Access the full text.
Sign up today, get DeepDyve free for 14 days.
571111 TAM0010.1177/1758834015571111Therapeutic Advances in Medical OncologyE Jokinen and JP Koivunen research-article2015 Therapeutic Advances in Medical Oncology Review Ther Adv Med Oncol MEK and PI3K inhibition in solid tumors: 2015, Vol. 7(3) 170 –180 DOI: 10.1177/ rationale and evidence to date © The Author(s), 2015. Reprints and permissions: http://www.sagepub.co.uk/ E. Jokinen and J.P. Koivunen journalsPermissions.nav Abstract: PI3K-AKT-mTOR and Ras-Raf-MEK-ERK are the most commonly altered oncogenic pathways in solid malignancies. There has been a lot of enthusiasm to develop inhibitors to these pathways for cancer therapy. Unfortunately, the antitumor activities of single-agent therapies have generally been disappointing, excluding B-Raf mutant melanoma and renal cell cancer. Preclinical studies have suggested that concurrent targeting of the PI3K-AKT-mTOR and Ras-Raf-MEK-ERK pathways is an active combination in various solid malignancies. In the current work, we review the preclinical data of the PI3K and MEK dual targeting as a cancer therapy and the results of early-phase clinical trials, and propose future directions. Keywords: inhibition, MEK, PI3K, solid tumors, targeted therapy Correspondence to: Introduction 2007; Motzer et al. 2008]. The effectiveness of Elina Jokinen, PhD Constitutive activation of oncogenic pathways single pathway inhibition could be suppressed by Department of Oncology and Radiotherapy, Oulu occurs in cancers with very high frequency. It is de novo dependency on multiple signaling path- University Hospital, Oulu, thought to be a central factor behind the hall- ways or feedback activation of other signaling Finland, PB20, 90029 OYS elinamirjamijokinen@ marks of cancer phenotype such as cell cycle pro- pathways [Faber et al. 2009; Chandarlapaty et al. gmail.com gression, inhibition of apoptosis and 2010]. This has led to studies combining phos- Jussi P. Koivunen, MD, PhD reprogramming of metabolism. In solid malig- phoinositide-3 kinase (PI3K)/AKT/mTOR and Department of Oncology and Radiotherapy, Oulu nancies, the PI3K-AKT-mTOR and Ras-Raf- Raf/mitogen-activated protein kinase kinase University Hospital, Oulu, MEK-ERK pathways are thought to play a (MEK) inhibitors. Finland central role in transmitting these oncogenic sig- nals. Genetic alterations, such as receptor muta- tions or amplifications, or mutations in PI3K pathway intermediate signal transducers, can lead to the PI3Ks are divided by structure, regulation and constitutive activation of the pathways. The high lipid substrate specificity into three subclasses: frequency of genetic alterations causing constitu- class I, II and III. Of the three subclasses of PI3K, tive activation of PI3K-AKT or Ras-Raf-MEK- class I has been related to cancer. Class I PI3K is ERK and the addiction of cancer cells on their composed of a p110 catalytic subunit, coded by signals have generated enthusiasm to develop PIK3CA, PIK3CB, PIK3CD and PIK3CG genes, pathway inhibitors. and a p85 regulatory subunit. The kinase activity of p110 is regulated downstream of receptor Considering the central role of the pathways in tyrosine kinases (RTKs) by the binding of tyros- transmitting upstream oncogenic signals, inhibi- ine-phosphorylated proteins to its regulatory p85 tion of these pathways could be an effective ther- subunit, resulting in reduction of its autoinhibi- apy in various cancer genotypes [Hanahan and tory activity. Furthermore, p110 has a Ras motif Weinberg, 2011]. Despite the undisputed ration- in which activated Ras proteins can bind leading ale of this approach, the clinical activity of the sin- to optimal lipid kinase activity. Activated PI3K gle pathway inhibition has been limited to B-Raf produce phosphatidylinositol (3,4,5)-trisphos- inhibition in B-Raf mutant melanoma and mam- phate (PIP ), which can activate downstream tar- malian target of rapamycin (mTOR) inhibition in gets, such as AKT which further activates mTOR renal cell cancer [Solit et al. 2006; Hudes et al. [Engelman et al. 2006]. 170 http://tam.sagepub.com E Jokinen and JP Koivunen The PI3K-AKT-mTOR signaling pathway has an malignancies but have shown clinical activity in a important role in cell survival and proliferation, limited number of indications. Everolimus has and constitutive activation of the pathway is com- been approved for renal cell cancer (RCC) and mon in solid malignancies. The hyperactivation of neuroendocrine cancers as a single-agent and for the pathway is most frequently caused by the acti- breast cancer in combination with hormonal ther- vation of RTKs, somatic mutations in specific apy [Motzer et al. 2008; Yao et al. 2011; Baselga signaling components of the pathway such as et al. 2012]. Another mTOR inhibitor, temsiroli- PIK3CA, or the loss of PTEN tumor suppressor. mus, has also been approved to the treatment of PIK3CA mutations are frequently detected in RCC [Hudes et al. 2007]. In contrast to most endometrial, breast, lung, urinary tract, colorec- solid malignancies, activation of hypoxia-induci- tal, gliomas and gastric cancers [Cancer Genome ble factor 1-alpha (HIF-1α) is a central oncogenic Atlas Research Network, 2008, 2012a, 2012b, event in RCC, which might relate to the clinical 2012c, 2013, 2014a, 2014b, 2014c] while ampli- activity of the rapalogs in this disease. mTOR is fication of the gene is commonly seen in head and an indirect upstream regulator of HIF-1α and neck, lung, gastric and ovarian cancers [Lin et al. therefore may represent an easier clinical target 2005; Cancer Genome Atlas Research Network, for mTOR inhibition compared with most solid 2011, 2012a, 2014a]. Genetic PTEN loss occurs malignancies in which the RTK pathways recurrently in endometrial, prostate, gliomas, upstream of mTOR dominate in oncogenesis lung and breast cancers [Cancer Genome Atlas [Bellmunt et al. 2013]. Research, 2008, 2012, 2012, 2013; Bismar et al. 2011] (Table 1, Figure 1). Numerous clinical trials are testing the safety and efficiency of dual PI3K-mTOR, PI3K and AKT inhibitors in various solid malignancies. Treatment PI3K-pathway inhibitors with these inhibitors seems to be well tolerated Several small molecular inhibitors targeting the with dose-limiting toxicities (DLTs) from hyper- PI3K pathway have been developed, including glycemia, rash, gastrointestinal side effects and dual PI3K–mTOR-, PI3K-, AKT- and mTOR stomatitis [Dienstmann et al. 2011; Bran and Siu, inhibitors. Allosteric, rapalog analog mTOR 2012]. Various methods have been used to ana- inhibitors were the first to enter clinical trials but, lyze the dose–effect relationship of these drugs, unfortunately, showed activity only in a limited such as the measurement of biomarkers pAKT number of solid tumor types. Limited clinical and pS6 in cells and tissues. With the maximum activity of the rapalogs has been suggested to tolerated dose (MTD), inhibitors induce 30–90% result from various drug-induced feedback loops inhibition of pAKT and/or pS6 biomarkers leading to mTORC2-IRS-1 mediated hyperacti- [Bendell et al. 2011; Moreno Garcia et al. 2011]. vation of PI3K-AKT [O’Reilly et al. 2006]. Based on early-phase clinical trials, the single- Rapalog-induced PI3K-AKT hyperactivation has agent clinical activity of PI3K-mTOR, PI3K and led to the development of upstream inhibitors, AKT inhibitors seems to be modest in unselected which could overcome the hyperactivation and solid malignancies. The drugs appear to be more have more clinical activity [Wander et al. 2011]. active in tumors bearing genetic alterations in These upstream inhibitors include dual PI3K- PIK3CA or PTEN, but no clear correlation mTOR (BEZ235), pan PI3K (GDC-0941 and between genotype and response have been estab- BKM120) and AKT inhibitors (MK-2206, lished [Brachmann et al. 2009; Weigelt et al. GDC-0068) (Table 2). Of these drugs, most fall 2011]. General lack of clinical benefit could result into adenosine triphosphate (ATP) competitive from various factors, such as limited target inhibi- class, while both ATP and non-ATP competitive tion or feedback activation of other pathways. As AKT inhibitors do exist. Recently, a newer gen- low as 30% target inhibition has been reported in eration of inhibitors has entered clinical trials clinical trials and it is unlikely that this magnitude including isoform specific PI3K inhibitors and of inhibition could result in clinical activity. ATP competitive inhibitors of mTORC1/2. Furthermore, feedback activation of other path- ways has been reported in preclinical studies but Currently, only the rapalog inhibitors of the general landscape of this phenomenon has not PI3K-AKT-mTOR pathway have entered the been characterized. Preclinical studies with PI3K clinic for the treatment of cancer. Rapalogs have and AKT inhibitors have shown that these previously been widely tested in various solid drugs induce feedback activation of the http://tam.sagepub.com 171 Membrane Therapeutic Advances in Medical Oncology 7(3) growth factors extracellular-signal-regulated kinase (ERK) and human epidermal growth factor receptor (HER) family, which at least partly results in decreased anticancer activity [Chandarlapaty et al. 2010; RTK Serra et al. 2011]. Similar feedback activation PI3Ki PI3K Ras phenomenon of the ERK and HER family has DUALi NF1 also been recapitulated in clinical trials of the PTEN AKTi AKT Raf RAFi agents [Yan et al. 2013]. mTORi mTOR MEK MEKi Ras-Raf-MEK-ERK pathway The binding of cytokines, growth factors or mito- ERK gens to their receptors causes activation of the Shc/Grb2/SOS coupling complex, which in turn cell proliferation, growth and survival stimulates the inactive Ras (H-, K, N-isotypes). Stimulated Ras exchanges guanosine diphos- Figure 1. Signaling of PI3K-AKT-mTOR and Ras-Raf- phate (GDP) to guanosine triphosphate (GTP) MEK-ERK pathways and inhibitors currently tested in cancer clinical trials. Signaling molecules are drawn and undergoes conformational change to active with circles and inhibitors with boxes. Stars indicate stage. Ras proteins have an intrinsic GTPase signaling molecules that are frequently altered in activity, required for inactivation of the GTP solid malignancies. bound active state. The GTPase activity of Ras AKTi, AKT inhibitor; DUALi, dual PI3K; ERK, extracellular- protein is strongly enhanced by binding of signal-regulated kinase; MEKi, mitogen-activated protein kinase kinase inhibitor; mTORi, mammalian GTPase proteins such as nuclear factor 1 (NF1). target of rapamycin inhibitor; NF1, neurofibromin; PI3Ki, Ras proteins have multiple downstream targets, phosphoinositide-3 kinase inhibitor; RAFi, RAF inhibitor; with Raf kinase being the most well character- RTK, receptor tyrosine kinase. ized. Raf kinase activates MEK1/2, which cata- lyzes activation of ERK1/2. Activation of ERK1/2 further phosphorylates a series of downstream targets involved in various cellular processes such Inhibitors of Ras-Raf-MEK-ERK pathway as proliferation, differentiation, cell survival and Considerable efforts have been made to develop angiogenesis. Ras targeted drugs, such as farnesyl transferase inhibitors, but so far clinical trials have been Ras-Raf-MEK-ERK hyperactivation is very fre- unsuccessful [Sousa et al. 2008]. Various Raf quently evident in solid malignancies and is inhibitors have been developed, of which sorafenib transforming in multiple cancer models. In can- was the first to enter the clinic in the treatment of cers, activation of Ras-Raf-MEK-ERK occurs renal and hepatocellular cancers [Escudier et al. through various mechanisms. Hyperactivation of 2007; Llovet et al. 2008]. Even though sorafenib RTKs can induce constitutive activation of the is an inhibitor of wildtype and mutant B-Raf, its pathway. Furthermore, genetic alterations in the clinical activity is limited in B-Raf driven cancers pathway members are common mechanisms such as melanoma [Eisen et al. 2006; Gollob et al. behind the hyperactivation. Ras mutations or 2006]. It is likely that the MTD of sorafenib does loss of NF1 cause activation of the pathway at not lead to meaningful inhibition of mutant B-Raf the level of Ras. Ras mutations are commonly [<10% by phosphorylated ERK (pERK) assess- found in pancreatic, colorectal, lung, and endo- ment] [Davies et al. 2012]. metrial cancers [Almoguera et al. 1988; Cancer Genome Atlas Network, 2012b, 2013, 2014c] The clinical activity of sorafenib in renal and and NF1 mutations are seen in gliomas, sarco- hepatocellular cancers is largely related to the mas and lung cancers [Cancer Genome Atlas other kinases inhibited by it such as VEGFR2-3, Research, 2008, 2014c; Barretina et al. 2010]. FGFR-1, PDGFR-b, Flt-3 and c-Kit [Wilhelm Mutations in the Raf genes are another level et al. 2008]. Later, inhibitors preferentially target- where genetic constitutive activation occurs ing mutant B-Raf, such as vemurafenib and dab- repeatedly, while MEK or ERK mutations are rafenib, were developed. These inhibitors induce rare. Raf mutations are frequent in melanoma marked downregulation of B-Raf signaling and and thyroid cancers [Davies et al. 2002; Kimura >80% inhibition of pERK in tumors correlates et al. 2003] (Table 1, Figure 1). with responses [Bollag et al. 2010]. Dabrafenib 172 http://tam.sagepub.com E Jokinen and JP Koivunen Table 1. Frequencies of genetic alterations leading to activation of PI3K-AKT-mTOR and RAS-Raf-MEK-ERK pathways in specific solid cancers. Gene Cancer Alteration PI3KCA Mutation Endometrial 51% Breast 25% Lung (SQC) 16% Urinary tract 15% Glioma 15% Colon 12% Gastric 12% PI3KCA Amplification Head and neck 43% Ovarian 18% Lung (SQC) 12% Gastric 12% PTEN Inactivation Endometrial 59% Prostate 42% Glioma 36% Lung (SQC) 15% Breast 10% K-RAS Mutation Pancreatic 95% Colorectal 43% Lung (AC) 32% Endometrial 18% NF1 Inactivation Glioma 14% Sarcoma 11% Lung (AC) 11% B-Raf Mutation Melanoma 66% Thyroid (PAP) 36% AC, adenocarcinoma; PAP, papillary; SCC, squamous cell carcinoma. and vemurafenib have significant clinical activity keratoachantomas or even leukemias. These clini- in B-Raf mutant melanoma and have been cal manifestations strongly associate with pres- approved for this indication [Chapman et al. ence of Ras mutations in them [Callahan et al. 2011; Hauschild et al. 2012]. Conversely, mutant 2012; Su et al. 2012]. B-Raf targeted inhibitors paradoxically hyperacti- vate MEK/ERK signaling in cells with wildtype MEK inhibitors are classed into ATP competitive protein and therefore do not have antitumor and ATP non-competitive agents. Most of the activity in B-Raf wildtype tumors. Paradoxical known MEK inhibitors fall into the latter cate- hyperactivation of signaling by B-Raf inhibitors gory, and hence are not directly competing for the has been linked to increased C-Raf dimerization- ATP-binding site but bind to a unique allosteric mediated activation of downstream targets MEK/ binding site adjacent to the ATP binding site. ERK [Poulikakos et al. 2010]. Paradoxical MEK/ Thus, non-ATP competitive MEK inhibitors ERK hyperactivation in patients undergoing have high specificity for the target [Wallace et al. B-Raf inhibitor therapy can be manifested 2005]. CI-1040 was the first allosteric MEK1/2 by presentation of skin squamous cell cancers, inhibitor to enter clinical trials. CI-1040 was http://tam.sagepub.com 173 Therapeutic Advances in Medical Oncology 7(3) shown to be well tolerated but possessed insuffi- 2011]. The interaction of the parallel PI3K-AKT- cient antitumor activity in patients [Rinehart et al. mTOR and Ras-Raf-MEK-ERK signaling path- 2004]. Most of the newer MEK inhibitors follow- ways is thought to explain the inefficiency of ing CI-1040 share a similar chemical structure single-agent treatments in most malignancies and but are more potent inhibitors of the target. These gives rationale for the concurrent targeting of inhibitors include agents such as trametinib, selu- both pathways. metinib, GDC-0973 and BAY869766. Treatment with these agents seems to be well tolerated with To date, most efforts to dual target the PI3K- DLTs from rash, diarrhea and periferial edema AKT-mTOR and Ras-Raf-MEK-ERK pathways [Adjei et al. 2008; Bennouna et al. 2011; Infante have been carried out with a combination of PI3K et al. 2012; Lorusso et al. 2012]. Various methods and MEK inhibitors. However, many studies with have been used to analyze the dose–effect rela- mTOR and AKT inhibitors do also exist. PI3K tionship of the drugs, such as measurement of targeting agents are theoretically the most potent pERK levels in normal tissues and tumors. With inhibitors of the pathway since they lack the down- MTDs, these inhibitors induce >80% reduction stream feedback activation of mTOR inhibitors of pERK levels in various tissues [Adjei et al. [O’Reilly et al. 2006]. Furthermore, PI3K is also 2008; Weekes et al. 2013]. Even if true with B-Raf known to have other important, cancer-associated inhibitors, it is unknown whether MEK inhibitor downstream targets than AKT making PI3K induced target inhibition of >80% surpasses the inhibitors more potent than AKT inhibitors from required level for cytotoxicity. that perspective [Vasudevan et al. 2009]. Allosteric MEK inhibitors are highly specific for their targets MEK inhibitors are currently tested as single and are known to also induce high inhibition of agents or in combination with other targeted the pathway in patients [Adjei et al. 2008; Weekes agents or chemotherapy. Single-agent activity has et al. 2013], while the current RAF inhibitors are been seen in Ras and Raf mutant patients [Zimmer significantly less effective in inhibiting down- et al. 2014]. The most promising results are in stream signaling in tumors with wildtype Raf. thyroid cancer in combination with I-131 and in B-Raf mutant melanoma in combination with Preclinical models have shown that dual targeting B-Raf inhibitors [Flaherty et al. 2012; Ho et al. with PI3K and MEK inhibitors possesses antitu- 2013] (Table 2). mor activity in various cancer models and geno- types. Activity is also seen in difficult to treat genotypes, such as Ras mutant tumors and basal Dual inhibition like breast cancer [Engelman et al. 2008; Hoeflich Even though PI3K-AKT-mTOR and Ras-Raf- et al. 2009; Sos et al. 2009]. Even though Ras is MEK-ERK are the most commonly altered sign- generally thought to be a direct upstream activa- aling pathways in solid malignancies, the clinical tor of Raf, responses to MEK inhibitors are vari- efficiency of single pathway inhibitors has gener- able in Ras mutant cancer models. Variable ally been disappointing with some exclusion such responses could be explained by Ras mediated as B-Raf mutant melanoma or renal cell cancer. activation of other signaling pathways such as Cancers can be de novo dependent concurrently PI3K and, therefore, dual targeting with MEK on these two parallel pathways and cross-signal- and PI3K inhibitors is potentially more effective ing of the pathways is also evident, such as direct [Mendoza et al. 2011]. Some models, such as K-Ras mediated activation of PI3K [Faber et al. basal-like breast cancer, have shown that treat- 2009; Chandarlapaty et al. 2010]. Many in vivo ment with MEK inhibitors induces feedback acti- and in vitro studies have shown that PI3K-AKT- vation of RTKs which signal downstream to PI3K mTOR and Ras-Raf-MEK-ERK pathways regu- and therefore, dual targeting is more active late each other’s activity through feedback [Hoeflich et al. 2009]. Many preclinical studies mechanisms, i.e. MEK inhibition causes PI3K/ have investigated predictive factors for dual PI3K Akt activation via ERBB receptors [Turke et al. and MEK inhibitor therapy [Sos et al. 2009; 2012]. Furthermore, PI3K-AKT-mTOR and Jokinen et al. 2012]. Even though some studies Ras-Raf-MEK-ERK have also been shown to have suggested that responses are commonly seen coregulate shared downstream targets including in some cancer genotypes, no clear predictive forkhead box transcription factor class O (FOXO), markers have been found. Existence of measura- Bcl-2-associated death promoter (BAD) and gly- ble predictive factors would make the design and cogen synthase kinase 3 (GSK3) [Mendoza et al. execution of clinical trials more fruitful. 174 http://tam.sagepub.com E Jokinen and JP Koivunen Table 2. Examples of PI3K and MEK inhibitors tested in clinic. PI3K inhibitors GDC-0941 GDC-0941 is a potent and selective, orally bioavailable class I PI3K inhibitor which inhibits the growth of a wide of human tumor cell lines. The compound has shown strong inhibition of the growth of IGROV-1 ovarian cancer and U87M glioblastoma xenografts. GDC-0941 has already been tested in clinical trials. BEZ235 The imidazo[4,5-c]quinoline derivative BEZ235 is a dual PI3K-mTOR inhibitor. The compound blocks abnormal PI3K activation causing G1 cell cycle arrest. In vivo studies showed BEZ235 to be well tolerated and suitable for combinations studies. The inhibitor has entered clinical trials. BKM120 The 2-morpholino pyrimidine derivative BKM120 is a selective pan-PI3K inhibitor which inhibits all four class I isoforms of PI3K. The compound preferentially inhibits tumor cells with PIK3CA mutations over KRAS and PTEN mutated tumor cells. The biologic, pharmacological and preclinical safety profile of BKM120 h support its clinical use and the compound has been tested in phase II clinical trials with cancer patients. MEK inhibitors CI-1040 CI-1040 (PD-184352) is a specific small-molecule drug that inhibits MEK1/MEK2. The compound has been suggested to act as an allosteric inhibitor of MEK since it is known not to compete with the binding of ATP or protein substrates. CI-1040 blocks ERK phosphorylation and inhibits the growth of multiple human tumor cell lines as well as tumor growth in xenograft models. It has been shown that the cell growth inhibitory effect of CI-1040 is rapidly reversed after it is removed from the growth media. CI- 1040 was the first MEK inhibitor to enter clinical trials where it was shown to be well tolerated, but was found to have insufficient antitumor activity in patients. PD-0325901 is a CI-1040 derivate with higher solubility and improved pharmacological properties currently being tested in clinical trials. GDC-0973 GDC-0973 (XL518) is methanone derived, potent, orally bioavailable MEK1/2 inhibitor which has a 100-fold MEK1 selectivity over MEK2. The compound has shown tumor growth inhibition in BRAF and KRAS mutated cancer cell lines in vivo and in vitro. Trametinib Trametinib (GSK1120212), a selective MEK1 and MEK2 inhibitor, dephosphorylates phosphorylated MEK and also stabilizes the unphosphorylated form, thereby blocking the downstream signaling. At the moment, it is the only MEK inhibitor, which has showed promising results in phase III clinical trials. Pimasertib Pimasertib (AS703026, MSC1936369B) is an orally available, selective second generation inhibitor of MEK1/2 which binds ATP non-competitively to the distinctive allosteric site of MEK1/2 causing G0–G1 cell cycle arrest and apoptosis. Pimasertib has been shown to cause significant tumor regression in human multiple myeloma and KRAS mutated colorectal cancer xenografts. ATP, adenosine triphosphate; ERK, extracellular-signal-regulated kinase; MEK, mitogen-activated protein kinase kinase; PI3K, phosphoinositide-3 kinase. Dual inhibition: clinical studies adverse events (AEs) of therapy include diarrhea, Numerous early-phase clinical studies investigat- rash, fatigue, vomiting and hyperglycemia. ing dual PI3K and MEK targeting are ongoing Unfortunately, the rate of response seems to be and some results have been presented in meetings quite low with an overall response rate of 4.7% during the past three years. Because of the hetero- and a disease control rate of 19.2% when all four geneity and early-phase nature of the studies, it is presented studies are combined. As in the pre- difficult to derive a strong conclusion of the activ- clinical models, no clear predictive factors for ity in various cancer subgroups and general tox- response have been identified. However, responses icities. Therefore, we review four different phase I have been seen in Ras or Raf mutated cancers, and clinical studies and present their results. later studies have enriched these patient populations. Generally, combined PI3K and MEK inhibitor therapy seems to be feasible with manageable The clinical activity of GDC-0973 (MEK1/2 safety and toxicity profile. The most common inhibitor) and GDC-0941 (class I PI3K inhibitor) http://tam.sagepub.com 175 Therapeutic Advances in Medical Oncology 7(3) was studied in 78 patients with advanced solid treated once a day with BYL719 (80–270 mg) tumors. The patients received daily both GDC- and twice a day with 30 or 45 mg of binimetinib. 0973 and GDC-0941 with a 21 day on, 7 days off Common AEs were diarrhea, nausea, vomiting, schedule or with an intermittent schedule where reduced appetite, rash, pyrexia, fatigue and hyper- GDC-0973 was dosed on days 1, 4, 8, 11, 15 and glycemia. Confirmed partial responses were seen 18 with a 28 day cycle and GDC-0941 daily with with three of the four patients with K-Ras-mutant a 21/7 schedule. DLTs included elevation of grade ovarian cancer, with one patient with N-Ras 3 lipase and grade 4 creatine phosphokinase mutated melanoma. Stable disease status lasting (CPK). Reported AEs were diarrhea, fatigue, nau- for over 6 weeks was seen with 18 patients [Juric sea, rash, vomiting, dysgeusia, decreased appetite et al. 2014]. and CPK elevation. The combination therapy was found to be well tolerated, with toxicities compa- rable with those reported with single agents in Further directions phase I trials. Higher drug doses were tolerated A strong rationale and preclinical results have with the intermitted dosing schedule. Partial established the groundwork for the clinical devel- responses were seen in three patients: one patient opment of dual PI3K and MEK inhibitor therapy with B-Raf mutated melanoma, one with B-Raf to solid malignancies. However, early-phase clini- mutated pancreatic cancer, and one with K-Ras cal trials presented to date have only shown mod- mutated endometrial cancer. Stable disease last- est activity of the combination. There is room for ing over 5 months was observed in five patients new ideas and approaches to improve the antitu- [Lorusso et al. 2012]. mor activity and tolerability of the dual PI3K and MEK inhibitor therapy. Daily dosing of BKM120 (pan-PI3K inhibitor) and trametinib (MEK inhibitor) was evaluated Optimal dosing schedule for PI3K and MEK with 49 patients with advanced Ras- or B-Raf- inhibitor combination is unknown. Current clini- mutant cancers. The combination was found to cal dosing regimens of PI3K and MEK inhibitors be safe on patients. Grade 3 DLTs included sto- mainly investigate longer drug exposures with matitis, dyspahgia, ejection fraction decrease, subtotal suppression of their targets. It is likely CPK elevation, nausea and anorexia. The most that more robust target inhibition with higher commonly observed AEs were listed as dermati- drug doses would have more antitumor activity. It tis, diarrhea, nausea, vomiting, rash, asthenia, is possible that higher drug doses with short expo- increase in CPK, loss of appetite, pyrexia, stoma- sure times could be clinically as tolerable as longer titis and hyperglycemia. Partial responses were exposure with lower doses, but poses more antitu- seen in three patients with K-Ras mutant ovarian mor activity. Some preclinical studies have shown cancer and stable disease was observed with two that short, alternative dosing schedules of PI3K patients with B-Raf mutated melanoma [Bedard and MEK inhibitor combination can be as effec- et al. 2012]. tive as longer exposures [Hoeflich et al. 2012; Jokinen et al. 2012]. Furthermore, it is currently In another combination study, 49 patients were unknown if both PI3K and MEK inhibitors need treated with the pan-PI3K inhibitor copanlisib to be administered concurrently for the same and the MEK inhibitor refametinib. DLTs period of time, or if either of the drugs could be included grade 3 aspartate transaminase (AST) used intermittently. There is some preclinical evi- or alanine aminotransferase (ALT) elevation, dence for the intermittent uses of the drugs and hyperglycemia, hypertension, diarrhea, mucositis some clinical studies have also tried to answer this and rash. The most common AEs included diar- question [Hoeflich et al. 2012; Jokinen et al. 2012; rhea, nausea, hyperglycemia, fatigue, rash, ano- Lorusso et al. 2012]. We believe that studying rexia and hypertension. One partial response was numerous, highly dual PI3K and MEK therapy- seen in a patient with endometrial cancer, and sensitive preclinical models could be a useful way stable disease lasting >4 cycles in 9 patients to figure out the optimal dosing schedules from [Ramanathan et al. 2014]. the perspective of cytotoxicity. The dosing regi- mens could be brought forward for early-phase The combination of BYL719 (PI3Kα inhibitor) clinical trials to test their safety and toxicity. and binimetinib (MEK inhibitor) was studied in patients with advanced solid tumors with Ras or An open question remains if the current PI3K B-Raf mutations. A total of 58 patients were and MEK inhibitors are optimal drugs for cancer 176 http://tam.sagepub.com E Jokinen and JP Koivunen therapy. Allosteric MEK inhibitors are well-toler- activation of the other pathways or shared down- ated agents and do induce robust target inhibition stream targets. Therefore, a strong rationale for with daily MTD [Adjei et al. 2008; Weekes et al. dual PI3K and MEK inhibitor therapy for cancer 2013]. Based on the clinical trials, current PI3K treatment exists. Various preclinical models have inhibitors are only able to induce modest target identified this therapy to be efficient in various can- inhibition (30–90% inhibition of pAKT and pS6 cers and genotypes. Currently, numerous early- biomarkers) with daily MTD [Bendell et al. 2011; phase clinical trials investigating dual PI3K and Moreno Garcia et al. 2011]. Isoform-specific MEK therapy are ongoing. Clinical results pre- PI3K inhibitors could have less off-tumor effects sented to date have shown that dual targeting is fea- and, therefore, enable more robust target inhibi- sible with manageable toxicity and safety profile, tion with MTD. Promising phase I results of this but the rate of response is quite low. It is possible approach (PI3Kα and MEK inhibitor combina- that efficiency of dual targeting could be increased tion) have recently been presented [Juric et al. by alternative dosing schedules, newer generations 2014]. Furthermore, it is possible that blocking of agents, intelligent drug combinations, or enrich- other members of the PI3K-AKT-mTOR path- ing the patients with predictive factors. way could be more effective and tolerable than inhibiting PI3K itself. Funding The study was supported by Cancer Society of Enriching the therapy-sensitive patients in clinical Northern Finland, Emil Aaltonen Foundation, trials is an efficient way to speed up the develop- Finnish Foundation for Tuberculosis Resistance, ment of an agent or a combination. Despite the Finnish Oncological Society, Oulu University robust preclinical and clinical efforts, no clear Hospital, Orion-Farmos Science Foundation, predictive factors for dual PI3K and MEK inhibi- Sigrid Juselius Foundation and Thelma Mäkikyrö tor therapy have been identified [Sos et al. 2009; foundation. Bedard et al. 2012; Jokinen et al. 2012; Lorusso et al. 2012]. From this perspective, it is unlikely Conflict of interest statement that an easily measurable predictive factor to the The authors declare no conflicts of interest in dual therapy will be identified. preparing this article. Cytotoxic drug combinations are the backbone of the current cancer therapy. Much less is known about targeted therapy combinations. It is possi- References Adjei, A., Cohen, R., Franklin, W., Morris, ble that combining either cytotoxics or targeted C., Wilson, D., Molina, J. et al. (2008) Phase I agents to PI3K and MEK therapy could increase pharmacokinetic and pharmacodynamic study of the number of tumors sensitive to them. As an the oral, small-molecule mitogen-activated protein example, novel preclinical works presented kinase kinase 1/2 inhibitor AZD6244 (ARRY-142886) recently that antitumor activity of a MEK or in patients with advanced cancers. J Clin Oncol 26: mTOR inhibitors can be markedly increased by 2139–2146. blocking of BCL-XL anti-apoptotic protein by Almoguera, C., Shibata, D., Forrester, K., Martin, J., BH3 mimetic drugs [Corcoran et al. 2012; Faber Arnheim, N. and Perucho, M. (1988) Most human et al. 2014]. We feel that large-scale preclinical carcinomas of the exocrine pancreas contain mutant modeling would be appropriate to try to identify C-K-Ras genes. Cell 53: 549–554. the best combinations for ‘dual or triple inhibi- tion’ therapy. However, it remains highly chal- Barretina, J., Taylor, B., Banerji, S., Ramos, A., Lagos- Quintana, M., Decarolis, P. et al. (2010) Subtype- lenging to bring two or three investigational specific genomic alterations define new targets for agents concurrently to clinical trials. soft-tissue sarcoma therapy. Nat Genet 42: 715–721. Baselga, J., Campone, M., Piccart, M., Burris, H., Summary 3rd, Rugo, H., Sahmoud, T. et al. (2012) Everolimus PI3K-AKT-mTOR and Ras-Raf-MEK-ERK are in postmenopausal hormone-receptor-positive advanced breast cancer. N Engl J Med 366: 520–529. the most commonly altered signaling pathways in solid malignancies, but the clinical efficiency of Bedard, P., Tabernero, J., Kurzrock, R., Britten, C., single-pathway inhibitors have generally been dis- Stathis, A., Perez-Garcia, J. et al. (2012) A phase Ib, appointing. Failure of the single-pathway targeting open-label, multicenter, dose-escalation study of the could result from various factors, such as feedback oral PAN-PI3K inhibitor BKM120 in combination http://tam.sagepub.com 177 Therapeutic Advances in Medical Oncology 7(3) with the oral MEK1/2 inhibitor GSK1120212 in Cancer Genome Atlas Network (2012c) patients (pts) with selected advanced solid tumors. Comprehensive molecular portraits of human breast J Clin Oncol 30(Suppl.): abstract 3003. tumours. Nature 490: 61–70. Bellmunt, J., Teh, B., Tortora, G. and Rosenberg, J. Cancer Genome Atlas Research Network (2013) (2013) Molecular targets on the horizon for kidney Integrated genomic characterization of endometrial and urothelial cancer. Nat Rev Clin Oncol 10: 557–570. carcinoma. Nature 497: 67–73. Cancer Genome Atlas Research Network (2014a) Bendell, J., Rodon, J., Burris, H., De Jonge, M., Comprehensive molecular characterization of gastric Verweij, J., Birle, D. et al. (2011) Phase I, dose- adenocarcinoma. Nature 513: 202–209. escalation study of BKM120, an oral pan-class I PI3K inhibitor, in patients with advanced solid tumors. Cancer Genome Atlas Research Network (2014b) J Clin Oncol 30: 282–290. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature 507: 315–322. Bennouna, J., Lang, I., Valladares-Ayerbes, M., Boer, K., Adenis, A., Escudero, P. et al. (2011) A Cancer Genome Atlas Research Network (2014c) phase II, open-label, randomised study to assess the Comprehensive molecular profiling of lung efficacy and safety of the MEK1/2 Inhibitor AZD6244 adenocarcinoma. Nature 511: 543–550. (ARRY-142886) versus capecitabine monotherapy in Chandarlapaty, S., Sawai, A., Scaltriti, M., Rodrik- patients with colorectal cancer who have failed one Outmezguine, V., Grbovic-Huezo, O., Serra, V. et al. or two prior chemotherapeutic regimens. Invest New (2010) AKT inhibition relieves feedback suppression Drugs 29: 1021–1028. of receptor tyrosine kinase expression and activity. Bismar, T., Yoshimoto, M., Vollmer, R., Duan, Q., Cancer Cell 19: 58–71. Firszt, M., Corcos, J. et al. (2011) PTEN genomic Chapman, P., Hauschild, A., Robert, C., Haanen, J., deletion is an early event associated with ERG gene Ascierto, P., Larkin, J. et al. (2011) Improved survival rearrangements in prostate cancer. BJU Int 107: with vemurafenib in melanoma with BRAF V600E 477–485. mutation. N Engl J Med 364: 2507–2516. Bollag, G., Hirth, P., Tsai, J., Zhang, J., Ibrahim, Corcoran, R., Cheng, K., Hata, A., Faber, A., Ebi, P., Cho, H. et al. (2010) Clinical efficacy of a RAF H., Coffee, E. et al. (2012) Synthetic lethal interaction inhibitor needs broad target blockade in BRAF- of combined BCL-XL and MEK inhibition promotes mutant melanoma. Nature 467: 596–599. tumor regressions in KRAS mutant cancer models. Brachmann, S., Hofmann, I., Schnell, C., Fritsch, Cancer Cell 23: 121–128. C., Wee, S., Lane, H. et al. (2009) Specific apoptosis Davies, H., Bignell, G., Cox, C., Stephens, P., induction by the dual PI3K/mTOR inhibitor NVP- Edkins, S., Clegg, S. et al. (2002) Mutations of the BEZ235 in HER2 amplified and PIK3CA mutant BRAF gene in human cancer. Nature 417: 949–954. breast cancer cells. Proc Natl Acad Sci U S A 106: 22299–22304. Davies, M., Fox, P., Papadopoulos, N., Bedikian, A., Hwu, W., Lazar, A. et al. (2012) Phase I study Brana, I. and Siu, L. (2012) Clinical development of the combination of sorafenib and temsirolimus in of phosphatidylinositol 3-kinase inhibitors for cancer patients with metastatic melanoma. Clin Cancer Res treatment. BMC Med 10: 161. 18: 1120–1128. Callahan, M., Rampal, R., Harding, J., Klimek, V., Dienstmann, R., Rodon, J., Markman, B. and Chung, Y., Merghoub, T. et al. (2012) Progression of Tabernero, J. (2011) Recent developments in anti- RAS-mutant leukemia during raf inhibitor treatment. cancer agents targeting PI3K, AKT and mTORC1/2. N Engl J Med 367: 2316–2321. Recent Pat Anticancer Drug Discov 6: 210–236. Cancer Genome Atlas Research Network (2008) Eisen, T., Ahmad, T., Flaherty, K., Gore, M., Kaye, Comprehensive genomic characterization defines S., Marais, R. et al. (2006) Sorafenib in advanced human glioblastoma genes and core pathways. Nature melanoma: a phase II randomised discontinuation 455: 1061–1068. trial analysis. Br J Cancer 95: 58–586. Cancer Genome Atlas Research Network (2011) Engelman, J., Chen, L., Tan, X., Crosby, K., Integrated genomic analyses of ovarian carcinoma. Guimaraes, A., Upadhyay, R. et al. (2008) Effective Nature 474: 609–615. use of PI3K and MEK inhibitors to treat mutant Cancer Genome Atlas Research Network (2012a) KRAS G12D and PIK3CA H1047R murine lung Comprehensive genomic characterization of squamous cancers. Nat Med 14: 1351–1356. cell lung cancers. Nature 489: 519–525. Engelman, J., Luo, J. and Cantley, L. (2006) The Cancer Genome Atlas Network (2012b) evolution of phosphatidylinositol 3-kinases as Comprehensive molecular characterization of human regulators of growth and metabolism. Nat Rev Genet colon and rectal cancer. Nature 487: 330–337. 7: 606–619. 178 http://tam.sagepub.com E Jokinen and JP Koivunen Escudier, B., Eisen, T., Stadler, W., Szczylik, C., Juric, D., Soria, J., Sharma, S., Banerjee, U., Azaro, Oudard, S., Siebels, M. et al. (2007) Sorafenib in A., Desai, J. et al. (2014) A phase 1b Dose-escalation advanced clear-cell renal-cell carcinoma. N Engl J study of BYL719 plus binimetinib (MEK162) in Med 356: 125–134. patients with selected advanced solid tumors. J Clin Oncol 32:5s (Suppl.): abstract 9051. Faber, A., Coffee, E., Costa, C., Dastur, A., Ebi, H., Hata, A. et al. (2014) mTOR inhibition Kimura, E., Nikiforova, M., Zhu, Z., Knauf, J., specifically sensitizes colorectal cancers with KRAS Nikiforov, Y. and Fagin, J. (2003) High prevalence or BRAF mutations to BCL-2/BCL-Xl inhibition by of BRAF mutations in thyroid cancer: genetic suppressing MCL-1. Cancer Discov 4: 42–52. evidence for constitutive activation of the RET/PTC- RAS-BRAF signaling pathway in papillary thyroid Faber, A., Li, D., Song, Y., Liang, M., Yeap, B., carcinoma. Cancer Res 63: 1454–1457. Bronson, R. et al. (2009) Differential Induction of apoptosis in HER2 and EGFR addicted cancers Lin, S., Liu, C., Ko, S., Chang, H., Liu, T. and following PI3K inhibition. Proc Natl Acad Sci U S A Chang, K. (2005) Copy number amplification of 106: 19503–19508. 3Q26-27 oncogenes in microdissected oral squamous Flaherty, K., Infante, J., Daud, A., Gonzalez, R., cell carcinoma and oral brushed samples from areca Kefford, R., Sosman, J. et al. (2012) Combined BRAF chewers. J Pathol 206: 417–422. and MEK inhibition in melanoma with BRAF V600 Llovet, J., Ricci, S., Mazzaferro, V., Hilgard, P., mutations. N Engl J Med 367: 1694–1703. Gane, E., Blanc, J. et al. (2008) Sorafenib in advanced Gollob, J., Wilhelm, S., Carter, C. and Kelley, S. hepatocellular carcinoma. N Engl J Med 359: 37–390. (2006) Role of RAF kinase in cancer: therapeutic Lorusso, P., Shapiro, G., Panday, S., Kwak, E., Jones, potential of targeting the RAF/MEK/ERK signal C., Belvin, M. et al. (2012) A first-in-human phase transduction pathway. Semin Oncol 33: 392–406. Ib study to evaluate the MEK inhibitor GDC-0973, Hanahan, D. and Weinberg, R. (2011) Hallmarks of combined with the pan-PI3K Inhibitor GDC-0941, cancer: the next generation. Cell 144: 646–674. in patients with advanced solid tumors. J Clin Oncol 30(Suppl.): abstract 2566. Hauschild, A., Grob, J., Demidov, L., Jouary, T., Gutzmer, R., Millward, M. et al. (2012) Dabrafenib in Mendoza, M., Er, E. and Blenis, J. (2011) The BRAF-mutated metastatic melanoma: a multicentre, RAS-Erk and PI3K-mTOR pathways: cross-talk and open-label, phase 3 randomised controlled trial. compensation. Trends Biochem Sci 36: 320–328. Lancet 380: 358–365. Moreno Garcia, V., Baird, R., Shah, K., Basu, B., Ho, A., Grewal, R., Leboeuf, R., Sherman, E., Pfister, Tunariu, N., Blanco, M. et al. (2011) A phase I study D., Deandreis, D. et al. (2013) Selumetinib-enhanced evaluating GDC-0941, an oral phosphoinositide-3 radioiodine uptake in advanced thyroid cancer. N Engl kinase (PI3K) inhibitor, in patients with advanced J Med 368: 623–632. solid tumors or multiple myeloma. J Clin Oncol Hoeflich, K., Merchant, M., Orr, C., Chan, J., 29(Suppl.): abstract 3021. Den Otter, D., Berry, L. et al. (2012) Intermittent Motzer, R., Escudier, B., Oudard, S., Hutson, administration of MEK inhibitor GDC-0973 plus T., Porta, C., Bracarda, S. et al. (2008) Efficacy PI3K inhibitor GDC-0941 Triggers robust apoptosis of everolimus in advanced renal cell carcinoma: a and tumor growth inhibition. Cancer Res 72: 210–219. double-blind, randomised, placebo-controlled phase Hoeflich, K., O’Brien, C., Boyd, Z., Cavet, G., III trial. Lancet 372: 449–456. Guerrero, S., Jung, K. et al. (2009) In vivo antitumor activity of MEK and phosphatidylinositol 3-kinase O’Reilly, K., Rojo, F., She, Q., Solit, D., Mills, G., inhibitors in basal-like breast cancer models. Clin Smith, D. et al. (2006) mTOR inhibition induces Cancer Res 15: 4649–4664. upstream receptor tyrosine kinase signaling and activates AKT. Cancer Res 66: 1500–1508. Hudes, G., Carducci, M., Tomczak, P., Dutcher, J., Figlin, R., Kapoor, A. et al. (2007) Temsirolimus, Poulikakos, P., Zhang, C., Bollag, G., Shokat, K. and interferon alfa, or both for advanced renal-cell Rosen, N. (2010) RAF inhibitors transactivate RAF carcinoma. N Engl J Med 356: 2271–2281. dimers and ERK signalling in cells with wild-type BRAF. Nature 464: 427–430. Infante, J., Fecher, L., Falchook, G., Nallapareddy, S., Gordon, M., Becerra, C. et al. (2012) Safety, Ramanathan, R., Von Hoff, D., Eskens, F., pharmacokinetic, pharmacodynamic, and efficacy data Blumenschein, G., Richards, D., Renshaw, F. et al. for the oral MEK inhibitor trametinib: a phase 1 dose- (2014) A Phase 1b trial of PI3K inhibitor copanlisib escalation trial. Lancet Oncol 13: 773–781. (BAY 80-6946) combined with the allosteric-MEK Jokinen, E., Laurila, N. and Koivunen, J. (2012) Inhibitor refametinib (BAY 86-9766) in patients with Alternative dosing of dual PI3K and MEK inhibition advanced cancer. J Clin Oncol 32:5s(Suppl.): in cancer therapy. BMC Cancer 12: 612. abstract 2588. http://tam.sagepub.com 179 Therapeutic Advances in Medical Oncology 7(3) Rinehart, J., Adjei, A., Lorusso, P., Waterhouse, D., Wallace, E., Lyssikatos, J., Yeh, T., Winkler, J. and Hecht, J., Natale, R. et al. (2004) Multicenter phase II Koch, K. (2005) Progress towards therapeutic small study of the oral MEK inhibitor, CI-1040, in patients molecule MEK inhibitors for use in cancer therapy. with advanced non-small-cell lung, breast, colon, and Curr Top Med Chem 5: 215–229. pancreatic cancer. J Clin Oncol 22: 4456–4462. Wander, S., Hennessy, B. and Slingerland, J. Serra, V., Scaltriti, M., Prudkin, L., Eichhorn, P., (2011) Next-generation mTOR inhibitors in Ibrahim, Y., Chandarlapaty, S. et al. (2011) PI3K clinical oncology: how pathway complexity informs inhibition results in enhanced HER Signaling and therapeutic strategy. J Clin Invest 121: 1231–1241. acquired ERK dependency in HER2-overexpressing Weekes, C., Von Hoff, D., Adjei, A., Leffingwell, breast cancer. Oncogene 30: 2547–2557. D., Eckhardt, S., Gore, L. et al. (2013) Multicenter Solit, D., Garraway, L., Pratilas, C., Sawai, A., Getz, phase I trial of the mitogen-activated protein kinase G., Basso, A. et al. (2006) BRAF mutation predicts 1/2 inhibitor BAY 86-9766 in patients with advanced sensitivity to MEK inhibition. Nature 439: 358–362. cancer. Clin Cancer Res 19: 1232–1243. Sos, M., Fischer, S., Ullrich, R., Peifer, M., Weigelt, B., Warne, P. and Downward, J. (2011) Heuckmann, J., Koker, M. et al. (2009) Identifying PIK3CA mutation, but not PTEN loss of function, genotype-dependent efficacy of single and combined determines the sensitivity of breast cancer cells to PI3K- and MAPK-pathway inhibition in cancer. Proc mTOR inhibitory drugs. Oncogene 30: 3222–3233. Natl Acad Sci U S A 106: 18351–18356. Wilhelm, S., Adnane, L., Newell, P., Villanueva, A., Sousa, S., Fernandes, P. and Ramos, M. (2008) Llovet, J. and Lynch, M. (2008) Preclinical overview Farnesyltransferase inhibitors: a detailed chemical of sorafenib, a multikinase inhibitor that targets both view on an elusive biological problem. Curr Med Chem RAF and VEGF and PDGF Receptor tyrosine kinase 15: 1478–1492. signaling. Mol Cancer Ther 7: 3129–3140. Su, F., Viros, A., Milagre, C., Trunzer, K., Bollag, Yan, Y., Serra, V., Prudkin, L., Scaltriti, M., Murli, G., Spleiss, O. et al. (2012) RAS mutations in S., Rodriguez, O. et al. (2013) Evaluation and clinical cutaneous squamous-cell carcinomas in patients analyses of downstream targets of the AKT inhibitor treated with BRAF inhibitors. N Engl J Med 366: GDC-0068. Clin Cancer Res 19: 6976–6986. 207–215. Yao, J., Shah, M., Ito, T., Bohas, C., Wolin, E., Van Turke, A., Song, Y., Costa, C., Cook, R., Arteaga, C., Cutsem, E. et al. (2011) Everolimus for advanced Asara, J. et al. (2012) MEK inhibition leads to PI3K/ pancreatic neuroendocrine tumors. N Engl J Med 364: AKT activation by relieving a negative feedback on 514–523. ERBB receptors. Cancer Res 72: 3228–3237. Zimmer, L., Barlesi, F., Martinez-Garcia, M., Vasudevan, K., Barbie, D., Davies, M., Rabinovsky, Dieras, V., Schellens, J., Spano, J. et al. (2014) Phase R., McNear, C., Kim, J. et al. (2009) AKT- I expansion and pharmacodynamic study of the Visit SAGE journals online independent signaling downstream of oncogenic oral MEK inhibitor RO4987655 (CH4987655) in http://tam.sagepub.com PIK3CA mutations in human cancer. Cancer Cell 16: selected patients with advanced cancer with RAS-RAF SAGE journals 21–32. mutations. Clin Cancer Res 20: 4251–4261. 180 http://tam.sagepub.com
Therapeutic Advances in Medical Oncology – SAGE
Published: Feb 24, 2015
Keywords: inhibition; MEK; PI3K; solid tumors; targeted therapy
Access the full text.
Sign up today, get DeepDyve free for 14 days.