Abstract
Aim: Mitotane is used in adrenal cancer as adjuvant therapy, monotherapy or combined with other cytotoxic agents in advanced disease, but only 30% of patients respond. The aim of this study was to define the structural requirements for drug activity and to develop analogs with improved adrenalytic action. Materials and Methods: Nine analogs of [1-(2-chlorophenyl)-1-(4-chlorophenyl)-2,2dichloroethane] (o,p'-DDD) were tested by measuring suppression of cortisol secretion and the presence of inflammatory changes in the dog adrenal and inhibition of cell proliferation and cortisol production by NCI-H295 human adrenal cancer cells. Results: In addition to mitotane, o,p'-DDClBr and o,p'-DDBr2, were active in vitro and in vitro: Their effects were comparable to that of o,p'-DDD when tested at 50 μM concentration, but o,p'DDBr2 was significantly more active at the lower 20 μM concentration. Conclusion: A dihalogenated methine carbon is required for adrenalytic activity. A change in the aromatic portion of the mitotane molecule causes loss of activity. Because of its greater activity at lower concentrations, o,p'-DDBr2 has potential application in the treatment of patients with adrenal cancer.
Adrenal cortical carcinomas are rare, highly malignant tumors that account for only 0.2% of cancer related deaths (1). Their incidence has been estimated at two per million people annually. About half of these tumors produce hormonal and metabolic syndromes that lead to their discovery. The other half is silent and is discovered when metastases develop or when the primary tumor becomes large enough to produce abdominal symptoms (1).
Mitotane is an adrenalytic drug that has been used for several decades in the treatment of patients with metastatic adrenal cortical carcinoma (ACC), either as monotherapy or in combination with other chemotherapeutic drugs (2-10). Renewed interest in the drug has led to additional clinical studies in recent years (11-16) but there is no consensus regarding its efficacy (17). Retrospective studies show that mitotane is effective as adjuvant therapy after surgical removal of the primary tumor in patients with stages I and II ACC (18) and a prospective control study (ADIUVO; University of Turin, Italy. Clinical Trial Government Identifier NCT 00777244) is currently underway. Mitotane has selective action on the adrenal cortex and when given in low doses, it affects predominantly the zona fasciculata and reticularis while sparing the glomerulosa (4). This specificity is related to its mechanism of action, which involves its transformation into active metabolites that covalently combine with specific targets in the cells leading to toxicity (19). There is evidence that mitotane is transformed to an acyl chloride via mitochondrial P450–mediated hydroxylation (20-22) and that the acyl-chloride covalently combines with specific bionucleophiles within the adrenal cortical cell for the adrenalytic effect to take place (23). In a series of reports (2), mitotane has been associated with partial or complete response in only 33% of patients with adrenal cancer. It is possible that although normal adrenal glands are consistently destroyed by mitotane, adrenal tumors vary in their ability to effect metabolic transformation, thereby expressing variable sensitivity to mitotane. Mitotane causes significant systemic toxicity in therapeutically effective doses. The adverse effects of mitotane are dose-dependent and usually intolerable when doses exceed 6 g daily, a dose that may be required to achieve therapeutic blood levels of 14–20 μg/dl. Treatment with mitotane inhibits hormone production and eventually causes necrosis of the contralateral non-tumorous adrenal gland.
The structure of mitotane [1-(2-chlorophenyl)-1-(4-chlorophenyl)-2,2dichloroethane] (o,p'-DDD) (Figure 2) comprises of an aromatic portion with chlorine substitutions in specific positions and a dichloromethane aliphatic portion. The objective of the present study was to study which of these structural features is required for the drug's activity and to develop agents with greater adrenalytic potential by a) substituting Cl at the β carbon with other halogens and b) substituting aromatic Cl in different positions in order to produce compounds that are more efficiently bioavailable to the adrenal cortex and more actively metabolized.
Materials and Methods
Compounds. Two analogs of o,p'-DDD, m,p'-DDD and p,p'-DDD, were purchased or synthesized as outlined in the literature (24). Other analogs were synthesized in our laboratory according to Scheme I (Figure 1a) for aliphatic substitutions, and Scheme II (Figure 1b) for aromatic substitutions. Nine analogs of o,p'-DDD were tested for purity and were shown to be greater than 95% pure by Nuclear Magnetic Resonance (NMR) and elemental analysis. Their structure is shown in Figure 2.
Hormone assays. Cortisol was measured by solid phase 125I radioimmunoassay with a sensitivity of 0.2 μg/dl (5.5 μM/dl) and a specificity of 97%. Plasma adrenocorticotropin (ACTH) levels were determined by radioimmunoassay by the method of Vague et al. (25), using a very sensitive and specific antibody at a dilution of 1:800,000. This antibody has low cross reactivity with other peptides and is capable of detecting ACTH in plasma, in concentrations of 6 pg/ml or more.
In vivo studies in dogs. o,p'-DDD or its analogs were screened for in vitro pharmacological activity in dogs. We used dogs because of their known sensitivity to o,p'-DDD. For the majority of the analogs one dog was used but multiple blood samples were drawn from each animal. The drugs were administered orally in sesame oil solution in capsule form to normal conditioned male mongrel dogs (Marshall Farms, North Rose, New York, USA) weighing 12-15 kg. The drug dose was 47 mg/kg daily for 6 days. Dogs receiving vehicle alone were used as controls. Blood samples for cortisol and ACTH were obtained every three hours from 8:00 to 20:00 h. Adrenal cortical functional reserve was tested with an ACTH stimulation test. Blood samples were obtained before and at 30, 60 and 90 min following iv injection of synthetic ACTH 1-24 (Cortrosyn). The dogs were adrenalectomized under anesthesia and subsequently euthanized on the day following the last drug dose. Samples of adrenal glands were fixed in 10% formalin and embedded in paraffin. Hematoxylin-eosin (H&E) stained sections were evaluated for evidence of tissue damage.
In vitro studies in adrenal cancer. The adrenalytic potential of mitotane and its analogs in human adrenal cancer was compared in vitro using a steroid hormone-secreting human adrenocortical carcinoma cell line, NCI-H295 (26) (American Type Culture Collection, Manassas, VA, USA). Cells were maintained at 37°C with 5% CO2 in RPMI 1640 medium with 5 μg/ml (34.7 μM/ml) insulin from bovine pancreas, 10 μg/ml (0.1 μM/ml) human transferrin, 2 mM glutamine, 10 mM hydrocortisone, 10 mM estradiol, 50 mM sodium selenite, 100 units/ml penicillin, 0.1 mg/ml streptomycin, 250 mg/ml amphotericin B and 5% fetal bovine serum. Cells were harvested for sub-culturing and for experiments using 0.05% porcine trypsin and 0.02% EDTA in Hank's Balanced Salt Solution. All medium components and trypsin/EDTA were obtained from Sigma Chemical Company (St. Louis, MO, USA). For evaluation of cell growth inhibition, 10,000 cells/well were distributed into 96-well plates and incubated for 7 days. The drugs were then dissolved in ethanol and added to the wells to obtain the desired concentrations. Incubation was continued for an additional 7 days, after which the medium was sampled for determination of cortisol concentration, using the assay described for serum. Cell growth was evaluated by measuring the relative protein content of the cell cultures using the sulforhodamine B assay (27), on the day the drugs were added and on the seventh day of drug treatment. The cortisol concentration and the protein content of the wells to which the drugs had been added were expressed as a percentage of the results obtained from control wells without drug.
Studies of cell proliferation were carried out in duplicate and cortisol concentration in the media was assayed in quadruplicate. Testing of the various analogs was repeated in multiple assays (2-10 times) at a concentration of 50 μM and the results were pooled. In addition, we conducted a head-to-head comparison of all the analogs in a single assay. Intra-assay coefficient of variation for cell proliferation was 4.4% and for cortisol was 12.6%; interassay coefficient of variation for cell proliferation was 7.5% and for cortisol was 18%. Dose-response curves were calculated for the most active analogs at drug concentrations of 5, 20, 35 and 50 μM. These assays were validated by comparison with other adrenal inhibitors without adrenalytic effect. While markedly suppressing cortisol, metyrapone and aminoglutethimide had no effect on cell proliferation.
Animal studies were conducted in the Unit for Laboratory Animal Medicine of the University of Michigan (ULAM) and were approved by the University Committee on Use and Care of Animals (UCUCA).
Statistical analysis. For the purpose of statistical analysis, each of the duplicate wells was considered a replicate and each of the quadruplicate cortisol samples was also considered a replicate. Using a general linear mixed model with clustered units (replicated wells or samples clustered within assay to account for correlation within well or sample), we tested the null hypotheses that percentage change in cell proliferation and cortisol secretion under each analog was significantly different from the percentage change in cell proliferation and cortisol secretion under o,p-DDD. We also used a general linear model to test whether the percentage change in cell proliferation or cortisol secretion, at four concentrations of analogs (5, 20, 35 and 50 μM),was significantly different from the percentage change under o,p'-DDD at the same four concentrations. All testings were performed at a nominal significance level of 0.05.
Results
In vivo studies in normal dogs. The ability of the o,p'-DDD analogs to suppress adrenal function was compared in vitro. Only analogs 2, 3, 4, 6 and 7 caused suppression of cortisol secretion and concomitant increase in ACTH levels (Figure 3). As shown in Figure 4, analogs that actively suppressed cortisol secretion also abolished the adrenocortical response to ACTH. The histological changes observed in dog adrenals following treatment with mitotane and its analogs was expressed as the ratio of zona glomerulosa to zona fasciculata plus reticularis and as an indicator of adrenal cortical atrophy induced by the mitotane analogs. A higher value indicates greater atrophy. The thickness of the zona glomerulosa was standardized at 0.2 mm for all compounds. We used the zona glomerulosa for reference because mitotane selectively causes necrosis of the zona reticularis and fasciculata but spares the glomerulosa (4, 28) and patients treated with mitotane do not require mineralocorticoid replacement except for cases of chronic, long standing treatment. The adrenalytic effect of the most active compounds was characterized by atrophy of the zona fasciculata and reticularis (high zona glomerulosa/zona fasciculata plus zona reticularis ratio) and an intense inflammatory reaction with leukocyte infiltration. The density of inflammatory cells was graded from 1 to 4+. Of these, o,p'-DDClBr with a ratio of 0.33 and o,p'-DDBr2 with a ratio of 0.28, were the most active with 2-3+ inflammatory response; in contrast, o,p'DDD had a ratio of 0.16 and 2+ inflammatory response. In contrast, inactive compounds did not cause significant histological changes.
In vitro cancer studies. The suppression of cell proliferation and cortisol secretion by o,p'-DDD analogs was tested in multiple assays on adrenocortical carcinoma cells in vitro. Drugs were used at a concentration of 50 μM in 2% Bovine Serum Albumin (BSA). Using a linear mixed effects model with cell proliferation or cortisol as the outcome variable and the analog as the sole covariate, o,p'-DDClBr and o,p'-DDBr2 were comparable in activity with o,p'-DDD while the other analogs were significantly less active (p<0.05). Of interest was the ability of some of these analogs to suppress cortisol secretion without influencing cell growth. This was most notable with compound 9, NH2DiBr, which did not affect cell growth but markedly suppressed cortisol secretion. In a head-to-head comparison with o,p'-DDD (Figure 5), the two bromine substituted analogs exhibited comparable activity at a dose of 50 mM but in a dose response comparison (Figure 6) o,p'-DDBR2 appeared to be significantly more active (p<0.05) at lower concentrations. The half maximal inhibitory concentration (IC50) for all analogs was also calculated. For each compound, the cell proliferation and the cortisol concentration as percentage that of the control were determined at 5, 20, 35 and 50 μM in two to four independent experiments. The mean values obtained at each concentration were plotted against the log of the concentration-response curve, and the IC50 values were determined graphically from this curve. IC50 values for cell proliferation were 16, 19 and 12 μM for o,p'-DDD, o,p'-DDCl,Br and o,p'-DDBr2, respectively and for cortisol production 12, 16 and 11 for the three analogs respectively. In contrast, IC50 for the other six analogs were 25-36 μM for cell proliferation and 27-37 for cortisol production. The exception was mNH2, p'Cl-DDBr2 whose IC50 was 25 μM for cell proliferation and <5 for cortisol production. We also tested as control three well-known adrenal inhibitors (ketoconazole, aminoglutethimide and metyrapone in doses up to 100 μM, without effect on cell proliferation but with very low IC50 values (1.25-2) for suppression of cortisol secretion.
Discussion
Nelson and Woodard (29) first reported that commercial DDD can produce adrenocortical atrophy and others reported (30, 31) that DDD inhibits 17-hydroxycorticosteroids in dogs. Subsequently, mitotane, the o,p'-isomer in the mixture was found to be active, as an adrenalytic agent, and useful in the treatment of patients with adrenal cancer. The drug has been tried in a patient with a malignant Leydig cell tumor (32) with significant response, but there have not been any reports of mitotane activity against other solid tumors. This specificity for the adrenal cortex and for adrenal carcinoma may indicate a requirement for biotransformation of the drug for activity in ways that differ from those taking place in extra-adrenal sites (33, 34). Early studies by Martz and Straw (23) showed that dog adrenal mitochondria metabolize mitotane to reactive products that covalently bind to mitochondrial macromolecules. The level of metabolism and covalent binding of 14C o,p'-DDD was species-specific, with the dog being the most sensitive and the rat the least sensitive species. A review of 26 DDD analogs tested in dogs suggested that cytotoxic adrenal effects were most consistently associated with a dihalogenated ethane structure. Substitutions in the benzene rings were less critical in determining adrenalytic effects (35-40). Using a cell line derived from a feminizing adrenal cortical carcinoma, Fang (41) tested several analogs of o,p-DDD and concluded that a dichloro- or trichloroethylene structure was essential for cytotoxic activity while chlorine substitutions on the phenyl rings appeared to be unimportant. Studies from our laboratory differ from those conclusions. Using a methylated analog of o,p'-DDD, we have demonstrated in vitro that C-H needs to be present at the C-l position of the ethane portion of o,p'-DDD in order to obtain adrenalytic effects (42). When the hydrogen atom is substituted by a methyl group, the drug loses its adrenalytic effect. The drug is also inactive if hydrogen is replaced by chlorine as in the trichlorinated compound DDT. Studies from our laboratories have demonstrated metabolic oxidation of both the aromatic and aliphatic portions of o,p'-DDD in rats and humans (43, 44). In addition, in vitro studies have shown aliphatic oxidation in bovine adrenal incubation as well as aromatic hydroxylation and side chain oxidation in perfusion studies of dog adrenals (45). Incubation studies with dog adrenal homogenates support the presence of these metabolites and indicate that the direct oxidation of the dichloromethyl moiety of o,p'-DDD is more important than the oxidation of the unsaturated intermediate in the formation of DDA and its derivatives (33). The process of bioactivation of mitotane within the adrenal cortex may be similar to that of other chloroethane compounds. In the case of dihalo substituted carbons, acyl halides would be the anticipated product. Studies in mice and hamsters suggest this mechanism for the formation of DDA from 1,1-dichloro-2,2bis(p-chlorophenyl) ethane (p,p'-DDD) (47). Another example of this type of metabolic processing, is the metabolic activation of chloramphenicol (CAP), a compound with partial structural homology with o,p'DDD (47) and a similar mechanism, which appears to apply to compounds such as halothane and chloroform. Based on o,p'-DDD studies reported in the literature and our own work, it is likely that the adrenalytic action of o,p'-DDD is linked to the ability of the dichloromethyl group to become metabolically oxidized to an active species, possibly an acyl chloride, which covalently binds to metabolically important macromolecules in the adrenal cortex. This metabolic oxidation takes place in the mitochondria and is mediated by a cytochrome P450 hydroxylating enzyme, most likely a P450c11 isozyme involved in xenobiotic metabolism (48).
An objective of the present series of studies was the synthesis of agents with improved adrenalytic action. Our previous studies with mitometh (42) strongly suggested that the action of o,p'-DDD involves metabolic conversion of the dichloromethyl moiety to an acylated species but nothing is known about the influence of the type of halogen on this process. We addressed this question by synthesizing bromochloro and dibromo analogs and comparing them with o,p'-DDD.
Another property of o,p'-DDD which limits its action is its bioavailability. Gastrointestinal absorption is low and variable depending on the vehicle in which it is administered. As a consequence, blood levels achieved are very low, and large doses (4-10 g/day) are required in order to reach therapeutic drug concentrations. This poor absorption could be attributed to its extremely low water solubility. It is well known that for maximum absorption, a balance between lipophilicity and hydrophilicity is required. In order to address the question of bioavailability, we attempted to synthesize analogs which are less lipophilic by replacing aromatic chlorine (n Cl=0.76) with amino groups (n NH3=−1.23). We chose amino groups not only because of their hydrophilic nature but also because an aminophenyl moiety is a common feature of several drugs with adrenal suppressive activity (aminoglutethimide, amphenone B).
Nine analogs of o,p'-DDD were tested. In vivo, five were active and suppressed adrenal function within 48 hours of the onset of treatment. The brominated compounds were the most active: cortisol decreased to 1-2 μg/dl (27.59-55.18 μM), ACTH increased to 10-fold baseline levels and the cortisol response to ACTH stimulation was abolished. Histologically, early inflammatory cellular infiltrates were noted in the innermost levels of the adrenal cortex. In contrast, substitutions on the aromatic portion of the molecule resulted in variably lower activity. This is at variance with early reports (34-40) suggesting that aromatic substitutions were not critical for activity. The position of the chlorine substitutions in the aromatic portion of the molecule alters the activity of the drug without alterations in the aliphatic portion of the molecule, the one postulated to be the active site for formation of the acyl chloride, the required intermediate. The possibility exists that these aromatic substitutions reduce the adrenal uptake of the compound. When given in equimolar doses to other active compounds, p,p'-DDD and NH2Cl-DDD resulted in much lower adrenal concentration. When the drugs were used directly in contact with adrenal cells in vitro, the in vitro data were confirmed for the dibromo- and the bromochloro-derivatives as well as for the m,p'-DDD. However, in contrast to the in vitro data, mitometh suppressed both cortisol and cell doubling in vitro. Since mitometh is not metabolized to the acyl chloride, we need to consider a different mechanism, possibly production of free radicals, for the in vitro action of mitometh. Studies from our laboratory (unpublished) indicate that this effect of mitometh can be reversed with tocopherol acetate, an antioxidant, in a dose-dependent fashion.
We conclude from these studies that a dihalogenated methine carbon is required for both in vitro and in vitro adrenalytic activity of o,p'-DDD. While in vitro activity does not always translate into in vitro activity, o,p-DDBr2 could have potential application in the treatment of adrenal cancer. While comparable with the effects of o,p'-DDD at higher concetrations, o,p'-DDBr2 is significantly more active at lower concentration and it may have advantage over mitotane in terms of a lower toxicity:activity ratio.
Acknowledgements
We are grateful to M. Daniel Schteingart for his assistance with the illustrations.
This study was upported by grant NIH 5 R01 CA 37794-0
- Received April 2, 2012.
- Revision received April 28, 2012.
- Accepted April 30, 2012.
- Copyright© 2012 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved