25,000 mg) in the study by McDiarmid and colleagues than dosages reported by Neal et al. in four of seven sampling periods, which ranged from 44,500 mg in a 40 hour period to 74,000 mg in an 80 hour period.Health Effects of Occupational Exposure
to Antineoplastic Drugs:
An Integrative Research Review
Prepared for the Occupational Disease Panel (Industrial Diseases Standards Panel)
Ministry of Labour
Ontario, Canada
Jeanne Beauchamp Hewitt, RN, PhD
University of Wisconsin--Milwaukee
January 13, 1997
Chemical carcinogen: A chemical carcinogen is a chemical agent that causes any of four conditions: (a) an increased incidence of cancer compared to controls; (b) other types of tumors that are not present in controls; (c) shortened latency to tumor formation; or (c) an increased number of tumors in the same subject compared to controls (Williams & Weisburger, 1986).
Genotoxic agent: A genotoxic agent is one that can interact by binding covalently with genetic material (Williams & Weisburger, 1986). Genotoxic agents can be either direct or indirect agents. Broadly speaking, genotoxic agents include both mutagenic and clastogenic agents.
Mutagen: A mutagen is a chemical or physical agent that produces base- pair substitutions or small additions or deletions of one or more base pairs in genetic material (Thilly & Call, 1986, p. 176).
Clastogen: A clastogen is an agent that can cause one of two types of structural changes. A clastogen can cause breaks in chromosomes that result in the gain, loss, or rearrangements of chromosomal segments. A clastogen can also cause sister chromatid exchanges, which are "homologous chromatid strand interchanges and reunions [that occur] during DNA replication" (Thilly & Call, 1986, p. 181).
Background and Scope of the Review
Antineoplastic drugs (ANDs) have been in clinical use for five decades. The carcinogenic properties of many of these drugs have been studied extensively in animal models and in follow-up studies of patient populations. Comparatively less research has evaluated other adverse effects such as liver and reproductive system toxicities in laboratory animals or epidemiological studies of patient populations. Concern about the potential adverse health effects of ANDs on health care workers (HCWs) was generated when Falck and colleagues (1979) reported that oncology nurses' urine was significantly more mutagenic than that of office controls. The relatively recent interest in occupational health consequences of AND exposures and the lack of existing exposure records for health care workers (HCWs) has limited both the number of studies, and in some instances, the quality of studies, available to evaluate adverse health effects associated with occupational exposure to ANDs among HCWs. While the emphasis of this integrative research review is on occupational epidemiological studies of HCWs, the toxicity of the drugs in animal models and in patient populations will be summarized because these data are essential to draw conclusions about causal relationships.
Inception of Antineoplastic Drug Therapy
The use of antineoplastic drugs (ANDs) to treat cancer is based on the observation that soldiers in World War I who survived the very acute phase of mustard gas attacks suffered severe health effects including chemical burns and neutropenia (Gilman & Philips, 1946). Neutropenia indicated that mustard gas was selectively cytotoxic to the rapidly proliferating bone marrow cells, which led researchers to hypothesize that related chemicals might be useful in the treatment of cancer. This hypothesis was tested in the United States during World War II using two chemical analogues of mustard gas: sulfur mustard and nitrogen mustard. Nitrogen mustard was found to be superior to sulfur mustard, and was particularly effective in inducing remission in Hodgkin's disease (Gilman & Philips, 1946; Rhoads, 1946). Nitrogen mustard also produced short-term benefits for lung cancer patients (Rhoads, 1946). The findings that chemical analogues differed in their effectiveness and that cancer remissions could be induced by cytotoxic drugs spurred the search for other ANDs.
By 1960, 20 ANDs had been developed for clinical use (Krakoff, 1991) (Figure 1). These drugs represented five classes of ANDs: (a) alkylating agents, (b) antimetabolites, (c) mitotic spindle inhibitors, (d) antitumor antibiotics, and (e) hormones, as well as miscellaneous drugs that are not readily classifiable. Although many more drugs have been subsequently synthesized and tested, many are structural analogues of the drugs developed in the 1940s and 1950s (Figures 2 and 3). While these drugs were developed for the purpose of systemically treating cancer, practitioners soon began to use a number of ANDs to treat other unremitting chronic diseases such as rheumatoid arthritis, systemic lupus erythematosus, and polycythemia vera (Baker et al., 1987; Elliott et al., 1982; Kinlen, 1985; Plotz et al., 1979; Thiede et al., 1964).
Antineoplastic Drug Classification and their Mechanisms of Action and Toxicity
Alkylating drugs. Nitrogen mustard and other alkylating agents, or their reactive intermediates, form covalent bonds with deoxyribonucleic acid (DNA), ribonucleic acid (RNA), and protein to form a DNA adduct in which a methyl or ethyl group is added to the DNA (Thilly & Call, 1986). DNA adducts are believed to play a major role in mutagenesis and clastogenesis (Thilly & Call, 1986), as well as in carcinogenesis (Williams & Weisburger, 1986). DNA adducts are formed at a number of reactive sites on nucleotide bases. Common locations include the N-7 and O-6 of guanine. In particular, alkylation at N-7 is believed to be associated with mutagenesis and carcinogenesis (Williams & Weisburger, 1986).
Nitrosoureas. Nitrosoureas appear to function as alkylating agents, as well as through other mechanisms such as carbamoylation. Nitrosoureas are highly lipophilic, which allows them to readily cross lipophilic membranes such as those found in the central nervous system and skin (Knobf et al., 1984).
Antimetabolites. Antimetabolites generally alter the synthesis of DNA or RNA in one of two ways. Antimetabolites that are structural analogues of nucleotides are incorporated into cell components as if they were the essential pyrimidine or purine, and as a consequence, disrupt the synthesis of nucleic acids (Knobf et al., 1984). Other antimetabolites disrupt essential enzymatic processes of metabolism. An example is the folate antagonist, 5-fluorouracil, which disrupts vital folic acid metabolism.
Antitumor antibiotics. Antitumor antibiotics, like other antibiotics, are derived from microorganisms. However, they display toxicities such as myelosuppression and cardiotoxicity that make them unsuitable as general antibiotics. Most antitumor antibiotics intercalate between DNA base pairs (Fischer & Knobf, 1989) and disturb the synthesis and/or function of nucleic acids (Knobf et al., 1984). However, a different mechanism is ascribed to bleomycin. Bleomycin apparently binds to DNA and results in single-strand breaks and double-strand scissions, thereby disrupting DNA synthesis (Fischer & Knobf, 1989). Doxorubicin not only intercalates between base pairs, but also alkylates macromolecules (Gianni, Corden, & Myers, 1983). Daunorubicin, doxorubicin, and their derivatives, belong to a subclass of antitumor antibiotics called anthracyclines.
Mitotic spindle inhibitors. In general, mitotic spindle inhibitors bind to microtubular proteins, a process which halts cell replication at metaphase (Knobf et al., 1984). However, at high concentration, nucleic acid and protein synthesis is suppressed. Mitotic inhibitors include the vinca alkaloids (vincristine and vinblastine) and the epipodophyllotoxins (teniposide and etoposide). Teniposide apparently causes single-strand breaks in DNA, while etoposide impairs the transportation and incorporation of nucleosides into RNA and DNA.
Hormone-related drugs. Hormones or hormone-blocking agents either exert a corticosteroid effect, such as prednisone, or manipulate the hormone environment in hormone-responsive tumors. For example, the antiandrogenic agent, flutamide, which is used to treat prostate cancer, is believed to block androgen receptor sites. Similarly, the antiestrogenic agent, tamoxifen, binds to intracellular estrogen receptors, then enters the nucleus where the tamoxifen-estrogen-receptor complex inhibits DNA and protein synthesis (Fischer & Knobf, 1989).
Miscellaneous antineoplastic drugs. A variety of ANDs with diverse chemical properties and mechanisms of action are grouped in the miscellaneous category. An example is L-asparaginase, an enzyme that interferes with the synthesis of asparagine-containing proteins (Fischer & Knobf, 1989).
Toxicities of Antineoplastic Drugs
Mutagenicity. All classes of ANDs have been shown to be mutagenic in in vitro assays including those drugs shown in Table 1.
Reproductive toxicity. An array of adverse reproductive outcomes have been induced in laboratory animals and observed in patient populations who have undergone AND therapy. For example, in murine models, nitrogen mustard (at 0.08 mg administered intraperitoneally [i.p.]) produced temporary infertility and teratogenesis (Auerbach & Falconer, 1949). In another study, single i.p. doses of nitrogran mustard (< 0.7 mg/kg) or other alkylating agents (triethylene melamine, < .55 mg/kg; thiotepa, < 5.0 mg/kg; chlorambucil, < 10 mg/kg; myleran, < 34 mg/kg) on day 12 of gestation in Wistar rats induced gross anomalies and fetal growth retardation (Murphy et al., 1958). In a multigenerational study that involved chronic dosing of rats with cyclophosphamide (1.0 or 3.0 mg/kg/day per oral feeding), no increased risk of malformation was observed (Botta et al., 1974). Both doses after 12 weeks produced leukopenia. The 3.0 mg/kg/day dosing schedule was associated with significant fetal growth retardation and decrease in birth weight, as well as a progressive increase in fetal resorption rates in subsequent generations. An increase in sterility was noted among those in the third generation who received the 3.0 mg/kg/day dosing schedule. In the single generation component of the study by Botta et al., female rats fed 4.5 mg/kg of cyclophosphamide for 7 days showed no reproductive effects. However, 20 days of treatment produced a significant increase in resorptions and a striking reduction in birthweight.
In a study by Thompson et al. (1978), the anthracyclines, daunorubicin and doxorubicin, induced a significant increase in major soft tissue and skeletal anomalies in rats. This effect was observed for doxorubicin at doses ranging between 1.0 and 2.0 mg/kg/day i.p. administered at various times during gestation, and for daunomycin at doses ranging between 1.0 and 4.0 mg/kg/day administered i.p. (specific doses of daunomycin or timing during gestation were not stated). No teratogenic effects were noted in rabbits administered either doxorubicin (0.2 to 0.6 mg/kg intravenous) or daunomycin (0.15 to 0.6 mg/kg intravenous). In the same study, 0.6 mg/kg of doxorubicin was an effective abortifacient in rabbits. In addition, various dosing regimens of doxorubicin and daunomycin were associated with increased rates of resorption and decreased fetal weight in rats, but not in rabbits. In a dominant lethal study, doxorubicin administered to male mice at doses of 3.0, 6.0, and 12 mg/kg, induced infertility and a decreased number of implants and living embryos in pregnant females (Generoso et al., 1989). In another study, antimetabolites were usually teratogenic to rat fetuses (Murphy et al., 1958). The litter LD-50 for methotrexate was very low, 0.4 mg/kg administered on day 12 of gestation. In the surviving fetuses exposed to methotrexate, most were normal upon visual inspection. Single doses of thiadiazole (50 to 200 mg/kg), 6-aminonicotinamide (8 mg/kg), and azaserine (3.0 mg/kg) induced gross soft tissue and skeletal anomalies in rat fetuses, while 6-diazo-5-oxy-L-norleucine administered at 0.2 mg/kg induced fetal resorption.
In case reports and case series of patient populations, methotrexate has been clearly linked to severe, and frequently multiple, birth defects (Milunsky et al., 1968; Powell & Ekert, 1971; Warkany, 1986). Methotrexate also has been associated with low birth weight (Milunsky et al., 1968). In another case report, multiple severe birth defects were observed in the electively aborted fetus of a women who received 500 mg. of fluorouracil late in the first trimester (Stephens et al., 1980). In a report on new (7) and published case studies (51), 56% of the births were premature and 8% experienced intrauterine growth retardation (Reynoso et al., 1987). A single infant had three major anomalies and later developed two de novo primary malignancies. In this case series, most of the ANDs were antimetabolites or the anthracyclines, daunorubicin or doxorubicin. Follow-up studies also have reported increased risk of birth defects in the offspring of female patients (Holmes & Holmes, 1978; Mulvihill et al., 1987; Ross, 1976) and wives of male patients (Green et al., 1991), as well as increased risk of fetal loss in wives of male patients (Holmes & Holmes, 1978). In addition, both ovarian failure and azoospermia or oligospermia have been associated with AND therapy (Chapman et al., 1979; DaCuhna et al., 1983). In the study by Chapman and colleagues, complete ovarian failure occurred in 49% of the women treated with methotrexate, vincristine, procarbazine, and prednisone (MOPP regimen) or variations of MOPP that contain cyclophosphamide or doxorubicin. Partial ovarian failure (menstrual cycles ranging from 22 days to 12 months) occurred in 34% of the women studied. Importantly, the effect of complete or partial ovarian failure was greater in women who over 30 years of age at the time of therapy.
Carcinogenicity. Numerous ANDs have been found to be carcinogenic in animal models (Weisburger, 1977). Indications that AND-therapy was associated with second primary malignancies were initially reported in case reports and case series. In the mid-1970s, large-scale follow-up studies of cancer patients undergoing AND therapy were first reported in the literature. These reports strongly and consistently associated several alkylating agents with the development of leukemia and myelodysplastic syndromes (Bartolucci et al., 1983; Bersagel et al., 1979; Boice et al., 1983; Boivin et al., 1984, 1995; Chak et al., 1984; Coleman et al., 1987; Curtis et al., 1984; Devereaux et al., 1990; Haas et al., 1987; Hancock et al., 1988; Pedersen-Bjergaard et al., 1985a, 1985b, 1993; Pedersen-Bjergaard & Larsen, 1982; Tester et al., 1984; Tucker et al., 1988). Studies that evaluated histological subtypes found that acute non-lymphocytic leukemia predominated. Case series and longitudinal studies also specifically implicated cyclophosphamide (and chlornaphazine used in Europe) with the development of bladder cancer (Baker et al., 1987; Dale & Smith, 1974; Elliott et al., 1982; Fairchild et al., 1979; Kinlen, 1985; Pearson & Soloway, 1978; Plotz et al., 1979; Wall & Clausen, 1975).
Across studies, the latency pattern associated with alkylating agents on the development of leukemia generally showed an increased risk one year after the initiation of therapy, with increasing risk during the first ten years. The incidence of AND therapy-related leukemia seemed to peak around five years post-therapy and plateaued or decreased thereafter. The latency period for bladder cancer risk associated with cyclophosphamide has usually been several years or longer.
As practitioners' experience with AND therapy has increased, two other patterns have emerged. First, alkylating agents have been shown to be associated with increased risk for the development of non-Hodgkin's lymphoma (NHL) and a variety of solid tumors including melanoma, and cancers of the lung, stomach, bone, connective tissue, thyroid, oral-pharynx, and liver (Boivin et al., 1995; Coleman et al., 1987; Kaldor et al., 1992; Kinlen, 1985; Matteson et al., 1991; Tucker et al., 1988). Second, there is increasing evidence that single- or multiple-agent AND regimens even without alkylating agents are leukemogenic. Studies have implicated anthracycline antibiotics (e.g., doxorubicin, epi-doxorubicin, mitoxantrone), nitrosoureas (e.g., lomustine and semustine), and epipodophyllotoxins (e.g., etoposide and teniposide) as being leukemogenic (Boice et al., 1983, 1986; Devereaux et al., 1990; Hawkins, 1991; Pederson-Bjergaard et al., 1991, 1993; Ratain et al., 1987; Verdeguer et al., 1992; Whitlock et al., 1991; Zeimet et al., 1992).
The magnitude of risk for leukemia associated with AND-therapy have ranged from approximately a two-fold increased risk associated with single-agent therapy (vinblastine or chlorambucil) for Hodgkin's disease (Boivin et al., 1995) to more than 100-fold for various drug therapies (Bartolucci et al., 1983; Bersagel et al., 1979; Boivin et al., 1984; Chak et al., 1984; Einhorn et al., 1982; Johnson et al., 1986; Pedersen-Bjergaard & Larsen, 1982; Pedersen-Bjergaard et al., 1991; Tucker et al., 1988). Because some cancers such as Hodgkin's disease are treated with ANDs, radiotherapy, or both, it is possible to evaluate the independent or interactive effects of radiotherapy and AND-therapy. For example, Tucker and colleagues (1988) found that radiotherapy alone increased the risk of leukemia 11-fold (95% CI = 1.2, 38.4). Radiotherapy and adjuvant AND therapy combined was associated with a relative risk of 117 (95% CI = 69, 185), while AND therapy alone was associated with a relative risk of 130 (95% CI = 26, 380). In a study of patients treated for Hodgkin's disease, Kaldor and colleagues (1992) found that AND therapy alone doubled the risk of lung cancer compared to those who received only radiotherapy or a combination of AND therapy and radiotherapy. Based on animal data and studies of patient populations, the International Agency for Research on Cancer (IARC) has concluded that a number of ANDs are human carcinogens (IARC, 1975, 1979).
Other toxicities. In animal models and patient populations, various ANDs have been associated with myocardial or cardiac conduction defects; pulmonary fibrosis; gastrointestinal disturbances including stomatitis, nausea, vomiting, and diarrhea; changes in liver function; neurotoxicity; myelosuppression; and alopecia (hair loss) (Buckner et al., 1989; D'Alessandro et al., 1983; Coleman et al., 1984; Drago & Goldman, 1980; Henderson et al., 1982; Kimler et al., 1984; Kojima et al., 1994; Kris et al., 1985; Lindpaintner et al., 1986; Martoni et al., 1985; McCarthy & Skillings, 1992; Shaw et al., 1989; Shinar & Hasin, 1984; Tamatsu et al., 1984; Tan et al., 1987; Unverferth et al., 1983; Vorobiof et al., 1983).
Occupational Exposure to Antineoplastic Drugs in Health Care Workers
The research has demonstrated clear patterns of toxicities associated with ANDs in experimental or observational studies of patients undergoing AND therapy. However, studies of toxicities in occupational groups with potential exposure to ANDs are complicated by a number of factors including the general lack of sensitive and specific measures of biological absorption of ANDs, the "mixture" of exposures to various ANDs in combination with workplace and lifestyle exposures, the lack of existing exposure histories on HCWs, or the availability of readily-assembled cohorts with appropriate internal or external comparison groups. Clearly, the routes and mechanisms of exposure to AND in occupational groups differs substantially from therapy-related exposures. Yet, they are critical for understanding potential health risks.
Routes and Mechanisms of Occupational Exposure
Occupational exposure to ANDs occurs primarily through inhalation, skin and mucous membrane contact, and inadvertent ingestion (OSHA, 1986, 1995). A number of manipulations involved in preparing and administering ANDs can generate drug aerosols (Hewitt et al., 1993; Hoy & Stump, 1984; Kleinberg & Quinn, 1981; Neal et al., 1983). For example, without a special device to equalize pressure, a needle that is inserted and then removed from a vial to mix or dispense drugs generates an aerosol (Hoy & Stump, 1984).
Many other procedures also can result in personal and environmental contamination. These include breaking ampules to reconstitute or dispense ANDs, breaking needles and crushing syringes, priming intravenous tubing in clinical areas, removing air from intravenous tubing, improperly disposing of ANDs or AND contaminants, and the clean-up and disposal of spills (OSHA, 1986; OSHA, 1995). In addition, urine, feces, vomitus, blood, and other body products from patients undergoing AND therapy are highly contaminated with the parent drug(s) and their metabolite(s) (Falck et al., 1979; Venitt et al., 1983) so that direct contact with these body fluids/excreta is a source of exposure to ANDs for HCWs. Moreover, when surfaces are contaminated with drugs or AND-contaminated body fluids/excreta and then dry, particles of the drug can become airborne and present an inhalation hazard.
An additional mechanism of exposure can occur through inadvertent ingestion of ANDs (OSHA, 1986, 1995). When food or beverages are prepared, stored, or consumed in work areas, they may easily become contaminated with airborne particles of ANDs (Hewitt, 1986; Neal et al., 1983). Likewise, hands, and even cosmetics, chewing gum, or other incidental items, can be contaminated and result in the inadvertent ingestion of these drugs (Hewitt et al., 1993).
The study by Falck et al. (1979), which showed a statistically significant increase in urine mutagenicity in oncology nurses, first alerted HCWs to the possible genotoxic effects of these drugs at occupational levels of exposure. Following this landmark study, a number of surveys were conducted in the early 1980s which indicated that it was common practice for nurses and others to (a) prepare ANDs on countertops in clinical areas with high traffic, (b) discard AND-contaminated equipment with other trash, (c) dispose of unused ANDs in sinks, and (d) use little or no protective equipment (e.g., gloves, disposable impermeable gowns, and appropriate ventilation hoods) when working with ANDs (Crudi et al., 1982; Neal et al., 1983; Rogers & Emmett, 1987; Stellman et al., 1984). Surveys have typically relied on self-reported behaviors. However, three studies that also included direct observation of AND handling practices found that unacceptable work practices and critical incidents occurred. These studies reported that food items were stored and food and beverages were consumed in drug preparation areas (Hewitt, 1986; Neal et al., 1983). During drug preparation and transport, spillage of ANDs was observed, as well as leaks and breakage of syringes and drug containers (Newman et al., 1994). During AND administration, AND leakage was observed when intravenous tubing was primed, when air was removed from parenteral systems, and when intravenous lines were connected or disconnected from patients (Newman et al., 1994).
Although various guideline for safe handling of ANDs have been promulgated (American Society of Hospital Pharmacists, 1990; OSHA, 1986, 1995) and some improvements have been noted, adherence to safe work practices continues to be a challenge (Barry & Booher, 1985; Christensen et al., 1990; Cloak et al., 1985; Hewitt, 1986; LeRoy et al., 1983; Neal et al., 1983; Rogers & Emmett, 1987; Stajich et al., 1986; Stellman et al., 1984; Valanis & Browne, 1985; Valanis et al., 1991, 1992; Valanis & Shortridge, 1987). Although training personnel regarding ANDs and safety aspects of handling these drugs are considered key elements (OSHA, 1986, 1995), no studies have evaluated the quality or frequency of training, and few have evaluated adherence to OSHA or other comprehensive guidelines (Hewitt, 1986; Valanis et al., 1991, 1992).
According to OSHA guidelines, the two key elements for safely handling ANDs are the use of a vertical laminar air flow (VLAF) hood and proper training (OSHA, 1986). The VLAF hood, if working properly, directs airflow from the room, through the opening where the preparer is situated, and across the drugs being prepared. The drug contaminated air is then vented through a plenum at the back of the preparation area and through a high efficiency particulate air (HEPA) filter, where the filtered air is either recirculated into the room or vented to the outside. Although the VLAF hood is deemed by OSHA to be essential, some studies have found that nurses were less likely than pharmacists to have access to VLAF hoods (Hewitt, 1986; Valanis et al., 1991, 1992). Many surveys indicated that, compared to pharmacists, nurses were less likely to use personal protective equipment (e.g., surgical latex gloves and impermeable disposable gowns), particularly for administering ANDs or when contact with ANDs or body products (e.g., urine, feces, vomitus) occurs (Hewitt, 1986; Stellman et al., 1984; Valanis et al., 1987, 1991; Valanis & Shortridge, 1987). As practitioners became more aware of potential hazards and recommended guidelines, it is possible that some response bias may have occurred in more recent studies. Consequently, surveys that have relied solely on questionnaire data may be somewhat optimistic about actual practice and adherence to recommended guidelines such as those published by OSHA (1986, 1995). Several studies have noted that, independent of the use of protective equipment, self-reported skin contact with ANDs was associated with a significantly elevated risk of symptoms characteristic of AND exposure (Valanis et al., 1992, 1993) and with menstrual cycle dysfunction (Shortridge et al., 1995).
Physical Detection of Skin and Environmental Contamination
Nurses have reported that skin contact occurs in virtually all manipulations involved in handling ANDs (Cloak et al., 1985). Despite the importance of skin contact as a route of biological absorption of these drugs, verification of skin contact has rarely been reported. However, among nurses and physicians who handled the nitrosourea, carmustine, skin hyperpigmentation due to direct contact with the drug was confirmed independently (Frost & DeVita, 1966).
Environmental contamination with ANDs has been demonstrated in a number of studies, and could result in biological absorption of the drugs through inhalation, skin absorption, or inadvertent ingestion. In a simulated procedure, 9 trials were conducted each using 5 10-mL ampules of fluorouracil (Kleinberg & Quinn, 1981). The ampules containing fluorouracil were snapped open and the contents transferred by syringe to empty vials. Air samples were taken within the hood close to where the manipulations took place. Fluorouracil was measured using ultraviolet spectrophotometry, which had a detection limit of 0.2 µ/mL. The drug was recovered in 5 of 9 trials, ranging from a low of 0.02µ/L to 0.11 µg/L.
Two other studies used dyes to simulate AND admixture procedures. Stellman (1987) used a saline-soluble fluorescent dye in which environmental contamination was evaluated under ultraviolet light. Mixing the dye in three vials of saline produced aerosolized droplets on the work surfaces and resulted in the contamination of the external surface of the syringe and vial. Similarly, Hoy and Stump (1984) used brilliant green dye to determine the extent to which aerosolized drug (dye) was dispersed when a 20-gauge 1.5-inch needle was withdrawn from a vial under positive pressure. These studies demonstrated that considerable aerosolization occurred during the simulated admixture procedure.
In another study, VanRaalte and colleagues (1990) used visible light-stimulated fluorescence to evaluate personal and environmental exposure to doxorubicin. The limit of detection of visible light-stimulated fluorescence was estimated to be near the low nanogram/high picogram amount. The researchers detected skin contamination on two nurses, although neither was aware that skin contamination had occurred. One nurse had apparently contaminated the skin just above the wrist (gloves had been worn, but not a protective gown) 4 hours earlier either when she instilled doxorubicin through a catheter into the patient's bladder or when handling this patient's soiled linens. The second nurse had a small spot of doxorubicin on a thumbnail detected prior to gloving for a procedure. Although this method only detects doxorubicin (and most likely analogue drugs such as daunorubicin and idarubicin), it suggests that the same type of contamination could occur with other ANDs. In this same study, VanRaalte and colleagues demonstrated that doxorubicin was stable on various surfaces for at least 7 weeks. Doxorubicin fluorescence also was evident in various sections of one hood used by pharmacists to prepare the drug, on the countertop where ANDs were administered, on the top surface of an AND waste disposal container, and on various surfaces in two of three clinics surveyed.
Measurement of contamination of ambient air with various ANDs has also been reported. Neal et al. (1983) tested air samples using high performance liquid chromatography (HPLC) to detect methotrexate, fluorouracil, doxorubicin, and cyclophosphamide, which were prepared on a countertop in the medication preparation room. The detection of fluorouracil was confirmed using mass spectrometry (MS). Detection limits ranged from a low of 0.06 ng for fluorouracil to 30 ng for cyclophosphamide. Samples collected over a period of 40 or 80 hours were taken at breathing zones in the drug preparation rooms where ANDs were prepared on countertops and also in a room about 50 feet away from the drug preparation room. Three air samples were analyzed for the presence of methotrexate and doxorubicin, which were used less frequently than the other two drugs. Neither methotrexate or doxorubicin was detected. However, fluorouracil was detected by HPLC in 9 out of 14 samples. The quantity detected ranged from a low of 0.26 ng/m3 to 82.26 ng/m3. The presence of fluorouracil was confirmed by MS in most, but not all, samples. In a modified replication study that also used HPLC for analysis, McDiarmid and others (1986) were unable to detect fluorouracil in two air-pump filtered samples of 56 hours (7 working shifts) and 95 hours (10 working shifts) duration, respectively. In contrast to the study by Neal and others, sampling took place in the breathing zone of pharmacists who used a vertical laminar air flow (VLAF) hood to prepare ANDs. The limit of detection was 0.2 ng/m3 in this study. In addition to the use of an appropriate type of hood, the total dosage of fluorouracil prepared also was less (25,000 mg) in the study by McDiarmid and colleagues than dosages reported by Neal et al. in four of seven sampling periods, which ranged from 44,500 mg in a 40 hour period to 74,000 mg in an 80 hour period.
Using HPLC, McDevitt et al. (1993) examined air and surface wipe samples for evidence of contamination with cyclophosphamide in a hospital pharmacy and clinic. Of the 34 air samples from the pharmacy that were analyzed, 91% were below the limit of detection (1.25 µg/mL). Two samples were taken from the inside the VLAF hood (0.218 µmg/mL ± 0.035 and 0.407 µg/mL ± 0.044) and one immediately outside the hood (0.296 µg/mL ± 0.041). In the clinic air samples, all 39 were below the limit of detection. In surface wipe samples, which ranged from 100 to 900 cm², the limit of detection ranged from 0.003 to 0.025 µg/cm². Of the 34 wipe samples in the pharmacy, 18% yielded detectable levels of cyclophosphamide. Of the 42 wipe samples in the clinic, 14% were above the level of detection. Positive wipes, ranging from 0.005 to 0.035 µg/cm², were found inside the VLAF hood (2), on the pharmacy floor (1), on the countertops in the pharmacy (3) and nurses' station (1), on a video display terminal at the nurses' station (1), and on the sink (2) and countertop (2) in the clinic treatment area.
Physical Detection of Biologically Absorbed Antineoplastic Drugs
Several physical detection methods have been employed to detect AND exposures among HCWs. Venitt et al. (1984) used platinum, a molecular component of cisplatin, as a marker for occupational exposure. They used atomic-absorption spectrometry to measure urinary platinum levels for eight nurses and two pharmacists. None of the nurses or pharmacists had detectable levels of urinary platinum, which had a limit of detection of approximately 20 Because only single spot urine samples were obtained on Friday afternoons at the end of a 5-day work week, the volume of urine obtained may have been inadequate for the method.
A sensitive method (nanomolar levels) for detecting methotrexate has been reported for recovering the drug from urine of oncology nurses who prepared this drug or who cared for patients being treated for osteosarcoma with a high-dose (20 g.) methotrexate regimen (Mader et al., 1993). Using HPLC and urine specimens collected over three time periods of 12 hours, the researchers detected methotrexate in both the nurses who gave patient care only and in those who prepared the drug using thick latex gloves, reusable coat, and "dust-proof" mask. Higher levels of methotrexate (e.g., 143 ng/mL of urine) were recoverable in the urine of nurses who prepared the drug compared to nurses who solely gave patient care.
Hirst et al., (1984) used gas chromatography/mass spectroscopy (GC/MS) to detect cyclophosphamide in samples of urine from nurses who prepared and administered this drug without gloves or a VLAF hood. Thirty-four percent of multiple samples provided by five nurses indicated occupational exposure had occurred, ranging from 0.83 µg/ml to 16.47 µg/ml. Urinary levels of cyclophosphamide did not correlate with doses handled, but the lack of correlation may be due to the collection of spot urine samples at varying intervals. Using GC/MS, but with a 24-hour urine sample collected on the fourth workday, Evelo et al., (1986) detected a correlation between the number of doses of cyclophosphamide handled and the level of urinary excretion. Each of the 20 nurses handled this drug from 1 to 30 times over a 4-day period. The limit of detection was 0.5 µg/24 hour urine specimen. Five nurses, all smokers, had levels of cyclophosphamide that ranged between 0.7 µg/24 hours and 2.5 µg in a 24 hour urine specimen. Since smoking would not confound these findings as it could in genotoxicity assays, Evelo et al. (1986) concluded that it may be coincidental that only smokers had measurable urinary levels of cyclophosphamide or it may be that smoking increased the biological absorption of the drug through hand-to-mouth behavior.
Sessink and others (1994) also measured urinary cyclophosphamide excretion using GC/MS analysis. Eleven of 17 Dutch workers who handled ANDs contributed 79 urine samples. Eleven Czech workers who handled ANDs contributed 68 urine samples and an additional 147 samples were analyzed from Czech workers who were not involved in handling ANDs or cleaning up spills. Their results showed detectable levels in 3 Dutch workers who handled ANDs (0.1 µg/24 hours and 0.5 µg/24 hours in non-smokers and 0.3 µg/24 hours in a smoker). Of 11 Czech workers whose urinary cyclophosphamide excretion levels were measured, no cyclophosphamide was detected in 3 workers, while values ranged between 0.1 µg/24 hours and 2.9 µg/24 hours in the other 8 workers. The limit of detection of this method was approximately 0.25 ng/ml urine. One Czech nurse, a non-smoker, prepared ANDs, but not cyclophosphamide, yet the urinary level of cyclophosphamide was 0.3 µg/24 hours. The highest level of urinary cyclophosphamide was found in another nurse who prepared and administered ANDs, including cyclophosphamide, and who smoked 10 cigarettes per day. Detectable levels were also found in one technician (non-smoker) and a cleaning woman (smoker) who removed waste after administration of cyclophosphamide.
Mutagenicity and Clastogenicity Studies in Health Care Workers
In addition to using direct physical detection measures, researchers have used various indirect methods to evaluate occupational exposure to ANDs. Some of the more commonly reported methods include urine mutagenicity assays (e.g., the Ames Salmonella assay) and clastogenicity assays such as SCEs, micronuclei, and other chromosomal aberrations (gaps, breaks, dicentrics, and translocations). A review of studies that evaluated mutagenicity and/or clastogenicity in HCWs who handled ANDs found that various methods have yielded both positive and negative findings (Hewitt, 1992) and are not discussed in detail here. However, it is important to note that false positive findings are most likely to result from uncontrolled confounding by extraneous factors such as cigarette smoking (Hewitt, 1992). False negative findings can be attributable to a number of factors including study design issues such as inadequate sample size and uncontrolled confounding (Hewitt, 1992), as well as technical problems (e.g., the use of relatively insensitive assays, biological sampling that does not correspond to the drug distribution pattern consistent with routes of occupational exposure, and using spot urine samples) (Hewitt, 1992; Newman et al., 1994). Notwithstanding these problems, a number of methodologically sound studies demonstrate that regardless of whether protective equipment (e.g., safety hoods and gloves) was used for handling ANDs, genotoxic endpoints were found to be significantly elevated in HCWs. Several of these studies are highlighted below.
Rogers and Emmett (1987) examined urine mutagenicity in oncology nurses who handled ANDs compared to non-oncology nurses using the Ames Salmonella assay (strains TA100 and TA98). Each oncology nurse provided a 24 hour urine specimen after working 3 or more consecutive days and again at the end of a minimum of two days away from work. Twenty four hour urine specimens were collected from comparison nurses at the end of a work week. None of the nurses who prepared ANDs (n = 28) used a VLAF hood. Only 20% of all oncology nurses (n = 59) used gloves most of the time when handling ANDs. Positive urine mutagenicity was found for all oncology nurses (X² = 25.0 (df = 1), p = .0001) and for only non-smoking oncology nurses (X² = 35.2 (df = 1), p = .0001) compared to their non-oncology counterparts. Among oncology nurses, urine mutagenicity was significantly greater after a continuous work period than after a two day work free period both for all nurses (X² = 33.2 (df = 1), p = .0001) and for non-smoking nurses (X² = 40.3 (df = 1), p = .0001). Among oncology nurses, greater mutagenicity was positively associated with handling more doses of ANDs (X² = 16.1 (df = 3), p = .001). Urine mutagenicity was also significantly greater for all nurses who prepared and administered ANDs compared to nurses who only administered these drugs (X² = 6.9 (df = 1), p = .0086), and when analysis was limited to only non-smoking oncology nurses (X² = 9.0 (df = 1), p = .0027) [these two X² were calculated based on data provided].
Benhamou et al. (1988) compared SCE frequencies and chromosomal aberrations in peripheral lymphocytes of 29 oncology nurses who prepared and administered ANDs without safety measures and those of office controls matched on age and sex, but not smoking habits. After adjustment for age and smoking habits, no difference was found in mean SCE frequencies between oncology nurses and the office controls. However, after adjusting for smoking and the use of medications within 6 months, the mean SCE frequency correlated with the total number of ANDs handled (p = .05). No difference between groups was detected for chromosomal gaps, breaks, dicentrics, or translocations. In their companion report on urine mutagenicity, Benhamou et al. (1986) found that urine mutagenicity in the 11 non-smoking oncology nurses was significantly elevated (p = .05) compared to the non-smoking age- and sex-matched controls when Salmonella typhimurium TA98 strain was used with or without exogenous liver enzymes (S-9 mix). Comparisons using the TA100 or TA1535 strains showed no difference between groups.
Thiringer et al. (1991) conducted one of the largest biological monitoring studies of occupational AND exposure. They examined micronuclei and SCE frequencies in peripheral lymphocytes, urine mutagenicity, and the excretion of urinary thioethers in oncology nurses (n = 60) and a referent group individually matched on age, sex, and smoking habits. Of the 60 nurses who prepared and administered ANDs, 91% reportedly always used safety hoods for preparing the drugs and 98% used gloves consistently. These oncology nurses on average prepared four doses of ANDs weekly. Blood and spot urine samples were collected concurrently after a few days of work, as well as after having at least two days away from work (pre-exposure). Using the highest values, pre- and post-exposure comparisons for oncology nurses showed no difference in urinary thioether excretion or in urine mutagenicity using either Salmonella typhimurium or Escherichia coli. Compared to referents, no increase among oncology nurses was noted in the frequency of micronuclei, levels of urinary thioethers, or urine mutagenicity tested with the E. coli WP2 uvrA strain. However, despite the use of basic protection measures by most oncology nurses, they experienced significantly increased urine mutagenicity using the S. typhimurium TA98 strain (p < .01) and elevated SCE frequencies (p < .05) compared to referent nurses.
McDiarmid and colleagues (1992) did not observe a correlation between the duration of preparing ANDs (measured in hours) and baseline SCE frequencies in 34 pharmacy personnel who prepared ANDs in VLAF hoods. They used a technique in which peripheral lymphocyte cultures received challenge doses (0.1µg/ml and 0.25 µg/ml) of phosphoramide mustard (PM), a metabolite of cyclophosphamide. PM-induced SCE frequencies were highly correlated with lifetime AND handling history (mean = 939.0 ± S.E. 246.3 hours) and with hours of drug handling per week in the pharmacy workers' present position (mean = 1.04 ± S.E. 0.21). For lifetime duration of handling of ANDs, correlations were 0.63 (p = .0001) for the low dose challenge and 0.67 (p = .0001) for the high dose challenge. Similar correlations were found between AND handling in the present position and PM-induced SCE frequencies.
Sessink and colleagues (1994) studied clastogenic effects in Dutch and Czech nursing and pharmacy personnel who handled ANDs using VLAF hoods and protective gloves and clothing. Sessink et al. found significantly increased chromosomal aberrations in these workers compared to other HCWs who did not handle ANDs. Dutch AND exposed workers had significantly greater breaks per cell than unexposed workers (p = .04), whereas Czech workers who handled ANDs had both significantly greater breaks per cell (p = .04) and percent aberrations (p = .01). The findings indicated a synergistic effect between smoking and exposure to ANDs.
Case Reports of Acute Exposure
Reynolds et al. (1982) described signs and symptoms associated with the admixture of amsacrine that is consistent with the side effects of the drug. They reported that seven pharmacy personnel experienced nausea, light-headedness, headache, and lethargy following the preparation of amsacrine without protective equipment or while using a horizontal laminar air flow (HLAF) hood, which directs the air current over the product towards the worker and the work environment. One pharmacy technician developed an urticarial rash on the arms and torso within minutes of handling amsacrine under a HLAF hood, and the same type of rash recurred on two other occasions when gloves or other personal protective equipment were used. A pharmacist also experienced severe nausea and vomiting four hours after a vial of amsacrine shattered, which resulted in contamination of skin (washed immediately) and clothing (not changed for four hours). These events occurred in three geographically distinct areas in California and stopped once it became standard practice to prepare the drugs in a VLAF hood and to use gloves and other personal protective equipment.
In another report, a technician who worked repeatedly with doxorubicin and daunorubicin without gloves or other protective clothing experienced a severe contact dermatitis (Reich & Bachur, 1975). Areas affected by the dermatitis were clearly fluorescent under ultraviolet light, indicating that contamination of the skin with these drugs was persistent. The contact dermatitis partially healed during periods when the technician wore gloves. However, for those periods when he did not consistently use gloves, the dermatitis worsened. Reich and Bachur also noted that the dermatitis was exacerbated when work involved handling both doxorubicin or daunorubicin and solvents.
Two other instances of acute exposure during AND handling have been reported in the literature (McDiarmid & Egan, 1988). In the first case report, a 22 year old male pharmacy intern exhibited symptoms of allergy within seconds of admixing vincristine. He experienced shortness of breath, chest tightness, sneezing, "hot flashes" without elevated body temperature, and mild periorbital edema. The symptoms were treated with several doses of diphenhydramine taken at the time of emergency treatmemt and the next day. In the second case report, a 24 year old female oncology nurse was preparing to administer an intravenous dose of 30 mg. dose of carmustine in 250 ml. of 5% dextrose solution when the intravenous tubing dislodged from the intravenous bag. The entire solution spilled on her right arm and right leg, as well as on her clothing. Although the nurse washed her arm and leg, the clothing was not changed during the shift. Approximately four hours after this incident occurred, the nurse experienced intense gastrointestinal symptoms consisting of eructation, and abdominal cramping and very severe diarrhea, followed several hours later by a copious emesis. She sought medical attention the next day at which time the gastrointestinal symptoms had abated without treatment. Follow-up over the next six weeks showed no evidence of abnormal findings including hematological values. In both instances, the symptoms were consistent with biological absorption of the specific AND.
Adverse Health Effects of Antineoplastic Drugs on Health Care Workers
Sotaniemi et al. (1983) described three cases of chronic irreversible liver fibrosis that occurred in nurses who prepared and administered ANDs for 6, 8, and 16 years duration, respectively, without using protective equipment. Although total bilirubin, albumin, and thrombotests were within normal limits, liver enzymes were elevated in one nurse and arylhydrocarbon hydroxylase was elevated in two nurses. Cytochrome P-450 was elevated in one nurse, below normal in another, and within normal limits for the other nurse. Two subjects prepared approximately 20 doses of ANDs per month for the 2 years prior to the liver biopsy. The third nurse, who worked for 16 years in oncology before retiring, had handled approximately 600 doses per year. All three nurses were nonsmokers and drank only a few alcoholic beverages per year or none at all. Each had negative viral antibody titres.
In patient populations, methotrexate is known to cause liver fibrosis or cirrhosis, and has been known to occur even when liver function tests remain within normal limits (Menard et al., 1980). Several other ANDs including 6-mercaptopurine, azathioprine, 5-fluorouracil, and cytosine arabinoside have been associated with liver toxicity as well. Importantly, Menard and colleagues noted that "the prevalence of cirrhosis and fibrosis appeared significantly greater in subjects given frequent small doses of methotrexate than in those given intermittent large doses" (p. 143). Occupational exposures are assumed to consist of low "doses" of various ANDs over the course of employment activities that involve handling ANDs without adequate protection of the worker and work environment. Although no other reports of liver fibrosis of HCWs who handled ANDs were found in the literature, the clinical and histological findings of the three nurses in this case series reported by Sotaniemi et al. (1983) suggest that the irreversible liver fibrosis was due to occupational exposures to ANDs.
Shortridge and colleagues (1995) examined menstrual cycle dysfunction defined as amenorrhea of at least 3 months duration, cycles of less than 25 days or more than 31 days, or long (> 7 days) or short (< 2 days) menses. A convenience sample of members of the Oncology Nursing Society (n = 982) and the American Nurses Association (n = 897) under age 46 participated in separate mail surveys. No significantly increased risk of menstrual cycle dysfunction was noted for any exposure category (past or current exposure to excreta, past handling of ANDs, or current administration, or both preparation and administration of ANDs). However, menstrual cycle dysfunction was more than 3 times as likely in nurses between 30 and 45 years of age, inclusive, who currently administered ANDs compared to nurses who either never handled ANDs or did so only in the past (odds ratio = 3.43; 95% Confidence Interval [CI] = 1.6, 7.3). No statistically significant association was found between the use of protective equipment (e.g., gloves, VLAF hood) and menstrual cycle dysfunction (Shortridge, 1988).
Hemminki and colleagues (1985) were the first to report reproductive effects of ANDs and other workplace exposures (anesthetic gases, x-rays, sterilants, disinfecting soaps, and shiftwork) (Table 2). They found an unstable two-fold increased risk of birth defects associated with occasional first trimester exposure to ANDs, defined as nurses preparing less than one AND dose per week. Nurses who prepared at least one AND dose weekly, however, experienced a statistically significant four-fold increased risk of birth defects in offspring. McDonald and colleagues (1988a) found a statistically significant two-fold increased risk of birth defects among offspring of nurses and physicians who administered ANDs during pregnancy. However, Skov and colleagues (1992) found no increased risk of malformation.
Increased risk of early fetal loss associated with AND exposure was found to be significantly increased in several studies (Selevan et al., 1985; Taskinen et al., 1986; Rogers & Emmett, 1987), but not in other studies (Hemminki et al., 1985; McDonald et al., 1988b; Stucker et al., 1990) (Table 2). In the study by Skov and colleagues (1992), increased risk was detected only in the low exposure group (prepared and/or administered one AND per week during pregnancy).
Two studies evaluated premature birth and low birth weight (Skov et al., 1992; Stucker et al., 1993) (Table 2). No difference in prematurity was noted between births that occurred to nurses exposed to ANDs during pregnancy (Skov et al., 1992) or before or during the pregnancy (Stucker et al., 1993). However, an increased risk of low birth weight was noted among oncology nurses in both studies, although neither estimate reached statistical significance. In the study by Stucker et al., (1993), a deficit in birthweight of 56 g., adjusted for multiple risk factors including smoking during pregnancy, was observed among oncology nurses compared to nurses in other departments. Although only adjusted for pregnancy-related factors, Skov and colleagues (1992) similarly found a deficit in birthweight of 58 g. among oncology nurses.
In a study of operating room nurses, Saurel-Cubizolles and colleagues (1993) detected an eleven-fold increased risk of ectopic pregnancy among operating room nurses who also had exposure to ANDs during the first trimester. The risk estimate was adjusted for relevant factors except a history of sexually transmitted diseases (STDs). Nonetheless, it seems implausible that STDs would be a confounder in this study as it would require that a history of an STD would be associated with exposure to ANDs as well as with ectopic pregnancy.
Occupational reproductive epidemiological studies are very challenging to conduct, and therefore, are easily subject to methodological flaws. The studies available for review with respect to AND exposure are no exception. The strongest studies from a methodological perspective are those that evaluated outcomes following fetal exposure, namely the study of congenital malformations by Hemmminki et al. (1985) and the study of early fetal loss by Selevan et al. (1985). Both studies were conducted in Finland using computerized national registries of health care personnel and health outcome data (congenital malformation, hospital discharge, and polyclinic data) and each study ascertained exposure to ANDs during the first trimester, which is the relevant window of exposure for these two outcomes in relation to fetal exposure (Selevan & Lemasters, 1987).
To have adequate power to evaluate congenital malformations, Hemminki and colleagues (1985) selected subjects from all Finnish hospitals from those who worked on pediatric, gynecological, cancer, and pulmonary units where ANDs were likely to be used. Then, to eliminate recall bias, they obtained data on AND preparation or administration from nurse managers and logs of drugs prepared, which were available at two of the hospitals. These data were collected corresponding to the first trimester for the pregnancy of interest. Information on occupational exposures were collected blind to exact study purpose and outcome status. A ratio of three controls per case matched on age and hospital employment was used. In this case-control study, data analysis retained the matching factors and controlled for other potential confounders through conditional logistic regression analysis. At the time the study was conducted, in general, measures to minimize occupational exposures to ANDs were not used. A limitation of the study design which used nurse managers as the respondents was that data on lifestyle and reproductive factors could not be obtained. Handling ANDs less than weekly was associated with an unstable two-fold increased risk of congenital malformations, whereas handling them at least weekly was associated with a 4.7-fold statistically significant increased risk.
It is noteworthy that no increased risk of fetal loss was detected in the study by Hemminki et al. (1985), which reported an increased risk for congenital malformations associated with AND exposure. The "negative" finding with respect to fetal loss was likely to be influenced by the selection of hospital units (anesthesia surgery, intensive care, operating room, and internal medicine) that had a low probability of having personnel exposed to ANDs.
As a follow-up to the study by Hemminki et al. (1985), Selevan and colleagues (1985) first selected hospitals that used 100 g. or more of cyclophosphamide per year or at least 200 g. of all ANDs per year as reported to the Finnish Institute for Occupational Health in 1979. The sample of nurses used for this matched case-control study was restricted to those female nurses who were 40 years of age or younger in 1980 and who worked on units where ANDs may have been used. They considered nurses to be exposed to ANDs only if they prepared ANDs, a conservative criterion which included all but three nurses who exclusively administered ANDs. Data were obtained using a mail questionnaire in which nurses provided data on the number and types of ANDs prepared on a monthly basis during a two year time period that included the pregnancy of interest. Consequently, by using nurses as the respondents, it also was possible to obtain data on other lifetime occupational exposures, as well as lifestyle and reproductive factors. Both cumulative exposure and first trimester exposure models were examined initially, but only first trimester exposure was provided in the report as cumulative exposure was unrelated to fetal loss in this study. Conditional logistic regression was used to simultaneously adjust for age and hospital of employment and other factors. The findings showed a 2.3-fold increased risk of early fetal loss. The ANDs which were reported to be used during pregnancy by at least 10 nurses were cyclophosphamide, doxorubicin, fluorouracil, and vincristine (Table 3). Only fluorouracil was not associated with an increased risk of fetal loss in both of the models examined.
Two analytical studies and one case report were identified in the English literature that evaluated cancer outcomes in HCWs who were occupationally exposed to ANDs. Skov et al. (1990) conducted a nested case-control study to evaluate physicians' risk of leukemia and NHL. Cases and referents were identified from a cohort of 21,781 physicians who were members of the Danish Medical Association between 1965 and 1988. Since data on actual occupational exposure to ANDs were not available, Skov and colleagues defined exposure as working for 6 months or more in a department where patients are treated with ANDs, which was available from a description of professional activities contained in several sources. The earliest exposure occurred in 1950. A total of 20 physicians were newly diagnosed with leukemia and 25 with NHL. The conditional logistic regression estimate of the relative risk (RR) for developing leukemia was 2.9 (95% Confidence Interval (CI) = 0.5, 16.0). The RR for NHL was 0.7 (95% CI = 0.1, 4.3). The latency period from the time of first exposure to the development of leukemia and NHL ranged from 7 to 33 years. Average length of follow-up was not reported, nor were latency periods reported separately for the two outcomes.
In the second study, Skov and colleagues (1992) conducted an historical cohort study to evaluate the risk of leukemia in female oncology nurses. The nurses who worked in one of five oncology departments during the study period were identified through hospital administration records. Standardized incidence ratios were used to compare nurses' risk to the general population of women in Denmark. In four of the oncology departments, nurses began preparing or administering ANDs in the early 1970s, while in the other department, nurses began handling ANDs in 1976. The risk of leukemia among the 794 Danish oncology nurses who prepared or administered ANDs was 10.7 (95% CI = 1.3, 38.5) based on two cases of leukemia. The risk estimate was based on 5,636 women-years of observation, an average of 7 years per nurse. One nurse prepared an average of 5 treatments of ANDs on a weekly basis for four years beginning in 1974. In 1977, she developed Hodgkin's disease, which was treated with radiotherapy. Acute myelogenous leukemia developed only a few months later. The other nurse had prepared about five treatments of ANDs per week for a four month period during 1982. She was diagnosed with chronic myelogenous leukemia in 1987. Skov and colleagues noted that AND handling practices changed in 1980 following the report by Falck et al. (1979) that showed increased urine mutagenicity among oncology nurses. However, no information was provided on the extent to which protective equipment such as VLAF hoods, gloves, and other equipment was available, nor were training and compliance with safe work practices addressed.
A single case report of bladder cancer in a HCW with occupational exposure to ANDs was identified in the published literature. Levin et al. (1993) reported that a female non-smoking vegetarian pharmacist developed bladder cancer that was diagnosed at age 39. Twelve years prior to the diagnosis, she prepared approximately two to three doses of ANDs daily over a 20 month period. The most commonly prepared drugs in her practice were cyclophosphamide, fluorouracil, methotrexate, doxorubicin, and cisplatin. She was required by her employer to use a HLAF hood to prepare ANDs. The HLAF hood, unlike the VLAF hood directs the air flow from the back of the hood, over the drugs being prepared, through the opening where the preparer is situated, and into the environment. Thus, the HLAF hood would be expected to have increased the personal exposure of this pharmacist and further resulted in environmental contamination. Except for preparing ANDs, no other risk factors for bladder cancer could be identified. Because bladder cancer occurs predominantly in older white men and in cigarette smokers, this case is particularly noteworthy.
With respect to the study of leukemia among nurses occupationally exposed to ANDs (Skov et al., 1992), the relative risk of leukemia may be biased upward if the chronic myelogenous leukemia in the one nurse actually was unrelated to the short-term AND exposure as reported. However, even if only one case was relevant, the estimated risk (observed/expected) would be 1/0.19, or 5-fold. Compared to physicians (Skov et al., 1990), oncology nurses' risk, therefore, would be approximately double.
Alternatively, the findings of increased risk of leukemia among HCWs (Skov et al., 1990, 1992) may be considered conservative for a number of reasons. Imprecision in classification on exposure (or outcome) would bias the findings towards unity. In fact, Skov et al. (1990) used a proxy measure of exposure to ANDs for physicians, based on having worked for 6 months or longer on oncology units. It is not known from the paper whether physicians typically handled ANDs as part of their responsibilities or merely worked on units where these drugs were handled. Nonetheless, as expected, the risk estimate is less than that for oncology nurses for whom exposure was actually recorded or where estimates were possible. Notably, risk of leukemia for both physicians and nurses was found to be at the low end of the range found for patient populations who developed AND-therapy related leukemia.
Among the studies reviewed, the most methodologically sound were those by Hemminki et al. (1985) with respect to congenital malformations, Selevan et al. (1985) regarding early fetal loss, and Skov et al. (1990, 1992) with respect to a range of reproductive outcomes and leukemia. The studies by McDonald et al. (1988a, 1988b), Stucker et al. (1990, 1993), and Saurel-Cubizolles et al. (1993) were by design exploratory or preliminary in nature. In several reproductive studies the confidence intervals included unity so that chance may have explained the findings. In most of the cross sectional studies on reproductive outcomes, selection bias may have occurred. (It is unlikely that selection bias would have operated in the essentially population-based studies by McDonald and colleagues). Information bias and confounding were not significant problems in the studies reviewed. However, random misclassification on exposure may have reduced the risk estimate in several studies. Overall, the case series on liver toxicity, and the studies on adverse reproductive outcomes and cancer in relation to occupational exposure to ANDs were thorough, well-conducted case series and formal epidemiological studies. The study on menstrual cycle dysfunction (Shortridge et al., 1995) is consistent with laboratory data and with ovarian failure noted in women over age 30 treated with certain AND regimens. Altogether, these studies provide solid data on a range of adverse health effects that are, with the exception of ectopic pregnancy, well-documented in laboratory studies and patient populations.
In the studies reviewed that showed elevated risks, moderate to strong risks were observed for fetal loss, congenital malformations, and leukemia. Temporality was established for liver fibrosis, leukemia, and bladder cancer. However, the criterion of temporality was satisfied less consistently in the reproductive studies as cross-sectional and case-control study designs were used most frequently. In addition, in some instances, the most relevant period of exposure for a given reproductive outcome (Selevan & Lemasters, 1987) was not used.
The study findings regarding fetal loss and leukemia in HCWs were largely consistent with each other and very biologically plausible. The findings also were consistent with laboratory studies and epidemiological studies of patient populations in a number of ways. All outcomes addressed in the studies reviewed except ectopic pregnancy have been demonstrated in animal models and humans. The magnitude of risk for leukemia corresponded to the lower risk estimates in patient populations. The routes of exposure and mechanisms of action are consistent with industrial hygiene studies of exposure and laboratory studies of liver toxicity, reproductive outcomes, and cancer. Although attempts to measure dose-response were made in some studies, there were obvious problems including imprecision in measurement of exposure to ANDs (e.g., nurse managers' estimates of levels of exposure) and small sample sizes for the relatively rare outcomes examined. These issues were particularly evident for reproductive outcomes.
In summary, there is adequate limited evidence to link occupational exposure to ANDs -- preparing or administering ANDs, handling AND contaminated excreta, or working in environments where work practices for handling ANDs are inadequate -- with liver fibrosis, congenital malformations, fetal loss, leukemia, and bladder cancer. At this time, there is only preliminary evidence with respect to other reproductive outcomes (e.g., low birth weight, ectopic pregnancy, menstrual cycle dysfunction). Clearly, further research is needed on liver fibrosis, reproductive outcomes, and site-specific cancers in HCWs who have occupational exposure to ANDs in the interest of public health and sound policy-making.
References
American Society of Hospital Pharmacists. (1990). ASHP technical assistance bulletin on handling cytotoxic and hazardous drugs. American Journal of Hospital Pharmacy, 47, 1033-1049.
Association of Community Cancer Centers. (1992). Compendia-Based Drug Bulletin, Spring. Rockville, MD: Association of Community Cancer Centers.
Auerbach, C., & Falconer, D. S. (1949). A new mutant in the progeny of mice treated with nitrogen mustard. Nature, 163, 678-679.
Baker, G. L., Kahl, L. E., Zee, B. C., Stolzer, B. L., Agarwal, A. K., Medsger, T. A., Jr. (1987). Malignancy following treatment of rheumatoid arthritis with cyclophosphamide. American Journal of Medicine, 83(1), 1-9.
Banerjee, A., & Benedict, W. F. (1979). Production of sister chromatid exchanges by various cancer chemotherapeutic agents. Cancer Research, 39, 797-799.
Barry, L., & Booher, R. B. (1985). Promoting the responsible handling of antineoplastic agents in the community. Oncology Nursing Forum, 12, 41-46.
Bartolucci, A. A., Liu, C. Durant, J. R., & Gams, R. A. (1983). Acute myelogenous leukemia as a second malignant neoplasm following the successful treatment of advanced Hodgkin's disease. Cancer, 52, 2209-2213.
Benedict, W. F., Baker, M. S., Haroun, L., Choi, E., & Ames, B. N. (1977). Mutagenicity of cancer chemotherapeutic agents in the Salmonella/microsome test. Cancer Research, 37, 2209-2213.
Benhamou, S., Callais, F., Sancho-Garnier, H., Min, S. et al. (1986). Mutagenicity in urine from nurses handling cytostatic agents. European Journal of Cancer and Clinical Oncology, 22, 1489-1493.
Benhamou, S., Pot-Deprun, J., Sancho-Garnier, H., & Chouroulinkov, I. (1988). Sister chromatid exchanges and chromosomal aberrations in lymphocytes of nurses handling cytostatic agents. International Journal of Cancer, 41, 350-353.
Bersagel, D. E., Bailey, A. J., Langley, G. R., MacDonald, R. N., White, D. F., & Miller A. B. (1979). The chemotherapy of plasma-cell myeloma and the incidence of acute leukemia. New England Journal of Medicine, 301, 743-748.
Boice, J. D., Jr., Greene, M. H., Killen, J. Y., Jr., Ellenberg, S. S., Keehn, R. J., McFadden, E., Chen, T. T., & Fraumeni, J. F., Jr. (1983). Leukemia and preleukemia after adjuvant treatment of gastrointestinal cancer with semustine (methyl-CCNU). New England Journal of Medicine, 309, 1079-1084.
Boivin, J. F., Hutchison, G. B., Lyden, M., Godbold, J., Chorosh, J., & Schottenfeld, D. (1984). Second primary cancers following treatment of Hodgkin's Disease. Journal of the National Cancer Institute, 67, 751-760.
Boivin, J. F., Hutchison, G. B., Zauber, A. G., Bernstein, L., Davis, F. G., Michel, R. P., Zanke, B., Tan, C. T. C., Fuller, L. M., Mauch, P., & Ultmann, J. E. (1995). Incidence of second primary cancers in patients treated for Hodgkin's Disease. Journal of the National Cancer Institute, 87, 732-741.
Botta, J. A., Jr., Hawkins, H. C., & Weikel, J. H., Jr. (1974). Effects of cyclophosphamide on fertility and general reproductive performance of rats. Toxicology and Applied Pharmacology, 27, 602-611.
Brusick, D. J., Simmon, V. F., Rosenkranz, H. S., Ray, V. A., & Stafford, R. S. (1980). An evaluation of the Escherichia coli WP2 and WP2 uvrA reverse mutation assay. Mutation Research, 76, 169-190.
Buckner, J. C., Edmonson, J. H., Ingle, J. N., & Schaid, D. J. (1989). Evaluation of menogaril in patients with metastatic sarcomas and no prior chemotherapy exposure. American Journal of Clinical Oncology, 12(5), 384-386.
Chak, L. Y., Sikic, B. I., Tucker, M. A., Horns, R. C., Jr., & Cox, R. S. (1984). Increased incidence of acute nonlymphocytic leukemia following therapy in patients with small cell carcinoma of the lung. Journal of Clinical Oncology, 2, 385-390.
Chapman, R. M., Sutcliff, S. B., & Malpas, J. S. (1979). Cytotoxic-induced ovarian failure in women with Hodgkin's disease: Hormone function. Journal of the American Medical Association, 242, 1877-1881.
Christensen, C. J., Lemasters, G. K., & Wakeman, M. J. A. (1990). Work practices and policies of hospital pharmacists preparing antineoplastic agents. Journal of Occupational Medicine, 32, 508-512.
Cloak, M. M., Connor, T. H., Stevens, K. R., Theiss, J. C., Alt, J. M., Matney, T. S., & Anderson, R. W. (1985). Occupational exposure of nursing personnel to antineoplastic agents. Oncology Nursing Forum, 12, 33-39.
Coleman, M. P., Bell, C. M. J., & Fraser, P. (1987). Second primary malignancy after Hodgkin's disease, ovarian cancer and cancer of the testis: A population-based cohort study. British Journal of Cancer, 56, 349-355.
Coleman, R. E., Maisey, M. N., Knight, R. K., & Rubens, R. D. (1984). Mitoxantrone in advanced breast cancer--a phase II study with special attention to cardiotoxicity. European Journal of Cancer and Clinical Oncology, 20, 771-776.
Crudi, C. B., Stephens, B. L., & Maier, P. (1982). Possible occupational hazards associated with the preparation/administration of antineoplastic agents. National Intravenous Therapy Association, 5, 264-266.
Curtis, R. E., Hankey, B. F., Myers, M. H., & Young, J. L., Jr. (1984). Risk of leukemia associated with the first course of cancer treatment: An analysis of the Surveillance, Epidemiology, and End Results Program Experience. Journal of the National Cancer Institute. 72, 531-544.
D'Alessandro, N., Gebbia, N., Crescimanno, M., Flandina, C., Leto, G., Tumminello, F. M., & Messina, L. (1983). Effects of Amsacrine (m-AMSA), a new aminoacridine antitumor drug, on the rabbit heart. Cancer Treatment Reports, 67, 467-474.
Dale, G. A., & Smith, R. B. (1974). Transitional cell carcinoma of the bladder associated with cyclophosphamide. Journal of Urology, 112, 603-604.
Devereux, S., Selassie, T. G., Hudson, G. V., Hudson, B. V., & Linch, D. C. (1990). Leukemia complicating treatment for Hodgkin's disease: The experience of the British National Lymphoma Investigation. British Medical Journal, 301, 1077-1080.
Drago, J. R., & Goldman, L. B. (1980). Evaluation on nonhormonal cytotoxic chemotherapy in the Nb Rat (Pr-90) prostatic carcinoma. Cancer, 45, 757-763.
Einhorn, N., Eklund, G., Franzen, S., Lambert, B., Lindsten, J., & Soderhall, S. (1982). Late side effects of chemotherapy in ovarian carcinoma. Cancer, 49, 2234-2241.
Elliott, R. W., Essenigh, D. M., & Morley, A. R. (1982). Cyclophosphamide treatment of systemic lupus erythematosus--Risk of bladder cancer exceeds benefits. British Journal of Medicine, 284, 1160-1161.
Evelo, C. T. A., Bos, R. P., Peters, J. G. P., & Henderson, P. T. (1986). Urinary cyclophosphamide assay as a method for biological monitoring of occupational exposure to cyclophosphamide. International Archives of Occupational and Environmental Health, 58, 151-155.
Fairchild, W. V., Spence, C. R., Solomon, H. D., & Gangai, M. P. (1979). The incidence of bladder cancer after cyclophosphamide therapy. Journal of Urology, 122, 163-4.
Falck, K., Grohn, P., Sorsa, M., Vainio, H, Heinonen, E., & Holsti, L. R. (1979). Mutagenicity in urine of nurses handling cytostatic drugs. Lancet, 1, 1250-1251.
Fischer, D. S., & Knobf, M. T. (1989). The cancer chemotherapy handbook (3rd ed.). Chicago: Year Book Medical.
Frost, P., & DeVita, V. T. (1966). Pigmentation due to a new antitumor agent. Archives of Dermatology, 94, 265-268.
Generoso, W. M., Cain, K. T., Hughes, L. A., & Foxworth, L. B. (1989). A restudy of the efficacy of adriamycin in inducing dominant lethals in mouse spermatogonia stem cells. Mutation Research, 226, 61-64.
Gianni, L., Corden, B. J., & Myers, C. E. (1983). The biochemical basis of anthracycline toxicity and antitumor activity. In E. Hodgson, J. R. Bend, & R. M. Philpot (Eds.), Reviews in biochemical toxicology (pp. 1-82). New York: Elsevier.
Gilman, A., & Philips, F. S. (1946). The biological actions and therapeutic applications of the b-chlorethyl amines and sulfides. Science, 103(2675), 409-415.
Graf, U., Frei, H., Kagi, A., Katz, A. J., & Wurgler, F. E. (1989). Thirty compounds tested in the Drosophila wing spot test. Mutation Research, 222, 359-373.
Green, D. M., Zevon, M. A., Lowrie, G., Seigelstein, N., & Hall, B. (1991). Congenital anomalies in children of patients who received chemotherapy for cancer in childhood and adolescence. New England Journal of Medicine, 325, 141-146.
Haas, J. F., Kittelmann, B., Mehnert, W. H., Staneczek, W., Mohner, M., Kaldor, J. M., & Day, N.(1987). Risk of leukaemia in ovarian tumour and breast cancer patients following treatment by cyclophosphamide. British Journal of Cancer, 55, 213-218.
Hancock, S. L., Hoppe, R. T., Horning, S. J., & Rosenberg, S. A. (1988). Intercurrent death after Hodgkin disease therapy in radiotherapy and adjuvant MOPP trials. Annals of Internal Medicine, 109, 183-189.
Hawkins, M. M. (1991). Second leukaemia after epidodophyllotoxins. Lancet, 2, 1408.
Hemminki, K., Kyyronen, P., & Lindbohm, M.-L. (1985). Spontaneous abortions and malformations in the offspring of nurses exposed to anaesthetic gases, cytostatic drugs, and other potential hazards in hospitals, based on registered information of outcome. Journal of Epidemiology and Community Health, 39, 141-147.
Henderson, B. M., Dougherty, W. J., James, V. C., Tilley, L. P., & Noble, J. F. (1982). Safety assessment of a new anticancer compound, Mitoxantrone, in beagle dogs: Comparison with doxorubicin. I. Clinical observations. Cancer Treatment Reports, 66, 1139-1143.
Hewitt, J. B. (1986). Protocols used in the preparation, administration, and disposal of Adriamycin. Unpublished masters' thesis, University of Wisconsin-Milwaukee, Milwaukee, WI.
Hewitt, J. B., Misner, S. T., & Levin, P. F. (1993). Health hazards of nurses: Identifying workplace hazards and reducing risks. AWOHHN's Clinical Issues in Perinatal and Women's Health Nursing.
Hirst, M., Tse, S., Mills, D. G., Levin, L., & White, D. F. (1984). Occupational exposure to cyclophosphamide. Lancet, 1, 186-188.
Holmes, G. E., & Holmes, F. E. (1978). Pregnancy outcome of patients treated for Hodgkin's disease: A controlled study. Cancer, 41, 1317-1322.
Hoy, R. H., & Stump, L. M. (1984). Effect of air venting filter device on aerosol production from vials. American Journal of Hospital Pharmacy, 41, 324-326.
International Agency for Research on Cancer (IARC). (1975). IARC monograph on the evaluation of the carcinogenic risk of chemicals to humans, Vol. 9, Some aziridines, N-, S-, and O-mustards and selenium. Lyons, France: IARC.
International Agency for Research on Cancer (IARC). (1979). Chemicals and industrial processes associated with cancer in humans. IARC Monographs, Vols. 1-20. Lyons, France: IARC.
Johnson, D. H., Porter, L. L., List, A. F., Hande, K. R., Hainsworth, J. D., & Greco, A. F. (1986). Acute nonlymphocytic leukemia after treatment of small cell lung cancer. American Journal of Medicine, 81, 962-968.
Kaldor, J. M., Day, N. E., Bell, J., Clarke, E. A., Langmark, F., Karjalainen, S., Band, P. Pedersen, D., Choi, W., Blair, V., Henry-Amar, M., Prior, P., Assouline, D., Pompe-Kirn, V., Cartwright, R. A., Koch, M., Arslan, A., Fraser, P., Sutcliffe, S. B., Host, H., Hakama, M., & Stovall, M. (1992). Lung cancer following Hodgkin's disease: A case-control study. International Journal of Cancer, 52, 677-681.
Kimler, B. F., Mansfield, C. M., Svoboda, D. J., & Cox, G. G. (1984). Ultrastructural evidence of cardiac damage resulting from thoracic irradiation and anthracyclines in the rat. International Journal of Radiation Oncology, Biology, and Physics, 10, 1465-1469.
Kinlen, L. J. (1985). Incidence of cancer in rheumatoid arthritis and other disorders after immunosuppressive treatment. American Journal of Medicine, 78(Suppl1A), 44-49.
Kleinberg, M. L., & Quinn, M. J. (1981). Airborne drug levels in a laminar-flow hood. American Journal of Hospital Pharmacy, 38, 1301-1303.
Knobf, M. K., Fischer, D. S., & Welch-McCaffrey, D. (1984). Cancer chemotherapy: Treatment and care (2nd ed.). Boston: G. K. Hall Medical.
Kojima, S., Hzuka, H., Yamaguchi, , H., Tanuma, S.-I., Kochi, M., & Ueno, Y. (1994). Antioxidative activity of benzylideneascorbate and its effect on Adriamycin-induced cardiotoxicity. Anticancer Research, 14, 1875-1880.
Krakoff, I. H. (1991). Cancer chemotherapeutic and biologic agents. CA-A Cancer Journal for Clinicians, 41, 264-278.
Kris, M. G., Gralla, R. J., Kelsen, D. P., & Casper, E. S. (1985). Phase II trial of oral 4- demothoxydaunorubicin in patients with non-small cell lung cancer. American Journal of Clinical Oncology, 8, 377-379.
LeRoy, M. L., Roberts, M. J., & Theisen, J. A. (1983). Procedures for handling antineoplastic injections in Comprehensive Cancer Centers. American Journal of Hospital Pharmacy, 40, 601-603.
Levin, L. I., Holly, E. A., & Seward, J. P. (1993). Bladder cancer in a 39-year-old female pharmacist [letter]. Journal of the National Cancer Institute, 85(13), 1089-1091.
Lindpaintner, K., Lindpaintner, L. S., Wentworth, M., & Burns, C. P. (1986). Acute myocardial necrosis during administration of amsacrine. Cancer, 57, 1284-1286.
Mader, R. M., Rizovski, B., Steger, G. G., Rainer, H., Proprentner, R., & Kotz, R. (1995). Determination of methotrexate in human urine at nanomolar levels by high-performance liquid chromatography with column switching. Journal of Chromatography, 613, 311-316.
Martoni, A., Pacciarini, M. A., & Pannuti, F. (1985). Activity of 4-demethoxydaunorubicin by the oral route in advanced breast cancer. European Journal of Cancer and Clinical Oncology, 21, 803-806.
Matteson, E. L., Hickey, A. R., Maguire, L., Tilson, H. H., & Urowitz, M. B. (1991). Occurrence of neoplasia in patients with rheumatoid arthritis enrolled in a DMARD registry. Journal of Rheumatology, 18, 809-814.
McCann, J., Spingarn, N. E., Kobori, J., & Ames, B. N. (1975). Detection of carcinogens as mutagens: Bacterial tester strains with R factor plasmids. Proceedings of the National Academy of Science. 72, 979-983.
McCarthy, G. M., & Skillings, J. R. (1992). Orofacial complications of chemotherapy for breast cancer. Oral Surgery Oral Medicine and Oral Pathology, 74, 172-178.
McDiarmid, M., & Egan, T. (1988). Acute occupational exposure to antineoplastic agents. Journal of Occupational Medicine, 30(2) 984-987.
McDiarmid, M. A., Egan, T., Furio, M., Bonacci, M., & Watts, S. R. (1986). Sampling for airborne fluorouracil in a hospital drug preparation area. American Journal of Hospital Pharmacy, 43, 1942-1945.
McDiarmid, M. A., Kolodner, K., Humphrey, F., Putman, D., Jacobson-Kram, D. (1992). Baseline and phosphramide mustard-induced sister-chromatid exchanges in pharmacists handling anti-cancer drugs. Mutation Research, 279, 199-204.
McDevitt, J. J., Lees, P. S. J., & McDiarmid, M. A. (1993). Exposure of hospital pharmacists and nurses to antineoplastic agents. Journal of Occupational Medicine, 35, 57-60.
McDonald, A. D., McDonald, J. C., Armstrong, B., Cherry, N. M., Cote, R., Lavoie, J., Nolin, A. D., & Robert, D. (1988a). Congenital defects and work in pregnancy. British Journal of Industrial Medicine, 45, 581-588.
McDonald, A. D., McDonald, J. C., Armstrong, B., Cherry, N. M., Cote, R., Lavoie, J., Nolin, A. D., & Robert, D. (1988b). Fetal death and work in pregnancy. British Journal of Industrial Medicine, 45, 148-157.
Menard, D. B., Gisselbrecht, C., Marty, M., Reyes, F., & Dhumeaux, D. (1980). Antineoplastic agents and the liver. Gastroenterology, 78, 142-164.
Milunsky, A., Graef, J. W., & Gaynor, M. F. (1968). Methotrexate-induced congenital malformations. Journal of Pediatrics, 72, 790-795.
Mohn, G. R., & Ellenberger, J. (1980). Appreciation of the value of different bacterial test systems for detecting and for ranking chemical mutagens. Archives of Toxicology, 46, 45-60.
Mulvihill, J. J., McKeen, E. A., Rosner, F., & Zarrabi, M. H. (1987). Pregnancy outcome in cancer patients. Cancer, 60, 1143-1150.
Murphy, M. L., Moro, A. D., & Lacon, C. (1958). The comparative effects of five polyfunctional alkylating agents on the rat fetus, with additional notes on the chick embryo. Annals of the New York Academy of Science, 68, 762-782.
Neal, A. D., Wadden, R. A., & Chiou, W. L. (1983). Exposure of hospital workers to airborne antineoplastic agents. American Journal of Hospital Pharmacy, 40, 597-601.
Newman, M. A., Valanis, B. G., Schoeney, R. S., & Que Hee, S. (1994). Urinary biological monitoring markers of anticancer drug exposure in oncology nurses. American Journal of Public Health, 84, 852-855.
Occupational Safety and Health Administration. (1986). OSHA work practice guidelines for personnel dealing with cytotoxic (antineoplastic) drugs. Washington, DC: U.S. Department of Labor.
Occupational Safety and Health Administration. (1995). Controlling occupational exposure to hazardous drugs. OSHA Instruction TED 1.15. Washington, DC: Office of Science and Technology Assessment.
Pearson, R. M., & Soloway, M. S. (1978). Does cyclophosphamide induce bladder cancer? Urology, 11, 437-47.
Pedersen-Bjergaard, J., Ersboll, J., Sorensen, M., Keiding, N., Larsen, S. O., Philip, P, Larsen, M. S., Schultz, H. & Nissen, N. I. (1985a). Risk of acute nonlymphocytic leukemia and preleukemia in patients treated with cyclophosphamide for non-Hodgkin's lymphomas. Annals of Internal Medicine, 103, 195-200.
Pedersen-Bjergaard J., Hansen, S. W., & Larsen, S. O. (1991). Increased risk of myelodysplasia and leukemia after etoposide, cisplatin, and bleomycin for germ-cell tumours. Lancet, 338, 359-363.
Pedersen-Bjergaard J., & Larsen, S. O. (1982). Incidence of acute nonlymphocytic leukemia, preleukemia, and acute myeloproliferative syndrome up to 10 years after treatment of Hodgkin's Disease. New England Journal of Medicine, 307, 965-971.
Pedersen-Bjergaard, J., Osterlind, K., Hansen, M., Philip, P., Pedersen, A. G., & Hansen, H. H. (1985b). Acute nonlymphocytic leukemia, preleukemia, and solid tumors following intensive chemotherapy of small cell carcinoma of the lung. Blood, 66, 1393-1397.
Pedersen-Bjergaard, J., Philip, P., Larsen, S. O., Andersson, M., Daugaard, G., Ersboll, J., Hansen, S. W., Hou-Jensen, K., Nielsen, D., Sigsgaard, T. C., Specht, L., & Osterlind, K. (1993). Therapy-related myelodysplasia and acute myeloid leukemia. Cytogentic characteristics of 115 consecutive cases and risk in seven cohorts of patients treated intensively for malignant diseases in the Copenhagen Series. Leukemia, 7, 1975-1986.
Plotz, P. H., Klippel, J. H., Decker, J. L., Grauman, D., Wolff, B., Brown, B. C., & Rutt, G. (1979). Bladder complications in patients receiving cyclophosphamide for systemic lupus erythematosus or rheumatoid arthritis. Annals of Internal Medicine, 91, 221-223.
Powell, H. R., & Ekert, H. (1971). Methotrexate-induced congenital malformations. Medical Journal of Australia, 2, 1076-1077.
Reynolds, R. D., Ignoffo, R., Lawrence, J., Torti, F. M., Koret, K., Anson, N., & Meier, A. (1982). Adverse reactions to AMSA in medical personnel (letter). Cancer Treatment Reports, 66, 1885.
Reynoso, E. E., Shepherd, F. A., Messner, H. A., Farquharson, H. A., Garvey, M. B., & Baker, M. A. (1987). Acute leukemia during pregnancy: The Toronto Leukemia Study Group experience with long-term follow-up of children exposed in utero to chemotherapeutic agents. Journal of Clinical Oncology, 5, 1098-1106.
Rhoads, C. P. (1946). Nitrogen mustards in the treatment of neoplastic disease. Journal of the American Medical Association, 131, 656-658.
Rogers, B. (1987). Work practices of nurses who handle antineoplastic agents. AAOHN Journal, 35(1), 24-31.
Rogers, B., & Emmett, E. A. (1987). Handling antineoplastic agents: Urine mutagenicity in nurses. Image: Journal of Nursing Scholarship, 19, 108-113.
Ross, G. T. (1976). Congenital anomalies among children born of mothers receiving chemotherapy for gestational trophoblastic neoplasms. Cancer, 37, 1043-1047.
Saurel-Cubizolles, M. J., Job-Spira, N., & Estryn-Behar, M. (1993). Ectopic pregnancy and occupational exposure to antineoplastic drugs. Lancet, 341, 1169-1171.
Selevan, S. G., & Lemasters, G. K. (1987). The dose-response fallacy in human reproductive studies of toxic exposures. Journal of Occupational Medicine, 29, 451-454.
Selevan, S. G., Lindbohm, M.- L., Hornung, R. W., & Hemminki, K. (1985). A study of occupational exposure to antineoplastic drugs and fetal loss in nurses. New England Journal of Medicine, 313, 1173-1177.
Sessink, P. J. M., Cerna, M., Rossner, P., Pastorkova, A., Bavarova, H., Frankova, K., Anzion, R. B. M., & Bos, R. P. (1994). Urinary cyclophosphamide excretion and chromosomal aberrations in peripheral blood lymphocytes after occupational exposure to antineoplastic agents. Mutation Research, 309, 193-199.
Shaw, N. J., Tweedale, P. M., & Eden, O. B. (1989). Pulmonary function in childhood leukaemia survivors. Medical and Pediatric Oncology, 17, 149-154.
Shinar, E., & Hasin, Y. (1984). Acute electrocardiographic changes induced by amsacrine. Cancer Treatment Reports, 68, 1169-1172.
Shortridge, L. A. (1988). The effects on menstrual function associated with occupational exposure to antineoplastic drugs. Unpublished dissertation. Cincinnati, OH, University of Cincinnati.
Shortridge, L. A., Lemasters, G. K., & Valanis, B. (1995). Menstrual cycles in nurses handling antineoplastic drugs. Cancer Nursing, 18, 439-443.
Skov, T., Lynge, E., Maarup, B., Olsen, J., Rorth, M., & Wintherreik, H. (1990). Risks for physicians handling antineoplastic drugs [letter]. Lancet, 2, 1446.
Skov, T., Maarup, B., Olsen, J., Rorth, M., & Wintherreik, H., & Lynge, E. (992). Leukemia and reproductive outcome among nurses handling antineoplastic drugs. British Journal of Industrial Medicine, 49, 855-861.
Sotaniemi, E. A., Sutinen, S., Arranto, A.J., Sutinen, S., Sotaniemi, K. A., Lehtola, J., & Pelkonen, R. (1983). Liver damage in nurses handling cytostatic agents. Acta Medica Scandinavica, 214, 181-189.
Stajich, G. V., Barnett, C. W., Turner, S. V., & Henderson, C. (1986). Protective measures used by oncologic office nurses handling parenteral antineoplastic agents. Oncology Nursing Forum, 13, 47-49.
Stellman, J. M., Aufiero, B. M., & Taub, R. N. (1984). Assessment of potential exposure to antineoplastic agents in the health care setting. Preventive Medicine, 13, 245-255.
Stellman, J. M. (1987). The spread of chemotherapeutic agents at work: Assessment through simulations. Cancer Investigation, 5(2), 75-81.
Stephens, J. D., Golbus, M. S., Miller, T. R., Wilber, R. R., & Epstein, C. J. (1980). Multiple congenital anomalies in a fetus exposed to 5-fluorouracil during the first trimester. American Journal of Obstetrics and Gynecology, 137, 747-749.
Stucker, I., Caillard, J.-F., Collin, R., Gout, M., Poyen, D., & Hemon, D. (1990). Risk of spontaneous abortion among nurses handling antineoplastic drugs. Scandinavian Journal of Work and Environmental Health, 16, 102-107.
Stucker, I., Mandereau, L., & Hemon, D. (1993). Relationship between birthweight and occupational exposure to cytostatic drugs during or before pregnancy. Scandinavian Journal of Work and Environmental Health, 19, 148-153.
Tamatsu, H., Nakazawa, M., Imai, S., & Watari, H. (1984). 31P-topical Nuclear magnetic resonance (31P-TMR) studies of cardiotoxic effects of 5-fluorouracil (5-FU) and 5'-deoxy-5-fluorouridine (5'-DFUR). Japanese Journal of Pharmacology, 34, 375-379.
Tan, C. T. C., Hancock, C., Steinherz, P., Bacha, D. M., Steinherz, L., Luks, E., Winick, N., Meyers, P., Mondora, A., Dantis, E., Niedzwiecki, D., & Stevens, Y.-W. (1987). Phase I and clinical phamacological study of 4-demthoxydaunorubicin (Idarubicin) in children with advanced cancer. Cancer Research, 47, 2990-2995.
Taskinen, H., Lindbohm, M.-L., & Hemminki, K. (1986). Spontaneous abortions among women working in the pharmaceutical industry. British Journal of Industrial Medicine, 43, 199-205.
Tester, W. J., Kinsella, T. J., Waller, B., Makuch, R. W., Kelley, P., Glatstein, E., & DeVita, V. T. (1984). Second malignant neoplasms complicating Hodgkin's disease: The National Cancer Institute experience. Journal of Clinical Oncology, 2, 762-9.
Thiede, T., Chievitz, E., & Christensen, B. C. (1964). Chlornaphazine as a bladder carcinogen. Acta Medica Scandinavica, 175, 721-725.
Thilly, W. G., & Call, K. M. (1986). Genetic toxicology. In C. D. Klaassen, M. O. Amdur, & J. Doull (Eds.), Cassarett and Doull's Toxicology (3rd ed.) (pp. 174-194). New York: Macmillan.
Thiringer, G., Granung, G., Holmen, A., Hogstedt, B., Jarvholm, B., Jonsson, D., Persson, L., Wahlstrom, J., & Westin, J. (1991). Comparison of methods for the biomonitoring of nurses handling antitumor drugs. Scandinavian Journal of Work and Environmental Health, 17, 133-138.
Thompson, D. J., Melello, J. A., Strebing, R. J., & Dyke, I. L. (1978). Teratogenicity of Adriamycin and Daunomycin in the rat and rabbit. Teratology, 17, 151-158.
Tripathy, N. K., Wurgler, F. E., & Frei, H. (1990). Genetic toxicity of six carcinogens and six non-carcinogens in the Drosophila wing spot test. Mutation Research, 242, 169-180.
Tucker, M. A., Coleman, C. N., Cox, R. S., Varghese, A., & Rosenberg, S. A. (1988). Risk of second cancers after treatment for Hodgkin's disease. New England Journal of Medicine, 318, 76-81.
Tuffnell, P. G., Gannon, M. T., Dong, A., DeBoer, G., & Erlichman, C. (1986). Limitations of urinary mutagen assays for monitoring occupational exposure to antineoplastic drugs. American Journal of Hospital Pharmacy, 43, 344-348.
Unverferth, D. V., Unverferth, B. J., Balcerzak, S. P., Bashore, T. A., & Niedhart, J. A. (1983). Cardiac evaluation of mitoxantrone. Cancer Treatment Reports, 67, 343-350.
Valanis, B., & Browne, M. (1985). Use of protection by nurses during occupational handling of antineoplastic drugs. National Intravenous Therapy Association, 8, 218-222.
Valanis, B., Hertzberg, V., & Shortridge, L. (1987). Antineoplastic drugs: Handle with care. AAOHN Journal, 35, 487-492.
Valanis, B., McNeil, V., & Driscoll, K. (1991). Staff members' compliance with their facility's antineoplastic drug handling policy. Oncology Nursing Forum, 18, 571-576.
Valanis, B., & Shortridge, L. (1987). Self-protective practices of nurses handling antineoplastic drugs. Oncology Nursing Forum, 14, 23-27.
Valanis, B., Vollmer, W. M., & Glass, A. G. (1993). Acute symptoms associated with antineoplastic drug handling among nurses. Cancer Nursing, 16, 288-295.
Valanis, B., Vollmer, W. M., Labuhn, K., Glass, A. G., & Corelle, C. (1992). Antineoplastic drug handling protection after OSHA guidelines: Comparison by profession, handling activity, and work site. Journal of Occupational Medicine, 34, 149-155.
VanRaalte, J., Rice, C., & Moss, E. (1990). Visible-light system for detecting doxorubicin contamination on skin and surfaces. American Journal of Hospital Pharmacy, 47, 1067-1074.
Venitt, S., Crofton-Sleigh, C., Hunt, J., Speechley, V., & Briggs, K. (1984). Monitoring exposure of nursing and pharmacy personnel to cytotoxic drugs; Urinary mutation assays and urinary platinum as markers of absorption. Lancet, 1, 74-76.
Verdeguer, A., Ruiz, J. G., Ferris, J., Esquembre, C., Tasso, M. J., Fernandez, J. M., Prieto, F., & Castel, V. (1992). Acute nonlymphoblastic leukemia in children treated for acute lymphoblastic leukemia with an intensive regimen including teniposide. Medical and Pediatric Oncology, 20, 48-52.
Vorobiof, D. A., Iturralde, M., & Falkson, G. (1983). Amsacrine cardiotoxicity: Assessment of ventricular function by radionuclide angiography. Cancer Treatment Reports, 67, 1115-1117.
Wall, R. L., & Clausen, K. P. (1975). Carcinoma of the urinary bladder in patients receiving cyclophosphamide. New England Journal of Medicine, 293, 271-273.
Warkany, J. (1986). Aminopterin and methotrexate: Folic acid deficiency. In J. L. Sever, & R. L. Brent (Eds.), Teratogen update: Environmentally induced birth defect risks (pp. 39-43), New York: Alan R. Liss.
Weisburger, E. K. (1977). Bioassay program for carcinogenic hazards of cancer chemotherapeutic agents. Cancer, 40, 1935-1949.
Whitlock, J. A., Greer, J. P., & Lukens, J. N. (1991). Epipodophyllotoxin-related leukemia: Identification of a new subset of secondary leukemia. Cancer, 68, 600-604.
Williams, G. M., & Weisburger, J. H. (1986). Chemical carcinogens. In C. D. Klaassen, M. O. Amdur, & J. Doull (Eds.), Cassarett and Doull's Toxicology (3rd ed.) (pp. 99-173). New York: Macmillan.
Zeimet, A. G., Thaler, J., Abfalter, E., Marth, C., Dapunt, O. (1992). Secondary leukaemias after etoposide. Lancet, 2, 379-380.
| Alkylating Agents | Antimetabolites |
| Busulfan (Myleran)
Carboplatin (Paraplatin) Chlorambucil Cisplatin Cyclophosphamide (Cytoxan) Dacarbazine (DTIC-Dome) Dianhydrogalactitol * Dibromodulcitrol * Estramustine Phosphate Ifosphamide Mechlorethamine (Nitrogen Mustard) Melphalan (Phenylalanine Mustard) Thiotepa Uracil Mustard |
5-Azacytidine *
Bisantrene * Cytarabine (Cytosine Arabinoside) Dichloromethotrexate * Floxuridine (FUDR, 5-Fluorodeoxyuridine) 5-Fluorouracil (5FU) Hydroxyurea 6-Mercaptopurine (6MP) Methotrexate (Amethopterin) 6-Thioguanine Pentostatin * Pibobroman Tegafur * Triazinate * Trimetrexate Glucuronate * |
| Antibiotics | Mitotic Inhibitors |
| Aclarubicin *
Bleomycin Dactinomycin (Actinomycin D) Daunorubicin Doxorubicin (Adriamycin) Epirubicin * Esorubicin * Idarubicin Menogaril * Mitomycin C Plicamycin (Mithramycin) |
Etoposide (VP-16, VePesid)
Teniposide (VM-26) Vinblastine Vincristine Vindesine * |
| Nitrosoureas | Miscellaneous |
| Carmustine (BCNU)
Chlorozotozin (DCNU) * Lomustine (CCNU) Semustine (methyl-CCNU) Streptozocin |
Aminoglutethimide
Amsacrine (M-AMSA) Diaziquone * L-Asparaginase Levamisole * Methyl-GAG * Mitotane Mitoxantrone (Novantrone) Procarbazine (Matulane) Razoxane * |
| Figure 2. List of most antineoplastic drugs in current use or under investigation (*). List compiled from data in D. S. Fischer & T. M. Knobf (1989), The cancer chemotherapy handbook (3rd ed.), Chicago: Year Book Medical; and Association of Community Cancer Centers (Spring, 1992), Compendia-based drug bulletin, Rockville, MD: Author. |
|
| Buserelin
Conjugate Equine Estrogen (Premarin) Cortisone Chlorotriansene (Tace) Dexamethasone (Decadron) Diethylstilbestrol (DES) Ethinyl Estradiol (Estinyl) Fluoxymesterone (Halotestin) Flutamide Goserelin Acetate (Zoladex) Hydroxyprogesterone Caproate (Delalutin) Leuprolide Medroxyprogesterone Acetate (Provera) Megestrol Acetate (Megace) Prednisone Tamoxifen (Nolvadex) Testolactone (Teslac) Testosterone |
||
| Figure 3. | List of some hormone and hormone blocking agents in current use in the treatment of various cancers. List compiled from data in D. S. Fischer & T. M. Knobf (1989), The cancer chemotherapy handbook (3rd ed.), Chicago: Year Book Medical; and Association of Community Cancer Centers (Spring, 1992), Compendia-based drug bulletin, Rockville, MD: Author. |
|
| Table 1. Examples of Findings from Mutagenicity Testing of
Antineoplastic Drugs Using Three Types of Assays1 |
|||
| Antineoplastic Drug | Assays | ||
| Salmonella | Escheria coli | Drosophila
wing spot |
|
| Nitrogen mustard | +2, +3 | +6 | |
| Sulfur mustard | +6 | ||
| Busulfan | +7 | ||
| Melphalan | +2 | +8 | |
| Chlorambucil | +8 | ||
| Thiotepa | +2 | ||
| Cyclophosphamide | +2, +3 | +5, +6 | +7 |
| Ifosphamide | +2 | +5 | |
| Chlornaphazine | +2 | ||
| Dacarbazine | +3 | ||
| Cisplatin | +2, ++3 | +6 | |
| Carmustine | +3 | ||
| Procarbazine | +5 | ||
| Daunorubicin | +2 | +7 | |
| Doxorubicin | +2, ++3 | ||
| Dactinomycin | -2 | ||
| Bleomycin | -2 | ||
| Mitomycin | +4, +5 | +5, +6 | |
| Methotrexate | -2 | ||
| 6-Mercaptopurine | +2 | ||
| Cytosine arabinoside | +7 | ||
| 5-Fluorouracil | +7 | ||
| 5-Fluorodeoxyuridine | +7 | ||
| 1 Findings are described qualitatively as being positive (+) or negative (-); in the study by Tuffnell et al.,
the highest mutagenicity levels are indicated as being more positive (++) than other ANDs tested. 2 Benedict et al., 1977; 3 Tuffnell et al., 1986; 4 McCann et al., 1975; 5 Mohn & Ellenberger, 1980; 6 Brusick et al., 1980; 7 Graf et al., 1989; 8 Tripathy et al., 1990 |
|||
| Table 2. Studies of Adverse Reproductive Outcomes in Workers Occupationally Exposed to Antineoplastic Drugs |
||||||
| Author
(Year of Publication) |
Study Design, Methods, and
Sample |
Estimates of the Relative Risk (95% Confidence Interval) | ||||
| Birth Defects | Fetal Loss | Birth Weight | Prematurity | Other | ||
| Hemminki et al. (1985) | Case-control
(matched 1:3) Early fetal losses and congenital malformations identified through national computerized records; outcomes linked to nurses who worked in all general hospitals in Finland identified through a central registry of health personnel Estimates of 1st trimester exposure to ANDs provided by nurse managers and logs (at 2 hospitals) of AND prepared or prepared and administered |
Adjusted for age, and occupational
exposures1 Prepared ANDs < weekly 2.0 (0.7, 5.3) [based on 11 exposed cases] Prepared ANDs  weekly
4.7 (1.2, 18.1) [based on 8 exposed cases] |
Early fetal loss
Adjusted for age and occupational exposures2 [based on a total of 11 cases exposed to ANDs] Prepared ANDs < weekly 1.1 (0.6, 1.8) Prepared ANDs weekly
0.8 (0.3, 1.7) |
|||
| Selevan et al. (1985) | Case-control
(matched 1:3) Early fetal losses identified through national computerized hospital discharge registry and polyclinic data in Finland and linked to nurses who worked in 17 hospitals with high volume use of ANDs Exposure defined as preparing ANDs during 1st trimester; mail questionnaire used to obtain data from nurses |
Early fetal loss
Adjusted for age and occupational and other exposures3 2.30 (1.20, 4.39) [based on 18 cases exposed to ANDs during first trimester] |
||||
| Taskinen et al. (1986) | Case-control
(matched 1:3) Early fetal losses identified through national computerized hospital discharge registry and polyclinic data in Finland and linked to pharmaceutical workers exposure to ANDs and other pharmaceuticals during the 1st trimester was supplied by occupational health nurse or physician |
Early fetal loss
(1973-1980) Adjusted for age 2.8 (0.6, 14.0) [based on 3 exposed cases and 3 exposed controls] Drugs produced were primarily cyclophosphamide, and to a lesser extent, methotrexate, busulphan, chlorambucil, and mercaptopurine |
||||
| Rogers & Emmett (1987) | Cross-sectional
Oncology nurses in the Baltimore metropolitan area (n = 126) and community health nurses (n = 117) were surveyed; two 24 hour urine specimens were analyzed for mutagenicity using the Ames Salmonella test |
Early fetal loss
Unadjusted 2.5 (p = .04) with age-adjustment, AND exposure associated with untoward pregnancy outcome (12 early fetal losses and 1 congenital malformation) remained statistically significant (p = .03) |
||||
| McDonald et al.
(1988a) |
Cross-sectional
Risk of congential defect was examined in 47,913 pregnancies that occurred to women in Montreal who worked at least 15 hours per week at the time of conception; case-control analysis for subgroups with specific chemical exposures including 152 pregnancies among nurses and physicians who administered ANDs during pregnancy |
8 birth defects
vs. 4.05 expected 1.98 (p = .05) (1 each: Down syndrome, hypospadias, diaphragmatic hernia, dislocated hip; 2 each: club foot, hernia) |
||||
| McDonald et al.
(1988b) |
Cross-sectional
Risk of early and late fetal loss in previous pregnancies examined in 22,613 women who were interviewed immediately following the termination of their current pregnancy There were 63 nurses and physicians who administered ANDs during the first 4 weeks after the first missed menstrual period |
Early fetal loss
Observed/ Expected 0.97 (n.s.) Late fetal loss (stillbirth) Observed/ Expected 0/0.48 (n.s.) |
||||
| Stucker et al.
(1990) |
Cross-sectional
Oncology nurses and unexposed nurses from one of three hospitals and a cancer center in France were interviewed; Nurses on eligible cancer units had prepared at least 10 vials per week per year, while comparison units consisted of medical, cardiology, and endocrinology units in the same hospitals Exposure was defined as "any pregnancy occurring after or during a period of exposure to cytostatics" (p. 103) |
Early fetal loss
Adjusted for age, pregnancy order, smoking 1.7 (1.0, 2.8) based on 139 exposed pregnancies |
||||
| Skov et al.
(1992) |
Historical cohort
Pregnancy outcomes of oncology nurses and unexposed nurses in Denmark were identified through computerized birth records, hospital discharge register and the register of congenital malformations and linked to nurses through hospital administrative records |
Adjusted for age
oncology nurses not handling ANDs 1.21 (0.42, 3.49) low exposure group -- intermediate exposure group 0.66 (0.15, 2.81) high exposure group 1.36 (0.59, 3.14) based on 13 malformations occurring to nurses employed in oncology departments during pregnancy |
Adjusted for age
oncology nurses not handling ANDs 0.67 (0.20, 2.26) low exposure group 3.74 (1.17, 12.0) intermediate exposure group 0.71 (0.22, 2.36) high exposure group 0.49 (0.19, 1.26) based on 15 fetal losses occurring to nurses employed in oncology departments during pregnancy |
Randomly selected pregnancy, adjusted for gestational age, pregnancy order, sex of neonate
mean birth weight in grams oncology nurses 3397g unexposed nurses 3455g p = 0.11 |
Randomly selected pregnancy
mean weeks gestation oncology nurses 39.64 unexposed nurses 39.69 |
Sex Ratio
boys:girls 1.12 (0.79, 1.61) Ectopic Pregnancy [raw data analyzed using EpiInfo 6.0] 1.16 (0.41, 3.26) |
| Stucker et al.
(1993) |
Cross-sectional
Oncology nurses and unexposed nurses from one of three hospitals and a cancer center in France were interviewed; eligible cancer units had been dispensed a minimum of 10 vials per week per year, while comparison units consisted of medical, cardiology, and endocrinology units in the same hospitals Exposure was defined as "any pregnancy occurring after or during a period of exposure to [ANDs]" (p. 149) |
Crude
mean birth weight in grams exposed pregnancies 3201g unexposed pregnancies 3286g Adjusted for gestational age, body mass index, maternal age, parity, cigarett consumption, and gender -56g (-155.1--43.1) none of the differences in birth weight were statistically significant at the p = 0.05 level |
Crude
mean weeks gestation exposed pregnancies 38.8 (SE = 0.2) unexposed pregnancies 39.2 (SE = 0.1) |
Small (at or below 10th percentile) for Gestational Age
exposed pregnancies 9.3% unexposed pregnancies 10.1% |
||
| Saurel-Cubizolles
et al. (1993) |
Cross-sectional
Operating room staff (except physicians) at 18 hospitals in Paris were interviewed prior to their annual occupational health examination; data were obtained in 1987-88 on 734 pregnancies that occurred after January 1, 1970 Exposure was defined as 1st trimester exposure to ANDs and the risk estimate was calculated for only those nurses who did not work in the operating room |
Ectopic
Pregnancy Crude exposure to ANDs during first trimester 7.0% not exposed to ANDs during first trimester 1.0% p = 0.001 Adjusted for pregnancy number, previous fetal loss, smoking, and working in operating room 11.4 (2.7, 47.6) |
||||
| 1 Adjusted for age and the following occupational exposures: anesthetic gases, sterilizing agents, x-rays, and hexachlorophene
2 Adjusted for age and the following occupational exposures: anesthetic gases, sterilizing agents, x-rays, hexachlorophene, telephone duty, and third shift 3 Adjusted for age and the following exposures: first trimester exposure to anesthetic gases and x-rays, previous fetal loss, previous induced abortion, alcohol consumption, use of oral contraceptives or intrauterine device at time of conception; smoking was not adjusted for because there was no difference in the distribution of smoking between cases and controls |
||||||
| Table 3. Association of First Trimester Exposure to Specific Antineoplastic
Drugs on Risk of Early Fetal Loss in Nurses1 |
||
| Odds Ratio (95% Confidence Interval) | ||
| Model 1: Drugs examined one at a
time, adjusted for other factors2 |
Model 2: Drugs examined
simultaneously, adjusted for other factors2 |
|
| Doxorubicin | 3.96 (1.31, 11.97) | 2.45 (0.61, 9.78) |
| Cyclophosphamide | 2.66 (1.25, 5.71) | 1.77 (0.47, 6.69) |
| Fluorouracil | 1.70 (0.55, 5.21) | 0.62 (0.15, 2.53) |
| Vincristine | 2.46 (1.13, 5.37) | 1.50 (0.51, 4.42) |
| 1 Data from Selevan et al. (1985).
2 Adjusted for first trimester anesthetic gas and x-ray exposure, previous fetal loss and induced abortion, alcohol consumption, and use of contraception (oral contraceptives or intrauterine devices) at conception |
||