Lung cancer remains a major human cancer site for which chemopreventive drugs are being developed. Synthetic and natural retinoids are among the drugs being evaluated for prevention and treatment of lung cancers. The steroid/thyroid receptor superfamily contains 2 types of nuclear retinoid receptors, the retinoic acid receptors (RAR) followed by retinoid X receptors (RXR).1, 2, 3, 4, 5 Retinoids in general exert their biological activity by binding to the nuclear receptors that in turn bind as either homodimers or heterodimers to retinoic acid response elements (RARE) to regulate gene transcription so as to ultimately result in the regulation of cell proliferation, apoptosis and differentiation. The ligand specificity of the retinoid receptors is quite distinct. While the RAR receptors interact with a defined set of genes via the RARE elements, RXR receptors form heterodiemers with various nuclear receptors e.g., PPARs, RARs, thyroxine receptor, and their various comodulators and corepressors. Thus, RXR agonists can modulate a striking number of genes that are themselves modulated by these receptors and their cognate ligands. Nonsmall-cell lung cancer (NSCLC) has been demonstrated to have decreased expression of RAR and RXR genes, especially RAR-β and RXR-β.6, 7 This would suggest that alteration in the activity of retinoids receptors plays an important part in lung carcinogenesis, and indicate the potential utility of retinoids in the prevention and treatment of lung cancer.
Targretin®, 4-[1-(3,5,5,8,8,-pentamethyl-5,6,7,8-tetrahydro-naphthalen-2-yl)-vinyl]-benzoic acid (Bexarotene, Ligand Pharmaceuticals Inc.; Fig. 1), is a synthetic RXR agonist with a chemical structure that is similar to 9-cis-retinoic acid.8 Targretin has limited affinity for the various RAR receptors. As a chemotherapeutic agent for NSCLC, targretin is being evaluated in various clinical trials especially in combination with other chemotherapeutic drugs.8 Results to date have indicated a usefulness of targretin as adjuvant and maintenance therapy in NSCLC patients. As a chemopreventive agent, targretin has proven to be highly effective in preventing and inducing the regression of mammary cancer in animal models.9, 10, 11 The therapeutic activity of targretin in NSCLC in humans and its ability to prevent mammary cancer in animals suggest that it could prevent lung cancer. In mice, lung tumors induced by carcinogens including 4-(methylnitrosoamino)-1-(3-pyridyl)-1-butanol (NNK) and vinyl carbamate have been used as models to screen and evaluate the efficacy of drugs to prevent lung cancer.12, 13 We report here that targretin is very effective in preventing lung tumors induced in mice by the two carcinogens, so as to warrant further development as a chemopreventive drug for lung cancer.
Material and methods
Chemicals and reagents
Vinyl carbamate (purity >99%) was purchased from Toronto Research Chemicals (North York, Ontario, Canada), carboxymethylcellulose (Catalog No. C-5678) from Sigma Chemical Co. (St. Louis, MO) and NNK from Chemsyn Service Laboratories (Lenexa, KS). Targretin was obtained from the National Cancer Institute, DCPC Repository (Rockville, MD). AIN-76A diet was obtained from Dyets (Bethlehem, PA).
Female Strain A mice (5–6-weeks-old) were purchased from Jackson Laboratories (Bar Harbor, ME) and placed on study when they were 7–8 weeks of age. The mice were housed in the AAALAC-accredited laboratory animal facilities of the Medical College of Ohio and the Ohio State University. AIN-76A diet and drinking water were provided ad libitum.
Experimental design: prevention of vinyl carbamate-induced lung tumors
In Experiment 1, the female mice were administered 20 mg/kg vinyl carbamate in saline by intraperitoneal injection once a week for 2 consecutive weeks. In the 3 experiments reported here, the carcinogens and targretin were administered in 0.2 ml/mouse vehicle. Starting at 2 weeks after the last dose of vinyl carbamate, the mice received 0 (2% carboxymethylcellulose vehicle), 10, 30, 100 and 300 mg/kg targretin by oral gavage 5-days/week and for a total of 22 weeks after which the mice were sacrificed (25 weeks from the first dose of vinyl carbamate). Thirty-six mice received the carboxymethylcellulose vehicle, 16 mice/group received 10 or 30 mg/kg targretin and 32 mice/group received 100 or 300 mg/kg targretin.
In the 3 experiments performed, the following procedure was used to evaluate the lungs tumors. At necropsy, the right lung was evaluated for the number of tumors and then frozen in liquid nitrogen for analysis of molecular end-points. The left lung was fixed in formalin overnight and transferred to 70% alcohol. The lungs were evaluated for the number and size of tumors prior to embedding in paraffin. Sections of lungs were stained with hematoxylin and eosin for histopathologic evaluation of the tumors.
In Experiment 2, female mice were administered 20 mg/kg vinyl carbamate by intraperitoneal injection once a week for 3 consecutive weeks. The total dose of vinyl carbamate was increased relative to Experiment 1 so as to increase the yield of tumors and the sensitivity of the assay to detect prevention by targretin. Two weeks later, some mice started to receive 200 mg/kg targretin or its carboxymethylcellulose vehicle by oral gavage 5-days/week from weeks 4–13, 4–19 and 4–25 after the first dose of vinyl carbamate. Each of these treatment groups contained 25 mice. Another treatment group containing 10 mice received targretin for weeks 23–25. All the mice in this experiment were killed at week 25.
Experimental design: prevention of NNK-induced lung tumors
In Experiment 3, the female mice were administered 100 mg/kg NNK in saline by intraperitoneal injection once a week for 3 consecutive weeks. Three weeks after the last dose of NNK, they were administered 0, 100 and 300 mg/kg targretin in carboxymethylcellulose by oral gavage 5 times a week for a total of 3 weeks. Each treatment group contained 20 mice. The mice were killed 26 weeks after the last dose of NNK, and the lungs evaluated for tumors.
Unless otherwise stated, the results were analyzed for statistical significance by a one-way analysis of variance (ANOVA) followed by the Bonferroni t-test.
In the 3 experiments reported here, targretin did not affect the body weight of the mice. However, when administered up until the mice were killed, targretin increased the liver to body weight ratio. For example in Experiment 2, the liver to body weight ratio was 4.29 ± 0.163, 4.78 ± 0.127, 10.4 ± 0.22 and 9.05 ± 0.89 for targretin administered from 4–13, 4–19, 4–25 and 23–25 weeks after the first dose of vinyl carbamate compared to 4.36 ± 0.125 and 3.92 ± 0.135 for the carboxymethylcellulose vehicle and diet controls, respectively (means ± SE). Thus, 2 weeks of treatment with targretin virtually doubled the liver weight, while cessation of treatment for 6 weeks resulted in the liver weight returning back to control level.
Prevention of vinyl carbamate-induced lung tumors
In Experiment 1, doses of 0, 10, 30, 100 and 300 mg/kg targretin were administered starting 2 weeks after vinyl carbamate and continued until the mice were killed. All the animals had at least 1 tumor, so that targretin did not alter the incidence of animals with tumors. Dose levels of 30 mg/kg or greater significantly reduced the multiplicity of mouse lung tumors by 22–33% (Fig. 2). The dose dependency was relatively shallow, and although the multiplicity of lung tumors was less in mice administered 300 mg/kg than in those administered 30 mg/kg targretin, it was not statistically significant.
Experiment 2 was performed to determine whether shorter duration of treatment with targretin as well as whether stopping treatment would be effective regimen in preventing lung tumors. Mice were administered 200 mg/kg targretin from 4–13, 4–19, 4–25 and 23–25 weeks after the first dose of vinyl carbamate. All the animals had at least 1 tumor, so that targretin did not alter the incidence of animals with tumors. Administering targretin from weeks 4–19 and 4–25 significantly decreased the multiplicity of lung tumors from 35.3 ± 1.43 (controls) to 29.1 ± 1.51 and 25.0 ± 0.93, respectively (Fig. 3). The treatment duration of 4–25 weeks was significantly more effective than the treatment for 4–19 weeks (p-value = 0.037). Less than 2% of the tumors were found to be adenocarcinomas. The low yield of adenocarcinomas did not vary significantly among the groups.
Using a similar experimental protocol, we have previously reported that short-term administration of budesonide, another chemopreventive agent that prevents lung tumors in mice, decreased the size of the tumors even when its administration ceased 15 weeks prior to killing.14 Targretin also decreased the size of the tumors (Fig. 4). The tumors from mice administered the control diet had an average diameter of 0.951 ± 0.0082 and an area of 0.551 ± 0.026. The longer duration of treatment for 21 weeks (4–25) that lasted until the mice were killed was more effective in reducing tumor size than a shorter duration of treatment for 15 (4–19) weeks that was terminated 6 weeks prior to killing. Treatment for only 9 (4–13) weeks followed by a holding period of 12 weeks was not effective in decreasing the size of the tumors. Interestingly, a short treatment duration of only 2 weeks from weeks 23–25 significantly decreased tumor size, being as effective as the 15 weeks of treatment that was terminated 6 weeks prior to killing.
Prevention of NNK-induced lung tumors
In Experiment 3, targretin was evaluated for prevention of NNK-induced mouse lung tumors. Three weeks after the last dose of NNK, targretin (100 and 300 mg/kg) was administered for only a total of 3 weeks followed by a holding period of 20 weeks. All the animals had at least 1 tumor except for 2 of the 20 mice that received 300 mg/kg targretin, which was not a statistically significant decrease in tumor incidence. Surprisingly, this very limited duration of treatment significantly reduced lung tumor multiplicity from 10.6 ± 1.13 to 6.38 ± 0.75 and 4.60 ± 0.70 for 0, 100 and 300 mg/kg targretin, respectively, (Fig. 5). The reduction in tumor multiplicity appeared to be dose-related, although the difference between the 2 dose levels of targretin was not statistically significant. The size of the tumors induced by NNK was small, being ≤1 mm in diameter and did not differ among the treatment groups. In the NNK model, limited treatment with targretin for only 3 weeks followed by a holding period of 20 weeks was sufficient to reduce the yield of mouse lung tumors.
The rationale for evaluating targretin for prevention of lung tumors in mice was its therapeutic activity in NSCLC in humans8 and its ability to prevent mammary cancer in animals.9, 10, 11 In the mouse lung tumor model using 2 different carcinogens, targretin decreased the tumor yield in a dose-dependent response. However, the dose-response relationship for prevention of lung tumors was very shallow. In the vinyl carbonate model, 300 mg/kg targretin appeared to cause a greater decrease in tumor multiplicity than the 30 mg/kg dose; however, it was not statistically significant (p-value > 0.05). Similarly in the NNK model, 300 mg/kg appeared to reduce the yield of lung tumors by a greater extent than 100 kg/kg, but again the difference was not statistically significant. In methylnitrosourea (MNU)-induced mammary cancer in rats, targretin also caused a shallow dose-response relationship with 6.7 and 60 mg/kg by oral gavage causing a 78 and 96% reduction in tumor multiplicity.11 In contrast, the dose-response relationship for another chemopreventive agent that has demonstrated efficacy in the mouse lung tumor model using both NNK and vinyl carbamate, i.e., budesonide, was much steeper with 0.6 < 1.2 < 2.4 mg/kg diet being statistically significant.14 Hence, for 2 highly effective chemopreventive agents, targretin and budesonide, their dose-response relationship appears to be different with the first being very shallow and the other being steep.
In the vinyl carbamate-induced mouse lung tumor model, targretin decreased the size of the tumors. In fact, administering targretin to mice with lung tumors for only 2 weeks prior to killing (weeks 23–25) decreased the size of the tumors by the same extent as treatment for 15 (4–19) weeks followed by a 6 weeks holding period. However, administering targretin for a longer duration of 21 (4–25) weeks was more effective than treatment for 2 weeks prior to killing or for 15 weeks ending 6 weeks prior to killing. The treatment regimen of 9 (4–13) weeks did not alter tumor size. The inability to alter tumors size of this treatment regimen could be due to its termination prior to the occurrence of vinyl carbamate-induced tumors in this model.13 The reduction in tumor size caused by just 2 weeks of treatment administered to mice with tumors would suggest a therapeutic effect on the tumors. A similar therapeutic response has been reported in the MNU-induced mammary model in which short-term treatment with targretin decreased the size of the tumors.11 Thus, targretin would appear to have a therapeutic effect for lung cancer in mice, supporting its use as such in clinical trials.
In the NNK-induced lung tumor model, a short duration of treatment for only 3 weeks early in the carcinogenic process (3–6 weeks after administering the NNK) significantly decreased the yield of tumors. In this study, targretin was administered for a total of 3 weeks followed by a holding period of 20 weeks. Since lung tumor multiplicity was decreased after the 20 weeks holding period, it would appear that in the NNK model, targretin had a large residual effect on tumor development. In the vinyl carbamate model, targretin was effective in decreasing tumor yield when administered 4–19 weeks after the carcinogen and followed by a 6-week holding period. However, it had a statistically less effect when administered for a longer duration of 21 weeks that lasted until the mice were killed. In the rat MNU-induced mammary cancer model, targretin's effect on tumor yield also demonstrated a limited residual effect after termination of treatment.11 That is when treatment ceased, mammary cancer multiplicity rapidly increased after a short delay. In contrast, in the NNK-induced lung tumor model targretin appeared to have a more significant residual effect especially on tumor multiplicity that was still decreased after a holding period of 20 weeks. The difference could be that the lung tumors induced by vinyl carbamate appear to more rapidly progress to cancer than those induced by NNK.13 Carcinomas begin to appear much more rapidly after vinyl carbamate, by half a year, than in the NNK model where a year or more is required. Thus, the vinyl carbamate lung model is more analogous to the rat mammary model where carcinomas also rapidly occur. This difference in the rate of progression to cancer might explain why targretin exhibited only a limited residual effect in both the rapidly progressing vinyl carbamate lung and MNU mammary models, while exhibiting a more extensive residual effect in the more slowly progressing NNK lung model.
Targretin caused a reversible increase in the liver to body weight ratio, found only when it was administered up to the time the mice were killed. The increase in liver weight would suggest that at the dose levels used targretin might alter its own metabolism and might be toxic. CYP26B is one of the retinoid metabolizing cytochrome P450s.15 Targretin increased the mRNA expression CYP26b in both lung and lung tumors obtained from Experiment 2 (unpublished results). Adverse events associated with retinoids including targretin include hyperlipidemia with elevation in both cholesterol and triglycerides.16 Targretin also increased the mRNA expression Apolipoprotein D in both lung and lung tumors obtained from Experiment 2 (unpublished results). Hence, at the dose levels used in our studies, targretin appears to induce its and other retinoids metabolism and to cause hyperlipidemia.
In summary, targretin was demonstrated to prevent lung tumors induced by 2 different carcinogens in mice, indicating it warrants further development as a chemopreventive agent for lung cancer. Targretin also demonstrated therapeutic activity in mouse lung tumors, as evident by its ability to reduce the size of established tumors that further supports its continuing development as a therapeutic agent for lung cancer.