**3. Overview of small molecule Cdk inhibitors: Selectivity and mode of action**

In the last decade, numerous of structurally diverse small molecules inhibitors of Cdk activity have been developed and evaluated *in vitro* and *in vivo*. Thereby, structural information of different Cdk, in complex with corresponding regulatory Cyclin subunits, and in association with inhibitors facilitated the development of new potent compounds with high specificity for Cdk versus other protein kinases (De Bondt et al., 1993; Jeffrey et al., 1995; Sridhar et al., 2006). Motivation of Cdk specific targeting was attributed to high cytotoxicities of first generation of unspecific kinase inhibitors, e. g., staurosporine targeting several protein kinase families, which limited clinical application (Sielecki et al., 2000). In addition, effort to the identification of Cdk-subtype selective inhibitors has been made to optimize therapeutic success and minimize side effects for cancer patients. In first preclinical and clinical trials of potent, pharmacological Cdk inhibitors only misleading results for tumor treatment were revealed due to non-specific and/or non-selective targeting (Shapiro, 2006). According to this, Cdk inhibitors were classified to their effects on Cdk family members as pan-Cdk or highly selective Cdk inhibitors. Nevertheless, the classification of Cdk inhibitors often reveals the subjective view of the authors, not least because of different experimental setup of Cdk affinity measurements, missing data of the selectivities to many kinases and diffuse boundaries between Cdk inhibitors with broad and narrow activity profile.

The major targets of pharmacological Cdk inhibitors are key enzymes regulating interphase (Cdk2, Cdk4, Cdk6) and mitotic Cdk1. But often also Cdk activating kinase Cdk7 and transcriptional Cdk, e. g., Cdk9 are affected.

All potent small molecule Cdk inhibitors interact with the catalytic active site of Cdk and compete with ATP or block ATP binding. Several review articles of the last decade well document the development and evaluation of dozens of Cdk inhibitors representing different chemical classes (Galons et al., 2010; Rizzolio et al., 2010; Senderowicz & Sausville, 2000; Sharma et al., 2008; Sridhar, 2006). About 25 of them reached clinical trials. However, several studies with Cdk inhibitors showing promising preclinical results for example AG-024322, AZD5438, R547, SNS-032, and ZK 304709 had to be terminated or were discontinued after phase I (Lapenna & Giordano, 2009) (www.clinical trials.gov).

This book chapter will focus on 9 promising compounds tested in clinical trials at the time: AT7519, BAY 1000394, flavopiridol, P1446A-05, P276-00, PD 0332991, PHA-848125, Rroscovitine, and SCH 727965 (Fig. 2, Table 2).

Universal abnormal regulation of Cdk pathways especially in G1/ S phase suggests involvement and importance of these kinases in carcinogenesis, despite of uncertain results concerning dependence of Cdk2, Cdk4 and Cdk6 on cell cycle progression in embryogenesis

In consequence, cell cycle regulating Cdks are attractive molecular targets for new (radio)pharmaceutical strategies in both cancer therapy and diagnosis, considering the heterogeneity of Cdk activity in different human tumor types. Compounds directly inhibiting the cell cycle machinery hold promise in restoring missing cellular Cdk regulators and arrest of proliferating cells, thus providing a non-genotoxic therapy modality. Conspicuous amplification of Cdk in tumor cells provides an opportunity for visualization

**3. Overview of small molecule Cdk inhibitors: Selectivity and mode of action**  In the last decade, numerous of structurally diverse small molecules inhibitors of Cdk activity have been developed and evaluated *in vitro* and *in vivo*. Thereby, structural information of different Cdk, in complex with corresponding regulatory Cyclin subunits, and in association with inhibitors facilitated the development of new potent compounds with high specificity for Cdk versus other protein kinases (De Bondt et al., 1993; Jeffrey et al., 1995; Sridhar et al., 2006). Motivation of Cdk specific targeting was attributed to high cytotoxicities of first generation of unspecific kinase inhibitors, e. g., staurosporine targeting several protein kinase families, which limited clinical application (Sielecki et al., 2000). In addition, effort to the identification of Cdk-subtype selective inhibitors has been made to optimize therapeutic success and minimize side effects for cancer patients. In first preclinical and clinical trials of potent, pharmacological Cdk inhibitors only misleading results for tumor treatment were revealed due to non-specific and/or non-selective targeting (Shapiro, 2006). According to this, Cdk inhibitors were classified to their effects on Cdk family members as pan-Cdk or highly selective Cdk inhibitors. Nevertheless, the classification of Cdk inhibitors often reveals the subjective view of the authors, not least because of different experimental setup of Cdk affinity measurements, missing data of the selectivities to many kinases and diffuse boundaries between Cdk inhibitors with broad and narrow activity

The major targets of pharmacological Cdk inhibitors are key enzymes regulating interphase (Cdk2, Cdk4, Cdk6) and mitotic Cdk1. But often also Cdk activating kinase Cdk7 and

All potent small molecule Cdk inhibitors interact with the catalytic active site of Cdk and compete with ATP or block ATP binding. Several review articles of the last decade well document the development and evaluation of dozens of Cdk inhibitors representing different chemical classes (Galons et al., 2010; Rizzolio et al., 2010; Senderowicz & Sausville, 2000; Sharma et al., 2008; Sridhar, 2006). About 25 of them reached clinical trials. However, several studies with Cdk inhibitors showing promising preclinical results for example AG-024322, AZD5438, R547, SNS-032, and ZK 304709 had to be terminated or were

discontinued after phase I (Lapenna & Giordano, 2009) (www.clinical trials.gov).

This book chapter will focus on 9 promising compounds tested in clinical trials at the time: AT7519, BAY 1000394, flavopiridol, P1446A-05, P276-00, PD 0332991, PHA-848125, R-

(Malumbres & Barbacid, 2009; Santamaria et al., 2007).

of tumors by means of positron emission tomography (PET).

profile.

transcriptional Cdk, e. g., Cdk9 are affected.

roscovitine, and SCH 727965 (Fig. 2, Table 2).

*pan-Cdk inhibitors* 

Fig. 2. Molecular structures of selected Cdk inhibitors tested in clinical trials at the time

Most of the Cdk inhibitors are less selective and affect several Cdk family members, not only resulting in cell cycle arrest via blocking of Cdk1, Cdk2, and Cdk4, but also in manipulation of RNA synthesis through targeting of transcriptional kinases Cdk7 and Cdk9. Combination of cell cycle kinase inactivation and Cdk9 inhibition has been shown to trigger cell death via promotion of apoptosis in tumor cells (Cai et al., 2006). Otherwise, toxic effects of transcriptional manipulation in non-tumor cells via inhibition of Cdk7 and Cdk9 could be crucial for therapeutic application of single agent Cdk inhibitors. In addition, it has to be considered, that unselective targeting of transcriptional Cdk, e. g., Cdk10 and Cdk11 would

Cyclin-Dependent Kinases (Cdk) as Targets for Cancer Therapy and Imaging 271

arrest, growth inhibition, and apoptosis in a large spectrum of human solid tumor cells and colon carcinoma xenograft models (Squires et al., 2009). Influence on transcriptional regulating Cdk, as well as GSK3 resulting in induction of apoptotic pathways was observed in multiple myeloma and leukemia cell lines (Santo et al., 2010; Squires et al.,

Optimization of selectivity and physicochemical properties of pyrazolo[4,3-*h*]quinazoline-3 carboxamide derivatives identified **PHA-848125** as potent Cdk inhibitor (IC50 < 400 nM) with inhibitory effects on cell cycle progression and Rb phosphorylation *in vitro* in a wide range of tumor cell lines (Brasca et al., 2009; Caporali et al., 2010). Among a panel of 43 other serine-threonine and tyrosine kinases only tropomyosin receptor kinase A (TRKA), linked to cancer cell survival, was inhibited by PHA-848125 in the same nanomolar range (IC50 53 nM). *In vivo* characterization of oral active compound PHA-848125 showed a dosedependent inhibition of human A2780 ovarian carcinoma xenograft tumor growth on mice up to 91%. Significant tumor growth inhibition was also observed in a K-Ras mutant lung adenocarcinoma transgenic mouse model, carcinogen-induced tumors, and in disseminated

Only a few details have been published for pan-Cdk inhibitor **BAY 1000394** with unknown structure targeting Cdk1, Cdk2, Cdk4 and Cdk9 with high affinity (IC50 < 10 nM) (Siemeister et al., 2010). Inhibition of cell proliferation was amongst others depicted for breast, cervical, and colorectal human tumor cell lines and described to be independent of Rb, p53 and tumor suppressor gene status. Suppression of Rb phosphorylation and growth inhibition after treatment with BAY 1000394 was observed in preclinical studies in a broad range of

Several compounds, particularly, members of chemical classes of benzothiadiazines, diarylureas, indolocarbazoles, oxindoles, pyrido[2,3-*d*]pyrimidines, thienopyrimidinhy drazones, thioacridones, and triaminopyrimidines were developed in the last decade and described with preferential selectivity to Cdk4 and Cdk6 (Graf, 2010; Lee & Sicinski, 2006). Cdk4 and Cdk6 reveal due to their high homology and identical substrate specificities comprehensive activities in the cell cycle. According to the occurrence in different tissues Cdk4 and Cdk6 compensate each other in function (Ciemerych & Sicinski, 2005; Meyerson & Harlow, 1994). **P1446A-05** and pyrido[2,3-*d*]pyrimidine **PD 0332991** seem to be the most promising compounds due to their inclusion in clinical evaluation. Unfortunately, no information about chemical structure and mechanism of Cdk4 selective inhibition is provided in the literature for P1446A-05. However, potency of PD 0332991 to inhibit Cdk4 and Cdk6 pathway and tumor cell progression has been extensively studied *in vitro*, as well as *in vivo* in mouse xenograft models (Baughn et al., 2006; Fry et al., 2004; Menu et al., 2008; Saab et al., 2006). In all studies an antiproliferative response to PD 0332991 treatment as a result of G1 phase arrest after inhibition of Cdk4- and Cdk6-mediated Rb phosphorylation was observed in lymphoma, myeloma, sarcoma, breast, lung, and colon carcinoma. Though, treatment efficiency of Cdk4/ Cdk6 selective inhibitors like PD 0332991 is limited to tumors with wild-type *rb* gene status. Furthermore, PD 0332991 does not induce apoptosis and seems to be ineffective in the majority of quiescent G0/ G1 arrested leukemic cells with primarily defects in apoptosis pathway, as it was demonstrated for chronic lymphocytic

primary leukemia models (Albanese et al., 2010; Degrassi et al., 2010).

2010).

tumor xenografts.

*Cdk4/ Cdk6 selective inhibitors* 

leukemia (Wesierska-Gadek et al., 2011).

diminish therapeutic effects, due to their relevance for tumor suppression and apoptosis (Chandramouli et al., 2007; Iorns et al., 2008).

Among the first compounds described with Cdk specificity flavopiridol and R-roscovitine as derivatives of natural products were developed. All other Cdk inhibitors are mainly synthetic compounds.

**Flavopiridol**, an alkaloid derivative and member of the flavone group, displayed antiproliferative and cytotoxic effects on tumor cells at nanomolar concentrations (Sedlacek et al., 1996). This observation was associated with cell cycle arrest through inhibition of Cdk and induction of apoptosis in human hematopoietic cell lines, breast, lung, as well as head and neck squamous cell carcinoma (Kaur et al., 1992; Konig et al., 1997; Parker et al., 1998; Patel et al., 1998). In spite of inconsistent data of IC50 and Kd values for different Cdk, flavopiridol showed high affinity to all Cdk with IC50 values below 400 nM and exhibit higher selectivity to Cdk9 (IC50 < 10 nM) (Chao et al., 2000; Sedlacek, 2001; Sedlacek, 1996). Cdk9, as well as Cdk7 inhibition lead to profound influence on cellular transcription, e. g., on mRNA transcripts for cell cycle regulators Cyclin D, antiapoptotic proteins Bcl-2 and Mcl-2, and NFB as well as p53 pathway (Lam et al., 2001; Lu et al., 2004). Independent from good antitumorigenic effects in preclinical studies, in 2008 only low specificity of flavopiridol for Cdk has been demonstrated due to nanomolar affinity to also 25 other protein kinases, like GSK3 and ERK (Karaman et al., 2008), maybe leading to discouraging results in some clinical trials (see next section).

Screening of about hundred compounds structurally related to flavopiridol identified **P276-00** as potent Cdk specific inhibitor with moderate selectivity for Cdk1, Cdk4, and Cdk9 (Joshi et al., 2007a). Similar to flavopiridol, P276-00 showed antiproliferative and proapoptotic activity in human breast, colon, lung, prostate carcinoma, and promyelocytic leukemia cell lines *in vitro* (Joshi et al., 2007b). Decreased Rb phosphorylation, G1/ G2 phase arrest, and caspase-dependent apoptosis could be observed in preclinical studies with multiple myeloma cells *in vitro* and *in vivo* (Manohar et al., 2011; Raje et al., 2009).

The purine derivative **R-roscovitine** inhibits Cdk1, Cdk2, Cdk5, Cdk7, and Cdk9 with selectivity to Cdk4 and Cdk6 (Meijer et al., 1997). In consequence, R-roscovitine lead to arrest of tumor cells in almost all cell cycle phases and affected cell proliferation, as it was demonstrated for over 60 human tumor cell lines (e. g., melanoma, lung, breast, colon carcinoma, leukemia). *In vivo* activity in mice bearing colorectal carcinoma xenografts, but also in hematopoietic progenitors (Mcclue et al., 2002; Raynaud et al., 2005; Song et al., 2007) and enhancement of antitumor effects of radiation and doxorubicin in combination with Rroscovitine was reported (Appleyard et al., 2009; Maggiorella et al., 2003). Like flavopiridol and other Cdk inhibitors with broad spectrum, several additional effects have been described for R-roscovitine *in vitro*: interruption of transcriptional elongation, interference with survival-associated pathways (IB kinase inhibition), induction of p53 phosphorylation and apoptosis (Alvi et al., 2005; Dey et al., 2008; Hahntow et al., 2004).

A novel potent Cdk inhibitor with striking similarity to R-roscovitine regarding both chemical structure and cytotoxic properties is pyrazolo[1,5-*a*]pyrimidine derivative **SCH 727965** (Parry et al., 2010). SCH 727965 inhibits Cdk1, Cdk2, Cdk5 and Cdk9 activity *in vitro* in low nanomolar concentrations (IC50 < 5 nM) and exhibited antiproliferative effects due to complete suppression of Rb phosphorylation and apoptosis induction, respectively.

Fragment-based screening techniques identified pyrazole-3-carboxamide **AT7519** as potent Cdk2 inhibitor and antiproliferative agent (Wyatt et al., 2008). AT7519 caused cell cycle arrest, growth inhibition, and apoptosis in a large spectrum of human solid tumor cells and colon carcinoma xenograft models (Squires et al., 2009). Influence on transcriptional regulating Cdk, as well as GSK3 resulting in induction of apoptotic pathways was observed in multiple myeloma and leukemia cell lines (Santo et al., 2010; Squires et al., 2010).

Optimization of selectivity and physicochemical properties of pyrazolo[4,3-*h*]quinazoline-3 carboxamide derivatives identified **PHA-848125** as potent Cdk inhibitor (IC50 < 400 nM) with inhibitory effects on cell cycle progression and Rb phosphorylation *in vitro* in a wide range of tumor cell lines (Brasca et al., 2009; Caporali et al., 2010). Among a panel of 43 other serine-threonine and tyrosine kinases only tropomyosin receptor kinase A (TRKA), linked to cancer cell survival, was inhibited by PHA-848125 in the same nanomolar range (IC50 53 nM). *In vivo* characterization of oral active compound PHA-848125 showed a dosedependent inhibition of human A2780 ovarian carcinoma xenograft tumor growth on mice up to 91%. Significant tumor growth inhibition was also observed in a K-Ras mutant lung adenocarcinoma transgenic mouse model, carcinogen-induced tumors, and in disseminated primary leukemia models (Albanese et al., 2010; Degrassi et al., 2010).

Only a few details have been published for pan-Cdk inhibitor **BAY 1000394** with unknown structure targeting Cdk1, Cdk2, Cdk4 and Cdk9 with high affinity (IC50 < 10 nM) (Siemeister et al., 2010). Inhibition of cell proliferation was amongst others depicted for breast, cervical, and colorectal human tumor cell lines and described to be independent of Rb, p53 and tumor suppressor gene status. Suppression of Rb phosphorylation and growth inhibition after treatment with BAY 1000394 was observed in preclinical studies in a broad range of tumor xenografts.

#### *Cdk4/ Cdk6 selective inhibitors*

270 Advances in Cancer Therapy

diminish therapeutic effects, due to their relevance for tumor suppression and apoptosis

Among the first compounds described with Cdk specificity flavopiridol and R-roscovitine as derivatives of natural products were developed. All other Cdk inhibitors are mainly

**Flavopiridol**, an alkaloid derivative and member of the flavone group, displayed antiproliferative and cytotoxic effects on tumor cells at nanomolar concentrations (Sedlacek et al., 1996). This observation was associated with cell cycle arrest through inhibition of Cdk and induction of apoptosis in human hematopoietic cell lines, breast, lung, as well as head and neck squamous cell carcinoma (Kaur et al., 1992; Konig et al., 1997; Parker et al., 1998; Patel et al., 1998). In spite of inconsistent data of IC50 and Kd values for different Cdk, flavopiridol showed high affinity to all Cdk with IC50 values below 400 nM and exhibit higher selectivity to Cdk9 (IC50 < 10 nM) (Chao et al., 2000; Sedlacek, 2001; Sedlacek, 1996). Cdk9, as well as Cdk7 inhibition lead to profound influence on cellular transcription, e. g., on mRNA transcripts for cell cycle regulators Cyclin D, antiapoptotic proteins Bcl-2 and Mcl-2, and NFB as well as p53 pathway (Lam et al., 2001; Lu et al., 2004). Independent from good antitumorigenic effects in preclinical studies, in 2008 only low specificity of flavopiridol for Cdk has been demonstrated due to nanomolar affinity to also 25 other protein kinases, like GSK3 and ERK (Karaman et al., 2008), maybe leading to discouraging

Screening of about hundred compounds structurally related to flavopiridol identified **P276-00** as potent Cdk specific inhibitor with moderate selectivity for Cdk1, Cdk4, and Cdk9 (Joshi et al., 2007a). Similar to flavopiridol, P276-00 showed antiproliferative and proapoptotic activity in human breast, colon, lung, prostate carcinoma, and promyelocytic leukemia cell lines *in vitro* (Joshi et al., 2007b). Decreased Rb phosphorylation, G1/ G2 phase arrest, and caspase-dependent apoptosis could be observed in preclinical studies with

The purine derivative **R-roscovitine** inhibits Cdk1, Cdk2, Cdk5, Cdk7, and Cdk9 with selectivity to Cdk4 and Cdk6 (Meijer et al., 1997). In consequence, R-roscovitine lead to arrest of tumor cells in almost all cell cycle phases and affected cell proliferation, as it was demonstrated for over 60 human tumor cell lines (e. g., melanoma, lung, breast, colon carcinoma, leukemia). *In vivo* activity in mice bearing colorectal carcinoma xenografts, but also in hematopoietic progenitors (Mcclue et al., 2002; Raynaud et al., 2005; Song et al., 2007) and enhancement of antitumor effects of radiation and doxorubicin in combination with Rroscovitine was reported (Appleyard et al., 2009; Maggiorella et al., 2003). Like flavopiridol and other Cdk inhibitors with broad spectrum, several additional effects have been described for R-roscovitine *in vitro*: interruption of transcriptional elongation, interference with survival-associated pathways (IB kinase inhibition), induction of p53 phosphorylation

A novel potent Cdk inhibitor with striking similarity to R-roscovitine regarding both chemical structure and cytotoxic properties is pyrazolo[1,5-*a*]pyrimidine derivative **SCH 727965** (Parry et al., 2010). SCH 727965 inhibits Cdk1, Cdk2, Cdk5 and Cdk9 activity *in vitro* in low nanomolar concentrations (IC50 < 5 nM) and exhibited antiproliferative effects due to complete suppression of Rb phosphorylation and apoptosis induction, respectively. Fragment-based screening techniques identified pyrazole-3-carboxamide **AT7519** as potent Cdk2 inhibitor and antiproliferative agent (Wyatt et al., 2008). AT7519 caused cell cycle

multiple myeloma cells *in vitro* and *in vivo* (Manohar et al., 2011; Raje et al., 2009).

and apoptosis (Alvi et al., 2005; Dey et al., 2008; Hahntow et al., 2004).

(Chandramouli et al., 2007; Iorns et al., 2008).

results in some clinical trials (see next section).

synthetic compounds.

Several compounds, particularly, members of chemical classes of benzothiadiazines, diarylureas, indolocarbazoles, oxindoles, pyrido[2,3-*d*]pyrimidines, thienopyrimidinhy drazones, thioacridones, and triaminopyrimidines were developed in the last decade and described with preferential selectivity to Cdk4 and Cdk6 (Graf, 2010; Lee & Sicinski, 2006). Cdk4 and Cdk6 reveal due to their high homology and identical substrate specificities comprehensive activities in the cell cycle. According to the occurrence in different tissues Cdk4 and Cdk6 compensate each other in function (Ciemerych & Sicinski, 2005; Meyerson & Harlow, 1994). **P1446A-05** and pyrido[2,3-*d*]pyrimidine **PD 0332991** seem to be the most promising compounds due to their inclusion in clinical evaluation. Unfortunately, no information about chemical structure and mechanism of Cdk4 selective inhibition is provided in the literature for P1446A-05. However, potency of PD 0332991 to inhibit Cdk4 and Cdk6 pathway and tumor cell progression has been extensively studied *in vitro*, as well as *in vivo* in mouse xenograft models (Baughn et al., 2006; Fry et al., 2004; Menu et al., 2008; Saab et al., 2006). In all studies an antiproliferative response to PD 0332991 treatment as a result of G1 phase arrest after inhibition of Cdk4- and Cdk6-mediated Rb phosphorylation was observed in lymphoma, myeloma, sarcoma, breast, lung, and colon carcinoma. Though, treatment efficiency of Cdk4/ Cdk6 selective inhibitors like PD 0332991 is limited to tumors with wild-type *rb* gene status. Furthermore, PD 0332991 does not induce apoptosis and seems to be ineffective in the majority of quiescent G0/ G1 arrested leukemic cells with primarily defects in apoptosis pathway, as it was demonstrated for chronic lymphocytic leukemia (Wesierska-Gadek et al., 2011).

Cyclin-Dependent Kinases (Cdk) as Targets for Cancer Therapy and Imaging 273

P276-00 administered along with radiation in head and neck squamous cell carcinoma patients, and in combination with gemcitabine in patients with advanced pancreatic cancer. A phase I clinical trial of **R-roscovitine** has been performed in 22 patients with advanced cancer utilizing several schedules resulting in only minor therapeutical success with hypokalemia, rash and fatigue as dose limiting side-effects (Benson et al., 2007). Further evaluation of R-roscovitine in patients with non-small cell lung cancer was described: Rroscovitine in combination with cisplatin and gemcitabine in phase I showed similar adverse events, e. g., hypokalemia, liver -glutamyltranspeptidase elevation, vomiting, and a 42.9% response rate in 14 patients (Siegel-Lakhai et al., 2005). A phase II study of oral Rroscovitine to treat non-small cell lung cancer has been terminated without any reports on available data. Antitumor activity and pharmacodynamic effects of R-roscovitine sequentially administered with sapacitabine was initiated in 2009 in patients with advanced

Safety and tolerability of **SCH 727965** was demonstrated in phase I clinical trials for multiple malignant indications including solid tumors, non-Hodgkin's lymphoma, multiple myeloma, and chronic lymphocytic leukemia (Shapiro et al., 2008). The drug was administered once every 21 days as a 2 hour intravenous infusion and dose limiting toxicity was neutropenia. To improve treatment success a phase I study currently recruiting patients is initiated in patients with advanced cancer for schedule adjustment (3 infusions ever 7 days in a 28 day cycle). First results in previously treated patients with chronic lymphocytic leukemia showed common toxicities, like fatigue, nausea, and diarrhea, but also evidence of therapeutic benefit of SCH 727965 (Flynn et al., 2009). In an active phase II clinical trial activity of SCH 727965 in patients with advanced breast cancer, non-small cell lung cancer, mantle cell lymphoma, or B-cell chronic lymphocytic leukemia each in comparison to standard treatment (capecitabine, erlotinib, bortezomib, or alemtuzumab) is determined. Furthermore, phase I/II studies including patients with malignant melanoma and multiple

Two phase I/II studies with **AT7519**, currently recruiting patients are under evaluation. Single AT7519 treatment of patients with advanced/metastatic solid tumors or refractory non-Hodgkin's lymphoma will provide data to applicable dose, pharmacokinetics, pharmacodynamics and side effects. First results of this study suggested evidence of clinical activity of AT7519, due to reduction of PCNA, a protein required for DNA replication, as a result of Cdk inhibition and an increase of apoptosis markers (Mahadevan et al., 2011). Mucositis, neutropenia, and reversible thrombocytopenia were identified as dose limiting toxicities. In another ongoing clinical trial, efficacy of either AT7519 alone or AT7519 in combination with proteasome inhibitor bortezomib is determined in patients with

**PHA-848125** was first evaluated in a multi center phase I study to investigate the safety and pharmacokinetics in patients with advanced/ metastatic solid tumors. Oral administration of PHA-848125 in different treatment schedules showed good tolerability with ataxia and tremors as dose-limiting side-effects and occasional hematological toxicities (Benouaich-Amiel et al., 2010; Cresta et al., 2010; Tibes et al., 2008). Similar results were found in a phase I study of PHA-848125 in combination with gemcitabine (Bahleda et al., 2010). Antitumor activity of PHA-848125 is currently assessed in phase II clinical trials in patients with recurrent or metastatic thymic carcinoma previously treated with chemotherapeutic agents.

myeloma, as well as plasma cell neoplasms will be performed.

previously treated multiple myeloma.

solid tumors.

### **4. Cdk inhibitors for cancer therapy**

Despite of promising preclinical data and numerous potent Cdk inhibitors in clinical trials, none of these molecules has already been approved as drug for cancer therapy. Nevertheless, evaluation of various Cdk inhibitors contributed to important information about bioavailability, pharmacokinetics, as well as, competing toxicities and lead to alterations in dosing schedule to reach a maximum in response. To date, monotherapy did not demonstrate convincing utility of a certain Cdk inhibitor for cancer treatment, although clinical trials of several hematologic malignancies have shown stable disease. Encouraging results depicting antitumor activity of Cdk inhibitors also in solid tumors have been obtained in various combination studies with cytostatic or cytotoxic agents. However, beneficial effects of Cdk inhibition alone and in combination with other chemotherapeutic agents to tumor treatment must be demonstrated in randomized clinical trials in the future. Information about current clinical trials of Cdk inhibitors were achieved from the webpage

*www.clinicaltrials.gov* and a summary is given in Table 2. **Flavopiridol** was the first Cdk inhibitor entering clinical trials. Since the early 1990s flavopiridol was extensively studied in patients with refractory neoplasms, a variety of solid tumors and hematopoietic malignancies. Clinical trials with single flavopiridol were completed for, e. g., advanced gastric cancer, endometrial adenocarcinoma, metastatic melanoma, multiple myeloma, and non-small cell lung cancer in phase II, but only unsatisfying results concerning antitumor activity and toxicities like myelosuppression, diarrhea as well as thromboses were observed (Burdette-Radoux et al., 2004; Dispenzieri et al., 2006; Grendys et al., 2005; Schwartz et al., 2001; Shapiro et al., 2001). Also, no response to flavopiridol administered as 24 hour continuous infusion in chronic lymphocytic leukemia patients was observed (Flinn et al., 2005). Detection of high serum protein binding of flavopiridol and collection of pharmacokinetic data contributed to the adjustment of treatment-schedule of flavopiridol. In consequence, hematopoietic malignancies showed encouraging responses to flavopiridol. Weekly application of flavopiridol as a single agent for 6 weeks in a phase I study of chronic lymphocytic leukemia patients achieved a median progression-free survival of 12 month (Byrd et al., 2007; Phelps et al., 2009). Addition of prophylactic corticosteroid dexamethasone to flavopiridol treatment improved tolerability in chronic lymphocytic leukemia patients (Lin et al., 2009). In combination with cytostatic drugs cytosine arabinoside and mitoxantrone in phase II clinical trials a 75% response rate of patients with acute myelogenous leukemia was observed (Karp et al., 2007). Combined treatment of flavopiridol with chemotherapeutic agents cisplatin, carboplatin, docetaxel, gemcitabine, irinotecan or paclitaxel, respectively, have shown response or at least stable disease in phase I trials (Bible et al., 2005; Fekrazad et al., 2010; Fornier et al., 2007; George et al., 2008). However, subsequent phase II clinical trial with flavopiridol in combination with docetaxel resulted again in disappointing activity and significant toxicity in patients with metastatic pancreatic adenocarcinoma (Carvajal et al., 2009). Currently ongoing and recruiting clinical studies focus on evaluation of flavopiridol (combined with chemotherapeutic agents like bortezomib, oxaliplatin, or respectively, lenalidomide) in patients with B-cell neoplasms, lymphoma, multiple myeloma, and germ cell tumors.

**P276-00** is currently in phase I/II clinical trials to determine anticancer activity in patients with advanced Cyclin D1 positive malignant melanoma, mantle cell lymphoma, squamous cell head and neck carcinoma, and multiple myeloma. A phase I study in patients with advanced refractory neoplasms has already been completed. No results have been provided in the literature yet. Combination studies currently recruiting patients are initiated for

Despite of promising preclinical data and numerous potent Cdk inhibitors in clinical trials, none of these molecules has already been approved as drug for cancer therapy. Nevertheless, evaluation of various Cdk inhibitors contributed to important information about bioavailability, pharmacokinetics, as well as, competing toxicities and lead to alterations in dosing schedule to reach a maximum in response. To date, monotherapy did not demonstrate convincing utility of a certain Cdk inhibitor for cancer treatment, although clinical trials of several hematologic malignancies have shown stable disease. Encouraging results depicting antitumor activity of Cdk inhibitors also in solid tumors have been obtained in various combination studies with cytostatic or cytotoxic agents. However, beneficial effects of Cdk inhibition alone and in combination with other chemotherapeutic agents to tumor treatment must be demonstrated in randomized clinical trials in the future. Information about current clinical trials of Cdk inhibitors were achieved from the webpage

**Flavopiridol** was the first Cdk inhibitor entering clinical trials. Since the early 1990s flavopiridol was extensively studied in patients with refractory neoplasms, a variety of solid tumors and hematopoietic malignancies. Clinical trials with single flavopiridol were completed for, e. g., advanced gastric cancer, endometrial adenocarcinoma, metastatic melanoma, multiple myeloma, and non-small cell lung cancer in phase II, but only unsatisfying results concerning antitumor activity and toxicities like myelosuppression, diarrhea as well as thromboses were observed (Burdette-Radoux et al., 2004; Dispenzieri et al., 2006; Grendys et al., 2005; Schwartz et al., 2001; Shapiro et al., 2001). Also, no response to flavopiridol administered as 24 hour continuous infusion in chronic lymphocytic leukemia patients was observed (Flinn et al., 2005). Detection of high serum protein binding of flavopiridol and collection of pharmacokinetic data contributed to the adjustment of treatment-schedule of flavopiridol. In consequence, hematopoietic malignancies showed encouraging responses to flavopiridol. Weekly application of flavopiridol as a single agent for 6 weeks in a phase I study of chronic lymphocytic leukemia patients achieved a median progression-free survival of 12 month (Byrd et al., 2007; Phelps et al., 2009). Addition of prophylactic corticosteroid dexamethasone to flavopiridol treatment improved tolerability in chronic lymphocytic leukemia patients (Lin et al., 2009). In combination with cytostatic drugs cytosine arabinoside and mitoxantrone in phase II clinical trials a 75% response rate of patients with acute myelogenous leukemia was observed (Karp et al., 2007). Combined treatment of flavopiridol with chemotherapeutic agents cisplatin, carboplatin, docetaxel, gemcitabine, irinotecan or paclitaxel, respectively, have shown response or at least stable disease in phase I trials (Bible et al., 2005; Fekrazad et al., 2010; Fornier et al., 2007; George et al., 2008). However, subsequent phase II clinical trial with flavopiridol in combination with docetaxel resulted again in disappointing activity and significant toxicity in patients with metastatic pancreatic adenocarcinoma (Carvajal et al., 2009). Currently ongoing and recruiting clinical studies focus on evaluation of flavopiridol (combined with chemotherapeutic agents like bortezomib, oxaliplatin, or respectively, lenalidomide) in patients with B-cell neoplasms, lymphoma, multiple myeloma, and germ cell tumors. **P276-00** is currently in phase I/II clinical trials to determine anticancer activity in patients with advanced Cyclin D1 positive malignant melanoma, mantle cell lymphoma, squamous cell head and neck carcinoma, and multiple myeloma. A phase I study in patients with advanced refractory neoplasms has already been completed. No results have been provided in the literature yet. Combination studies currently recruiting patients are initiated for

**4. Cdk inhibitors for cancer therapy** 

*www.clinicaltrials.gov* and a summary is given in Table 2.

P276-00 administered along with radiation in head and neck squamous cell carcinoma patients, and in combination with gemcitabine in patients with advanced pancreatic cancer. A phase I clinical trial of **R-roscovitine** has been performed in 22 patients with advanced cancer utilizing several schedules resulting in only minor therapeutical success with hypokalemia, rash and fatigue as dose limiting side-effects (Benson et al., 2007). Further evaluation of R-roscovitine in patients with non-small cell lung cancer was described: Rroscovitine in combination with cisplatin and gemcitabine in phase I showed similar adverse events, e. g., hypokalemia, liver -glutamyltranspeptidase elevation, vomiting, and a 42.9% response rate in 14 patients (Siegel-Lakhai et al., 2005). A phase II study of oral Rroscovitine to treat non-small cell lung cancer has been terminated without any reports on available data. Antitumor activity and pharmacodynamic effects of R-roscovitine sequentially administered with sapacitabine was initiated in 2009 in patients with advanced solid tumors.

Safety and tolerability of **SCH 727965** was demonstrated in phase I clinical trials for multiple malignant indications including solid tumors, non-Hodgkin's lymphoma, multiple myeloma, and chronic lymphocytic leukemia (Shapiro et al., 2008). The drug was administered once every 21 days as a 2 hour intravenous infusion and dose limiting toxicity was neutropenia. To improve treatment success a phase I study currently recruiting patients is initiated in patients with advanced cancer for schedule adjustment (3 infusions ever 7 days in a 28 day cycle). First results in previously treated patients with chronic lymphocytic leukemia showed common toxicities, like fatigue, nausea, and diarrhea, but also evidence of therapeutic benefit of SCH 727965 (Flynn et al., 2009). In an active phase II clinical trial activity of SCH 727965 in patients with advanced breast cancer, non-small cell lung cancer, mantle cell lymphoma, or B-cell chronic lymphocytic leukemia each in comparison to standard treatment (capecitabine, erlotinib, bortezomib, or alemtuzumab) is determined. Furthermore, phase I/II studies including patients with malignant melanoma and multiple myeloma, as well as plasma cell neoplasms will be performed.

Two phase I/II studies with **AT7519**, currently recruiting patients are under evaluation. Single AT7519 treatment of patients with advanced/metastatic solid tumors or refractory non-Hodgkin's lymphoma will provide data to applicable dose, pharmacokinetics, pharmacodynamics and side effects. First results of this study suggested evidence of clinical activity of AT7519, due to reduction of PCNA, a protein required for DNA replication, as a result of Cdk inhibition and an increase of apoptosis markers (Mahadevan et al., 2011). Mucositis, neutropenia, and reversible thrombocytopenia were identified as dose limiting toxicities. In another ongoing clinical trial, efficacy of either AT7519 alone or AT7519 in combination with proteasome inhibitor bortezomib is determined in patients with previously treated multiple myeloma.

**PHA-848125** was first evaluated in a multi center phase I study to investigate the safety and pharmacokinetics in patients with advanced/ metastatic solid tumors. Oral administration of PHA-848125 in different treatment schedules showed good tolerability with ataxia and tremors as dose-limiting side-effects and occasional hematological toxicities (Benouaich-Amiel et al., 2010; Cresta et al., 2010; Tibes et al., 2008). Similar results were found in a phase I study of PHA-848125 in combination with gemcitabine (Bahleda et al., 2010). Antitumor activity of PHA-848125 is currently assessed in phase II clinical trials in patients with recurrent or metastatic thymic carcinoma previously treated with chemotherapeutic agents.

Cyclin-Dependent Kinases (Cdk) as Targets for Cancer Therapy and Imaging 275


combination with bortezomib (phase I)



combination with gemcitabine) (phase I, II)





(phase II)

(phase I, II)

(phase I)











with sapacitabine) (phase I)

leukemia (phase I)

Table 2. Cdk inhibitors currently under clinical evaluation. Information to clinical trials were

obtained from *www.clinical trials.gov.* (GSK3: glycogen synthase kinase 3; TRKA:

agents) (phase I, II)

inhibitor targets (IC50) clinical trials in patients with

AT7519 Cdk2, Cdk5, Cdk9 (< 50 nM)

GSK3β (89 nM) (Squires, 2009)

flavopiridol (alvocidib)

P1446A-05 Cdk4 (n. a.)

BAY 1000394 Cdk1, Cdk2, Cdk4, Cdk9 (≤ 11 nM) (Siemeister, 2010)

> GSK3β (450 nM) (Sedlacek, 2001)

P276-00 Cdk1, Cdk4, Cdk9 (< 100 nM)

(Joshi, 2007a)

kinases (> 10 µM) (Fry, 2004)

PD 0332991 Cdk4, Cdk6 (≤ 15 nM)

PHA-848125 Cdk2 (45/363 nM)

nM)

R-roscovitine (seliciclib) (CYC202)

SCH 727965 (dinaciclib)

TRKA (53 nM) (Brasca, 2009)

(≤ 800 nM)

(≥ 14 µM)

(Parry, 2010)

tropomyosin receptor kinase A)

Cdk4, Cdk9 (< 50 nM)

Cdk2, Cdk6 (≤ 400 nM) Cdk7, GSK3β (2.8 µM)

Cdk1, Cdk4, Cdk6 (≤ 210 nM)

Cdk1, Cdk2, Cdk6, Cdk7 (≤ 400 nM)

(www.piramallifesciences.com)

other protein kinases (> 10 µM)

Cdk1, Cdk2, Cdk5, other protein

Cdk1, Cdk4, Cdk5, Cdk7 (< 400

Cdk1, Cdk2, Cdk5, Cdk7, Cdk9

Cdk4, Cdk6, other protein kinases

(Meijer, 1997; Popowycz et al., 2009)

Cdk1, Cdk2, Cdk5, Cdk9 (< 5 nM)

Clinical studies of pan-Cdk inhibitor **BAY 1000394** and Cdk4 selective inhibitor **P1446A-05** are in the beginning and thus less well documented. Both compounds are tested in first clinical trials at the moment to identify tolerated dose of the drug and to determine both limiting side effects and efficacy in patients with advanced solid tumors. Evaluation of P1446A-05 includes also hematologic malignancies.

**PD 0332991** was first tested in a phase I dose escalation trial in 57 patients with Rb positive advanced cancer (e. g., breast, colorectal cancer, melanoma) (O'dwyer et al., 2007). Doses from 25 mg to 150 mg of single agent PD 0332991 were administered orally for 21 days in a 28 days cycle. Stable disease was manifested for 9 patients and dose limiting toxicity was myelosuppression. Another phase I study assessing mechanism of action, safety, and pharmacodynamic effects of PD 0332991 was performed in patients with mantle cell lymphoma using fluorine-18-labeled fluorodeoxyglucose ([18F]FDG) and fluorothymidine ([18F]FLT) positron emission tomography imaging (Leonard et al., 2008). Numerous phase II clinical trials of PD 0332991 are ongoing and currently recruiting patients: with advanced non-small cell lung cancer, with refractory solid tumors, with recurrent Rb positive glioblastoma, and with metastatic liposarcoma. Evaluation of PD 0332991 in combination with letrozole for treatment of hormone-receptor positive advanced breast cancer, in combination with bortezomib for mantle cell lymphoma, as well as in combination with velcade and dexamethasone in patients with multiple myeloma is ongoing after encouraging antitumor activity was revealed in phase I clinical trials (Niesvizky et al., 2009; Niesvizky et al., 2010; Slamon et al., 2010).

### **5. Specific aspects regarding kinase inhibitor resistance**

An important problem for new approaches in cancer therapy, also resulting in significant limitations of the potential use of kinase inhibitors, is the resistance of tumor cells to cytotoxic anticancer drugs acquired during therapy. As a relevant example, resistance to kinase inhibitor imatinib selective for Bcl-Abl was observed in chronic myelogeneous leukemia and gastrointestinal stromal tumor patients (Bixby & Talpaz, 2011; Wang et al., 2011). Development of kinase inhibitor resistance is accomplished by increased expression and even more common by specific point mutations of Bcl-Abl oncogene. In consequence, association of inhibitor with kinase is prevented without rigorous effects on enzyme activity. Despite of no reported appearance of Cdk inhibitor resistance in clinical trials so far, essential side chains for Cdk inhibitor specificity and selectivity, but not for ATP binding were predicted. An aromatic amino acid in the conserved Cdk domain, e. g., phenylalanine-80 in Cdk2, provides a hydrophobic site for essential *van der Waals* contacts with the Cdk inhibitor (for example isopropyl of R-roscovitine and acetyl of PD 0332991) (Krystof & Uldrijan, 2010). In parallel to mutation-based steric hindrance of a aurora B inhibitor causing kinase resistance (Girdler et al., 2008), histidine residues in Cdk4 and Cdk6 have been predicted to contribute to selectivity of PD 0332991 and discriminate Cdk1 and Cdk2 (Lu & Schulze-Gahmen, 2006). Point mutations modifying these side chains could directly result in the loss of Cdk inhibitor binding. However, findings of sufficiency of Cdk1 to drive the mammalian cell cycle in mouse embryogenesis raise the questions if Cdk1 could substitute all the other interphase Cdks in human tumor cells anyway and whether a point mutation is likely to cause resistance to a pharmacological Cdk inhibitor (Santamaria, 2007).

Clinical studies of pan-Cdk inhibitor **BAY 1000394** and Cdk4 selective inhibitor **P1446A-05** are in the beginning and thus less well documented. Both compounds are tested in first clinical trials at the moment to identify tolerated dose of the drug and to determine both limiting side effects and efficacy in patients with advanced solid tumors. Evaluation of

**PD 0332991** was first tested in a phase I dose escalation trial in 57 patients with Rb positive advanced cancer (e. g., breast, colorectal cancer, melanoma) (O'dwyer et al., 2007). Doses from 25 mg to 150 mg of single agent PD 0332991 were administered orally for 21 days in a 28 days cycle. Stable disease was manifested for 9 patients and dose limiting toxicity was myelosuppression. Another phase I study assessing mechanism of action, safety, and pharmacodynamic effects of PD 0332991 was performed in patients with mantle cell lymphoma using fluorine-18-labeled fluorodeoxyglucose ([18F]FDG) and fluorothymidine ([18F]FLT) positron emission tomography imaging (Leonard et al., 2008). Numerous phase II clinical trials of PD 0332991 are ongoing and currently recruiting patients: with advanced non-small cell lung cancer, with refractory solid tumors, with recurrent Rb positive glioblastoma, and with metastatic liposarcoma. Evaluation of PD 0332991 in combination with letrozole for treatment of hormone-receptor positive advanced breast cancer, in combination with bortezomib for mantle cell lymphoma, as well as in combination with velcade and dexamethasone in patients with multiple myeloma is ongoing after encouraging antitumor activity was revealed in phase I clinical trials (Niesvizky et al., 2009;

An important problem for new approaches in cancer therapy, also resulting in significant limitations of the potential use of kinase inhibitors, is the resistance of tumor cells to cytotoxic anticancer drugs acquired during therapy. As a relevant example, resistance to kinase inhibitor imatinib selective for Bcl-Abl was observed in chronic myelogeneous leukemia and gastrointestinal stromal tumor patients (Bixby & Talpaz, 2011; Wang et al., 2011). Development of kinase inhibitor resistance is accomplished by increased expression and even more common by specific point mutations of Bcl-Abl oncogene. In consequence, association of inhibitor with kinase is prevented without rigorous effects on enzyme activity. Despite of no reported appearance of Cdk inhibitor resistance in clinical trials so far, essential side chains for Cdk inhibitor specificity and selectivity, but not for ATP binding were predicted. An aromatic amino acid in the conserved Cdk domain, e. g., phenylalanine-80 in Cdk2, provides a hydrophobic site for essential *van der Waals* contacts with the Cdk inhibitor (for example isopropyl of R-roscovitine and acetyl of PD 0332991) (Krystof & Uldrijan, 2010). In parallel to mutation-based steric hindrance of a aurora B inhibitor causing kinase resistance (Girdler et al., 2008), histidine residues in Cdk4 and Cdk6 have been predicted to contribute to selectivity of PD 0332991 and discriminate Cdk1 and Cdk2 (Lu & Schulze-Gahmen, 2006). Point mutations modifying these side chains could directly result in the loss of Cdk inhibitor binding. However, findings of sufficiency of Cdk1 to drive the mammalian cell cycle in mouse embryogenesis raise the questions if Cdk1 could substitute all the other interphase Cdks in human tumor cells anyway and whether a point mutation is

likely to cause resistance to a pharmacological Cdk inhibitor (Santamaria, 2007).

P1446A-05 includes also hematologic malignancies.

Niesvizky et al., 2010; Slamon et al., 2010).

**5. Specific aspects regarding kinase inhibitor resistance** 


Table 2. Cdk inhibitors currently under clinical evaluation. Information to clinical trials were obtained from *www.clinical trials.gov.* (GSK3: glycogen synthase kinase 3; TRKA: tropomyosin receptor kinase A)

Cyclin-Dependent Kinases (Cdk) as Targets for Cancer Therapy and Imaging 277

functional, non-invasive characterization and imaging of Cdk in human tumors. Particularly, radiolabeled Cdk4/Cdk6 inhibitors are of interest for the assessment of Cdk4/Cdk6 protein status and activity in human tumors in current translational cancer research. PET affords the opportunity of three-dimensional imaging and quantitation of physiological processes *in vivo*. Additionally, PET provides pharmacological data of radiolabeled Cdk4/Cdk6 inhibitors, which may help to understand mechanism of action *in* 

In this regard, new Cdk4/Cdk6 inhibitors derived from pyrido[2,3-*d*]pyrimidine lead structure (Vanderwel et al., 2005) were designed, synthesized and characterized in our institute for the first time. Evaluation of five compounds – CKIA, CKIB, CKIC, CKID, and CKIE – showed specific and selective inhibition of Cdk4/ Cdk6-mediated pathway, induction of G1 phase arrest and blocking of tumor cell proliferation in pharmacological concentrations (Graf et al., 2009; Graf, 2010). Most potent Cdk4/ Cdk6 inhibitors CKIA, CKIB, and CKIE were radiolabeled with positron-emitters iodine-124 [124I] or fluorine-18 [18F], respectively, and characterized concerning their radiopharmacological properties in cellular radiotracer uptake studies, biodistribution and, small animal PET studies (Graf, 2009; Graf, 2010; Koehler et al., 2010). Iodine-124 with a half-life of 4.18 days allows extended radiopharmacological evaluation. Nevertheless, minor positron emission (26%) and high positron energy are disadvantages leading to low resolution PET images. Most frequently used PET nuclide

*vivo* and estimate the applicability of the compounds for tumor therapy.

fluorine-18 exhibits nearly 97% positron emission with a half-life of 109.8 minutes.

of [18F]CKIE in human tumor xenografts in mice.

 li, liver, i, intestine, b, bladder, tu, human squamous cell carcinoma.

Fig. 3. PET studies with fluorine-18 radiolabeled pyrido[2,3-*d*]pyrimidine derivative CKIE

*In vitro* radiotracer uptake studies using [124I]CKIA, [124I]CKIB, and [18F]CKIE demonstrated substantial tumor cell uptake, an important prerequisite for PET studies, in NMRI nu/nu mice bearing the human squamous cell carcinoma tumor FaDu. Dynamic small animal PET studies demonstrated rapid clearance of [124I]CKIA, [124I]CKIB, and [18F]CKIE from the blood and fast hepatobiliary excretion. The half-life of radiotracer elimination from the blood was calculated to between 7.2 and 7.9 min. Only marginal tumor uptake of radiotracers [124I]CKIA and [124I]CKIB was observed. In the case of [18F]CKIE higher uptake was detected in the peripheral cell-cycle active region of the tumor one hour after intravenous injection (Fig. 3). However, the constant tumor-to-muscle ratio of 1.5 suggests a non-Cdk4- or non-Cdk6-mediated association

Referring to this, a critical view on the advantages and disadvantages of Cdk inhibitor selectivity for the success of therapeutic approaches should be given. Is development of exclusively selective Cdk inhibitors challenging tumor therapy?

First of all, due to high homology in Cdk active sites development of monospecific Cdk inhibitor is nearly impossible. This prediction becomes apparent by the given IC50 values in the list of potent Cdk inhibitors in clinical trials (Table 2). Only PD 0332991 exhibits high selectivity to Cdk4 and Cdk6 *in vitro*. But at certain pharmacological concentrations, there is no exclusion of influence on various members of Cdk family.

Cdk inhibitors with improved selectivity promise minimization of undesired side effects, since no effects on global transcriptional machinery (Cdk7, Cdk9) would be expected in healthy cells as it was described for pan-Cdk inhibitors. Certainly, selective inactivation of any Cdk would not be as efficient as inhibition of multiple Cdk, including Cdk1, and involves the risk of easy development of Cdk inhibitor resistance, especially by alterations in the primarily targeted pathway. The response on Cdk4 and Cdk6 selective inhibitor PD 0332991 is significantly linked to functional Rb, transcriptional repression through E2F and the ability to attenuate Cdk2 activity, otherwise resistance to PD 0332991 acquired in tumor cell lines (Dean et al., 2010). In contrast, high efficiency of treatment in Cdk4/ Cdk6 dependent tumors and even reversal of resistance to estrogen receptor pathway targeting with tamoxifen by combined therapy with PD 0332991 could be achieved (Finn et al., 2009).

Pan-Cdk inhibitors affect multiple pathways in tumor cells and attack heterogeneous malignant cells (e. g., chronic lymphocytic leukemia, multiple myeloma), hindering Cdk inhibitor resistance. But, one has to consider the consequences of pan-Cdk inhibitors on healthy proliferating cells, like hematopoietic and dermal cells *in vivo*, which are much more sensitive to blockage of global cell cycle machinery and regulators.

To emulate and characterize Cdk inhibitor resistance *in vitro*, ambitions to generate cellular models were described. An increased activity due to increased protein synthesis of Cyclin E and simple overexpression of Cdk1 was observed in ovarian and colorectal carcinoma cell lines with acquired resistance to flavopiridol (Bible et al., 2000; Smith et al., 2001). But overexpression of Cyclin E in the cells is not mandatory associated with decreased sensitivity to flavopiridol (Smith, 2001). Also no evidence to mutational modification of Cdk was given in the studies. In fact, resistance to flavopiridol and another Cdk inhibitor was related to increased expression or activity of ATP-binding cassette transporter ABCG2 minimizing accumulation of the inhibitors and responsible for multidrug resistance in certain cancer cells (Robey et al., 2001; Seamon et al., 2006). Thus, mechanisms of resistance to Cdk inhibitors described to date seem to be inconsistent in dependence of the genetic status of different cell lines. Decreased sensitivity to a certain drug is often a result of multiple alterations in the tumor cells and further studies will elucidate the appearance of Cdk-associated genetic changes as a consequence of long-term treatment with Cdk inhibitor. A hopeful therapeutic strategy to overcome development of inhibitor resistance is the combination of Cdk inhibitors with common chemotherapeutic agents for fast and effective elimination of malignant cells. Recent clinical studies with combined therapeutic regimes will clarify the benefit for the patients.

#### **6. Cdk inhibitors as molecular probes for cancer imaging**

The development of potent radiolabeled Cdk inhibitors, as radiotracers for tumor visualization using positron emission tomography (PET) is a novel approach allowing

Referring to this, a critical view on the advantages and disadvantages of Cdk inhibitor selectivity for the success of therapeutic approaches should be given. Is development of

First of all, due to high homology in Cdk active sites development of monospecific Cdk inhibitor is nearly impossible. This prediction becomes apparent by the given IC50 values in the list of potent Cdk inhibitors in clinical trials (Table 2). Only PD 0332991 exhibits high selectivity to Cdk4 and Cdk6 *in vitro*. But at certain pharmacological concentrations, there is

Cdk inhibitors with improved selectivity promise minimization of undesired side effects, since no effects on global transcriptional machinery (Cdk7, Cdk9) would be expected in healthy cells as it was described for pan-Cdk inhibitors. Certainly, selective inactivation of any Cdk would not be as efficient as inhibition of multiple Cdk, including Cdk1, and involves the risk of easy development of Cdk inhibitor resistance, especially by alterations in the primarily targeted pathway. The response on Cdk4 and Cdk6 selective inhibitor PD 0332991 is significantly linked to functional Rb, transcriptional repression through E2F and the ability to attenuate Cdk2 activity, otherwise resistance to PD 0332991 acquired in tumor cell lines (Dean et al., 2010). In contrast, high efficiency of treatment in Cdk4/ Cdk6 dependent tumors and even reversal of resistance to estrogen receptor pathway targeting with tamoxifen by combined therapy with PD 0332991 could be achieved (Finn et al., 2009). Pan-Cdk inhibitors affect multiple pathways in tumor cells and attack heterogeneous malignant cells (e. g., chronic lymphocytic leukemia, multiple myeloma), hindering Cdk inhibitor resistance. But, one has to consider the consequences of pan-Cdk inhibitors on healthy proliferating cells, like hematopoietic and dermal cells *in vivo*, which are much more

To emulate and characterize Cdk inhibitor resistance *in vitro*, ambitions to generate cellular models were described. An increased activity due to increased protein synthesis of Cyclin E and simple overexpression of Cdk1 was observed in ovarian and colorectal carcinoma cell lines with acquired resistance to flavopiridol (Bible et al., 2000; Smith et al., 2001). But overexpression of Cyclin E in the cells is not mandatory associated with decreased sensitivity to flavopiridol (Smith, 2001). Also no evidence to mutational modification of Cdk was given in the studies. In fact, resistance to flavopiridol and another Cdk inhibitor was related to increased expression or activity of ATP-binding cassette transporter ABCG2 minimizing accumulation of the inhibitors and responsible for multidrug resistance in certain cancer cells (Robey et al., 2001; Seamon et al., 2006). Thus, mechanisms of resistance to Cdk inhibitors described to date seem to be inconsistent in dependence of the genetic status of different cell lines. Decreased sensitivity to a certain drug is often a result of multiple alterations in the tumor cells and further studies will elucidate the appearance of Cdk-associated genetic changes as a consequence of long-term treatment with Cdk inhibitor. A hopeful therapeutic strategy to overcome development of inhibitor resistance is the combination of Cdk inhibitors with common chemotherapeutic agents for fast and effective elimination of malignant cells. Recent clinical studies with combined therapeutic regimes

The development of potent radiolabeled Cdk inhibitors, as radiotracers for tumor visualization using positron emission tomography (PET) is a novel approach allowing

exclusively selective Cdk inhibitors challenging tumor therapy?

no exclusion of influence on various members of Cdk family.

sensitive to blockage of global cell cycle machinery and regulators.

**6. Cdk inhibitors as molecular probes for cancer imaging** 

will clarify the benefit for the patients.

functional, non-invasive characterization and imaging of Cdk in human tumors. Particularly, radiolabeled Cdk4/Cdk6 inhibitors are of interest for the assessment of Cdk4/Cdk6 protein status and activity in human tumors in current translational cancer research. PET affords the opportunity of three-dimensional imaging and quantitation of physiological processes *in vivo*. Additionally, PET provides pharmacological data of radiolabeled Cdk4/Cdk6 inhibitors, which may help to understand mechanism of action *in vivo* and estimate the applicability of the compounds for tumor therapy.

In this regard, new Cdk4/Cdk6 inhibitors derived from pyrido[2,3-*d*]pyrimidine lead structure (Vanderwel et al., 2005) were designed, synthesized and characterized in our institute for the first time. Evaluation of five compounds – CKIA, CKIB, CKIC, CKID, and CKIE – showed specific and selective inhibition of Cdk4/ Cdk6-mediated pathway, induction of G1 phase arrest and blocking of tumor cell proliferation in pharmacological concentrations (Graf et al., 2009; Graf, 2010). Most potent Cdk4/ Cdk6 inhibitors CKIA, CKIB, and CKIE were radiolabeled with positron-emitters iodine-124 [124I] or fluorine-18 [18F], respectively, and characterized concerning their radiopharmacological properties in cellular radiotracer uptake studies, biodistribution and, small animal PET studies (Graf, 2009; Graf, 2010; Koehler et al., 2010). Iodine-124 with a half-life of 4.18 days allows extended radiopharmacological evaluation. Nevertheless, minor positron emission (26%) and high positron energy are disadvantages leading to low resolution PET images. Most frequently used PET nuclide fluorine-18 exhibits nearly 97% positron emission with a half-life of 109.8 minutes.

*In vitro* radiotracer uptake studies using [124I]CKIA, [124I]CKIB, and [18F]CKIE demonstrated substantial tumor cell uptake, an important prerequisite for PET studies, in NMRI nu/nu mice bearing the human squamous cell carcinoma tumor FaDu. Dynamic small animal PET studies demonstrated rapid clearance of [124I]CKIA, [124I]CKIB, and [18F]CKIE from the blood and fast hepatobiliary excretion. The half-life of radiotracer elimination from the blood was calculated to between 7.2 and 7.9 min. Only marginal tumor uptake of radiotracers [124I]CKIA and [124I]CKIB was observed. In the case of [18F]CKIE higher uptake was detected in the peripheral cell-cycle active region of the tumor one hour after intravenous injection (Fig. 3). However, the constant tumor-to-muscle ratio of 1.5 suggests a non-Cdk4- or non-Cdk6-mediated association of [18F]CKIE in human tumor xenografts in mice.

li, liver, i, intestine, b, bladder, tu, human squamous cell carcinoma. Fig. 3. PET studies with fluorine-18 radiolabeled pyrido[2,3-*d*]pyrimidine derivative CKIE

Cyclin-Dependent Kinases (Cdk) as Targets for Cancer Therapy and Imaging 279

transcription regulation and survival. *Blood,* Vol.105, No.11, pp. 4484-91 Appleyard, M. V.; O'Neill, M. A.; Murray, K. E.; Paulin, F. E.; Bray, S. E.; Kernohan, N. M.;

Bahleda, R.; Spreafico, A.; Soria, J.; Moldovan, C.; Belli, C.; Fiorentini, F.; Scaburri, A.;

Baughn, L. B.; Di Liberto, M.; Wu, K.; Toogood, P. L.; Louie, T.; Gottschalk, R.; Niesvizky, R.;

Benouaich-Amiel, A.; Mazza, E.; Massard, C.; Gaviani, P.; Fiorentini, F.; Scaburri, A.;

Benson, C.; White, J.; De Bono, J.; O'Donnell, A.; Raynaud, F.; Cruickshank, C.; McGrath, H.;

Bible, K. C.; Boerner, S. A.; Kirkland, K.; Anderl, K. L.; Bartelt, D., Jr.; Svingen, P. A.; Kottke,

Bible, K. C.; Lensing, J. L.; Nelson, S. A.; Lee, Y. K.; Reid, J. M.; Ames, M. M.; Isham, C. R.;

Bixby, D. & Talpaz, M. (2011). Seeking the causes and solutions to imatinib-resistance in

Blain, S. W. (2008). Switching cyclin D-Cdk4 kinase activity on and off. *Cell Cycle,* Vol.7,

Brasca, M. G.; Amboldi, N.; Ballinari, D.; Cameron, A.; Casale, E.; Cervi, G.; Colombo, M.;

8-{[4-(4-methylpiperazin-1-yl)phenyl]amino}-4,5-dihydro-1H-pyrazolo[4,3-

Burdette-Radoux, S.; Tozer, R. G.; Lohmann, R. C.; Quirt, I.; Ernst, D. S.; Walsh, W.;

dependent kinase inhibitor. *J Med Chem,* Vol.52, No.16, pp. 5152-63

Colotta, F.; Croci, V.; D'Alessio, R.; Fiorentini, F.; Isacchi, A.; Mercurio, C.; Moretti, W.; Panzeri, A.; Pastori, W.; Pevarello, P.; Quartieri, F.; Roletto, F.; Traquandi, G.; Vianello, P.; Vulpetti, A. & Ciomei, M. (2009). Identification of N,1,4,4-tetramethyl-

h]quinazoline-3-carboxamide (PHA-848125), a potent, orally available cyclin

Wainman, N.; Colevas, A. D. & Eisenhauer, E. A. (2004). Phase II trial of

xenograft model. *Int J Cancer,* Vol.124, No.2, pp. 465-72

tumors. *J Clin Oncol,* Vol.28, 15s (Suppl.), abstract 3063

every 21 days. *Br J Cancer,* Vol.96, No.1, pp. 29-37

flavopiridol. *Clin Cancer Res,* Vol.6, No.2, pp. 661-70

chronic myeloid leukemia. *Leukemia,* Vol.25, No.1, pp. 7-22

No.15, pp. 7661-7

abstract 2087

Vol.11, No.16, pp. 5935-41

No.7, pp. 892-8

p53-dependent apoptosis in B-CLL by down-regulation of genes involved in

Levison, D. A.; Lane, D. P. & Thompson, A. M. (2009). Seliciclib (CYC202, Rroscovitine) enhances the antitumor effect of doxorubicin in vivo in a breast cancer

Pacciarini, M. A.; Laffranchi, B. & Reni, M. (2010). Phase I study of the oral CDK-TRKA inhibitor PHA-848125 in combination with gemcitabine in advanced solid

Cho, H.; Ely, S.; Moore, M. A. & Chen-Kiang, S. (2006). A novel orally active small molecule potently induces G1 arrest in primary myeloma cells and prevents tumor growth by specific inhibition of cyclin-dependent kinase 4/6. *Cancer Res,* Vol.66,

Pacciarini, M. A. & Calvo, E. (2010). Phase I study of the oral CDK-TRKA inhibitor PHA-848125 in recurrent malignant glioma (MG). *J Clin Oncol,* Vol.28, 15s (Suppl.),

Walton, M.; Workman, P.; Kaye, S.; Cassidy, J.; Gianella-Borradori, A.; Judson, I. & Twelves, C. (2007). A phase I trial of the selective oral cyclin-dependent kinase inhibitor seliciclib (CYC202; R-Roscovitine), administered twice daily for 7 days

T. J.; Lee, Y. K.; Eckdahl, S.; Stalboerger, P. G.; Jenkins, R. B. & Kaufmann, S. H. (2000). Characterization of an ovarian carcinoma cell line resistant to cisplatin and

Piens, J.; Rubin, S. L.; Rubin, J.; Kaufmann, S. H.; Atherton, P. J.; Sloan, J. A.; Daiss, M. K.; Adjei, A. A. & Erlichman, C. (2005). Phase 1 trial of flavopiridol combined with cisplatin or carboplatin in patients with advanced malignancies with the assessment of pharmacokinetic and pharmacodynamic end points. *Clin Cancer Res,*

In conclusion, the short biological half-life in the blood and low tumor uptake of the studied radiolabeled pyrido[2,3-*d*]pyrimidines limit the clinical application of these Cdk4/ Cdk6 inhibitors as radiotracers for the characterization of Cdk4/Cdk6 in tumors by means of PET. Nevertheless, further development and evaluation of suitable radiolabeled Cdk inhibitors with optimized properties *in vivo* are still of outstanding interest for the prospective functional characterization of Cdk in tumors by means of PET.
