Background

Deoxycoformycin (pentostatin) is a natural product first isolated in 1974 from the culture of Streptomyces antibioticus (see Figure 50-9).²⁸³ Its structure mimics the transitional-state form of adenosine in an adenosine deaminase-catalyzed reaction, and it is one of the most potent inhibitors of adenosine deaminase (K_i 5 × 10^-10-10^-12 M depending on the source of the enzyme).²⁸⁴ Because adenosine deaminase is not essential for cell growth in culture, this compound did not show antitumor activity in preclinical screenings.

The initial clinical development of deoxycoformycin centered on its activity as an adenosine deaminase inhibitor for the potentiation of adenosine arabinoside, which also was deaminated by adenosine deaminase to yield less toxic compounds. During early Phase I studies, the profound lymphotoxic effect of deoxycoformycin was noted. Other studies described a congenital syndrome of severe combined immunodeficiency associated with low or undetectable levels of adenosine deaminase in lymphocytes,²⁸⁵ and these results suggested the importance of adenosine deaminase in lymphocyte function, leading to intensive development of deoxycoformycin as a single agent for the treatment of lymphoproliferative diseases.

The most responsive tumor identified is hairy cell leukemia, in which durable remissions are achieved in over 90% of patients with a relatively brief course of treatment.^286,287 Other responsive lymphoid diseases include chronic lymphocytic leukemia and prolymphocytic leukemia, mycosis fungoides, and acute T-cell leukemia/lymphoma.^288,289 Considerable variation exists in the susceptibility of patients to deoxycoformycin toxicity. This includes immunosuppression,^290,291 CNS disturbances, impaired renal function, conjunctivitis, and muscle and joint pain. Impaired renal function and poor performance status place patients at high risk for toxicity, even with low dosages of this drug.

Metabolism

Deoxycoformycin enters the cell through the facilitated-diffusion nucleoside carrier. It can be phosphorylated to mono-, di-, and triphosphate nucleotides, and significant incorporation into DNA, but not RNA, has been observed.²⁸⁶ Adenosine kinase and deoxycytidine kinase²⁸³ do not appear to be responsible for the initial phosphorylation, but reversal of the 5′-nucleotidase reaction is a potential basis for nucleotide formation. Definitive statements cannot be made about the enzymology of deoxycoformycin metabolism at this time.

Mechanisms of Action and Resistance

The primary site of action is the inhibition of adenosine deaminase. Because of the inhibition of adenosine deaminase in vivo, deoxyadenosine and adenosine cannot be catabolized efficiently. Consequently, deoxyadenosine-phosphorylated metabolites accumulate in many types of cells.²⁹² This imbalance in deoxynucleotide pools is known to be toxic to cells, and the antitumor activity of deoxycoformycin may result from the combination of direct effects of deoxycoformycin and its metabolites as well as the expanded pools of deoxyadenosine.

The failure of deoxyadenosine to accumulate in cultures treated with deoxycoformycin is the reason deoxycoformycin was not identified as a potential antitumor compound in cell culture systems. The degree of deoxyadenosine triphosphate (dATP) accumulation correlated well with cell death caused by deoxycoformycin. Thus, dATP, which is known to be an allosteric inhibitor of ribonucleotide reductase, could result in growth inhibition by the generation of an imbalance of deoxynucleotide triphosphate pools. However, additional sites of action for both deoxycoformycin and deoxyadenosine are suggested by the observation that deoxycoformycin and deoxyadenosine are cytotoxic to nondividing cells, which do not require the function of ribonucleotide reductase. One potential site is the depletion of nicotinamide adenine dinucleotide (NAD) in deoxycoformycin- and deoxyadenosine-treated cells. NAD is required for poly-ADP ribosylation, a reaction that is essential to maintain the integrity of DNA and its repair process. Depletion of NAD could reduce the capacity for DNA repair, a constant process in cells, and cause DNA breaks as well as cell death.^293,294

The second suggested site is inhibition of S-adenosyl homocysteine hydrolase by deoxyadenosine.^295,296 Inhibition of this enzyme decreases the capacity of cells to perform transmethylation, a reaction that is critical for certain macromolecular functions. This mechanism does not require deoxyadenosine to be phosphorylated, and it may play an important role in the toxicity of deoxycoformycin to nonproliferating tissues, such as in the liver and CNS. The activation of mitochondrial-dependent apoptosis through the interaction of Apf and dATP could also contribute to its activity.

Deoxycoformycin and deoxyadenosine also decrease ATP levels in some cell systems. In mice, hemolysis after treatment with deoxycoformycin is related to ATP depletion. Deoxycoformycin has also been shown to form phosphorylated metabolites that can be incorporated into DNA; whether these metabolites contribute to deoxycoformycin action, however, is not clear.²⁹¹

The mechanism of resistance to deoxycoformycin has not been defined because deoxycoformycin is not cytotoxic in cell culture. The action of deoxycoformycin in vivo results from the combined action of deoxycoformycin and deoxyadenosine, so the mechanism of cellular resistance to deoxyadenosine should be applicable. This could include adenosine kinase deficiency or altered quality or quantity of ribonucleotide reductase.

Background

In the search for more effective compounds than adenine arabinoside (ara-A, vidarabine), which has limited clinical usefulness because of its rapid deamination by adenosine deaminase, 2-fluoroadenosine arabinoside (9-β-d-arabinofuranosyl-2-fluoradenine) was synthesized. It has been found to be relatively resistant to adenosine deaminase and has impressive antitumor activities in vivo as well as in cell culture.²⁹⁷ Its limited solubility and consequent difficulties in formulation led to the synthesis of a prodrug, the 5′-monophosphate of 2-F-ara-A (Fludara IV).

Fludara IV entered clinical trials in 1982, and it is one of the most active agents in the treatment of chronic lymphocytic leukemia.^298,299 A high level of activity also has been observed in a variety of indolent lymphoproliferative neoplasms, including low-grade non-Hodgkin lymphoma, cutaneous T-cell lymphoma, macroglobulinemia, and hairy cell leukemia.^300–302 The dose-limiting toxicities during Phase I trials were myelosuppression and leukopenia. Delayed onset of severe neurotoxicity also was noted with doses of 96 mg/m²/d for 5 to 7 days. Other toxicities noted during Phase I trials included somnolence, mild to moderate nausea and vomiting, and rare but reversible interstitial pneumonitis. Fludara IV is converted by phosphatases to 2-F-ara-A within several minutes of injection; it is not further catabolized in plasma.²³⁷

Metabolism

Transport of 2-fluoroadenosine arabinoside (F-ara-A) into mouse L1210 cells is mediated by nonconcentrative, high- and low-affinity systems.³⁰³ In contrast to these leukemia cells, epithelial crypt cells from mouse intestine possess only a low-affinity system,³⁰⁴ and this difference in transport could be partly responsible for the favorable therapeutic index of 2-F-ara-A against sensitive tumor cells in mice. In future human studies, the potential role of transport systems in determining the sensitivity to 2-F-ara-A should be considered. Once 2-F-ara-A is taken up by cells, it is phosphorylated to 2-fluoroadenine arabinoside monophosphate (2-F-ara-AMP), not like ara-A as a substrate of adenosine kinase, but by cytoplasmic deoxycytidine kinase.³⁰⁵ Tumor cells lacking cytoplasmic deoxycytidine kinase are resistant to F-ara-A. Intracellular F-ara-AMP can be further phosphorylated to the diphosphate F-ara-ADP, but it is not clear which enzyme is responsible for this reaction. AMP kinases likely may be responsible for the further phosphorylation of F-ara-AMP to F-ara-ADP. Nucleoside diphosphate kinases may be the predominant enzyme species responsible for the formation of F-ara-ATP from F-ara-ADP. F-ara-ATP can be incorporated into DNA in competition with dATP by DNA polymerases. Although DNA polymerases α, β, Δ, and γ are all capable of using F-ara-ATP as a substrate, DNA polymerase α has a greater affinity for F-ara-ATP than do other DNA polymerases.^306,307 Once F-ara-AMP is incorporated into the terminus of the growing DNA chain, the next step of elongation is retarded, regardless of which DNA polymerase is employed.^307,308

In addition, F-ara-A also has been shown to be incorporated into RNA,^309,310 but which RNA polymerase is responsible has not been established. The incorporation of F-ara-A into poly (A1) RNA was 12-fold greater than that into poly (A) RNA. A summary of the metabolism of 2F-ara-A is shown in Figure 50-11.

Figure 50-11

Structure and metabolism of 2-fluoro-arabinosyl-adenine (2F-ara-A). 2F-ara-l represents the dominant inosine derivative. MP, DP, TP = mono-, di-, and triphosphate.

Investigations of F-ara-A as a modulator of ara-C therapy are currently underway. When F-ara-A is given before ara-C, an increase in the accumulation of ara-CTP occurs in leukemic lymphocytes.³¹¹ This modulation of ara-C anabolism probably results from an indirect effect of F-ara-CTP on deoxycytidine kinase that relates to a reduction in the deoxynucleotide pools regulating the enzyme. It also may reflect a direct effect by F-ara-CTP on the activity of deoxycytidine kinase.^311,312 The in vitro accumulation of ara-CTP also has been shown in the lymphocytes of patients with chronic lymphocytic leukemia treated with this sequential combination.³¹³ The results of a clinical study in individuals who are refractory to F-ara-A therapy show partial or minor responses in approximately 35% of patients.³¹³

Mechanism of Action

The major site of growth inhibition by F-ara-A is the inhibition of DNA synthesis. Treatment of cells with F-ara-A is associated with the accumulation of cells at the G₁-S-phase boundary and in the S phase; thus, it is a cell cycle S-phase-specific drug. Incorporation of the active metabolite F-ara-ATP retards DNA chain elongation. The degree of incorporation of the analog nucleotide depends not only on the type of DNA polymerase but also on the amount of intracellular dATP that competes with F-ara-ATP for incorporation. Among DNA polymerases in human cells, polymerase α, which is the critical enzyme in nuclear DNA synthesis, is more susceptible to the incorporation of F-ara-ATP. A consequence of this analog nucleotide incorporation is the retardation of DNA-chain elongation.³¹⁴

F-ara-ATP is also a potent inhibitor of ribonucleotide reductase, the key enzyme responsible for the formation of dATP. This causes a decrease of deoxynucleotides in 2F-ara-A- treated cells, which enhances the incorporation of F-ara-ATP into DNA. This may be considered to be “self-potentiation” of the inhibition of DNA synthesis by F-ara-ATP. In addition, F-ara-ATP was found to be an inhibitor of DNA primase, which is responsible for Okazaki fragment synthesis,¹⁸¹ another important step in DNA synthesis. The inhibition of RNA primer formation for DNA synthesis by F-ara-ATP was recently demonstrated as well,³¹⁵ but the inhibition of Okazaki fragment formation by F-ara-ATP could conceivably play a role in the inhibition of DNA synthesis by F-ara-A. In addition, F-ara-A can inhibit mitochondrial DNA synthesis at concentrations similiar to those that cause cytotoxicity; however, such inhibition does not affect cell growth for several cell generations.³¹⁵ Thus, the cytotoxicity of F-ara-A, which usually is estimated by the continuous exposure of cells to drugs for three to four generations, likely does not result from the inhibition of mitochondrial DNA. Also, it has been reported that incubation of normal lymphocytes for 24 hours with 10 μM, but not 1 μM, caused a decrease in both cytoplasmic NAD and ATP concentrations that could be correlated with a decrease in cellular viability.³¹⁶ The mechanism for the depletion of NAD and ATP by F-ara-A is not clear, and whether the inhibition of mitochondrial DNA synthesis by F-ara-A or depletion of NAD and ATP is responsible for the delayed onset of F-ara-A toxicity observed clinically has not yet been established.

Resistance to F-ara-A may occur because of decreased uptake, lack of deoxycytidine kinase, increased intracellular concentration of dATP, decreased susceptibility to the activity of ribonucleotide reductase, decreased affinity of DNA polymerase for F-ara-ATP, or increased efficiency of the removal of F-ara-ATP from the 3′ terminus where incorporated into DNA. The potential role of the 3′ and 5′ exonuclease activities of DNA polymerase D and other 3′ and 5′ exonuclease activities in removal of incorporated F-ara-AMP remains to be defined as a possible mechanism of resistance.

Background

The rationale for the development of 2-chlorodeoxyadenosine (Cl-dAdo, cladribine) was that the death of lymphocytes in patients with adenosine deaminase deficiency was associated with the accumulation of deoxynucleotides. This deoxyadenosine analog was selected for its resistance to adenosine deaminase. Its specific action on lymphoid cells is attributed to the high level of deoxycytidine kinase and low 5′-nucleotidase activity in these cells.^317–319 This compound is highly cytotoxic to a variety of cell lines in culture, and it has potent antileukemic activity in mice.^320,321 Cladribine was shown to have potent and lasting effects in the treatment of low-grade B-cell neoplasms, such as chronic lymphocytic leukemia, non-Hodgkin lymphoma, and hairy cell leukemia.^322–324 In addition, Cl-dAdo has demonstrated clinical activity against acute myeloid leukemia in children, including those with leukemic blast cells in the CNS³²⁵ and in T-cell lymphoproliferative disorders.³²⁶ The spectrum of clinical activity is similar to that of Fludara IV; however, a few patients who do not respond to F-ara-A are sensitive to Cl-dAdo.³²⁷ The major toxicity encountered is bone marrow suppression that is associated with severe infections. The degree of suppression relates to the rate of administration, cumulative dose, and tumor burden at the start of therapy.^322,328

Metabolism

The mechanism of transport for cladribine into a variety of human hematopoietic cell lines was explored using nucleoside transport inhibitors, such as dipyridamole and nitrobenzyl thioinosine (NBTI). The transport mechanism appears to be different in different cell lines, an observation based on their differential response to nucleoside transport inhibitors.³²⁹ Both NBTI-sensitive and NBTI-insensitive nucleoside transporters are involved. Once Cl-dAdo enters cells, it can be phosphorylated by deoxycytidine (dCyd) kinase to 2-chloro-deoxyadenosine mono (2Cl-dAMP),³³⁰ which subsequently is phosphorylated to 2-chloro-deoxyadenosine diphosphate (2Cl-dADP) and then to 2Cl-dATP. The enzymes involved, however, are not established. As 2Cl-dATP, it can be incorporated into DNA through the action of DNA polymerases by competing with dATP.³³⁰ The structure and metabolism of 2-chlorodeoxyadenosine are shown in Figure 50-12.

Figure 50-12

Structure and metabolism of 2-chloro-deoxyadenosine (2C1-dAdo). MP, DP, TP = mono-, di-, and triphosphate.

Mechanisms of Action and Resistance

2Cl-dAdo can inhibit DNA synthesis in growing cells as well as DNA repair in resting cells.³³¹ When growing cells were treated with 2Cl-dAdo, an accumulation of cells in the S phase was observed, suggesting that inhibition of DNA synthesis could be responsible for the cell-killing effect of the drug. The active metabolite is 2Cl-dATP, which can compete with dATP to be incorporated into the 3′-end of the growing DNA chain. Elongation beyond the incorporated analog was significantly retarded, and this could partly contribute to its inhibitory activity against DNA synthesis. Furthermore, 2Cl-dATP is a potent inhibitor of ribonucleotide reductase.³³² Levels of intracellular deoxynucleoside triphosphates were found to decrease in cells after exposure to 2Cl-dAdo,³³⁰ which also could contribute to its antitumor activity.

The mechanism of resistance is not clear, but it could be similar to that of 2F-ara-A. Although 2F-ara-A and 2Cl-dAdo share many similar features, there are differences in metabolism and mechanisms of action. Recently, it was suggested that Ca⁺⁺-sensitive mitochondrial events could play an important role in 2-Cl-dAdo cytotoxicity.^333,334

Background

Although hydroxyurea was first synthesized in 1869,³³⁵ its biologic activity was not recognized until 60 years later, when it was discovered that hydroxyurea could produce leukopenia, anemia, and megaloblastic changes in the bone marrow of rabbits.³³⁶ This simple molecule (Figure 50-13) has been evaluated in a number of types of cancer, but its principal uses are in myeloproliferative diseases. Currently, it is an initial therapy of choice for chronic myelogenous leukemia; it also is used as therapy for polycythemia vera and hypereosinophilic syndrome. Activity against solid tumors has been demonstrated, but in these cases, it generally is used in combination with other anticancer agents or with radiation.³³⁷ A recent report also indicated the ability of hydroxyurea to inhibit human immunodeficiency type I DNA synthesis in activated blood lymphocytes, either alone or in combination with zidovudine or dideoxyinosine, suggesting a possible antiviral application for this compound.³³⁸ However, clinical studies did not support this approach for the treatment of AIDS patients.

Figure 50-13

Structure of hydroxyurea.

Hydroxyurea can be taken orally, and the half-life in plasma is approximately 4 hours.³³⁹ It readily crosses the blood-brain barrier. It is excreted predominantly in urine, but the interpatient variability is significant. The full extent and significance of hydroxyurea metabolism in humans has not been well established. It can be degraded by intestinal bacterial urease to form hydroxylamine, which can interact with acetylcoenzyme A to form acetohydroxamic acid; this metabolite is found in the plasma of patients receiving hydroxyurea therapy.³⁴⁰

The dose-limiting toxicity of hydroxyurea is myelosuppression. This results from inhibition of DNA synthesis in bone marrow. Toxicity begins within 2 to 5 days, and its duration is short once the drug is discontinued. Gastrointestinal side effects frequently are seen but rarely require discontinuation of therapy at the doses commonly used. Some dermatologic changes, such as hyperpigmentation, can also occur in patients after extended therapy.³³⁷

Mechanism of Action and Resistance

Hydroxyurea is considered to enter cells by passive diffusion.³⁴¹ It inhibits cellular DNA synthesis through the inhibition of ribonucleotide reductase, which is the key enzyme responsible for the synthesis of deoxynucleotides, the building blocks of DNA³⁴². The substrates for this reaction include the four ribonucleoside diphosphates and the diphosphonucleotides of fluorouridine, azacytidine, and thioguanosine, namely 5-FUDP, 5-aza-CDP, and 6-thio-GDP. The activity of ribonucleotide reductase is highly regulated by the intracellular concentration of ribonucleoside and deoxyribonucleoside triphosphates. Two models, sequential and intercalating, have been proposed for the interplay of ribonucleotide reductase and deoxynucleoside triphosphates.³⁴³ The metabolites of deoxynucleoside analogs, such as 2F-ara-ATP and ara-ATP, are potent inhibitors of this enzyme as well. The activity of this enzyme plays a key role in controlling the intracellular concentrations of deoxynucleotide triphosphates; thus, it can influence the activation or incorporation of deoxynucleoside antimetabolites, such as ara-C, 5-FUdR, and 2F-ara-A, into DNA.

Inhibition of ribonucleotide reductase by hydroxyurea would not affect the incorporation of these antimetabolites and, therefore, could potentiate their action. Ribonucleotide reductase is composed of two types of protein subunits, M1 and M2. These two proteins are coded by two different chromosomes. M1 protein, which is coded by chromosome 11 and has a molecular weight of 170 kD, does not vary with cell cycle and is responsible for the interaction with nucleotides.^344,345 M2 protein, which is coded by a gene on chromosome 2 in close proximity to the ornithine decarboxylase gene, has a molecular weight of 88 kD and fluctuates throughout the cell cycle, with peak activity in the S phase. The alteration of ribonucleotide reductase activity through the cell cycle is primarily controlled by the amount of M2 protein that binds a stoichiometric amount of iron and a stable organic free radical localized to a tyrosine residue.^346–348 Hydroxyurea inhibits ribonucleotide reductase through the inactivation of the tyrosyl free radical on the M2 subunit. This inactivation can be partially prevented by ferrous iron.³⁴⁹ The required concentration of hydroxyurea to inhibit human ribonucleotide reductase by 50% is approximately 0.5 μM.

Because of the inhibition of ribonucleotide reductase by hydroxyurea, pools of deoxynucleotide triphosphates decrease, with concomitant inhibition of DNA synthesis. The cytotoxicity of hydroxyurea is dose- and time-dependent. Most cells are accumulated in the S phase and at the G₁-S-phase boundary under the influence of hydroxyurea.^350,351

Cells can become resistant to hydroxyurea because of increased ribonucleotide reductase activity, primarily resulting from increased levels of M2 protein. Levels of M1 protein increase only when high levels of resistance to hydroxyurea are generated. These increases of M1 or M2 proteins generally reflect the overexpression of the proteins because of gene amplification.^352–356 It should also be noted that a new ribonucleotide reductase with different M2 subunit has been identified. Recently, a human KB cell line—resistant to hydroxyurea because of gene amplification of the M2 subunit, increased concentrations of M2 mRNA and protein, and increased ribonucleotide reductase activity—was found to express collateral sensitivity to 6-thioguanine. The mechanism responsible for this supersensitivity is believed to be an elevated conversion of 6-thioguanine ribonucleotide to its deoxyribonucleotide form.³⁵⁷ As previously hypothesized, alternating the use of hydroxyurea with antimetabolites, such as 6-thioguanine, warrants further clinical exploration in the treatment of cancer.³⁵⁸