal tissue tolerance limits the dose ofantifolates that can be delivered. Dose-limitingtoxicity can be related to folate levels, and folatesupplementation increases the maximum tol-erated dose of methotrexate, lometrexol andpemetrexed11. Folate supplementation wasfound to have no effect on antitumour activityin animal models and there is no evidence thatfolate supplementation reduces anti-tumouractivity in mesothelioma6. However, there is noa priori reason that folate supplementation willalso result in tumour protection, and this con-cept needs exploration in further clinical studies.

How well are the mechanisms of the anticanceractivity of antifolates understood, and what arethe implications for their clinical application?In general, the effect of antifolates depends onthe particular enzymes that are inhibited withaccumulation of enzyme substrate and deple-tion of enzyme products. However, the mecha-nism by which these events result in cell deathare not completely understood. The deconvolu-tion of the mechanism of action of antifolates isfurther complicated by their metabolism in cellsto higher-order polyglutamate metabolites andthe potential for inhibition of multiple targets.

Of the new generation of antifolates, themechanism of action of TS inhibitors has beenstudied in the most detail. In general, TS inhi-bition results in changes in deoxynucleotidepools, leading to the formation of single anddouble-stranded DNA breaks, S-phase arrestand induction of apoptosis12. However. themechanisms by which these effects occur differbetween cell lines, and it is likely that similarlydiverse mechanisms occur in tumours in vivo.Recent research has indicated that the cyclindependent kinase inhibitor p21 modulates TSinhibitor activity13 and that FAS is required forthe induction of apoptosis in response to TSinhibitors14. Studies such as these indicate thatmuch remains to be elucidated about the mech-anisms of action of antifolates, and also provideinformation that might allow rational choices tobe made in choosing the best cytotoxic drugsand targeted agents to combine with antifolates.

Studies of resistant cell lines indicate thatresistance to many antifolates occurs by a rela-tively limited repertoire of mechanisms. Thesemechanisms include alterations in the functionof transport proteins such as the RFC, enzymesresponsible for maintaining polyglutamatehomeostasis, upregulation of target enzymesand changes in intracellular folate pools.However, this also suggests that if means ofmodulating these resistance mechanisms canbe found, sensitivity to these agents might bemaintained or restored.

Finally, an area requiring further under-standing, because opinion is divided, is the con-tribution that the -folate receptor might maketo the antitumour activity of newer antifolatesin those epithelial tumours that overexpress it;for example, ovarian cancer and mesothelioma.

What do you think will be the impact of newantifolates in the next 5 years?In addition to pemetrexed, there are otheragents in clinical development, such as plevit-rexed (BGC 9331, a non-polyglutamated TSinhibitor), OSI-79041 (a liposomal-encapsu-lated TS inhibitor), nolatrexed (a lipohilic,non-polyglutamated TS inhibitor) and -methylene-10-deazaaminopterin (a DHFRinhibitor) or licensed for use in some countries(raltitrexed; a polyglutamated TS inhibitor).The different classes of newer antifolates haveproperties that might be advantageous in somecircumstances. Unlike methotrexate, polygluta-mation of raltitrexed and pemetrexed not onlyresults in increased retention within cells butalso increases affinity for their target enzymes.Plevitrexed cannot be polyglutamated and OSI-79041 does not require polyglutamation for its

activity and thus will retain activity in cells thathave reduced FPGS or increased folylpolyglu-tamate hydrolase activity. Pemetrexed inhibitsnot only TS but also GARFT and this might beresponsible for its greater activity against sometumour types.

The main recent impact of antifolates inclinical practice has been the approval of peme-trexed in combination with cisplatin in MPMand pemetrexed in the second-line treatment ofNSCLC. Both pemetrexed and raltitrexed havebeen tested in combination with several cyto-toxic agents. It is possible that some of thesecombinations might result in clinically relevantsynergy. However, combinations need to beselected on the basis of an understanding ofmechanisms of action as discussed above.

David Gibbs is at Christchurch Hospital,Riccarton Avenue, Christchurch, New Zealand.Ann Jackman is at the Institute of Cancer Research,15 Cotswold Road, Sutton, Surrey, UK.Peter Kirkpatrick is at Nature Reviews DrugDiscovery.Correspondence to D. G & P. K.e-mails: [email protected];[email protected]:10.1038/nrd1731

1. Takimoto, C. H. New antifolates: pharmacology and clinicalapplications. The Oncologist 1, 6881 (1996).

2. Hanauske, A.-R. et al. Pemetrexed disodium: a novelantifolate clinically active against multiple solid tumors.TheOncologist 6, 363373 (2001).

3. Taylor, E. C. et al. A dideazatetrahydrofolate analoguelacking a chiral center at C-6, N-[4-[2-(2-amino-3,4 dihydro-4-oxo-7H-pyrrolo[2,3-d]pyrimidin-5-yl)ethyl]benzoyl]-L-glutamic acid, is an inhibitor of thymidylate synthase.J. Med. Chem. 35, 44504454 (1992).

4. Shih, C. et al. LY231514, a pyrrolo[2,3-d]pyrimidine-basedantifolate that inhibits multiple folate-requiring enzymes.Cancer Res. 57, 11161123 (1997).

5. FDA drug approvals list [online], (2004).

6. Vogelzang, N. J. et al. Phase III study of pemetrexed incombination with cisplatin versus cisplatin alone in patientswith malignant pleural mesothelioma. J. Clin. Oncol. 21,26362644 (2003).

7. Hanna, N. et al. Randomized phase III trial of pemetrexedversus docetaxel in patients with non-small-cell lung cancerpreviously treated with chemotherapy. J. Clin. Oncol. 22,15891597 (2004).

8. Johnston, P. G. et al. Thymidylate synthase gene andprotein expression correlate and are associated withresponse to 5-fluorouracil in human colorectal and gastrictumors. Cancer Res. 55, 1407-1412 (1995).

9. Farrugia, D. C. et al. Thymidylate synthase expression inadvanced colorectal cancer predicts for response toraltitrexed. Clin. Cancer Res. 9, 792-801 (2003).

10. Wang, Y. et al. Decreased expression of the reduced folatecarrier and folypolyglutamate synthetase is the basis foracquired resistance to the pemetrexed antifolate (LY231514)in an L1210 murine leukemia cell line. Biochem. Pharmacol.65, 1163-1170 (2003).

11. Niyikiza, C. et al. Homocysteine and methylmalonic acid:markers to predict and avoid toxicity from pemetrexedtherapy. Mol. Cancer Ther. 1, 545-552 (2002).

12. Touroutoglou N, & Pazdur, R. Thymidylate synthaseinhibitors. Clin. Cancer Res. 2, 227-243 (1996).

13. Geller, J. I. et al. P21Cip1 is a critical mediator of the cytotoxicaction of thymidylate synthase inhibitors in colorectalcarcinoma cells. Cancer Res. 64, 6296-6303 (2004).

14. Longley, D. B. et al. The roles of thymidylate synthase andp53 in regulating Fas-mediated apoptosis in response toantimetabolites. Clin. Cancer Res. 10, 3562-3571 (2004).

ANTIFOLATES | VIEW FROM THE CLINIC




The global oncology market generated salesof US $30.0 billion in 2003, and is forecast toincrease by 14.6% by 2008, resulting in salesof US $59.3 billion. FIG. 1 shows the globalmarket sizes for the key cancer indications.

Nat Rev Drug Dis_2005

Documents

Transcript of Nat Rev Drug Dis_2005