p21. p21 was the first CKI to be discovered. It is a member of the KIP family of proteins that nonspecifically inhibit cyclin-cdk complexes in the nucleus.⁶² It can inhibit the cyclin D/cdk4 and cyclin E/cdk2 complexes early in G1, and it can also inhibit the cyclin A/cdk2 complex later, prior to the S-phase/G2-phase transition.⁶³ p21 transcript is directly activated by p53 and seems to be a fundamental partner for p53. It contributes to p53-induced apoptosis⁶⁴ and blocks the G2/M transition through cdc2 inhibition, mediating the pro-apoptotic and cell cycle-arresting effects induced by p53 in response to genotoxic stimuli.⁶⁵ Recent studies have
shown potential new roles for p21 and p27, as inhibitors of apoptosis and so, as potential oncogenes. These roles have been related to their relocalization to the cytoplasm,⁶⁶,⁶⁷,⁶⁸,⁶⁹,⁷⁰ whereas the cell cycle inhibitory activity is dependent on the nuclear localization. Although p21 is infrequently expressed in normal lung,⁷¹ it is overexpressed in 35% to 51% of NSCLC cases, without significant differences between ADC and SCC.⁷¹,⁷²,⁷³,⁷⁴,⁷⁵ p21 expression has been correlated with better differentiation in NSCLC.⁷¹,⁷³,⁷⁴ In two studies adequately controlled for disease stage, p21 expression was associated with improved survival,⁷²,⁷⁵ whereas another study reached slightly different conclusions.⁷⁶ In an analysis combining p21 and TGF-β, an improved survival was observed in early stage NSCLC when p21 and TGF-β were in concordance, presenting both high or both low expression.⁷⁶ Absence of p21 expression has been correlated to a worse prognosis when associated with p53 positivity ⁷⁵ and absence of p16 expression.⁷⁷

p27. p27, another member of the KIP family, blocks the activity of cyclin D1-CDK4/6 and cyclin E-CDK2 and is an important negative regulator of the G1 to S transition.⁶⁶,⁷⁸ In the normal resting cell, concentration of p27 is high and declines in response to proliferative stimuli. p27 overexpression leads to cell cycle arrest, whereas its loss of expression may result in tumor development and/or progression. As described for p21, some studies have suggested p27 to be an oncogene by inhibiting apoptosis, and that this role is related to its cytoplasmic relocalization.⁶⁶,⁷⁹ In vivo studies have shown that p27 expression decreases progressively during lung carcinogenesis. ⁸⁰ Increased p27 proteolytic activity has been associated to low levels of p27.⁷⁷,⁸¹ Low p27 expression has been associated with higher stages of lung cancer,⁸² as well as poor survival in NSCLC, either by itself⁷⁷,⁸¹,⁸²,⁸³ and/or when associated to other abnormalities like p53 mutations and/or Rb loss,⁸⁴ Ras mutations,⁸⁵ high cyclin E levels,⁷⁹ high proliferative rates⁸⁴,⁸⁵ and aneuploidy.⁸⁴ A favorable response to chemotherapeutic agents, including drugs targeting EGFR and HER-2, has been correlated to increased p27 expression.⁸⁶,⁸⁷

p16. p16 is a member of the INK4 family of proteins that inhibits cdk4 and cdk6 activation through a competitive mechanism for cyclin D binding site and specifically during the G1 phase.⁸⁸ Loss of p16 has similar effects on G1 progression than overexpression of cyclin D and loss of Rb. The discovery of p16 overexpression in Rb −/− cells, which alters the relationships between cyclin D, cdk4, and cdk6, suggests the existence of a feedback mechanism between p16 and Rb.⁸⁹ Decreased p16 is one of the most frequent alterations in lung cancer. Interestingly, loss of p16 is usually found in NSCLC, whereas loss of Rb is found in SCLC. These changes in p16 and Rb seem to be mutually exclusive.⁹⁰,⁹¹,⁹²,⁹³ The strong inverse correlation is more evident in SCC and SCLC than in ADC.⁹⁴,⁹⁵

p14. p14 is an alternate transcription product from the INK4/ARF locus shared with p16.⁹⁶,⁹⁷,⁹⁸ By its ability to antagonize the MDM2-mediated ubiquitination and degradation of p53,⁹⁹,¹⁰⁰,¹⁰¹,¹⁰² it induces apoptosis¹⁰³,¹⁰⁴ and growth inhibition.⁹⁹ p14 is decreased in 50% of NSCLC,⁹⁷,¹⁰¹ where it has an inverse relationship with MDM-2 expression.¹⁰¹ p14 can be lost secondary to losses at the p16 locus, and also because of the gene promoter methylation.¹⁰³,¹⁰⁵

p53 TUMOR SUPPRESSOR. p53, mapped on chromosomic region 17p13, is a tumor suppressor gene that, after sublethal DNA damage, mediates cell cycle arrest in the G1 phase.¹⁰⁶,¹⁰⁷ In this pathway, p53 interacts with several genes. The first one is mdm2: Levels of p53 are kept low by its association with the mdm2 oncogene product, which binds p53 and shuttles it out of the nucleus for proteolytic degradation, establishing an autoregulated feedback circuit. The second gene target is gadd45, that belongs to a gene family implicated in cellular growth arrest.¹⁰⁸,¹⁰⁹ The third target is the gene coding for p21, which develops inhibitory action on multiple cyclin-cdk complexes and also complexes with PCNA, a protein playing a key role in DNA reparation processes.¹¹⁰ The fourth target is the gene coding for Bax, which promotes apoptotic mechanisms and forms a heterodimer with Bcl-2 gene.¹¹¹ As a key regulator of cell growth and cell death, p53 is activated by several kinases that regulate the DNA damage checkpoint following many environmental stimuli, including DNA-damaging agents such as ionizing radiation, and is crucial in preventing the propagation of mutations in normal cells.⁶⁷ Activated p53 induces cell cycle arrest (p21^Cip1/Waf1) to allow cells to repair the damage or apoptosis if the damage is too severe and/or irreparable.¹¹² During carcinogenesis, p53 is frequently inactivated by multiple mechanisms. The most common mechanism is mutation at the p53 gene, which occurs in more than 50% of all human cancers. In response to DNA damage, some p53 mutants show less capacity to bind and initiate transcription from their target genes, such as p21, Mdm2, Bax, and cyclin G, and so, some of the p53-mediated effects are blunted,¹¹³ resulting in insensitivity to growth-inhibitory signals as well as evasion of apoptosis. Mouse models have confirmed cooperation between mutated p53 and mutative active K- ras in NSCLC development¹¹⁴ and between p53 and Rb in SCLC development.¹¹⁵

p53 mutation is one of the most common genetic alterations in lung cancer, with about 70% of SCLC and 50% of NSCLC presenting mutations in one allele of p53, often accompanied with loss of the wild-type allele.¹¹⁶ In NSCLC, p53 mutations are more frequently found in SCC than ADC.¹¹⁷ The p53 mutational spectra of lung cancer are different from those of other cancers with an excess of G:C to T:A observed,¹¹⁸ these ones especially linked to exposure to tobacco carcinogens such as benzopyrene.¹¹⁹ Several studies investigated the association of p53 abnormalities with prognosis in lung cancer, with discordant results. Some of them showed an association of p53 overexpression (mutants form have a longer half-life and lead to the detection of high levels of protein by IHC) with shorter survival,¹²⁰,¹²¹,¹²² others failed to find such an association.¹²³,¹²⁴ Concurrent p53 and p16 abnormalities have been associated to a worse prognosis.¹²⁵,¹²⁶ Many reports have linked the presence of p53 mutations to resistance to DNA-damaging chemotherapeutics agents,¹²⁷ and this is not surprising as we know that most chemotherapeutics agents act by stimulating apoptosis.

Cell Cycle as Target for Combined Treatment with Radiotherapy Cellular radiosensitivity varies along the different phases of cell cycle. The majority of cells surviving after a first dose of radiation were most likely in a less sensitive (G1 phase) or in a resistant phase of the cell cycle (late S phase). Those cells will then progress into a more sensitive cell cycle phase, that is at or close to mitosis, which represents a more ideal time for delivering radiation therapy. Several cell cycle regulatory proteins are potential molecular targets for cancer therapy, and agents can be combined to radiation to enhance its effects on lung cancer. Efforts have been made to sensitize cancer cells to the cytotoxicity of DNA damage by anticancer agents as early as four decades ago with compounds such as caffeine, which resulted in abrogation of the G2 cell cycle checkpoint. Many malignant cells, including lung cancer, have defective G1 checkpoint mechanisms and depend far more on G2 checkpoint than normal cells. More recently, several potent agents have been studied in preclinical and clinical trials. Among them is flavopiridol, the first CKI to enter clinical trials as a potential anticancer agent that exerts both cytostatic and cytotoxic actions on cancer cells. Another well-known drug is UCN-01, a potent prototypical chk1 inhibitor, which function as a CKI at high concentrations. Such compounds seek to force cells to enter mitosis without allowing adequate time for DNA repair, increasing the likelihood that cell death will occur by accumulation of DNA lesions. Treatment strategies consisting in abrogation of G2 checkpoints can also be achieved with the use of microtubule-targeting compounds, such as taxanes and epothilones to stall cells in G2/M (by preventing mitosis, thereby trapping cells in the phase of the cell cycle where ionizing radiation have the greatest effects). Microtubule-stabilizing agents will be discussed as mitotic targets in combination with ionizing radiation. Finally, Aurora kinases, part of the spindle checkpoint, are recent agents that can also be used in combined anticancer therapy (Table 14.2).

Cyclin-Dependent Kinases (CDK) Modulators Cell cycle regulatory proteins such as CDKs are potential molecular targets for radiation therapy.¹²⁸,¹²⁹,¹³⁰ The rationale for targeting the cell cycle and the CDKs in lung cancer therapy has been based on the frequency of their perturbations in lung tumors (overexpression of cyclins, and/or absent or diminished levels of CKI).⁵⁴,⁵⁵,¹³¹,¹³² For example, overexpression of cyclin D1 and loss of p16 gene expression has been associated with the development of lung cancer (see previous discussion). These defects in tumor cells lead to uncontrolled proliferation as a result of the loss of checkpoint integrity. In addition, cell cycle arrest by CKI has been shown to induce apoptosis.¹,¹³³,¹³⁴,¹³⁵

FLAVOPIRIDOL. Flavopiridol is a semisynthetic flavone that directly competes with the ATP substrate and reversibly inhibits kinase activity of multiple classes of CDKs, including cdk1, cdk2, cdk4, cdk6, cdk7, and cdk9 at submicromolar concentrations (IC50 values of 100 to 400).¹³⁶ This pancyclin inhibitor causes arrest at both G1 and G2/M phases of the cell cycle by several mechanisms: direct inhibition of cdks 1, 2, and 4, and indirect inhibition by downregulating cyclin H-cdk7,¹³⁷,¹³⁸,¹³⁹,¹⁴⁰ as well as tumor growth arrest in most solid tumors and xenografts.¹⁴⁰,¹⁴¹,¹⁴²,¹⁴³,¹⁴⁴ Flavopiridol also promotes a decrease in the level of cyclin D1,¹³⁵ which is commonly overexpressed in many cancers including lung cancer where it has been described as a poor prognosis marker.⁵¹,⁵²,⁵³,⁵⁴,⁵⁵,¹³¹,¹⁴⁵,¹⁴⁶,¹⁴⁷,¹⁴⁸,¹⁴⁹ However, at considerably higher concentrations than necessary to inhibit CDKs, flavopiridol inhibits the activity of several other protein kinases¹⁵⁰,¹⁵¹ including signal transducing kinases protein kinase A (PKA), PKC, and Erk-1, the receptor tyrosine kinase EGFR, and receptor-associated protein kinases such as c-Src.¹⁵² In addition, although described as cytostatic, flavopiridol has been shown to be also cytotoxic to many lung cancer cell lines¹⁴³,¹⁴⁴,¹⁵³,¹⁵⁴ by induction of apoptosis except in the A549 lung carcinoma cell line.¹⁵⁴ Normally, DNA-damaging agents induce p53, which in turn transcriptionally induces p21 and Mdm-2, the later binds p53 and targets it for degradation.¹⁵⁵ Although upregulating p53, flavopiridol was shown to inhibit transcription of p21 and Mdm-2 and to inhibit cell proliferation in A549 lung cancer cells.¹⁵³

Despite promising preclinical data, flavopiridol has not demonstrated in trial significant clinical activity as a single agent in patients with metastatic lung cancer.¹²⁹ The unexpected and significant toxicity of this agent given as single cancer therapy in all these clinical trials have so far discouraged its use in monotherapy.

Another approach for the use of flavopiridol in anticancer therapy is to take advantage of its potential to augment cytotoxic actions of chemotherapeutic agents and radiation. This strategy of combining flavopiridol with chemotherapy has been investigated in several studies showing promising results. Flavopiridol enhanced the cytotoxic effects of many chemotherapeutic agents in vitro (12, 13, 14, 15, 16, and 17) as well as in vivo (8, 13, and 15), in a sequence-dependent manner. Similar results were observed in the combined approach with ionizing radiation both in vitro,¹⁵⁶,¹⁵⁷ and in vivo¹³⁰ in various human cancer types.¹⁵⁶,¹⁵⁸,¹⁵⁹,¹⁶⁰ Flavopiridol sensitized human cancers to radiation in a dose-dependent manner, by cell cycle redistribution, by inhibiting cellular repair from radiation damage, and possibly by effects on angiogenic factors. In addition, studies also demonstrated the effects of flavopiridol in enhancing apoptosis and tumor regression.¹⁶⁰,¹⁶¹,¹⁶²,¹⁶³ The therapeutic ratio of radiotherapy in the in vivo tumor models was increased by flavopiridol.¹³⁰,¹⁵⁶ More specifically, the radiosensitizing action of flavopiridol was recently determined in zebrafish embryos using cyclin D1 (CCND1) downregulation by antisense hydroxylprolyl-phosphono peptide nucleic acid oligomers compared to control.¹⁶⁴ This study demonstrated that the specific sensitizing effects of flavopiridol in response to radiation were in part caused by the inhibition of cyclin D1, one of its primary pharmacologic targets.¹⁵⁸ Another study analysed the sequence-dependent effects of flavopiridol when combined to radiation and showed that the maximum radiopotentiation and apoptosis were observed when the lung cancer cells were treated with the sequence of docetaxel, then radiation, and finally flavopiridol both in vitro and in vivo. Therefore, the combined, sequence-dependent strategy radiation/flavopiridol has the potential to enhance outcome in many types of cancer and needs to be further investigated in a well-designed clinical trial.

UCN-01. Staurosporine is a potent nonspecific protein and tyrosine kinase inhibitor (TKI). UCN-01 or 7-hydroxystaurosporine, an analogue of staurosporine, was found to be a nonspecific inhibitor of CDKs, protein kinase C (PKC) isoenzymes, and causes cell arrest in G 1 and G 2 phases in different cell types.¹⁶⁵ UCN-01 is a checkpoint protein kinase inhibitor that promotes cellular G 1 arrest by inhibiting the activity of CDK-1 and CDK-2, thereby downregulating cyclins and increasing the expression of CKI, p21. Interestingly, UCN-01 abrogates the G2 checkpoint induced by radiation, leading irradiated cells into early mitosis and the onset of apoptotic death.²⁹,¹⁶⁶ The mechanisms by which the cells enter mitosis include activation of cdc2 kinase, and a direct inhibiting effect on the G 2 -associated checkpoint protein kinase 1 (Chk1).¹⁶⁷,¹⁶⁸ UCN-01 inhibits the cdc25C regulatory pathway and interferes with DNA damage-mediated inhibition of cdc2-cyclin B1-kinase. It targets more strongly Chkl than Chk2 (IC50 — 20 nM vs. — 1 pM).²⁹,¹⁶⁸,¹⁶⁹ It has been shown that Chk1 is required for the radiation-induced G 2 checkpoint.¹⁷⁰,¹⁷¹ Chk1 may control the G₂ checkpoint via phosphorylation-mediated negative regulation of Cdc25A, Cdc25C, and Cdc25B.¹⁷⁰,¹⁷²,¹⁷³ Thus, the potential of Chk1 inhibitors would most likely be as sensitizer to other anticancer treatments such as radiotherapy because inhibiting Chk kinases forces the cell to enter mitosis with DNA damage accumulation. Cancer cells could likely be particularly sensitive to such treatment, since they commonly lack normal G 1 checkpoint control and may rely more on the S and G 2 checkpoints compared to normal cells.¹⁷⁴ In vitro, UCN-01 exhibits a potent radiosensitivity effect in mutated p53 carrying NSCLC cell models.¹⁷⁵,¹⁷⁶ Despite encouraging preclinical results, combination studies of UCN-01 with radiotherapy have not yet been opened, whereas UCN-01 alone has undergone phase I and II clinical trials in several cancer types.¹⁷⁷,¹⁷⁸,¹⁷⁹