A One-Year Neonatal Mouse Carcinogenesis Study of Quinacrine Dihydrochloride

Abstract

Quinacrine is an acridine derivative under investigation for its use in nonsurgical female sterilization. Safety issues regarding the carcinogenic potential of quinacrine have been raised because it is mutagenic and clastogenic in vitro. The objective of the study was to evaluate the carcinogenic potential of quinacrine dihydrochloride (quinacrine) in neonatal mice treated with single intraperitoneal doses on postpartum days 8 and 15 and observed for 52 weeks. Neonatal Crl: CD-1 mice of each sex were randomly allocated into four treatment groups (0, 10, 50, and 150 mg/kg), dosed twice with quinacrine suspended in carboxymethylcellulose, observed for 52 weeks post dose, and then euthanized, necropsied, and subjected to a full histopathological examination. In male mice, tumor incidence was not significantly increased at any site at any dose level. In female mice, the incidence of benign uterine endometrial stromal polyps was slightly greater at the mid and high dose (≥50 mg/kg), as was the incidence of endometrial hyperplasia. The incidence of polyps in these groups was not significantly greater than in controls by pair-wise comparison but was significantly greater (p = .042) by the linear trend test. The authors conclude that quinacrine administered twice to neonatal mice may have enhanced or accelerated the development of endometrial hyperplasia and uterine stromal polyps at higher doses. Because uterine stromal polyps are a commonly observed benign tumor in older mice, the significance of this finding is unclear and will require a weight of evidence evaluation for a conclusion on the carcinogenic potential of quinacrine.

Quinacrine dihydrochloride (atabrine, mepacrine, quinacrine; CAS no. 6151–30–0) is a chemical that has been widely used for several purposes. Quinacrine was used extensively during World War II as an oral antimalarial agent. It continued to be used for this indication until it was replaced by more potent and newer antimalarial drugs with better clinical profiles and its U.S. marketing application was discontinued in 1994. In 1964, an injectable form of quinacrine was approved by the U.S. Food and Drug Administration (FDA) (New Drug Application 15–052) for the treatment of recurrent ascites associated with several types of cancer, but the manufacturer withdrew it from the U.S. market in 1997. Currently there are no commercially available formulations of quinacrine in the United States or Canada. Quinacrine remains available by prescription from compounding pharmacies in the United States under the FDA Modernization Act of 1997, section 503A(b)(1)(A)(i), and it is used as a second-line treatment for autoimmune diseases such as systemic lupus erythematosus and giardiasis.

However, quinacrine continues to be used as an investigational drug for its potential use for other clinical indications (Lippes et al. 2003; Nakajima et al. 2004; Toubi et al. 2003). Of particular interest is the use of quinacrine in nonsurgical female sterilization procedures. Due to its potent irritant and sclerosing activity, quinacrine has been used as a contraceptive method where it causes female infertility after transcervical instillation. More than 100,000 women in developing countries have undergone this sterilization procedure (Kessel 1996), and physicians in the United States offer quinacrine sterilizations to private patients with individual prescriptions filled by compounding pharmacists (Whitney 2003).

Considerable controversy remains about the appropriateness and safety of quinacrine use for female nonsurgical sterilization procedures. Quinacrine is an acridine derivative that belongs to a group of DNA-intercalating agents, which bind to DNA and intercalate between DNA base pairs (Lerman 1963; Aslangolu and Ayne 2004). It is this interaction with DNA strands that has raised questions about the mutagenic and carcinogenic potential of quinacrine.

Several studies have shown that quinacrine is genotoxic in vitro in the Salmonella typhimurium assay with and without metabolic activation (Young et al. 1981; Blake et al. 1983; Zeiger et al. 1988). An 18-month study of quinacrine in albino rats of the Osborne-Mendel strain (Fitzhugh, Nelson, and Calvery 1945) showed that quinacrine administered in the low-and high-protein diet (100, 200, 400, and 800 ppm) produced marked liver cirrhosis, raising concern for quinacrine’s carcinogenic potential. Many of these older published studies do not meet current methodological standards. Recent studies, conducted to meet current methodological standards, confirmed that quinacrine is genotoxic in vitro. Clarke et al. (2001) published a battery of genetic toxicology studies verifying that quinacrine is mutagenic and clastogenic in vitro. Quinacrine was mutagenic in the S. typhimurium tester strains TA1537 with and without S9-metabolic activation, TA98 with S9-metabolic activation, in the Escherichia coli WP2 uvrA tester strain without S9-metabolic activation, and in the mouse lymphoma assay in the absence of S9-metabolic activation. This same study showed that quinacrine was not mutagenic in the S. typhimurium tester strains TA100 and TA1535 and in the mouse lymphoma assay with S9-metabolic activation.

Quinacrine was clastogenic in Chinese hamster ovary (CHO) cells, with and without S9-metabolic activation. In contrast, results from the in vivo mouse micronucleated erythrocyte (micronucleus) assay indicate that quinacrine is not clastogenic in vivo. Nevertheless, the positive mutagenicity and clastogenicity results from in vitro studies suggest a potential cancer risk to humans. In addition, although the reported human experience to date has not suggested any carcinogenic potential, the information provided in these studies is not sufficient to adequately determine the safety of quinacrine with respect to carcinogenic potential.

As part of a program to assess the carcinogenic potential of quinacrine, a neonatal mouse carcinogenicity study was conducted in which neonatal mice received intraperitoneal doses of quinacrine on postpartum days 8 and 15 and were observed for 1 year. The neonatal mouse assay was performed because it has been shown to identify transpecies genotoxic carcinogens from different chemical classes and has been frequently used to determine the metabolic activation pathways of a variety of DNA-reactive chemical carcinogens (Fu et al. 2000). The utility of the neonatal mouse model for detecting potential human carcinogens was evaluated extensively as part of a collaborative program sponsored by the International Life Sciences Institute (ILSI) to identify potentially useful alternatives to the conventional 2-year rodent bioassay (Robinson and MacDonald 2001; McClain et al. 2001). Pharmaceutical regulatory agencies in Europe (Committee for Human Medicinal Products, CHMP) and the FDA currently consider the neonatal mouse assay to be acceptable to evaluate the carcinogenic potential of compounds that are clearly or equivocally genotoxic (MacDonald et al. 2004).

This paper describes the results of a 52-week neonatal mouse study that was conducted to add to the weight of evidence in defining the toxicity and carcinogenic potential of quinacrine. Because this study was intended to support clinical trials and marketing applications for the use of quinacrine as a human drug, the work described herein was conducted in compliance with Good Laboratory Practice Regulations generally consistent with U.K. Statutory Instrument 1999 No. 3106 as amended in 2004, Organization for Economic Cooperation and Development (OECD) Principles of Good Laboratory Practice issued in 1998, and the FDA code of federal regulations 21CFR Part 58.

MATERIALS AND METHODS

Mice

All procedures carried out on live mice as part of this study were subject to the provisions of United Kingdom National Law, in particular the Animals (Scientific Procedures) Act of 1986. Crl:CD-1(ICR)BR mice were used for this study. A minimum of 36 females with litters were obtained from Charles River (UK) (Margate, England). The offspring were 3, 4, and 5 days post partum on arrival. A second group of eight females with litters were obtained from the same vendor for use as replacement animals in case of unexpected mortality. They arrived 3 ½ weeks after the other animals had received their first dose on day 8 post partum.

Housing

Mice with litters were housed in a single, exclusive room that was air-conditioned to provide a minimum of 15 air changes/hour. The temperature and relative humidity were maintained in the ranges of 19°C to 25°C and 40% to 70%, respectively. Fluorescent lighting was controlled automatically to give a cycle of 12-h light (0600 h to 1800 h) and 12-h dark. The females with litters were housed in solid-floored polypropylene cages with soft wood chips (Datesand Sale) as bedding. Each batch of wood chips was analyzed for contaminants. No contaminants were present in the wood chips at levels that might have interfered with achieving the objective of the study. Cages were cleaned and soft wood chips changed weekly.

Dosed pups were housed as mixed sex groups up to weaning on day 18 post partum. Dosed mice were then housed in stainless steel mesh cages of size 33 × 15 × 13 cm, floor area 495 cm² in groups of four per sex (i.e., each cage contained four males or four females) for the remainder of the study. Animals were singly housed when problems arose with fighting among multiple-housed animals. Each cage was provided with wooden chew blocks as a means of environmental enrichment/welfare improvement.

Animal Identification

Following weaning, all animals were identified by subcutaneous electronic implant.

Feed and Water

Mice had access ad libitum to SQC Rat and Mouse Breeder Diet No. 3, Expanded, (Special Diets Services, Witham) during the preweaning period and to SQC Rat and Mouse Diet No. 1, Expanded (Special Diets Services), thereafter. Water from the municipal supply was provided ad libitum via bottles that were rinsed and refilled daily. The water and diet were analyzed and found free of specific contaminants at levels that might have interfered with achieving the objective of the study.

Prior to allocation to dose groups, pups considered unlikely to survive and pups at the extremes of the weight range within each litter were humanely culled and euthanized so that all litters considered suitable for study allocation contained four male and four female pups. Pups then were then allocated by litter to groups of 28 mice/sex using a cluster randomization procedure. The entire litter was assigned to the same dose group. Thus, each dose group initially consisted of seven litters, each of which contained four males and four females to the dose group.

Test Materials

Quinacrine dihydrochloride dihydrate (98.86% pure), a yellow powder, was manufactured by Vipor Chemicals, India, and supplied by SiPharm Sisseln AG, Switzerland. Quinacrine was stored in a sealed brown glass container in a controlled environment, at room temperature (10°C to 30°C).

The control article and the vehicle used to prepare quinacrine dosing suspensions was 1% (w/v) carboxymethyl cellulose supplied by Sigma-Aldrich (Poole, UK) suspended in 0.9% saline supplied by Fresenius (Basingstoke, UK).

Dosing Suspensions

Quinacrine dosing suspensions were prepared under yellow fluorescent light on each day of administration by suspending quinacrine in 1% (w/v) carboxymethyl cellulose in 0.9% saline. Dosing suspensions were stored in amber glass bottles, protected from light. Dosing was completed within 6 h of preparation of the dosing formulations. The concentration of quinacrine in each dosing formulation was calculated on the basis of group mean body weights in each dose group on days 8 and 15 post partum, so that doses were administered at 10 μl/pup on day 8 and 20 μl/pup on day 15.

To verify the quinacrine concentration in dosing suspensions, duplicate samples (5 ml) were taken from suspensions prepared on each day of dosing and analyzed for quinacrine concentration. For control formulations only a single analysis was performed.

To verify that quinacrine dosing suspensions were homogenous as prepared, trial quinacrine dosing suspensions were prepared at concentrations of 4.0 and 75.0 mg/ml to cover the lowest and highest concentrations used in the study. Three samples were removed from the top and three from the bottom of each formulation and analyzed for quinacrine concentration.

To verify that quinacrine was stable as formulated and handled in this study, the trial formulations were allowed to stand in the testing facility’s Formulations Analysis Laboratory at room temperature in a brown glass bottle, protected from light, and then each formulation was sampled and analyzed for quinacrine concentration.

Study Design

Dosing

The neonatal mice were treated with a single intraperitoneal injection of quinacrine dihydrochloride dihydrate (quinacrine) suspended in methylcellulose on postpartum days 8 and 15, observed for 52 weeks post dose, and then euthanized, necropsied, and subjected to a full histopathologic examination. The nominal dose levels were 0 (methylcellulose vehicle), 10, 50, or 150 mg/kg and the dose volume was 10 to 20 μl/mouse.

When the high-dose group began treatment with a second dose of quinacrine at 150 mg/kg on day 15 post partum, there was more mortality than expected, so the following steps were taken:

The dose level for the second dose administered on postpartum day 15 was reduced to 100 mg/kg.

Backup litters were available for use in the event of increased mortality. Eight more male mice (two litters of four) were added to the high-dose and control groups at day 8 post partum to replace those that had died, so that sufficient number of males would enter the 52-week observation phase of the study to allow adequate evaluation of carcinogenic potential. The two additional male litters added to the high-dose group (group 4) received a dose of 150 mg/kg and 100 mg/kg on days 8 and 15 post partum, respectively.

Because some mice that had received a second dose of 150 mg/kg survived, the original high-dose group (group 4) ended up containing mice treated with two doses on postpartum days 8 and 15 of 150 mg/kg (group 4, subgroup A) and mice were treated with a dose of 150 mg/kg and a dose of 100 mg/kg, on postpartum days 8 and 15, respectively (group 4, subgroup B).

Postdose Culling of Litters and Final Group Sizes

One extra litter of dosed pups (males and females) was available for replacement of dead pups, to ensure that sufficient number of animals entered the study. Litters were weaned on post-partum day 18 or 21. Additional pups not used were culled. The addition of pups to the control and high-dose groups resulted in uneven group sizes. The final distribution of male and female mice that entered the 52-week observation period is summarized in Table 1.

Health Screening

Two litters containing four male and four female neonates per litter were euthanized soon after weaning. Mice were necropsied and examined for gross pathologic findings, and kidneys, liver, lungs, and gross abnormalities were processed and examined for histopathologic findings.

In Life Observations

Morbidity, Mortality, Clinical Signs, and Physical Examinations

Mice were examined twice daily, generally about 6 h apart, for signs of ill health or overt toxicity, and any observations were recorded. In addition, each mouse was given a detailed physical examination each week that included palpation for tissue masses. As quinacrine is known to be an irritant, particular attention was paid to the condition of the dosing sites.

Body Weight and Food Consumption

Body weights of all pups were recorded on day 5 post partum, and those of all dosed pups were recorded on days 8, 15, and 21 post partum and at approximately day 28 post partum. After day 28 post partum, individual body weights were recorded once every 2 weeks and at (or immediately before) necropsy.

The amount of food consumed by each cage of mice was determined 1 week in every 2, beginning the first full week after day 28 post partum. Weekly food consumption was calculated by weighing the food before placing in each cage and then weighing the residual diet after 7 days. Individual values for mice were calculated by dividing the total food consumed per cage by the number of mice in the cage.

Necropsy

Dosed pups found dead or moribund prior to weaning were given necropsied and all macroscopic findings were recorded. The carcass was retained in 10% neutral buffered formalin, but tissues were not processed for histopathologic evaluation. Mice found dead after weaning and prior to scheduled necropsy were necropsied and all macroscopic findings were recorded. Dependent upon the condition of the carcass, all protocol specified tissues, in particular liver, lungs, and gross abnormalities were processed and examined for histopathologic findings. Mice determined to be in moribund condition after weaning prior to scheduled necropsy were humanely euthanized by an intraperitoneal sodium pentobarbitone injection, followed by exsanguination and then necropsied, and all macroscopic findings were recorded. Blood samples were collected where possible and stored for examination if a blood disorder was suspected. All protocol-specified tissues were processed and examined for histopathologic findings.

Mice that survived to the scheduled terminal necropsy were fasted overnight and then humanely euthanized and exsanguinated. At necropsy, fasted body weights were recorded; blood smears and bone marrow smears were made. A full necropsy was performed under the general supervision of a pathologist and all gross pathologic findings were recorded. Selected organs were weighed, and tissues were collected and preserved for subsequent histopathologic evaluation. Brain, heart, kidney, liver, lungs, ovaries, and testes were dissected free from fat and other contiguous tissue and weighed before fixation. Paired organs were weighed together.

Histopathology

Preserved tissues were processed by conventional histologic techniques, sectioned, stained with hematoxylin and eosin, and examined by light microscopy, and histopathologic observations were recorded. A comprehensive list of organs and tissues from all mice were collected at necropsy and preserved in the appropriate fixative. These include eyes, optic nerves, harderian glands, skin, mammary glands, muscle (quadriceps), femur with bone marrow and articular surface, sciatic nerves, liver, spleen, pancreas, mesenteric lymph nodes, stomach, duodenum, jejunum, ileum, caecum, colon, rectum, gall bladder, adrenals, kidneys, testes and epididymides, ovaries, seminal vesicles, urinary bladder, prostate, uterus, vagina, salivary glands, mandibular lymph nodes, thymus, lungs, heart, aorta, trachea, oesophagus, tongue, thyroids and parathyroids, larynx, pituitary, brain, spinal cord (cervical, thoracic, and lumbar), lacrimal glands, zymbal glands, bronchial lymph nodes, trachea bifurcation, head, nasal turbinates, nasopharynx, tissue masses, animal identification site, and gross lesions.

The study pathologist examined the following tissues by light microscopy for neoplastic and non-neoplastic histopathologic findings:

All gross pathologic lesions and tissues from mice that died or were euthanized before the end of the study with the exception of mice found dead during dosing (days 8 and 15 post partum) and prior to weaning.

All tissues from control and high-dose mice euthanized at the end of the study.

Gross lesions and tissue masses from all mice.

Uterus from all female mice.

Liver and lung from all groups.

Histopathologic findings and selected slides were reviewed by external peer-review pathologists. The final histopathologic data reflect the consensus of the study pathologist and peer-review pathologists.

Selection of Dose Levels

The initial dose levels for each group were selected based on the results of two dose range–finding studies in which neonatal mice were dosed by intraperitoneal injection on days 8 and 15 post partum and subsequently monitored for a period of 4 weeks. Significant morbidity/mortality was observed at doses ≥189 mg/kg. On the basis of these results, the maximum tolerated dose level (MTD) for quinacrine in neonatal mice dosed by intraperitoneal injection on days 8 and 15 post partum was estimated to be 150 mg/kg. This was used as the high-dose level for the neonatal mouse carcinogenicity study. The intermediate-(50 mg/kg) and low- (10 mg/kg) dose levels for the carcinogenicity study were selected to give a reasonable spread between doses.

Statistical Analyses

Male and female data were analyzed separately. Body weight gains and necropsy body weights were analyzed using a mixed model analysis of variance (ANOVA), fitting group as a fixed effect and dam as a random effect. Pairwise comparisons with control were made using Dunnett’s test. A linear contrast was used to determine whether there was a relationship between increasing dose and response. Food consumption was analyzed using one-way ANOVA. Pairwise comparisons with control were made using Dunnett’s test. A significant trend (p < .05) was only reported where none of the pairwise comparisons was significant.

Organ weights were analyzed using a mixed-model analysis of covariance (ANCOVA), fitting necropsy body weight as covariate, group as a fixed effect, and dam as a random effect. This analysis depends on the assumption that the relationship between the organ weights and the covariate is the same for all groups (Covariate × Group interaction in the ANCOVA model), and the validity of this assumption was tested. Pairwise comparisons with control were made using Dunnett’s test.

Survival probability functions were estimated by the Kaplan-Meier technique. Survival curves were compared to the start of the terminal kill phase. Permutational tests for both an increasing and a decreasing dose response in mortality were performed across all groups using the starting dose levels as weighting coefficients, in accordance with the IARC annex (Peto et al. 1980). One directional pairwise tests of the treated groups against the control group were also performed.

The numbers of tumor-bearing animals were analyzed for tumor types found in at least three animals of the given sex. Analysis was done across groups. Tumors of similar histogenic origin were merged, as requested by the Pathologist. For tumor types examined in all groups, permutational tests for both an increasing and a decreasing dose response were performed across the groups using the starting dose levels as weighting coefficients, in accordance with the IARC annex. One directional pairwise permutational tests of the treated groups against the combined control groups were also performed. For tumor types examined only in control and high-dose, one directional pairwise permutational tests were performed.

Nonfatal tumors were analyzed using fixed intervals of 1 to 53 weeks and the terminal kill phase. The fatal and non-fatal results were combined in accordance with the IARC annex. No pathology was performed for the 11 male and 5 female decedents from group 4 whose deaths were associated with administration of the second dose. These animals were excluded from the survival analysis.

RESULTS

Health Screening of Mice

Among the mice culled for health screening purposes, there were no macroscopic findings in any tissue at necropsy and no histopathologic findings in the kidneys, liver, or lungs that suggested a disease process likely to affect the integrity of the study.

Dosing Suspension Analyses

Quinacrine dosing suspensions were considered homogeneous because the coefficient of variation (CV) of the measured concentrations was 6.0%, and all the homogeneity results were within ±10% of the mean. Quinacrine was considered stable in dosing suspensions during the 24-h storage period because measured concentrations did not change appreciably during that time. Measured concentrations were within 90% to 110% of nominal except for two low-dose replicates and one mid-dose replicate sample from suspensions prepared on day 1, for which results slightly greater than 110% were achieved. Quinacrine was not detected in the vehicle used to dose control mice.

Mortality

Of the original group of 28 males and 28 females dosed at 150/150 or 150/100 mg/kg on post natal day 15, 9 males and 5 females died during or shortly after dosing. Of the additional group of eight male mice dosed with 150/100 mg/kg on postnatal day 15, two died during dosing. All deaths occurred within 8 h post-dose, and all were considered to be quinacrine-related.

There was no indication for a quinacrine-related increase in post dosing/weaning morbidity and mortality. Of the mice that entered the 52-week observation period, there were five deaths, none of which were considered to be quinacrine-related: one control female, one low-dose male, one mouse of each sex in the mid-dose group, and one high-dose female.

Clinical Signs and Palpable Masses

The only clinical sign that was attributable to administration of quinacrine was hair loss at the injection site seen in approximately 48% of mice at 50/50 mg/kg and 89% of mice dosed at 150/150 or 150/100 mg/kg. All other clinical signs (e.g., fur staining, thinning fur, hair loss, incidence of injuries from fighting) were consistent with the expected pattern of clinical signs in laboratory-maintained mice of this strain and age and unrelated to quinacrine dose level, and were considered to be unrelated to quinacrine administration. The incidence of palpable masses was low and unrelated to quinacrine dose level.

Body Weight, Organ Weight, and Food Consumption

Quinacrine did not affect body weight or food consumption during the study, or the weight of the brain, heart, kidney, liver, lungs, ovaries, or testes at the end of the study.

Macroscopic Findings

There were no macroscopic findings at necropsy suggestive of adverse effects of quinacrine administration. Macroscopic findings were consistent with the normal spectrum of background lesions in this strain of mice and were generally similar in control and quinacrine-treated mice.

Non-Neoplastic Histopathologic Findings

Table 2 summarizes the non-neoplastic histopathologic findings of selected tissues of mice surviving the 52-week observation period. Non-neoplastic histopathologic findings were generally of a minor nature and consistent with the expected spectrum of findings in mice of this strain and age (Mohr et al. 1996). In males there were no non-neoplastic histopathologic findings considered to be quinacrine related. In females, there was a slightly greater incidence of endometrial hyperplasia in the uterus of intermediate and high-dose mice, when compared with the controls as shown in Table 2.

Neoplastic Histopathologic Findings

The spectrum of tumors in control and quinacrine-treated mice was generally consistent with that expected in mice of this strain and age (Table 3). In males, the incidence of tumors was generally similar across the groups, and there were no unusual neoplastic findings considered to be quinacrine related. In females, the incidence of tumors also was similar across groups, except for uterine stromal polyps that were noted in a few mice dosed with quinacrine in mid and high doses, as shown in Table 3. The increased incidence of stromal polyps was statistically significant (Table 4) by the linear trend test (p = .042) but not significant by pairwise comparison of either group with the controls (p = .054 and p = .11 at the mid- and high-dose levels, respectively).

DISCUSSION

The neonatal mouse carcinogenicity bioassay was first described in 1959 and has been extensively evaluated for its sensitivity to genotoxic carcinogens (McClain et al. 2001). The International Conference on Harmonization (ICH) of Technical Requirements for the Registration of Pharmaceuticals for Human Use listed the use of the neonatal mouse bioassay as an alternative model for evaluating the potential carcinogenicity of human drugs when used with one long-term (2-year) rodent carcinogenicity study.

A positive tumorigenicity response in the neonatal mouse bioassay is generally considered as evidence for a genotoxic carcinogenic effect. The greater sensitivity to genotoxic carcinogens in neonatal mice compared to adults is attributed to the greater rate of cell proliferation and DNA replication in young rapidly growing mice. A positive tumorigenic response in the neonatal mouse model usually occurs as an increased incidence of benign and malignant liver and lung tumors in mice of both sexes.

There are several factors that might have affected the response of neonatal mice to quinacrine in this study, such as the mouse strain used, the dose levels selected (particularly the highest dose level), age at dosing, and route of administration (Flammang et al. 1997). In addition, the chemical nature of the administered chemical may affect the absorption and distribution of the chemical to target and non target tissues.

In this study we used a standardized protocol developed by the ILSI collaborative project in 2001 to control some factors that affect the response of the neonatal mouse bioassay so that we could compare our results to those obtained with other drugs and chemicals. The ILSI protocol calls for using a high-dose level that is either the maximum feasible dose (MFD) or the MTD determined in a pair of dose range–finding studies. In our study, the high-dose level that we initially selected exceeded the MTD, based on the high acute mortality of males after the second dose on day 15 post partum. We also used the CD-1 strain of mice, which is one of the most commonly used strains in this bioassay and which is highly sensitive to the development of liver and lung tumors (Fu et al. 2000).

Quinacrine is rapidly absorbed after oral administration and is likely to be found in high concentrations in body tissues. Of importance to this study is that quinacrine has been shown to accumulate mostly in liver and lungs after oral ingestion or intravenous administration (Shannon et al. 1944; Dubin et al. 1982; Yung et al. 2004). High concentrations of quinacrine can also be found in kidneys and spleen after oral administration (Yung et al. 2004) and adrenal glands after intrauterine administration (Dubin et al. 1982). Based on this information, it seems likely that our dosing regimen produced substantial exposure to quinacrine in the target organs for this bioassay.

Because one of the purposes of this investigation was to evaluate the potential carcinogenic risk associated with the intrauterine administration of quinacrine to adult women as a means of nonsurgical sterilization, we evaluated the uterus as a possible target organ in addition to the lung and liver. In fact, the uterus was the only organ in which we saw a potential quinacrine-related effect, in the form of an increase in the relatively high background incidence of uterine endometrial hyperplasia in groups dosed ≥50 mg/kg. There also was a slightly increased incidence of endometrial stromal polyps in the same groups, suggesting that the two findings were related.

Endometrial hyperplasia is typified by a fairly uniform proliferation of endometrium that may be accompanied by cystic dilation of endometrial glands. Endometrial stromal polyps are pedunculated masses protruding into the uterine lumen consisting of loosely organized endometrial stromal cells with scattered cystic endometrial glands, and well-differentiated cuboidal epithelial cells on the surface. The relationship between endometrial hyperplasia and polyps may be more than histological. In animals, polyps almost always arise against a backdrop of hyperplasia, leading some investigators to suggest that polyps result when the hyperplastic endometrium is thrown into projecting folds by the limited space within the uterine body.

Uterine stromal polyps are not considered to be precancerous lesions in rodents. Both endometrial hyperplasia and endometrial stromal polyps are common findings in older mice (aged 18 to 24 months) and are characterized by proliferation of endometrial stromal cells (Mohr et al. 1996). The incidence of both findings increases with age and possibly with weight (Haseman and Johnson 1996; Maita et al. 1988; Maekawa and Maita 1996). In women over age 35, uterine endometrial polyps are common in asymptomatic premenopausal women and small polyps have a reasonable chance of regressing spontaneously (DeWaay et al. 2002).

In this study, the incidence of endometrial hyperplasia and stromal polyps across groups suggested that the increased incidence of both findings at higher dose levels was quinacrine-related. Based on data from conventional carcinogenicity studies, stromal polyps are one of the most common spontaneous tumors in 18- and 24-month old mice (Tamano et al. 1988; Turusov, Muñoz, and Dunn 1994). However, data on the incidence of stromal polyps in 1-year-old CD-1 mice are scarce. In neonatal mouse carcinogenicity studies conducted as part of the ILSI collaborative program to evaluate alternative models for carcinogenicity testing, no stromal polyps were observed in 156 1-year-old control mice (M. McClain, personal communication). In one neonatal mouse carcinogenicity study conducted at this testing facility in the same year, 1 of 14 control mice had a stromal polyp, for an incidence of 7%. Taken together, these historical data suggest that polyps can occur in 1-year-old mice but with an incidence that is less than in the mid and high dose groups in this study. Although these historical data represent a relatively small sample, and despite the fact that background tumor incidence is known to vary among studies (Maita et al. 1988; Ettlin and Prentice 2002), it seems reasonable to conclude that the incidence of polyps in this study is outside the range of expected variation.

Nonetheless, the significance of quinacrine treatment and the difference in incidence of hyperplasia and endometrial stromal polyps among groups of mice in this study remains unclear. The uterus of mice is undifferentiated at birth and genesis of endometrial glands is not observed until day 7 post partum (Gray et al. 2001). The possibility remains that the higher incidence of endometrial hyperplasia and consequently, endometrial stromal polyps is related to a disruption of uterine development in these mice as a result of quinacrine exposure during this window of opportunity. This mechanism of action would be particular to neonatal mice and irrelevant for adult women, as in adult women the endometrium is regenerated periodically after menstruation.

Even though this study revealed a lack of tumorigenic response to quinacrine in the usual target organs for the neonatal mouse bioassay (predominately liver and lung), the carcinogenic potential of quinacrine remains uncertain. The neonatal mouse bioassay would not be expected to reveal a positive tumorigenic response for chemicals that exert tumorigenicity through secondary mechanisms. Thus, the carcinogenic risk of quinacrine cannot be fully evaluated with one study alone. Instead, the weight of evidence from a package of information needs to be evaluated before a conclusion can be made. Of particular importance will be the results of an ongoing 2-year rat carcinogenicity study with direct intrauterine administration of quinacrine, as well as an evaluation of the human exposure information.