Other Toxicities

Hematopoietic and Neurologic Systems

N₂ O is the only anesthetic reported to produce hematologic toxicity and neurotoxicity with long-term administration. Both toxicities are the result of the interaction of N₂ O with vitamin B₁₂ and the disruption of several pathways involved in one-carbon chemistry. The biochemical basis of this effect is the oxidation of the cobalt in vitamin B₁₂ by a physicochemical reaction with N₂ O, which was previously discussed. The time required to produce megaloblastic hematopoiesis with N₂ O exposure varies among patients. In healthy patients undergoing routine surgery, mild megaloblastic bone marrow changes are not seen after 6 hours, but they are seen after about 12 hours of exposure to 50% N₂ O; after 24 hours of exposure, the changes are marked. ^[302] Complete bone marrow failure can be expected after several days of continuous exposure.^[303] Limited evidence suggests that N₂ O produces bone marrow changes earlier in seriously ill patients.^[304] Evidence also suggests that the bone marrow changes are preventable by pretreating patients with large doses of folinic acid.^[302] This drug is converted to the 5, 10-methylene tetrahydrofolate needed for thymidine synthesis ( Fig. 8-22 ). The neurologic disease, subacute combined degeneration of the spinal cord, develops only after several months of daily exposure to N₂ O. The symptoms and signs include numbness and paresthesia in the extremities, loss of balance and unsteady gait, impairment of touch, and muscle weakness. In the late 1970s, the disease was recognized to be similar to vitamin B₁₂ deficiency, but treatment with vitamin B₁₂ did not alleviate the symptoms or enhance recovery. ^[305] It occurs in those who abuse N₂ O on a long-term basis and rarely in individuals who work for many months in an environment grossly contaminated with the gas. Dental personnel who are occasionally exposed to waste N₂ O at levels greater than 1000 ppm in poorly ventilated dental operatories for long periods are particularly at risk. In one study, 3 of 20 dentists exposed to mean concentrations up to 4600 ppm had abnormal bone marrows.^[306] Personnel in modem-scavenged operating suites, however, are rarely exposed to such conditions and are not expected to have problems. Epidemiologic surveys confirm that dental, but not operating room, personnel have a higher incidence of neurologic disease, although exposure to waste N₂ O has not definitely been shown to be the cause.^[307] In humans and animals, there is an irreversible inactivation of the enzyme methionine synthase, which requires vitamin B₁₂ in the completely reduced form to act as its coenzyme. Methionine synthase catalyzes the conversion of methyltetrahydrofolate and homocysteine to tetrahydrofolate and methionine. Failure to produce these products has a number of biochemical consequences, including reduced synthesis of thymidine, which is an essential DNA base. The clinical syndrome associated with oxidation of vitamin B₁₂ is essentially the same as that seen in pernicious anemia: megaloblastic hematopoiesis and subacute combined degeneration of the spinal cord.

The time required for inactivation of methionine synthase depends on the species. In rats exposed to 50% N₂ O, the half-time of inactivation is about 5 minutes.^[308]

257

Figure 8-22 Conversion of methyltetrahydrofolate and homocysteine to tetrahydrofolate and methionine. (Adapted from Baden JM, Rice SA: Metabolism and toxicity of inhaled anesthetics. In Miller RD [ed]: Anesthesia, 5th ed. New York, Churchill Livingstone, 2000, p 147.)

Reproduction and Development

Inhaled anesthetics present the potential for adverse reproductive and developmental effects for patients administered inhaled anesthetics and health care personnel exposed to waste anesthetic gases. Exposure to waste anesthetic gases occurs because of leaks in the anesthetic system, principally from three sources: masks, high-pressure fittings, and exhalation valves. Good working practices and tracheal intubation results in lower exposure of waste anesthetic gases. Recognition of postoperative exposure to exhaled waste anesthetic gases in the postanesthesia care units (PACUs) and intensive care units (ICUs) represents a new challenge. To control occupational exposure to waste anesthetic gases, it is not sufficient to adopt scavenging systems in operating rooms; working environmental controls and occupational management are required throughout the perioperative process.

In the United States, at least 50,000 pregnant women undergo anesthesia and surgery during gestation for indications unrelated to pregnancy.^[310] Operations for ovarian cysts, acute appendicitis, mammary tumors, and repair of incompetent cervix are the most common causes. The risk of unexpected abortion or premature labor is higher after anesthesia. What is not immediately obvious is whether the patient's disease, surgery, anesthesia, or a combination of these factors is the precipitating cause. Perhaps an even greater concern is that anesthesia during pregnancy may lead to an increased incidence of congenital abnormalities in the offspring of the patient.

To assess the incidence of various anesthesia- and surgery-related hazards occurring during pregnancy, at least five major and several minor studies have been performed.^{[311] [312] [313] [314] [315] [316] [317]} The most extensive and thorough study links and analyzes data from three Swedish health care registries—the Medical Birth Registry, the Registry of Congenital Malformations, and the Hospital Discharge Registry—for the years 1973 to 1981.^[317] Among the 720,000 pregnant patients whose records were examined, there were 5405 operations performed. The adverse reproductive and developmental outcomes examined were stillbirths, early postnatal deaths, low birth weight, and congenital anomalies. Women who had an operation did not have an increased incidence of stillbirths or of congenital anomalies among offspring. The incidence of postnatal death within 7 days of delivery was increased when an operation occurred during the second or third trimester but not during the first trimester. The incidence of low birth weight and premature delivery was increased regardless of the trimester in which an operation occurred. The cause of these hazards was not determined, although the finding that no particular type of operation or anesthesia was associated with a higher incidence of adverse outcomes suggests that the patient's disease played the major role in determining the outcome.

The increasing popularity of in vitro fertilization has afforded some opportunity to study the effects of anesthetics on ova. Various anesthetic regimens have been examined for their effects on rates of fertilization and cleavage of oocytes, pregnancy, and carriage to term. Results for

258

Several studies have examined the effects of the inhaled anesthetics on sperm. In the only study of human sperm, semen was collected from 46 anesthesiologists who had worked for a minimum of 1 year in a scavenged operating suite.^[323] Semen from 26 residents beginning an anesthesia training program was used as controls. The concentration of sperm and the percentage of abnormally shaped sperm were not different between the two groups. Sperm collected from 13 of the 26 residents after they had been working for 1 year was not different from the first sample. Studies of mice and rats under various experimental conditions have provided both positive and negative results, but insufficient information is available to permit extrapolation to humans. Despite some suggestive animal data, inhaled anesthetics have not been shown to have significant effects on sperm in humans. However, the scope of animal and human

TABLE 8-8 -- Results of epidemiologic surveys of adverse reproductive effects among personnel exposed to waste anesthetic gases and their spouses

	Exposed Women		Wives of Exposed Men
Study (year)	Spontaneous Abortion	Major Anomaly in Offspring	Spontaneous Abortion	Major Anamoly in Offspring
Askrog and Harvald (1970)	Negative	—	Positive	—
Cohen et al (1971)	Positive	Negative	—	—
Knill-Jones et al (1972)	Negative	Positive	—	—
Rosenberg and Kirves (1973)	Positive	Negative	—	—
Corbett et al (1974)	—	Positive	—	—
American Society of Anesthesiologists (1974)	Positive	Positive	Negative	Positive
Knill-Jones et al (1975) (1)	Positive	Negative		Negative
Knill-Jones et al (1975) (2)	Positive	Positive	Negative	Positive
Cohen et al (1975)	Positive	Positive	Positive	Negative
Pharoah et al (1977)	Positive	Positive	—	—
From Baden JM, Rice SA: Metabolism and toxicity of inhaled anesthetics. In Miller RD (ed): Anesthesia, 5th ed. New York, Churchill Livingstone, 2000, p 147.

studies has been limited, and only moderate assurance of no effect has been achieved. Twenty reports have been published of epidemiologic surveys that examined the reproductive performance of operating room and dental personnel exposed to waste gases.^{[324] [325] [326] [327] [328] [329] [330] [331] [332] [333] [334] [335] [336] [337] [338] [339] [340] [341] [342] [343]} Table 8-8 lists the overall outcome of these surveys. An evaluation of the merits of each of these studies is beyond the scope of this chapter, which summarizes the outcomes, but a review of the studies is available elsewhere.^[344] In a review of many of the studies before 1978, an estimate was made of the relative risks for particular adverse health effects.^[345] The magnitude of the relative risk for spontaneous abortion among exposed women was approximately 1.3. The increase was reported to be consistent and statistically significant. For congenital abnormalities among offspring of anesthetic-exposed women, the relative risk was approximately 1.2. The overall data for wives of exposed men and for congenital abnormalities among their offspring were less consistent than the data for spontaneous abortion. For spontaneous abortion and congenital abnormalities, the increases observed were small and could not be attributed to a specific cause. Exposure to waste anesthetic gases, viruses, roentgenograms, a variety of chemicals other than anesthetics, or a combination of these factors might have accounted for the positive results. Most surveys had serious methodologic faults, including failure to verify the medical data supplied by respondents.

To resolve whether there were adverse effects that had not been appropriately acknowledged, in 1985, Buring and colleagues^[346] performed a meta-analysis of five studies.^{[325] [326] [327] [329] [338]} The investigators analyzed spontaneous abortion and congenital abnormalities among female physicians and nurses potentially exposed to waste anesthetics. Risk ratios were 1.3 (95% confidence interval [CI]: 1.2–1.4) for spontaneous abortion and 1.2 (95% CI: 1.0–1.4) for congenital abnormalities, both of which were statistically significant. However, response or

259

A retrospective cohort study design, which was subject to selection and recall bias, was used for most of the epidemiologic studies listed in Table 8-8 . One of the weakest links in the epidemiologic studies of adverse reproductive and developmental outcomes after occupational anesthetic exposure is the lack of quantification of the extent and duration of anesthetic exposure. Data generally were collected by self-administered questionnaire without verification of exposure or outcome. Because the extent and duration of exposure to waste anesthetics has decreased in recent years, it is difficult to draw conclusions on the likely adverse effects of current anesthetic exposure based on earlier studies at conditions of higher exposure. In the only three studies in which medical records were used to confirm medical data, negative results were obtained for a number of adverse reproductive effects, including spontaneous abortion.^{[334] [339] [340]}

In addition to problems in quantification of exposure, there is also a problem of identifying the cause of any congenital malformation that is observed. Although human development is remarkably consistent, it is not always perfect, and severe and trivial congenital malformations are found in 2% to 4% of all births in countries in which records are kept.^[347] There are many mechanisms involved in abnormal development ( Table 8-9 ). Some, such as mutations that result in specific biochemical abnormalities and chromosomal nondisjunction, are well established. Others, such as interference with abnormal cell membrane states, are less certain mechanisms of teratogenicity. Although the cause of most defects remains unknown, chemical teratogenesis in humans is

TABLE 8-9 -- Epidemiologic surveys for cancer incidence or deaths among personnel exposed to waste anesthetic gases

Studied Result	Population	Results	Study (year)
Deaths	ASA members	Negative	Bruce et al. (1968)
Incidence	Nurse anesthetists	3.3- fold increase for 1971	Corbett et al. (1973)
Incidence	ASA members	1.3- to 1.9-fold increase for women; negative for men	ASA (1974)
Deaths	ASA members	Negative	Bruce et al. (1974)
Deaths	Anesthetists	Negative	Doll and Peto (1977)
Deaths	Anesthetists and offspring	Negative	Tomlin (1979)
Incidence	Dental personnel	1.5-fold increase for women; negative for men	Cohen et al. (1980)
Deaths	Anesthetists	Negative	Neil et al. (1987)
Incidence	Operating room personnel	Negative	Cuirguis et al. (1990)
ASA, American Society of Anesthesiologists.
From Baden JM, Rice SA: Metabolism and toxicity of inhaled anesthetics. In Miller RD (ed): Anesthesia, 5th ed. New York, Churchill Livingstone, 2000, p 147.

well established. Because of the difficulty in identifying causes and investigating mechanisms in humans, sophisticated animal studies have been developed over the past 40 years for examining the teratogenicity of chemicals. Complete assessment of a chemical now involves examining its effect on mating behavior, fertility, embryonic and fetal wastage, congenital anomalies, and postnatal survival and behavior. Animal studies provide useful information, but they are limited because of the difficulty of extrapolating animal data to humans and the relatively low statistical power to detect the presence of a weak teratogen.

Developmental effects of the inhaled anesthetics have also been extensively studied in animals in an attempt to control conditions that cannot be controlled in epidemiologic studies. These studies tried to control for the many factors (e.g., genetic variability, environment, diet) that could contribute to the background incidence of congenital malformations in humans. Studies of the inhaled anesthetics on reproductive processes of experimental animals have been conducted at trace and subanesthetic concentrations to simulate occupational exposure and at anesthetic concentrations to simulate surgical exposure. The large number of such studies precludes a complete discussion in this chapter, but several reviews are available.^{[344] [348] [349]} In general, N₂ O is the only inhaled anesthetic that has been convincingly shown to be directly teratogenic to experimental animals, but only under experimental conditions that are considered extreme. High concentrations (50% to 75%) delivered to rats for 24-hour periods during organogenesis and low concentrations (0.1%) delivered to rats continuously throughout pregnancy result in an increased incidence of fetal resorptions and visceral and skeletal abnormalities. ^{[350] [351]} A particularly interesting abnormality produced by high N₂ O exposure is situs inversus, in which there is a disturbance of the left-right body axis, and organs such as the heart are on the wrong side of the body.^[352] Although effects in rodents are seen only after extended periods of continuous exposure, it is not known whether humans are more sensitive than rodents and would show effects after shorter periods of exposure.

260

The mechanisms that produce teratogenic effects with N₂ O are slowly being revealed. The former leading theory was that inhibition of methionine synthase causes critical shortages of intracellular thymidine (and therefore DNA), which is needed by the rapidly growing embryo. Decreased DNA synthesis and decreased total DNA content have been observed in embryos immediately after N₂ O exposure.^{[353] [354]} It appears that lack of methionine rather than the lack of thymidine plays the critical role in all the adverse reproductive effects other than situs inversus.^[355] An entirely different mechanism accounts for the situs inversus: stimulation of α₁ -adrenergic receptors by N₂ O.^[356] The effect is mediated by activation of calmodulin kinase-11 but not phosphokinase C.^[357]

N₂ O teratogenesis is being actively investigated because of its relevance to patient and occupational safety and because N₂ O can be used to determine the importance of vitamin B₁₂ , folates, one-carbon chemistry, and the adrenergic system in developmental processes. Halothane, enflurane, isoflurane, desflurane, and sevoflurane are also teratogenic to rodents, but only when they are administered at anesthetizing concentrations for many hours on several days during pregnancy. The consensus is that the teratogenic effects observed in animals exposed to the inhaled anesthetics are caused by physiologic changes associated with anesthesia, rather than by inherent teratologic properties of the anesthetics themselves. Nonetheless, all findings emphasize the potential for anesthetics to interfere with developmental processes, regardless of the mechanism.

Developmental toxicity also encompasses the cognitive and functional deficits that may occur in the absence of observable morphologic changes. Organ systems are most sensitive to chemical teratogens during periods of development (i.e., organogenesis). The central nervous system may be particularly vulnerable during the period of myelination. In humans, this period is from the fourth intrauterine month through the second postnatal year. A chemical or drug may produce behavioral teratogenesis if administered late in gestation or even after birth. Anesthetics have not escaped scrutiny as possible behavioral teratogens. A study of 1-day-old rats exposed to 0.75% or 1.0% halothane with or without 50% or 75% N₂ O for 6 hours demonstrated dose-dependent decrements in phosphokinase C activity in growth cone particles isolated from the forebrains.^[358] There were no effects at 0.5% halothane or 25% N₂ O. Although the anesthetic concentrations causing effects were high, the importance of phosphokinase C in signal transductions in developing brain and the potential disruption by anesthetics cannot be ignored. Although some rodent studies have shown behavioral deficits after exposure to the inhaled anesthetics,^{[359] [360] [361]} the mechanism for these changes is unknown, and the applicability to humans is unclear. In general, the anesthetic exposure conditions can be considered extreme relative to any reasonably expected human exposure.

Human studies have generally focused on long-term behavioral effects of maternal obstetric medication, including epidural anesthesia. Claims have been made that the medication given at delivery to the mother produces depressed motor skills and impaired language ability in the infant and child for several years. Such claims, however, are controversial. Behavioral abnormalities of offspring whose mothers received inhaled anesthetics at any time during delivery have not been well studied. Studies have not been done to assess neurobehavioral function of children of operating room personnel who have been exposed to waste anesthetic gases. Firm conclusions about the risk of the occurrence of behavioral teratogenesis among the offspring of exposed personnel or in exposed patients therefore await further investigation.

Long-term occupational exposure to trace concentrations of anesthetic gases is thought to have adverse effects on health of exposed personnel, and there is growing awareness and concern regarding occupational exposure to waste anesthetic gases in the PACUs. Several studies have documented excessive levels of waste anesthetic gases in poorly ventilated PACU areas.^{[362] [363] [364] [365]} However, none of these studies has documented any significant, long-term adverse health effects. ^{[362] [363]} An increased incidence of sister-chromatic exchanges has been found in operating room personnel in three of four studies.^{[366] [367] [368] [369]}

DNA, Genetics, and Neoplasia

The potential long-term adverse health effects from occupational exposure to trace concentrations of waste anesthetic gases have been recognized for more than 20 years. Even if inhaled anesthetics have a low potential for causing long-term toxicity, exposure of a large population may represent a considerable public health hazard. In the United States, about 50,000 hospital operating room personnel, including anesthesiologists, nurse anesthetists, and operating room technicians, are exposed daily to waste anesthetic gases.^[310] Surgeons, dental personnel, veterinarians, and their technical assistants have a variable but sometimes heavy exposure to waste anesthetics. The total number of exposed or potentially exposed personnel in the United States each year is about 225,000.^[310] Of particular concern are reports that inhaled anesthetics possess mutagenic and carcinogenic potentials.

Genotoxicity

Investigators have been interested in the mutagenic potential of inhaled anesthetics for several reasons. First, many chemical carcinogens are also mutagens, although the reverse is not necessarily true. Finding that a particular anesthetic is a mutagen also implies a potential for carcinogenesis that should be evaluated. Second, mutagens may pose a threat to the integrity of the human genome (i.e., the totality of genes and chromosomes). Mutations are heritable changes in genetic information. Unrepaired mutations in germ cells can be passed from generation to generation, and unrepaired mutations in somatic cells can result in diseases, including cancer. The four types of mutations traditionally recognized are base-pair mutations, frameshift mutations, deletions or rearrangements of chromosomal segments, and nondisjunction of chromosomes between daughter cells. A fifth and novel type of mutation has been recognized that involves continuous repeats of three nucleotides at a specific site in a gene.^[370]

261

In vitro assays for mutagenicity require much less time and expense than in vivo tests for carcinogenicity, and these assays have become popular screening methods for detecting carcinogens. A wide variety of test systems has been used to examine the mutagenicity of inhaled anesthetics, including assays with bacteria, yeast, mammalian cells in culture, and intact mammals.

Extensive work has been done with the Ames Salmonella/mammalian hepatic microsome system, which uses several strains of histidine-dependent Salmonella typhimurium as tester organisms. This system is a well-validated assay for mutagenicity and often is regarded as the standard against which other systems are compared. It has been used to test most currently and formerly used anesthetics. Only divinyl ether and fluroxene give unequivocally positive mutagenic responses, whereas trichloroethylene gives a weak mutagenic response. Other anesthetics, including N₂ O, halothane, enflurane, isoflurane, sevoflurane, and desflurane are not mutagenic when tested under a wide variety of experimental conditions and anesthetic concentrations. The finding that some mutagenic anesthetic metabolites have a chemical structure reminiscent of vinyl chloride (CH₂ =CHCl) is consistent with the known high reactivity and mutagenicity of this class of chemicals. Of the many inhaled anesthetic metabolites that have been tested in the Salmonella assay, only 1,1-difluoro-2-bromo-2-chloroethylene (CF₂ =CBrCl) and 1,1-difluoro-2-chloroethylene (CF₂ =CHCl) have been found to be even weakly mutagenic. In general, results from numerous studies with other test systems have confirmed those from studies with Salmonella.^{[33] [369]} Although there are some anomalous results, the overwhelming evidence from in vitro tests indicates that all currently used and most previously used anesthetics are not mutagens. Compound A has been reported to induce sister chromatic exchanges in Chinese hamster ovary cells.^[371] Halothane, enflurane, isoflurane, and desflurane all inhibit the production of nitric oxide, N₂ O, nitric oxide synthase (NOS), and messenger RNA for NOS in a dose- and time-dependent manner in a marine macrophage.^[372]

Results of genotoxicity studies of humans exposed to inhaled anesthetics have generally been negative. In an early study from 1977,^[372] no significant difference in the number of chromosomal aberrations could be detected between the lymphocytes obtained from operating room nurses and those obtained from surgical outpatient nurses. Mutagenic activity was not detected in the Salmonella assay for urine of personnel working in scavenged or unscavenged operating rooms or for urine of anesthesiology residents during the first 15 months of training. ^[373] Lymphocytes from personnel exposed to waste anesthetic gases for durations up to 312 months had no higher incidence of chromosomal aberrations or sister chromatic exchange compared with lymphocytes from unexposed individuals.^[374] The latter study was an extension of previous negative studies by the same investigators of sister chromatic exchanges in lymphocytes of operating room personnel exposed to waste anesthetic gases and of patients anesthetized with halothane, enflurane, or fluroxene. In contrast to these negative results, studies in 1990^[375] and 1992^[364] showed cytogenic damage in operating room personnel exposed to waste anesthetic gases. A 1993 study^[366] showed DNA single-strand breaks in lymphocytes of patients immediately after anesthesia with isoflurane and N₂ O. Why these differences in study results exist is unknown, although positive results could be explained by multiple factors unassociated with the anesthetics. As with the teratogenicity and carcinogenicity studies described elsewhere in this chapter, genotoxicity studies of operating room personnel are more realistically studies of the work environment, which is composed of many chemical and physical agents as well as anesthetic gases.

Inhaled Anesthetics and Carcinogenicity

Chemical carcinogenesis is a multistage, multistep process leading to the formation of malignant tumors. There are two general classes of carcinogens: genotoxic and nongenotoxic. The former type directly interacts with DNA, and the latter type does not. Chemical carcinogenesis consists of three operationally defined stages: initiation, promotion, and progression.

Genotoxic carcinogens are often referred to as initiators because, in interacting with DNA, they initiate the process in normal cells. Initiation is most often accomplished in several steps, the first of which is the metabolic activation of the chemical to a reactive intermediate. The second step consists of the covalent binding of the reactive intermediate to DNA. One or more cell divisions are required to "set" or "fix" the mutation. Before the mutation is fixed, the DNA may still be repaired. Initiation is accomplished only when the mutation is fixed. Once initiation has occurred, there is still no guarantee that a cell with defective DNA will survive a sufficiently long period to proceed in the process of carcinogenesis because initiated cells are also subject to apoptosis.

Nongenotoxic carcinogens, unlike genotoxic carcinogens, do not initiate cells. Nongenotoxic agents act through other mechanisms to promote the growth of initiated cells; the term promoter is often applied to these carcinogens. These agents may interact with other cellular components to promote the growth of these cells. A common mechanism is increased rate of cellular proliferation, preventing the normal repair process. One thing in common with these agents is that the promotion of initiated cells into tumors requires the continued presence of the promoter. Without its presence, apoptosis results in the elimination of single initiated cells and natural regression of the preneoplastic tumors.

The terms initiator and promoter are used to indicate the state of carcinogenesis in which a chemical is first active. An initiator may also have activity as a promoter. The phase of progression that results in the appearance of clinically apparent tumors involves additional phenotypic and genotypic changes to cells, resulting in karyotypic instability. These mechanisms are not understood, and many cancers appear to occur over a prolonged period. Covalent binding of reactive intermediates to tissue macromolecules is presumed to be a necessary requirement for initiation of chemical carcinogenesis. Covalent binding, however, is not sufficient for initiation because the binding must be specific to a DNA site that would result in miscoding of a necessary gene product or

262

Structural evaluation of chemicals for similarities to known carcinogens is useful to identify potential mutagens, but minor changes in structure often have profound effects on mutagenic and carcinogenic activity. Methoxyflurane, enflurane, and isoflurane are α-haloethers, as are the nonanesthetic but carcinogenic chemicals bis(chloromethyl)ether, chloromethyl methyl ether, and bis(α-chloroethyl)ether. The anesthetics, however, are not mutagenic. Halothane is a halide, as are the animal carcinogens methyl iodide, butyl bromide, and butyl chloride. Halothane, however, is not mutagenic. Fluroxene and divinyl ether also contain the vinyl moiety similar to vinyl chloride, a human and animal carcinogen. Although the structural similarity between the anesthetics and several chemical carcinogens suggests an association between anesthetics and carcinogenicity, there are no epidemiologic or other data to support that suggestion. Despite the obvious advantage of surveying human populations to determine the carcinogenic risk of exposure to anesthetics, such surveys have provided little information on the carcinogenicity of specific anesthetics. The primary reason is that the doses of anesthetics to which surveyed individuals have been exposed have not been measured and, at best, have been estimated. Nonetheless, the studies performed should indicate whether waste anesthetics are associated with a higher incidence of cancer.

There have been several surveys of cancer incidence among operating room and dental workers^{[329] [335] [341] [376] [377] [378] [379] [380]} (see Table 8-9 ). The largest was an ASA-conducted retrospective survey^[48] of 595 operating room personnel working throughout the United States.^[333] A 1.3- to 1.9-fold increase in cancer incidence was found for female members of the ASA and the American Association of Nurse Anesthetists compared with unexposed control female groups. No increase in cancer incidence was seen among surveyed mates. In an earlier study, the incidence of cancer among 525 female nurse anesthetists in Michigan was compared with that of women participating in the Connecticut Tumor Registry.^[377] The nurse anesthetists had a higher incidence of malignancies diagnosed during 1971 than did all the women of Connecticut during the period of 1966 to 1969.

In another large study, a national survey of health among dental personnel was reported.^[336] The study population consisted of 30,650 dentists and 30,547 chairside assistants and was readily divided into those who used or did not use inhaled anesthetics to provide pain relief and sedation of patients. Otherwise, both groups did similar work under similar conditions. An estimate of anesthetic exposure was made by calculating the number of hours per week spent by each respondent in the dental operatory. About 80% of inhaled anesthetics users were exposed to N₂ O alone, whereas the remainder was also exposed to potent volatile anesthetics. The cancer incidence among female chairside assistants exposed to waste anesthetics for more than 8 hours per week was 50% greater than among those not exposed, although the increase was not statistically significant (P = .056). Analysis of various types of cancer showed that only cancer of the cervix occurred more frequently (P = .06) in exposed women than in unexposed women. However, a 1990 study of health effects associated with waste anesthetic exposure in Ontario hospital personnel did not show an increased cancer incidence for men or women compared with unexposed controls. ^[341]

Collectively, the foregoing studies of cancer incidence appear to suggest a small risk to women directly exposed to waste anesthetic gases. However, because of problems with study design and the small increase in cancer incidence observed, reviewers have generally been unconvinced of the existence of a hazard to humans.^{[345] [348] [349]} This lack of conviction is strengthened by the uniformly negative results from surveys of deaths from cancer^{[335] [376] [378] [379] [380]} (see Table 8-9 ). Problems in interpreting epidemiologic surveys led many investigators to turn to animal studies to provide information on the carcinogenic potential of specific anesthetics. In early studies, chloroform was found to be a rodent carcinogen when the drug was administered in very high doses by oral gavage, but this route of administration is not relevant to inhaled exposure of patients or operating room personnel. It is now recognized that hepatic tumorigenesis in rodents with chloroform is related to high-dose cytotoxicity and the subsequent increased rate of reparative proliferation. At lower doses, at which cytotoxicity is not present, tumors do not form. When halothane, methoxyflurane, enflurane, isoflurane, and N₂ O have been administered by inhalation and assessed in adequate studies, results for carcinogenicity have been uniformly negative.^{[381] [382] [383] [384] [385] [386]} In none of these studies was there evidence of other types of organ toxicity on histologic examination. The conclusion from animal and human studies is that there is no carcinogenic risk from exposure to the currently used inhaled anesthetics.

Modern fluorinated inhaled anesthetics are not mutagenic, nor are they carcinogenic. N₂ O in concentrations greater than 50% is teratogenic in animals, as is exposure to the fluorinated anesthetics for several hours during pregnancy in animals. The analyses of Burning and coworkers,^[346] Tannenbaum and Goldberg,^[387] and Axelsson and Rylander ^[338] demonstrate that flawed studies can lead to inaccurate conclusions. In the only long-term, prospective study with or without scavenging of waste anesthetic gases, Spence^[388] found no causal relationship between exposure to waste anesthetic gases and adverse health effects. In the dental studies, with no scavenging of waste anesthetic gases and higher levels of N₂ O, increased incidences of infertility and spontaneous abortions were observed. Neither animal nor epidemiologic studies have proved that adverse effects are associated with exposure to trace levels of waste fluorinated anesthetic gases such as halothane, enflurane, isoflurane, desflurane, or sevoflurane.