Cancer Chemotherapy & Biotherapy: Principles & Practices, 4th Edition

Carcinogenesis of Anticancer Drugs

Lawrence S. Blaszkowsky

Charles Erlichman

Although the potential of antineoplastic agents to induce new malignancies was suggested by Haddow1 in 1947 on the basis of the ability of chemical carcinogens to cause growth inhibition, convincing evidence for carcinogenic effects of these agents in humans has been reported only in the past 40 years. The major reasons for this belated recognition of the problem are the long latency periods seen for expression of drug-induced carcinogenicity in humans (3 to 4 years) and the brief survival of most patients treated with chemotherapy. Only in the past 4 decades have a sizable number of patients with advanced malignancy been cured by chemotherapy or been treated with adjuvant chemotherapy; thus, sufficient time has elapsed and sufficient numbers of individuals are now at risk for second tumors to be seen in clinically significant numbers. Although survival benefits will undoubtedly continue to accrue from the use of these agents and will probably outweigh the risks of second neoplasms, concern for this complication is likely to grow as the use of antineoplastic drugs gain wider use in adjuvant programs and in non-neoplastic conditions such as renal transplantation or autoimmune disease, in which long-term survival of a large fraction of the treated population is ensured. In these instances, the benefits and risks (including carcinogenicity) of antineoplastic agents must be considered.

Definition of the risk of carcinogenesis as the result of chemotherapy is a difficult task. Prediction of carcinogenicity at the experimental level depends on test systems that examine the ability of chemicals to cause mutation of bacteria or mammalian cells, malignant transformation of mammalian cells, chromosomal aberrations, or tumors in mice or rats. Such tests are subject to interspecies variability in drug metabolism and target-tissue kinetics, and to other host-specific factors that influence susceptibilities to tumor development. These factors make extrapolation of the quantitative risk to humans a difficult, if not impossible, task. Second, the immune status of the patient is believed to play an important role in determining carcinogenicity, as indicated by the increased risk of lymphoid and cutaneous neoplasms in patients receiving immunosuppressive therapy; in cancer patients, the immune system is suppressed both as a result of the neoplastic process and as a consequence of therapy. This immunosuppression undoubtedly influences the risk of carcinogenesis but is not duplicated in test systems. Finally, the assessment of risk in humans at present is based partly on analyses of retrospective series, which often give incomplete information regarding key parameters of treatment (dose, duration) and which lack a control or untreated population. Such a control population is particularly important in risk assessment because an increased incidence of second tumors, such as acute myelocytic leukemia in patients with Hodgkin's disease, may exist in the absence of treatment. More recently, analyses of randomized trials that compare adjuvant chemotherapy regimens with no additional treatment have been undertaken with respect to incidence of second malignancies.2, 3, 4, 5, 6 The information derived from these studies clarifies some of the confounding variables mentioned.

With these limitations in mind, this chapter considers available information concerning the carcinogenic potential of antitumor agents. This discussion examines the common pharmacologic properties shared by antineoplastic agents and classic carcinogens, specific predictions of carcinogenicity based on nonhuman test systems, and clinical evidence for an increased risk of second neoplasms in patients receiving these agents.

RELATIONSHIP OF ANTINEOPLASTIC AGENTS TO CHEMICAL CARCINOGENS

The chemical induction of cancer in animals is thought to involve a multistage process with a long latency period. This process can be initiated by a variety of chemical structures that have at least one common thread in their mode of action: an interaction with DNA.7, 8, 9 Initiation results from irreversible genetic alterations, such as mutations or deletions in DNA.10 One of the most carefully studied systems of tumor induction is the induction of skin cancer in mice and rabbits by alkylating agents, polycyclic hydrocarbons, and ethyl carbamate (Fig. 5.1). Repeated applications of these agents over long periods result in the development of benign or malignant tumors. Exposure to these compounds in limited doses, however, causes morphologic changes in the epithelium but does not result in tumors unless this stage of initiation is followed by the introduction of a promoter, such as a phorbol ester, an ingredient of croton oil. Promoters are not carcinogenic by themselves but lead to tumor production if applied after the initiating agent. Treatment with a specific promoter before exposure to the initiating agent does not result in tumor formation. This stage of promotion occurs over weeks and months and is reversible in its early stages. Promotion involves changes not in DNA structure but in the expression of the genome mediated through promoter-receptor interaction. The binding of promoter to receptor alters the expression of genes downstream. For example, estrogens and androgens may act as promoters by binding to the estrogen and androgen receptors, respectively, in liver or mammary tissue. Promoters such as phorbol-12-myristate-13-acetate have a variety of biologic actions; they alter differentiation, cause changes in cell surface glycopeptides, alter various metabolic activities, and suppress immune surveillance of tumors by cytotoxic macrophages and natural killer cells.11 The final stage of progression is irreversible and is characterized by karyotypic instability and malignant growth. Thus, carcinogenesis is a multistep process that may be arrested at intermediate stages, that requires a long latency period for induction, and that can be influenced by, if it does not require, a promoting agent.

Figure 5.1 Chemical structures of three carcinogenic agents.

The existence or identity of an associated promoter has not been established for well-documented carcinogens in humans. For cancer patients, induction of second tumors may require not only an initiator (a DNA damaging agent), but also a promoter, a function that may be fulfilled by a second chemotherapeutic agent, by radiotherapy, or by a disease-related abnormality in metabolism or immune function.

Chemical carcinogens show a diversity of structures but share important metabolic features. Most are inert and require microsomal metabolic activation to positively charged (or electrophilic) intermediates that react with DNA bases. The primary sites for attack of DNA are relatively electron-rich (or nucleophilic) sites, such as the N-7 position of guanine12, 13 (see Chapter 11). This characteristic of carcinogens—namely, microsomal metabolism to an electrophilic intermediate that attacks DNA—is shared by certain antineoplastic agents such as cyclophosphamide, procarbazine, and mitomycin C and is essential in the antineoplastic action of these drugs. Other agents, such as L-phenylalanine mustard and nitrogen mustard, do not require metabolic activation to form alkylating species. Carcinogenicity has also been ascribed to ionizing irradiation, which produces free radicals, such as superoxide or hydroxyl radicals. A number of antitumor drugs have the same ability to promote formation of reactive oxygen intermediates; such agents include those that possess quinone functional groups (doxorubicin hydrochloride and plicamycin [mithramycin]) and those that bind electron-donating heavy metals (such as bleomycin sulfate and hydroxyurea). Four antineoplastic agents suspected as carcinogens and their probable carcinogenic intermediates are given in Figure 5.2; the varied chemical features of their reactive intermediates are illustrated.

Host factors, including enzymes such as glutathione S-transferases (GST), detoxify potentially mutagenic and toxic DNA-reactive electrophiles. Functional polymorphisms of GSTP1 (codon 105 Val allele) are associated with a higher risk of treatment (chemotherapy)-related acute myelogenous leukemia (AML), and not to radiation-induced or de novo AML. This risk is particularly relevant in patients receiving agents that are substrates for GSTP1.14 For example, in a report of 44 patients with breast cancer who had received chemotherapy and/or radiation therapy and subsequently developed AML/MDS (myelodysplasia), 55% had combined deletions of the glutathione S-transferase polymorphisms GSTM1 and GSTT1; both polymorphisms are associated with diminished enzymatic activity. This is in contrast to 8.8% of patients in the control group with AML/MDS. An insufficient detoxification of cyclophosphamide is the proposed mechanism.15 Other polymorphisms of drug-metabolizing enzymes, including cytochrome P450 3A4, NAD(P)H:quinine oxidoreductase and myeloperoxidase, may be markers of susceptibility to genotoxicity.16 Direct genotoxicity may not be the sole explanation for drug-induced carcinogenesis. Short and long interspersed elements (SINEs and LINEs) comprise one-quarter of the human genome and are spread throughout the genome through retrotransposition. In retrotransposition, an element is transcribed into RNA and then converted back to DNA by reverse transcription. The copied DNA is then reinserted into a new location. Genotoxic agents and gamma irradiation have recently been shown to induce SINE transcription and reverse transcriptase activity. This observation suggests that genotoxic exposure may lead to genomic mutation through both DNA damage and through potentially mutagenic mobile elements in the genome.17

Figure 5.2 Antineoplastic agents with reactive intermediates.

The identification of oncogenes and suppressor genes has added another variable to the equation. Their role in carcinogenesis is being pursued aggressively, and several possible mechanisms of actions have been proposed. The loss of one allele in a tumor suppressor gene such as p53 can potentially increase the risk of a drug-induced mutation in the other allele and development of the malignant phenotype. Oncogenes can be activated by a variety of mechanisms summarized in Table 5.1. Point mutations, chromosomal translocations, and gene amplification can alter expression of these genes. Just as altered oncogene expression and mutation occur with exposure to potential carcinogens, exposure to carcinogenic antitumor drugs likely alters oncogene expression and increases the risk of second malignancies. Most Nitrosomethylurea-induced mammary tumors contain an activated ras oncogene with a substitution of adenine for guanine in the 12th codon. This change is consistent with methylation of the oxygen in the 6 position of guanine, which would result in the replacement of guanine by adenine on DNA replication.18 Such studies bring together environmental and genetic factors in cancer causation.

TABLE 5.1 ONCOGENE ACTIVATION

Alteration

Effect

Base mutation in coding sequence

New gene product with altered activity

Base deletion in noncoding sequence

Altered regulation of normal gene product

Chromosomal translocation

Altered message and level of expression

Gene amplification

Increased gene expression

Adapted with permission from Pitot HC. The molecular biology of carcinogenesis. Cancer 1993;72:962–970.

 

TESTING OF ANTINEOPLASTIC AGENTS FOR CARCINOGENIC POTENTIAL

In view of the damaging effects of many antineoplastic agents on DNA and the suggestive clinical evidence of their carcinogenicity, application of methods for determining carcinogenic potential before widespread use of new agents in humans has become imperative. An ideal test system would be simple, rapid, inexpensive, and yet specific for carcinogens and sensitive to modestly potent agents. Unfortunately, the various methods available, ranging from in vitro bacterial mutagenesis assays to long-term studies in rodents, all have recognized drawbacks.19

Five types of test systems for carcinogen exposure are available. Mutagenesis assays such as the Ames test attempt to quantify the frequency with which a chemical induces mutational events based on the assumption that mutagenicity correlates with the likelihood of causing cancer in animals. The underlying premise is that carcinogenesis is the product of a mutational event that can be expressed in the short term as a change in biochemical features of a test organism. Cytogenetic studies attempt to correlate drug-induced chromosomal aberrations such as sister chromatid exchanges (SCEs) with carcinogenicity. Although certain characteristic karyotypic changes are associated with specific malignancies, such as the Ph1 chromosome with chronic myelogenous leukemia, cytogenetic abnormalities have proven neither necessary nor sufficient causes of neoplastic transformation. Tests of oncogenesis in tissue culture are based on the hypothesis that agents that produce neoplastic transformation in culture are likely to be carcinogenic in the whole animal. Like the Ames assay of bacterial mutagenesis, this system entails the assumption that the drug concentration, duration of exposure, and metabolism of the suspected carcinogen are relevant to the in vivo situation, but this assumption is of uncertain validity, and metabolic information is not available for many of the compounds tested. Carcinogenicity studies explore the tumorigenic potential in animals and attempt to predict the risk in humans. In vivo mammalian studies are usually conducted in rodents over extended periods and at great expense. The primary drawbacks of this system are the known species, sex, and age dependencies of drug metabolism in rodents and the lack of pharmacologic information that would allow an extrapolation of results from rodents to humans. The fifth approach, a measure of carcinogen exposure, uses detection of carcinogen-macromolecular adducts or somatic gene mutation in either target tissue or peripheral blood elements in animals or in man.

Mutagenesis Assays

Among the many mutagen-testing systems, the Ames test satisfies the requirements of simplicity and rapid return of results, and in addition, appears to possess high specificity for carcinogens, although certain exceptions have been identified. This test uses specific strains of Salmonella typhimurium that are histidine-requiring mutants.20 Exposure of these strains to the suspected mutagen in a histidine-free medium leads to growth of revertant mutants if the appropriate mutation is induced. Small amounts of chemicals (less than 1 mg) can be used, and results are obtained in approximately 2 days. For agents that require metabolic activation (as do many carcinogens), rat or human liver microsomes can be added to the test plates.

In extensive testing of a wide variety of agents previously documented to be carcinogens and noncarcinogens, 90% of the known carcinogens gave positive results in the Ames assay, and 87% of the noncarcinogens were inactive.21, 22 These findings suggest that the system has a high degree of specificity and sensitivity. Many of the antineoplastic agents in use today have been examined in the Ames system,23, 24, 25, 26 and some of the results are incorporated in Table 5.2. Most antimetabolites and the vinca alkaloids give negative results in both the Ames test and in vivo systems, whereas alkylating agents and many antitumor antibiotics give positive results in both assay systems. Both procarbazine hydrochloride and dactinomycin (actinomycin D), however, are carcinogenic in animals but give negative results in the Ames test. In the case of procarbazine, this discrepancy may be due to the failure of the test system to simulate the metabolism of procarbazine as it occurs in vivo. The agent 6-mercaptopurine, which has been reported to be carcinogenic in animals, shows weakly mutagenic results in the Salmonella system.

TABLE 5.2 RESULTS OF TESTING ANTINEOPLASTIC AGENTS IN THREE SYSTEMS FOR CARCINOGENICITY

Agent

Ames Test

SCEs

Animal Studies

Mechlorethamine hydrochloride

+

+

+

Cyclophosphamide

+

+a

+

Melphalan

+

+

+

Thiotriethylene phosphoramide (thiotepa)

+

+

+

Chlorambucil

NR

+

+

Procarbazine hydrochloride

-

+a

-

Lomustine (CCNU)

NR

+b

-

Doxorubicin hydrochloride

+

+

+

Streptozotocin

+

NR

+

Bleomycin sulfate

-

+c

-

Dactinomycin (actinomycin D)

-

±

+

Mitomycin C

+

+

+

Dacarbazine (DTIC)

NR

-b

+

Cisplatin

+

+

NR

5-Fluorouracil

-

NR

NR

6-Mercaptopurine

+

-

+

Cytosine arabinoside (ara-C)

-

NR

-

Vincristine sulfate

-

±

-

Vinblastine sulfate

-

NR

+

Pemetrexed

-

NR

-

Oxaliplatin

+

NR

NR

Methotrexate sodium

+

aDrug must be activated.
bTest done on patient lymphocytes after treatment with agent.
cConcentration giving positive results also causes significant numbers of other chromosomal aberrations.
+, positive result reported in at least one study; -, no positive result reported; ±, slight decrease over control (which is of unknown significance); NR, no result reported.

From the foregoing analysis, the Ames test would appear to be an excellent screening procedure but one with obvious false-negative results. An analysis of the Ames test results by Rinkus and Legato27 indicates that the false-negative rate is particularly high for specific chemical classes. At least seven classes of agents known to contain carcinogenic compounds are poorly detected in the Ames system, including azo compounds, carbonyl, hydrazine, chloroethylene, steroid, and antimetabolite structures. In some cases, known carcinogens such as urethane, probably cannot be metabolized to their carcinogenic form in the test system.

Another assay approach based on mutations measures the mutation frequency in the hypoxanthine-guanine phosphoribosyltransferase (HGPRT) gene.28 This technique can be used in vitro in mammalian cells and in vivo in patient samples. Assessment of mutation frequency at baseline and after treatment, and comparison between control groups and populations treated with chemotherapy have been carried out.29, 30 Whether these assays are predictive of increased malignant risk is yet to be determined.

In vivo mutational assays have been developed using transgenic rodent models.31 These models are composed of an altered genomic sequence that is inheritable, often the Escherichia coli lacI (lac repressor) or lacZ (β-galactosidase) genes. Animals are treated with the potentially carcinogenic agent and after sufficient time has passed to fix DNA adducts as mutations, genomic DNA is extracted, and the target gene is isolated by such methods as magnetic affinity capture. The transgenic model allows for rapid assessment of tissue-specific mutation after chemical treatment. This may focus subsequent clinical monitoring on specific organs. As with other in vivo studies, factors such as drug pharmacokinetics, DNA repair, animal age, diet, strain, sex, drug dose, and dosing duration influence the results.

Assay of Sister Chromatid Exchanges

Chromosomal damage resulting from exposure to chemical substances in vitro or in vivo has been used as an index of mutagenic or carcinogenic potential for many years but has required significant skill in recognizing the many different possible aberrations. Assay of SCE, a type of chromosomal study that detects the exchange of small DNA fragments between sister chromatid pairs, has considerable appeal because relatively few cells need to be examined, exchanges can be visualized easily, and the system is sensitive to small amounts of chemicals. The exchange is symmetric and does not alter the overall chromosomal morphology.32

The ability of various chemotherapeutic agents to induce SCE indicates that this technique might be useful as an assay for mutagenesis and ultimately carcinogenesis, but limitations of its potential have also become clear.32, 33 Ionizing radiation, known to be a potent mutagen and carcinogen, causes only slight increments in SCE; these changes are minimal in comparison with other chromosomal damage, including breaks, deletions, and other aberrations induced at the same dose level. On the other hand, ultraviolet light evokes dramatic increases in SCE frequency. Alkylating agents and some DNA intercalators induce a high frequency of SCE in addition to other chromosomal damage. Cyclophosphamide induces SCEs only after microsomal activation.34 Among the antimetabolites, methotrexate, which is not carcinogenic in laboratory animals, or humans, has been reported to induce SCE, but 6-mercaptopurine, a suspected carcinogen, does not cause these chromosomal abnormalities.35

The use of SCE has particular appeal because the effects of chemotherapeutic agents can be assessed in vivo by performing this test on peripheral lymphocytes from patients receiving antineoplastic therapy. Studies of lymphocytes from patients before and at intervals after chemotherapy have shown a marked increase in SCEs after the administration of lomustine (CCNU), dacarbazine, and mitomycin.36, 37, 38 Whether such increases in SCE frequency reflect the likelihood of carcinogenicity is still unclear.

Cell Culture Systems

Cell culture systems also have been advocated for the testing of carcinogenicity. Morphologic transformation of cells in culture and the ability of these cells to produce tumors when implanted in animals have been the primary criteria used for carcinogenicity. Three major test systems, which use hamster embryo cells, fibroblasts from the ventral prostate, or 3T3-like cells, have been applied to the screening of environmental carcinogens.39, 40 Using all three lines, investigators have shown a good quantitative correlation between transformation in vitro and in vivo carcinogenesis, although the number of antineoplastic agents tested has been limited. Mammalian cell culture systems, however, are subject to many of the same problems as those of bacterial mutagenesis assays discussed previously, including the need to activate compounds to reactive intermediates. An additional problem pertinent to these three systems is the use of cells of nonhuman and nonepithelial origin. Finally, tumors resulting from the implantation of transformed cells are sarcomas, and thus may not reflect the potential of the tested agent to cause tumors in epithelial cells or in humans.

The results of testing antineoplastic drugs in cell transformation systems have not correlated well with tests of carcinogenicity in experimental animals.41, 42Carcinogenic alkylating agents (melphalan and thiotriethylene phosphoramide [thiotepa]) increased the transformation frequency of C3H/10T1/2 cells, and dactinomycin and bleomycin showed a concentration-dependent increase in transformation frequency. These results are consistent with the known carcinogenicity of these agents. However, methotrexate also caused a concentration-dependent increase in transformation but at a relatively low frequency, whereas two other antimetabolites, 5-fluorodeoxyuridine and arabinosylcytosine, produced transformation in synchronized cells exposed during the S phase of the cell cycle. None of these antimetabolites has proved to be carcinogenic in animals or humans.

Cultured human tissue and cells may be used for carcinogenesis studies.43 Studies in these systems overcome some of the drawbacks of using nonhuman systems. Drug metabolism to the ultimate carcinogen, uptake of drug into human cells, the identification of specific DNA adducts, and the presence of DNA repair systems more closely approach the in vivo situation. However, since aspects of each of these processes differ among various human tissues, it is uncertain that any one test system can predict for results in people.

Animal Studies

The classic yardstick for assessing carcinogenicity has been the ability of the suspected agent to induce tumors in laboratory animals. These studies, although the most direct and reliable source of experimental information, are fraught with difficulties, including high cost, interspecies variability in susceptibility to carcinogens, and the long time required to obtain results. In addition, efforts must be made to design protocols of drug administration that mimic the intensity and duration of exposure found in humans, a problem compounded by differences in drug metabolism and pharmacokinetics in humans and rodents. A definite advantage of the bioassay system in intact animals is the preservation of the role of the immune system in determining the outcome. This factor is obviously missing in any of the in vitro assays.

The results of various bioassays of antineoplastic agents are recorded in Table 5.2.44, 45, 46, 47 Some results are conflicting and seem to depend on the age, sex, and species of animal used in the test. In general, however, most alkylating agents and antitumor antibiotics are carcinogenic in animals, whereas antimetabolites, including methotrexate, cytosine arabinoside (ara-C), and hydroxyurea, give negative results. Drug combinations have received only limited testing in bioassay systems.48 Tests of the combination of prednisone and azathioprine, commonly used in organ transplantation, showed a decrease in time before tumor appearance compared with azathioprine alone. With other combinations (e.g., prednisone plus CCNU, ara-C plus CCNU, and prednisone, vincristine sulfate, and cyclophosphamide), the median time before tumor appearance was longer than with the alkylating agent alone. Of the 10 combinations studied, 4 resulted in slightly higher tumor incidence than controls, whereas 6 caused fewer tumors than did the individual drugs.

Molecular and Biochemical Assays

Advances in the detection of carcinogen-molecular adducts and somatic gene mutations have opened the opportunity to study carcinogen exposure in humans.49, 50 The polymerase chain reaction and DNA sequencing enable rapid assessment of oncogene and tumor suppressor gene mutations in small patient samples. The use of 32P-postlabeling thin-layer chromatography and autoradiography assays, enzyme-linked immunosorbent assays, synchronous fluorescence spectroscopy, and gas chromatography/mass spectroscopy has made it feasible to detect low levels of adducts in human samples. Carcinogen-DNA adducts, exposure to chemicals, and carcinogenicity have been correlated with each other; but in the past, the low levels of adducts present in human samples limited the conventional assay systems. Enzyme immunoassays combined with synchronous fluorescence spectroscopy have increased sensitivity and specificity for polycyclic aromatic hydrocarbon–DNA adducts. High-pressure liquid chromatography or immunoaffinity chromatography in combination with 32P-postlabeling assay or immunoassay can be used to detect alkyl adducts in the human tissue with assay detection limits ranging from 1 to 600 adducts per 108 nucleotides, depending on assay and tissue examined. Such assays make it feasible to perform epidemiologic studies in patients receiving chemotherapy.

Proteomics may ultimately prove to be a more reliable and cost-effective tool to predict carcinogenicity. Studies have identified proteomic changes that occur as cells become cancerous.51, 52, 53 Consequently, the identification of such changes of the proteome on exposure of the cell to the investigational agent, would raise concern regarding its oncogenic potential.

CLINICAL STUDIES IMPLICATING ANTINEOPLASTIC AGENTS IN CARCINOGENESIS

Although experimental evidence demonstrating the carcinogenic potential of many antineoplastic agents was abundant, the clinical evidence of this problem was slower to appear. The fact that the rate of development of “secondary” cancers in patients with malignant lymphoma, pediatric cancers, ovarian cancer, and breast cancer is higher than that seen in an age-matched normal population has become clear. Many good reviews of this topic are now available in the medical literature.54, 55, 56, 57 Reports of second tumors in patients with prior histories of cancer comes from a variety of sources. Initial reports were mainly anecdotal and thus did not allow an analysis of factors that might be important. Data reported more recently have come from hospital-based, national, and international tumor registries and from longer follow-up of chemotherapy and hormonal therapy studies. The use of longer-term clinical trial data has the advantage that the initial cohort and treatment are tightly controlled. This provides a better analysis of how different drugs and treatments would impact the risk of second cancers. The use of clinical trial data for this purpose is somewhat limited by patient numbers, which rarely exceed 1,000. Registries, on the other hand, can have several thousand or tens of thousands of patients and thus allow a better assessment regarding less common second cancers such as acute leukemia or sarcoma.

Determining treatment and outcome from registries can be labor-intensive, however. One method that is used to identify treatment factors involved in the development of new cancers from a registry is referred to as a “nested” case-control study. In this approach, patients in the registry who develop a second cancer are compared with others who did not. These comparisons have provided a better estimate of the risks and the factors that influence the development of second cancers. Clinical information about the total dosages of drugs, concomitant therapy, and the duration of treatment is important in estimating risk. For some drugs such as the alkylating agents or etoposide, a threshold exists above which the risk of neoplasia rises sharply. Such thresholds have been previously identified in experimental carcinogenesis and in the induction of SCEs. Duration of treatment may also have a bearing, because a brief but intense exposure to a cytotoxic agent may be less carcinogenic than long-term low-dose exposure.

Another issue in assessing the true risk of second cancers from cytotoxic agents is the existence of other factors that may also influence their development. An underlying increased incidence of second malignancy is found independent of therapy in patients with retinoblastoma, Wilms' tumor, multiple myeloma, Hodgkin's disease, and other tumors such as those associated with the hereditary nonpolyposis colorectal cancer syndrome. Other therapies used to treat the cancer, particularly radiation therapy, also impact the development of secondary cancers. An increase in solid tumors after therapy for Hodgkin's disease and testicular cancer is most likely related to radiation rather than chemotherapy. In many reports, combination treatment regimens or regimens using irradiation and chemotherapy were used. Thus, the carcinogenic effects cannot necessarily be ascribed to one compound of the regimen with certainty, although the use of the nested case-control method may allow conclusions to be drawn regarding the carcinogenicity of different components of the regimen.

Interpretation of studies in this area must also take into consideration the statistical methods used to assess relative risk.58 The use of a person-years-of-risk analysis assumes that the yearly incidence of second malignancies is constant for the entire follow-up period and does not allow for the fact that a patient must live a certain time through the latency period for the occurrence of a second malignancy. Such an analysis allows a reasonable estimate of the carcinogenic effects of a single therapy, but its use when comparing two treatments biases results against the treatment that leads to a longer survival. Many studies compare the risk of cancer in the treated group with that of an age-matched cohort in the normal population to determine a relative risk. For a tumor that is uncommon in this age-matched population, an increase in the relative risk of fivefold to 10-fold sounds impressive but may only translate into a problem for fewer than 1% of patients who received therapy. On the other hand, small increases in relative risk for the more common solid tumors such as lung or breast cancer translate into a much greater problem in terms of absolute risk. This is the case with treatments for Hodgkin's disease as described later. One method that is useful in determining the overall impact of a secondary cancer in a population is to describe it in terms of the number of new cancers that occur per 10,000 patients treated.

Based on information currently available, one can attempt to categorize antineoplastic agents into high, moderate, low, and unknown risk groups on the basis of their oncogenic potential in humans (Table 5.3). The primary basis for this classification is reports of second malignancy in patients treated for both hematologic and solid tumors, with additional information coming from trials of cytotoxic agents in patients with immune diseases or after organ transplantation. Given that the latency period for the development of secondary cancers can range from 1 year (e.g., for etoposide-induced leukemias) to 20 years for solid tumors, the risk for many newer agents such as paclitaxel, docetaxel, irinotecan hydrochloride, gemcitabine hydrochloride, pemetrexed and oxaliplatin cannot yet be properly determined. Furthermore, the impact of the new targeted agents such as imatinib, gefitinb, and erlotinib on formation of secondary malignancies is unclear. Neither of the epidermal growth factor receptor tyrosine kinase inhibitors gefitinib nor erlotinib have demonstrated genotoxic potential with in vitro or in vivo assays. Imatinib, the tyrosine kinase inhibitor targeting C-kit, has been shown to induce benign and malignant tumors of preputial/clitoral gland, kidney and urinary bladder in rats. This has not been demonstrated in humans.59 A true assessment of agents primarily used in palliative therapy is also difficult because most patients may not survive long enough for problems such as second cancers to manifest.

The development of a new cancer can occur many years after treatment of the initial cancer. This means that large numbers of patients and long follow-up are required to define the risk of carcinogenesis and to understand which drugs and schedules are the probable causes. Some investigators have used preneoplastic lesions as markers of carcinogenicity to provide an earlier estimate of the risk. For example, a small group of patients with breast cancer who had been randomized previously to receive adjuvant chemotherapy or oophorectomy underwent cytologic and colposcopic screening of the uterine cervix.60 The results were compared with those for 79 controls with no known breast malignancy. Significantly more breast cancer patients who had received chemotherapy had cervical intraepithelial neoplasia (P <.01) than did controls; the proportion of breast cancer patients in the oophorectomy group who had cervical intraepithelial neoplasia did not differ significantly from the proportion in the control group. The incidence of chromosome abnormalities and structural chromosome changes in ovarian cancer patients treated with melphalan was higher than in a control group.61 For patients receiving both melphalan and radiation therapy, the frequency of chromosomal aberrations was even higher. Whether these chromosomal changes act as a marker for subsequent development of secondary leukemia is not yet known. In children with hematologic cancer who had previously received chemotherapy and cranial irradiation, the total-body mole counts were compared with those of their siblings. The median number of moles was 20.0 in the patient group (n = 79) and 11.0 in the healthy siblings (n = 88).62 In another study, a total-body count of melanocytic nevi in children receiving treatment for hematologic cancer was carried out before therapy and repeated 3 years later. Total-body nevus counts were significantly increased 3 years after treatment.63 To what degree these results predict subsequent cancer development is yet unknown. With increasing knowledge of progression from benign to neoplastic growth in diseases such as colorectal and pancreatic cancer, however, assessment of precursor lesions may be a useful way to evaluate risk.

TABLE 5.3 CATEGORIZATION OF ANTINEOPLASTIC AGENTS ACCORDING TO CARCINOGENIC RISK IN HUMANS

High Risk

Moderate Risk

Low Risk

Unknown

Melphalan

Doxorubicin hydrochloride

Vinca alkaloids

Bleomycin sulfate

Mechlorethamine hydrochloride

Thiotriethylene phosphoramide (thiotepa)

Methotrexate sodium

Taxanes

Nitrosoureas

Cytosine arabinoside (ara-C)

Busulfan

Etoposide

Gemcitabine hydrochloride

Teniposide

Irinotecan hydrochloride

Azathioprine

Mitoxantrone hydrochloride

 

Cyclophosphamide

5-Fluorouracil

Pemetrexed

Procarbazine hydrochloride

L-Asparaginase

Oxaliplatin

Dacarbazine (DTIC)
Cisplatin

Carboplatin

SECOND MALIGNANCIES IN SPECIFIC POPULATIONS OF CANCER PATIENTS

Pediatric Patients

Long-term survival is now possible for many patients with pediatric malignancies. This group of patients is followed closely for the development of late complications from treatment. Some pediatric tumors such as retinoblastoma have been associated with genetic abnormalities that may predispose to other cancers.64 Overall, the risk of developing a second cancer 20 years after childhood cancer has been estimated at 8 to 20%.65, 66 The Childhood Cancer Survivor Study analyzed the risk of second malignancies in 14,000 five-year survivors who received their treatment between 1970 and 1986. The relative risk for a second malignancy was as follows: non-Hodgkin's lymphoma, 3.2; leukemia, 5.7; and Hodgkin's disease, 9.7.67 One consistent finding has been an association between treatment with the epipodophyllotoxins, etoposide, or teniposide, and secondary AML, often with monocytic features. One series examined 205 children with acute lymphoblastic leukemia (ALL) who were treated with a four-drug induction consisting of prednisone, L-asparaginase, vincristine, and daunorubicin hydrochloride followed by maintenance therapy with oral 6-mercaptopurine, methotrexate, L-asparaginase, etoposide, and cytarabine. The etoposide was given twice weekly. The risk of secondary AML at 4 years was 5.9 ± 3.2%. Because none of these children received alkylating agent therapy or irradiation, etoposide was most likely responsible for these secondary leukemias.68 Risk factors for secondary AML in 734 consecutively treated children with ALL who attained complete remission and received maintenance treatment with epipodophyllotoxins were reported by Pui et al.69 Secondary AML was diagnosed in 21 of the 734 patients, and the overall cumulative risk at 6 years was 3.8% (range, 2.3 to 6.1%). For the subgroups treated twice weekly or weekly with etoposide or teniposide, the risk of AML at 6 years was 12.3%, whereas for the subgroups treated with these drugs only during remission induction, or every 2 weeks during maintenance treatment, the risk was 1.6%. In their analysis, the schedule of the epipodophyllotoxin administration was important, whereas the cumulative dose of drug did not appear to influence the risk of secondary leukemia. At the Dana-Farber Cancer Institute, no epipodophyllotoxin was used in their regimens. They reviewed 752 children with ALL who entered complete remission after induction therapy. Only two had developed AML after a median follow-up of 4 years.70 In a review of all ALL patients treated at the Dana-Farber Cancer Institute, the risk of a second malignancy was 2.7%, but the risk of other adverse events, including relapse, death, or induction failure, was 31%.71 Clinical and cytologic findings in epipodophyllotoxin-induced leukemia are a short latency period (mean, 24 to 36 months) between the completion of treatment and the development of AML, a Fab M-4 or M-5 subtype, a translocation of the MLL gene at chromosome band 11q23, and a poor response to treatment.72 Studies have shown an association between the breakpoints in the MLL gene and DNA topoisomerase II cleavage sites that are stimulated by etoposide. Secondary leukemia due to alkylating agents is characterized by a different phenotype with a longer latency period, antecedent myelodysplasia, and deletions of chromosomes 5 or 7.54

In view of this apparent increased risk of leukemia with epipodophyllotoxins, the National Cancer Institute Cancer Therapy Evaluation Program has instituted a monitoring plan for secondary leukemias after treatment with these agents. One report73 from this program analyzed 12 cooperative group clinical trials (11 in the pediatric population) that used cumulative doses of etoposide ranging from less than 1.5 g/m2 to more than 3.0 g/m2.74 The risk of developing a secondary leukemia at 6 years was 3.3%, 0.7%, and 2.2% in the dose ranges of less than 1.5 g/m2, 1.5 to 2.99 g/m2, and more than 3.0 g/m2, respectively. Their overall conclusions were that, at doses of less than 5 g/m2, only a minor risk of secondary leukemia is found. The risk of leukemia in patients receiving etoposide is probably influenced by other agents used in the regimens, particularly alkylating agents and other topoisomerase inhibitors. Relatively high rates of secondary leukemia have been reported in small series after the use of intensive treatments for pediatric tumors with poor prognosis that included both topoisomerase II inhibitors and alkylating agents.75

The development of secondary solid tumors in pediatric cancer patients is an issue of growing concern. The Roswell Park Cancer Institute reviewed the courses of 1,406 patients younger than 20 years of age who were treated over a 30-year period.76 The actuarial risk of a second malignant tumor 25 years after diagnosis was 5.6%. Prior therapy with carmustine and doxorubicin were the only factors that were significantly associated with the risk of a second malignant tumor. In Italy, a registry of all patients with childhood cancer who achieved complete remission was followed for a median time of 52 months after treatment. Twenty secondary malignancies occurred, which included nine hematologic malignancies (four AML, two chronic myelogenous leukemia, three non-Hodgkin's lymphoma), eight central nervous system tumors (all in patients given central nervous system radiation), and three other solid tumors.77 Others have reported the occurrence of unusual tumors such as squamous cell cancers of the skin occurring in teenagers who have previously received therapy for AML.78 The risk of specific types of second tumors appears to be a function of the types of treatment used, the sites of irradiation, and undoubtedly, the nature of the underlying malignancy.

In a follow-up of 674 patients treated in the German Ewing's sarcoma studies, the cumulative risk of a second malignancy was 0.7%, 2.9%, and 4.7% after 5, 10, and 15 years, respectively. The time until the development of myelodysplasia/leukemia was 17 to 96 months and until development of solid tumors was 82 to 136 months.79 Of 397 patients with Ewing's sarcoma treated at the Mayo Clinic, 26 patients (6.5%) had 29 malignancies. The mean age was 16 years and the interval from diagnosis of the sarcoma and a second malignancy averaged 9.5 years (range, 1 to 32.5 years). The cancers consisted of 12 sarcomas, 9 carcinomas, and 8 hematologic malignancies. The hematologic malignancies occurred at a mean of 4.8 years (range, 1.7 to 12.9 years) and sarcomas occurred after a mean of 10.9 years (range, 1.5 to 32.5 years).80 The importance of the development of second malignancy must be interpreted in relation to the risks of failure of therapy of the primary cancer. In the analysis of the German Ewing's sarcoma trials, second malignancies accounted for only 3 of the 328 deaths in this population; the remainder were due to Ewing's sarcoma.

The Memorial Sloan Kettering Cancer Center (MSKCC) group reported 14 second malignancies in 509 patients with osteosarcoma treated on 6 different clinical trials. Chemotherapy agents included high-dose methotrexate, doxorubicin, bleomycin, cyclophosphamide, dactinomycin, vincristine, cisplatin, and ifosfamide. The median age at diagnosis of the osteosarcoma was 16.6 years (range, 3.1 to 74.4 years), and time interval from osteosarcoma diagnosis and secondary malignancy was 5.5 years (range, 1.3 to 13.1 years). The most common secondary malignancy was in the CNS (four anaplastic gliomas, one meningioma, high-grade glioma and a maxillary astrocytoma). There were two cases of AML and one case each of MDS, Non-Hodgkins lymphoma (NHL), high-grade pleomorphic sarcoma, leiomyosarcoma, fibrosarcoma, breast cancer, and mucoepidermoid carcinoma. The overall 5- and 10-year cumulative incidences of secondary malignancies were 1.4 ± -1.1% and 3.1 ± -1.8%. The standardized incidence ratio for the cohort was 4.6% (95% confidence interval [CI], 2.53–7.78; P = 0.00001).81

A review from Stanford of 694 children with Hodgkin's disease showed a risk of both solid tumors and hematologic malignancies similar to that reported for adults with this disease (discussed later). Of note, the actuarial risk at 20 years in men was 10.6% and in women it was 15.4% because of the additional risk of breast cancers occurring within the radiation field.82 Similar observations were made by the Late Effects Study Group, which found the relative risk for second tumors in children treated for Hodgkin's disease to be 18.5. The cumulative risk of any second malignancy was 10.6% at 20 years, increasing to 26.3% at 30 years. Solid malignancies occurred in 7.3% at 20 years and 23.5% at 30 years and breast cancer was the most common malignancy, with a standardized incidence ratio of 56.7. Forty-seven percent received mantle radiation alone and 57% received combined modality therapy. The incidence of breast cancer at age 40 was 13.9%, rising to 20.1% at age 45.83 The risk of breast cancer is increased when the patient is of pubertal age at the time of radiation, and when the dose of radiation is higher.84, 85 Risk of leukemia was associated with use of alkylating agents and advanced stage at diagnosis. The second most common solid tumor developing in patients who have received radiation for Hodgkin's disease is thyroid cancer, with a relative risk of 36.67, 83 The risk of other epithelial malignancies such as colorectal and gastric cancer also seem to be increased, and occur at a younger age than the general population.83 Knowledge of these predispositions to secondary malignancies and toxicities have resulted in modification of the treatment in the pediatric population.86

Overall, the risk of AML peaks a few years after therapy, whereas the risk of a solid tumor increases with the length of follow-up. It is still too early to assess what additional risk this population will experience when they enter an age group in which the development of cancer is more common. In this setting, a modest increase in relative risk could translate into a substantial increase in the overall absolute risk of cancer. This has already been observed to some degree in the population treated for Hodgkin's disease.

Patients with hereditary retinoblastoma have a high incidence of second malignancies, in part because of their genetic predisposition, but this is exacerbated by the treatment. In a long-term follow-up study of 1604 1-year survivors of retinoblastoma diagnosed between 1914 and 1984, the relative risk of developing a second malignancy in the hereditary retinoblastoma population was 30. The cumulative risk of secondary malignancy diagnosis after 50 years was 51% in the hereditary retinoblastoma group and 5% in the nonhereditary group. The 50-year cumulative risk of a secondary malignancy in the previously radiated group was 58%, compared with 27% in those not receiving radiation therapy. Although second malignancies occurred at any radiation dose, there was a dose-response relationship, with a 12-fold increase at doses of 60 Gy or higher. These tumors are primarily soft tissue sarcomas and osteosarcomas and possess the RB1 mutations see in the primary tumor.87

Children treated for Wilms' tumor have an eightfold increased risk of developing a secondary malignancy: leukemia, lymphomas, and solid tumors. Treatment-related AML was diagnosed 1 to 6 years following initial therapy, as were lymphomas. Solid tumors consisted of sarcomas and cancers of the breast, thyroid, colon, liver, and parotid, in addition to brain, at a latency period of 3 to 21 years. Abdominal radiation therapy increased the risk twofold.88

Patients with Ovarian Cancer

Advanced ovarian cancer was treated initially with alkylating agents such as melphalan.89 Several reports have implicated alkylating agents (particularly melphalan and cyclophosphamide used as single agents) as a causative factor in the high incidence of AML in this group of patients.90, 91, 92, 93 A review of 5,455 cases of ovarian cancer revealed a 36.1-fold increased risk of acute leukemia compared with an age-matched control group. For patients surviving at least 2 years after the institution of therapy, the risk was 174.4-fold higher than that in the controls.94 Many patients with acute leukemia identified in this series also received radiotherapy alone or in combination with alkylating agents. Thus, determining which agent was responsible for the leukemia was impossible. An analysis of a large cohort of patients with ovarian cancer treated with melphalan or cyclophosphamide revealed a 93-fold increased risk of AML in women treated with chemotherapy.95 The risk was highest 5 to 6 years after the initiation of therapy and decreased thereafter. A dose-response relationship was apparent for melphalan and was suggested for cyclophosphamide. Melphalan was more likely to induce secondary leukemia than was cyclophosphamide. In an international collaborative group of cancer registries and hospitals, 114 cases of leukemia were identified after ovarian cancer.96 Chemotherapy alone was associated with a relative risk for leukemia of 12 compared with surgery alone, whereas radiotherapy alone did not produce a significant increase in risk. The risk of leukemia was greatest 4 to 5 years after chemotherapy and was increased for at least 8 years. Cyclophosphamide, chlorambucil, melphalan, thiotepa, and treosulfan were independently associated with significantly increased risks of leukemia. Chlorambucil and melphalan were the most leukemogenic. These studies support the clinical impression that a dose-response effect may exist, that the carcinogenic potential of all alkylating agents is not necessarily the same, and that the latency period is approximately 5 years. They also suggest that the risk for secondary leukemia does decrease after a period. The largest analysis of second tumors in ovarian cancer was done on nine population registries of the National Cancer Institute and Connecticut Tumor Registry.97 Researchers examined 32,251 women with ovarian cancer and found a relative risk of second cancers of 1.28 (95% CI, 1.21–1.35), with an excess of leukemia (relative risk [RR] = 4.1) and colorectal (RR = 1.4), bladder (RR = 2.1), and breast (RR = 1.2) cancers. The association with rectal and breast cancer was probably related to genetic predisposition; the risk of leukemia, to alkylating agents; and the risk of sarcomas and abdominal tumors, to previous radiation.

Ovarian cancer is now treated primarily with platinum-based regimens, and melphalan and chlorambucil are rarely used. The leukemogenic potential of cisplatin is assumed to be less than that for other alkylating agents. Anecdotal reports exist of patients developing acute non-lymphocytic leukemia (ANLL) after cisplatin therapy, but the relative risk of developing ANLL after cisplatin is not yet well known.98, 99 The newest agents for the treatment of ovarian cancer are paclitaxel and topotecan hydrochloride. It is too early to assess the carcinogenic potential of the topoisomerase I inhibitors and taxanes.

Patients with Breast Cancer

Breast cancer is another malignancy responsive to various cytotoxic and hormonal agents that are associated with an increased risk of secondary malignancies.100 Among patients receiving adjuvant chemotherapy for breast cancer, no increased risk of leukemia was identified in a group of 1,265 patients who received postoperative thiotepa (with or without radiotherapy), compared with untreated controls.101 The ongoing prospective adjuvant studies in breast cancer have addressed this question more definitively.2, 3, 102 The results of the National Surgical Adjuvant Breast and Bowel Program database analysis indicate that risk of leukemia in patients receiving melphalan-based adjuvant chemotherapy increases fivefold. An initial analysis of the Milan studies of cyclophosphamide, methotrexate, and 5-fluorouracil (CMF) adjuvant chemotherapy revealed no increased incidence of leukemia or other second malignancies.103 A more recent analysis of 2,465 patients with localized breast cancer treated in Milan from 1973 to 1990 revealed a 15-year cumulative risk of second cancers of 8.4% after local treatment only, 6.4% after CMF therapy, and 5.1% after doxorubicin-based chemotherapy. The relative risk for women receiving CMF treatment was 1.29.104 An analysis of 1,113 patients in Sweden treated with adjuvant CMF or radiation therapy did not demonstrate any increase in second cancers in the first 10 years of follow-up.105 Patients receiving chemotherapy actually had a lower rate of such cancers than those receiving radiation therapy.

The typical features of AML secondary to alkylating agent exposure include a latency period of 4 to 7 years, during which MDS often becomes apparent, deletions of the long arms 5 and/or 7 or loss of the whole chromosome, and an unfavorable response to chemotherapy. There is also a higher incidence of p53 mutations and microsatellite instability observed in therapy-induced myelodysplastic syndrome.106 Case reports have described the occurrence of a different type of AML with monocytic features associated with a translocation at 11q23 (the locus of the MLL gene) in patients who have received epirubicin hydrochloride-containing combination therapy for breast cancer.107 These cases occur after a brief latency period of 1 to 3 years rather than the more prolonged interval preceding AML induced by alkylating agents and are similar, if not identical, to the cases of leukemia associated with etoposide, another topoisomerase II inhibitor.73

The M.D. Anderson Cancer Center reviewed data on 1,474 patients treated on six adjuvant or neoadjuvant trials that included 5-fluorouracil, doxorubicin, and cyclophosphamide.108 The median follow-up was only 8 years, which is too short to evaluate risk of solid tumors. The 10-year estimated acute leukemia rate was 2.5% in patients who received both chemotherapy and radiation, and 0.5% in the chemotherapy-only group. This suggests that any leukemogenic risk from the use of anthracycline therapy is increased. A population-based cohort of 3,093 women in whom with breast cancer diagnosed were studied for the development of acute leukemia. Women who received chemotherapy and radiation had a standardized incidence ratio of 28.5. A dose-dependent increase in risk was observed in women treated with mitoxantrone and that the risk of leukemia was lower in the women receiving anthracyclines.109 Curtis et al.110reviewed the Surveillance, Epidemiology, and End Results database of 21,708 patients with breast cancer and found an 11.5 relative risk of developing secondary leukemias in patients treated with alkylating agents with or without radiation therapy as an adjuvant after a median follow-up of 4.2 years. In an attempt to assess the contributions of adjuvant radiotherapy, melphalan, or cyclophosphamide, Curtis and colleagues also reported a case-control study in a cohort of 82,700 women in whom breast cancer had been diagnosed.111 Results indicate a 2.4-fold increase in relative risk of leukemia after radiotherapy alone, a 10-fold increase after chemotherapy alone, and a 17.4-fold increase after a combination of the two. Melphalan was 10-fold more leukemogenic than cyclophosphamide, with little increase seen in the risk of leukemia after cumulative doses of cyclophosphamide of less than 20 g. The results from these analyses are consistent with data from the treatment of other malignancies. They do not rule out the possibility that second solid tumors that have a much longer latency period than leukemias may still develop.112 It has been suggested that postmastectomy irradiation increases the risk of lung cancer in smokers, and it is well established that radiation to the breast increases the risk of sarcomas, particularly angiosarcoma.113, 114, 115

Adjuvant therapy with tamoxifen citrate is now well established to improve relapse-free survival and overall survival in selected patients with breast cancer. A number of large studies randomizing women to receive tamoxifen or placebo after surgery have been completed. Longer follow-up on these patients has provided evidence about the influence of tamoxifen on the subsequent development of other malignancies. The short-term and long-term adverse effects of tamoxifen have been thought to be the result of its estrogenic effects. In postmenopausal women, tamoxifen treatment leads to endometrial hyperplasia and polyps.116 Tamoxifen also stimulates the growth of endometrial cancer in vitro.117 An association is found between tamoxifen and the development of endometrial cancer. A relative risk of 6.4 was found in a Scandinavian study in which 40 mg per day was used and was continued for 5 years.118 Other studies using lower tamoxifen dosages and a shorter duration of treatment have reported lower relative risks.119 Some have not reported any increased risk of endometrial cancer.

Not all studies, however, prospectively collected information on second primaries.120 The National Surgical Adjuvant Breast and Bowel Program reviewed 2,843 patients randomized to receive tamoxifen or placebo in their B-14 study.121 The relative risk of endometrial cancer in the tamoxifen-treated group was 7.5, and the overall annual hazard rate for the development of endometrial cancer was 1.6 per 1,000. In a meta-analysis of 32 randomized trials of tamoxifen versus a similar control arm including data from 52,929 patients, there was a significantly increased risk of developing endometrial cancers (RR = 2.7; 95% CI, 1.94–3.75) and gastrointestinal cancers (RR = 1.31; 95% CI, 1.01–1.69).122 If the estrogenic effects of tamoxifen cause endometrial cancer, those tumors that develop should be of low grade and have a relatively good prognosis. This assumption has been confirmed in some of the studies reported.123 Other studies have shown a distribution of grade and stage similar to that seen in nonhormonally induced cancers.124 In an analysis of 3,457 women with breast cancer, 53 subsequently developed endometrial cancer.125 Of these women, 15 had received tamoxifen and 38 had not. The number of high-grade cancers increased significantly in the tamoxifen-treated women, who also were more likely to die of their endometrial cancer. In a Japanese study, however, 825 women with primary breast cancer were followed prospectively with annual gynecologic examinations.126 Thirteen cases of endometrial cancer were discovered, but the incidence was no different in women who were and who were not taking tamoxifen. In a review of the Stockholm randomized trial of 2 years of adjuvant tamoxifen in postmenopausal women (n = 4,914; median follow-up of 9 years), an increased risk of endometrial cancer (RR = 4.1) and a decreased risk of contralateral breast cancers were noted.127 In addition, an increase in colorectal (RR = 1.9) and gastric (RR = 3.2) cancers was associated with the use of tamoxifen.

In summary, most studies have demonstrated that adjuvant tamoxifen leads to a higher rate of endometrial cancer. The highest relative risks are associated with higher dosages and a longer duration of therapy. The histopathologic features of tamoxifen-associated endometrial cancer are less clear because each reported series had only small numbers of such cancers. Tamoxifen can induce liver cancer in laboratory animals, but no increased incidence of primary liver cancer has been seen in the adjuvant breast studies. These tumors could well be missed because any tumor developing in the liver probably would be presumed to be a recurrence of the previous breast cancer.

Several studies have reported a reduction in the development of cancers in the contralateral breast with tamoxifen use.127, 128, 129, 130, 131 Either this could represent a reduction in the incidence of other breast cancers or could just be a manifestation of a reduction in the incidence of recurrence of the initial cancer within the contralateral breast. Reports have also appeared of reductions in cardiovascular mortality and increases in thromboembolic events when women take tamoxifen. An analysis of the impact of adjuvant tamoxifen on mortality was undertaken using published risks of endometrial cancers and thromboembolic events, as well as reductions in contralateral breast cancer and cardiovascular mortality.132 This analysis concluded that the overall impact of tamoxifen was favorable, with between 3 and 41 deaths avoided per 1,000 patients treated, depending on the age of the women being treated. The importance of breast cancer as a source of morbidity and mortality in women and the observations of reductions in contralateral breast cancers with adjuvant tamoxifen, have led to two breast cancer prevention trials in which healthy women were randomized to receive tamoxifen or placebo. In a prevention trial, the increased risks of adverse events such as second malignancies are more of a concern. This was all taken into account when these trials were developed; however, some reservations have been expressed about exposing women to an increased risk of endometrial cancer.133 No intervention is without risk; whether long-term tamoxifen usage leads to an overall health benefit to women can only be truly answered by these prevention trials.

Patients with Multiple Myeloma

Multiple myeloma, a disease commonly treated with single-agent alkylators such as melphalan, also has been associated with a high incidence of AML.134, 135,136 Because myeloma itself involves a bone marrow element, the possibility exists that a common process may be responsible for both diseases. However, the reported incidence of leukemia in patients with myeloma who do not receive alkylating therapy is no greater than expected for an age-matched population.137This suggests that the alkylating agents have contributed to the high incidence of leukemia. This contention is supported by a prospective trial of alkylating therapy for myeloma, which found that the actuarial risk of developing acute leukemia was 17.4% at 50 months, 214 times that expected.

Patients with Malignant Lymphoma

The incidence of second malignancies among patients with malignant lymphoma was no higher than expected during the era before intensive therapy.138 The use of combination chemotherapy and combined radiotherapy and chemotherapy has been associated with a high incidence of second malignancies, specifically AML and solid tumors.139, 140, 141, 142, 143 Many of these patients, however, would not have survived long enough to be exposed to the risk of a second malignancy before the introduction of intensive therapy. Many lymphoma patients have defective immune function, which may predispose them to a higher risk of cancer on exposure to an inciting agent. Mechlorethamine hydrochloride and procarbazine, components of nitrogen mustard, vincristine, procarbazine and prednisone (MOPP) combination chemotherapy for Hodgkin's disease, are potent carcinogens in animals.46 A case-control study of 1,939 patients treated for Hodgkin's disease in the Netherlands assessed factors influencing the development of acute leukemia.144 The cumulative dose of mechlorethamine was the most important factor. The use of lomustine was also associated with secondary leukemia, as was a requirement for a second course of chemotherapy. Overall, patients receiving chemotherapy had a 40-fold greater risk of leukemia than those receiving radiation therapy alone, whereas the use of combined-modality therapy did not increase the risk of leukemia beyond that seen with chemotherapy. Other analyses have similarly confirmed the importance of mechlorethamine, procarbazine, and nitrosoureas in the risk of second leukemia after treatment for lymphoma.145, 146 These studies also demonstrated that chemotherapy that did not include these three agents had a negligible risk of secondary leukemia.

Although many reports have been concerned with an increased risk of acute leukemia, solid tumors occur more frequently in patients with malignant lymphoma after intensive therapy.147, 148, 149, 150 Approximately one of patients with Hodgkin's disease develops a second cancer within 15 years of primary treatment.148, 151 Three-fourths of these are solid tumors, and the remainder are equally divided between leukemia and lymphoma. One hundred thirteen second cancers were seen in 2,846 British patients treated for Hodgkin's disease from 1970 to 1987.152 The relative risk compared with that of the general population for leukemia and non-Hodgkin's lymphoma was 16, but the chance of developing colon, lung, and thyroid cancer, as well as osteogenic sarcoma, was also higher. In a German series of over 1,500 patients with Hodgkin's disease treated with radiation therapy, with or without chemotherapy, from 1940 to 1991, the cumulative risk for malignancy was 1.5%, 4.2%, 9.4%, and 21% at 5, 10, 15, and 20 years, respectively.153 At the 20-year period, the risk for solid tumors, lymphoma, and leukemia was 19%, 1.9%, and 0.6%, respectively. Three-fourths of the solid tumors occurred within the radiation field. In patients receiving both chemotherapy and radiation therapy, the regimen of doxorubicin, bleomycin, vinblastine sulfate, and dacarbazine (ABVD) was associated with the highest risk. In another German report of 5,411 patients treated on one of three clinical trials for early, intermediate, and advanced Hodgkin's disease, 36 patients developed AML and 10 patients developed MDS. After a median observation time of 55 months, the incidence of AML/MDS was 1%. The prognosis was universally poor.154

In an intergroup trial of ABVD versus MOPP/ABV in 856 patients, secondary malignancies occurred in 18 patients receiving ABVD, 28 receiving MOPP/ABV, and 2 were initially treated with ABVD but subsequently received MOPP-containing regimens and radiation therapy before developing leukemia.155 A case-control study that compiled data on 19,046 patients with Hodgkin's disease treated between 1965 and 1994 demonstrated an increased risk of lung cancer for those receiving radiation at doses exceeding 5 Gy. In patients who were treated with alkylating agents and no radiation therapy, the risk of lung cancer was fourfold and the risk increased with the number of cycles administered.156 The risk of gastric cancer is increased twofold to 11-fold, and is highest for patients receiving combined modality therapy.156 Arseneau et al.142 reported a 23-fold increased risk of sarcoma after combined-modality therapy in patients with Hodgkin's disease. The overall risk of second malignancies increased 2.8-fold with intensive chemotherapy. In 885 women treated for Hodgkin's disease from 1961 to 1990, the relative risk of developing and dying from breast cancer was increased fourfold to fivefold.147 Although this is primarily the result of upper mantle irradiation, the concurrent use of chemotherapy further increased the relative risk.

Because combined-modality therapy exposes patients to a higher risk of neoplasm, a long-term assessment of its benefits and risks continues to be necessary. One interesting analysis examined 313 patients with early-stage Hodgkin's disease who received either full-dose radiation therapy or chemotherapy followed by a lower dose of involved-field radiation.157 The relative risk of a second cancer was 1.5 (95% CI, 0.6–3.5; P value not significant) in the combined-modality group but was 3.3 (95% CI, 2.2–5.3; P <.001) in the group receiving full-dose radiation.

Longer follow-up of patients receiving combination chemotherapy and radiotherapy for Hodgkin's disease has suggested that the increased risk of leukemia in this patient population may peak at between 3 and 9 years, followed by a decline thereafter.65, 139, 158, 159, 160 The risks for the development of solid tumors increase over time.160 Although the relative risk is highest for the development of leukemia and lymphoma, the twofold to threefold increase seen in the more common solid tumors accounts for most of the absolute increase in cancer cases in these patients.

Patients with non-Hodgkin's lymphoma are also at risk for developing second malignancies. The Groupe d'Etude des Lymphomes de l'Aduite (GELA) reported a 7-year cumulative incidence rate of 2.75% in 2,837 patients receiving doxorubicin, cyclophosphamide, vindesine, bleomycin and prednisone (ACVBP). Sixty-four of the 81 malignancies were solid tumors and 17 were hematologic malignancies. Age was the only risk factor on multivariate analysis. Considering all tumors, there was no increased risk of second cancers; however, in the male population there was an excess of lung cancer and MDS/AML, and in the female population there was an excess of MDS/AML.161 Up to 10% of patients with non-Hodgkin's lymphoma treated with either conventional-dose chemotherapy or high-dose chemotherapy and autologous stem cell transplantation may develop treatment-related MDS/AML within 10 years of primary therapy.162

Patients with Essential Thrombocythemia

Essential thrombocythemia (ET) is a relatively indolent disease that is not typically appreciated to transform into acute leukemia. The leukemogenic potential of hydroxyurea (HU) has long been questioned. In a study of 114 patients with ET, 56 patients were randomized to receive HU and 58 patients to no receive cytotoxic therapy; however, 50% of the control group subsequently did receive HU. Fifteen patients had received busulfan prior to randomization. At a median follow-up of 73 months, seven patients (13%) in the HU group ultimately developed AML/MDS or a solid tumor, compared with one (1.7%) in the control group (P = 0.032). Three of the 77 patients (3.9%) who had only received HU and 5 of the 15 patients (33%) who had previously received busulfan developed neoplasia.163

Patients with Gastrointestinal Cancer

Analysis of randomized trials of adjuvant methyl-CCNU in the management of patients with gastrointestinal cancers performed by Boice et al.4 has added more information regarding the leukemogenic potential of this treatment. The results of this analysis indicated that a 12.4 relative risk of leukemia exists in patients treated with methyl-CCNU. This risk seems to be dose-dependent when cumulative dose is considered. The latency period varies from 6 to 69 months and may continue to rise beyond that. Because the current data do not suggest a benefit in survival with such therapy, the leukemogenic risk has led to the removal of methyl-CCNU from adjuvant treatment regimens. The most important drug in adjuvant regimens for colorectal cancer is 5-fluorouracil, and it has not been associated with an increased risk of second cancers.

Patients with Testicular Cancer

More than 20 years have passed since cisplatin-based chemotherapy was first used for the treatment of advanced testicular cancer. This treatment has led to a large increase in the number of patients cured with chemotherapy, and reports about the long-term consequences of this therapy are only beginning to appear. In a group of 1,909 patients in The Netherlands diagnosed between 1971 and 1985, 78 second cancers occurred, or 1.6 times the number expected.164Significant increases were seen in gastrointestinal cancers (RR = 2.6) and leukemia (RR = 5.1). In this analysis, radiation therapy was the main contributing factor; patients treated with chemotherapy did not have an increased rate of second malignancies and actually had a decrease in the incidence of cancer in the contralateral testis. In a Norwegian series, the use of chemotherapy plus infradiaphragmatic radiation did increase the relative risk of second cancers over that seen with infradiaphragmatic radiation alone (RR = 1.3 versus 2.4).165 The highest risk was seen in patients who received both infradiaphragmatic and supradiaphragmatic radiation. An update of the Norwegian experience confirmed a modest increase in relative risk from the use of combined-modality therapy.166 The use of modern cisplatin-containing chemotherapy alone did not appear to increase the risk of a second cancer.

In a cohort of 1,025 German patients treated between 1970 and 1990, 224 received surgery only, 332 had radiation therapy, and 413 received chemotherapy, which in 293 cases included etoposide.167 The incidence of secondary neoplasms increased in patients receiving radiation therapy but not in those who received chemotherapy. The median follow-up in this review was relatively short (61 months). In a more recent review from France of 131 patients with seminoma, the relative risk of second tumors was not increased by infradiaphragmatic radiation. It was increased threefold, however, in patients receiving both infradiaphragmatic and supradiaphragmatic radiation, and it was increased 26-fold in the small number of patients who received chemotherapy plus radiation.168

No increases in second cancers have been reported after the use of cisplatin, vinblastine, and bleomycin (PVB) for testicular cancer.169 Etoposide is now used rather than vinblastine because a randomized trial demonstrated the improved effectiveness of cisplatin, etoposide, and bleomycin over PVB.170 The association between etoposide and secondary leukemia in the pediatric population led to a more detailed scrutiny of this relationship in patients with testicular cancer. Among 315 patients at Indiana University receiving etoposide, two cases of acute leukemia (0.63%) occurred.171 Of 340 patients treated with etoposide at Memorial Sloan-Kettering Cancer Center, two cases of acute leukemia also were seen.172 The overall conclusion of these and other reviews of etoposide use is that the dosages used in most germ cell cancer protocols are associated with a slightly increased risk of acute leukemia that is acceptable, given the benefits of etoposide-based therapy in treating this disease.173 It has been hypothesized that the total dose of epipodophyllotxin determines the risk of secondary AML, but researchers at the Cancer Therapy Evaluation Program (CTEP) have shown no such effect in patients receiving up to 5 g/m2 of etoposide.174

The largest review of second neoplasms in patients with testicular cancer includes data for almost 29,000 men in 16 different tumor registries.175 Overall, 1,406 second cancers were identified, yielding a relative risk of 1.43. An excess number of tumors were reported, including leukemias (RR = 3.07 to 5.20), melanoma (RR = 1.69), lymphoma (RR = 1.88), and a variety of gastrointestinal tumors (RR = 1.27 to 2.21). An analysis of the relationship between treatment and these new tumors revealed that the gastrointestinal tumors were associated with radiation therapy, whereas the secondary leukemia was associated with both radiation and chemotherapy.

Patients Receiving High-Dose Therapies

High-dose chemotherapy with autologous bone marrow transplantation (ABMT) or peripheral blood stem cell transplantation is being used with increased frequency for treating patients with hematologic malignancies and breast cancer. In this setting, very high doses of drugs are given over a short period of time, in contrast to the more conventional method of giving lower doses over a period of 4 to 12 months. The agents used differ slightly depending on the institution and tumor being treated, but commonly the oxazaphosphorine nitrogen mustards (cyclophosphamide and ifosfamide), carboplatin, and etoposide are used. The doses delivered with marrow rescue are threefold to sixfold higher than can be given with such support; thus, the total dose of drug delivered is similar to that given when such drugs are used in conventional regimens. In addition, patients frequently receive total body irradiation (TBI) as part of their preparative regimen. Myelodysplastic syndrome (MDS) and ANLL have been reported in patients who receive allogeneic, autologous, or peripheral blood transplantation for a variety of malignancies.176, 177 Most patients who have an ABMT, however, also receive other chemotherapy before this procedure, which confounds estimation of risk. A review of all 649 patients who received ABMT or peripheral blood stem cell transplantation at the University of Chicago from 1985 to 1997 revealed seven cases (1%) of MDS, ALL, or ANLL that were thought to be therapy-related.178 These occurred in five patients with Hodgkin's disease, one patient with non-Hodgkin's lymphoma, and one patient with breast cancer. The median latency period between initial standard-dose treatment of the cancer and development of leukemia/MDS was approximately 5 years, whereas the interval was less than 2 years from the high-dose therapy. In a retrospective analysis of 262 patients undergoing ABMT for non-Hodgkin's lymphoma at the Dana-Farber Cancer Institute from 1982 to 1991, the overall incidence of post-transplant MDS or AML was 7.6%, with a median onset of 31 months after transplant or 69 months after initial treatment of lymphoma. Variables predicting for development of MDS included prolonged interval between initial treatment and the transplant, increased duration of exposure to chemotherapy, and use of radiation therapy before transplant.179 Both of these studies suggest that conventional chemotherapy before the high-dose therapy was the more likely cause.

In a situation in which high-dose therapy is given repeatedly, however, the risk of secondary leukemia may become prohibitive. In a series of 86 patients with poor-risk solid tumors treated with repeated high doses of cyclophosphamide/ifosfamide, etoposide, and doxorubicin, the risk of ANLL at 24 months increased 5,000-fold.180 Cytogenetic analysis was consistent with leukemias induced both by alkylators and by etoposide. The risk of subsequent development of treatment-related MDS or AML may be greater for patients transplanted with CD34 + peripheral stem cells following chemotherapy priming, compared with patients receiving cells from the bone marrow without priming.181 The type and doses of alkylating agent used pretransplantation in addition to the dose of total body irradiation as part of the conditioning regimen may influence the risk of treatment-related MDS/AML. In a case-control study of 56 patients with MDS/AML and 168 matched controls within a cohort of 2,739 patients receiving autotransplants for Hodgkin's disease or non-Hodgkin's lymphoma, mechlorethamine less than 50 mg/m2 and more than 50 mg/m2 had a relative risk of 2.0 and 4.3, respectively, and chlorambucil given for fewer than 10 months and more than 10 months had a relative risk of 3.8 and 8.4, respectively, when compared with cyclophosphamide-based therapy. Total body irradiation at doses of 12 Gy or less had no influence on the development of AML/MDS, but doses of 1.2 Gy were associated with a relative risk of 4.6. Peripheral blood stems cells were associated with a nonsignificant risk of MDS/AML with relative risk of 1.8.182

Autologous stem cell transplant has been used for patients with breast cancer. As in other malignancies, the risk of AML/MDS appears to depend on the pretransplant chemotherapy. In a retrospective analysis of 364 patients with lymph node-positive breast cancer and who underwent autologous stem cell transplant, only one (0.27%) developed AML. The AML, which was FAB M4 with an 11q23 translocation, was diagnosed 18 months after receiving three cycles of epirubicin and cyclophosphamide.183 The Netherlands Working Party on Transplantation in Solid Tumors reported similar results. Eight hundred eighty-five patients were randomized to two cycles of 5-fluorouracil, epirubicin, and cyclophosphamide (FEC) versus high-dose chemotherapy and autologous stem cell transplantation after completing three cycles of FEC. The high-dose chemotherapy regimen consisted of cyclophosphamide, thiotepa, and carboplatin. At a median follow-up of 57 months, 15 patients in the conventional therapy arm and 21 patients in the high-dose chemotherapy arm developed a second malignancy. One patient in the conventional dose group had a diagnosis of MDS/AML.184 The Eastern Cooperative Oncology Group reported very different results. Five hundred eleven patients received six cycles of cyclophosphamide, doxorubicin, and 5-fluorouracil (CAF) and then randomized to no further therapy versus high-dose chemotherapy and stem cell transplantation. The high-dose chemotherapy consisted of cyclophosphamide and thiotepa. Similar to The Netherlands Study, there were more second malignancies in the high-dose chemotherapy group, 9 versus 15; however, no patients in the conventional therapy group had a diagnosis of MDS/AML, whereas 9 patients in the high-dose chemotherapy group developed MDS/AML.185

The University of Minnesota reported on the development of second malignancies in 3,372 patients who underwent stem cell transplants for various diseases from January 1974 through March 2001. There were 147 post-transplant malignancies in 137 patients; 24 of the malignancies were either nonmelanoma skin cancer or carcinoma in situ. This reperesented an 8.1-fold increased risk of post-transplant malignancy. The standardized incidence ratio (SIR) was 300 for MDS/AML, 54.3 for non-Hodgkin's lymphoma including post-transplant lymphoproliferative disorder, 14.8 for Hodgkin's disease, and 2.8 for solid tumors. For MDS/AML, the cumulative incidence plateaued at 1.4% by 10 years following transplant, but the cumulative incidence of developing a solid tumor did not plateau and was 3.8% at 20 years post-transplant.186 A higher incidence of solid malignancies was reported by the City of Hope in their analysis of 2,129 patients who had undergone bone marrow transplant for hematologic malignancies. The estimated cumulative probability for developing a solid cancer was 6.1% at 10 years. The risk was particularly elevated for liver cancer with an SIR of 27.7, cancer of the oral cavity with SIR of 17.4, and cervical cancer with SIR of 13.3. Both patients with liver cancer had hepatitis C infection and all patients with squamous cell carcinoma of the skin had chronic graft-versus-host disease. The risk was highest for survivors who were younger than 34 years at the time of transplant. Cancers of the thyroid, liver, and oral cavity occurred primarily in patients who had received total body irradiation.187 The development of solid tumors over a prolonged period warrants close long-term monitoring of these patients.

Patients who receive allogeneic transplants are also demonstrated to have an increased risk of solid tumors.177 One advantage in analyzing this population is the existence of good registries for many of the patients. In an analysis of 19,229 patients at 235 centers, the relative risk of solid tumors at 10 years was 8.3; the cumulative incidence was 2.2% at 10 years and 6.7% at 15 years. Solid tumors with a notable increase in risk included tumors of the skin, oral cavity, central nervous system, connective tissue, and liver. A younger age and higher dose of total-body irradiation predicted for a higher relative risk. The increased risk of skin and oral cavity tumors was primarily related to the presence of graft-versus-host disease. Osteosarcomas have been reported in four patients undergoing bone marrow transplant for ALL (three patients) and sickle cell disease (one patient). All four patients received alkylating agents and three received total body irradiation. The osteosarcoma arose at an age (adolescence) and site (around the knee) typical for the disease.188

Patients Receiving Cyclophosphamide Therapy

Bladder toxicity associated with the use of the oxazaphosphorine nitrogen mustards cyclophosphamide and ifosfamide has been long recognized.189 The acute cystitis is likely related to toxic metabolites and can be limited by the concomitant use of mesna. An increased number of reports have now been published of bladder cancer in patients who received long-term cyclophosphamide therapy.190, 191 The most common situations in which this occurs are in some pediatric protocols, in low-grade lymphomas, and in immunosuppressive therapy. A review of a cohort of 6,171 medium-term or long-term survivors of non-Hodgkin's lymphoma revealed 48 cases of urothelial cancer.191 Overall, a 4.5-fold increase in risk of bladder cancer was estimated from the use of cyclophosphamide; however, the cumulative dose was critical in determining risk. In patients who received more than 50 g of cyclophosphamide, the risk increased 15-fold, which translated to an absolute risk of 7% within 15 years of treatment. The long-term use of cyclophosphamide is now less common in treating pediatric and adult cancers; however, the risk of secondary urothelial cancer may be an important consideration in decisions about therapy in immunologic diseases.

Patients Receiving Immunosuppressive Agents

Cytotoxic agents such as azathioprine and cyclophosphamide are also immunosuppressive agents and have been used in the treatment of rheumatoid arthritis, scleroderma, Wegener's granulomatosis, nephrotic syndrome, and glomerulonephritis, as well as in the control of rejection in renal transplantation.192, 193, 194

Accumulated experience with these and other immunosuppressive agents suggests a different mechanism of tumor induction from that observed in patients treated for neoplastic conditions. Patients treated with immunosuppressive agents have a high incidence of malignant lymphomas, often with evidence of the presence of Epstein-Barr virus, which show a predilection for primary sites in the brain; this may be from long-term immunosuppression resulting in decreased immune surveillance. This state resembles the chronic immunodeficiency of certain inherited disorders, such as Wiskott-Aldrich syndrome, which is also associated with a high incidence of lymphomas.195

Nucleoside analogs are known to be potent immunosuppressors. They have been used in hematologic malignancies such as chronic lymphocytic leukemia (CLL) and hairy cell leukemia (HCL). A review of 2,014 patients treated by National Cancer Institute protocols with fludarabine for relapsed or refractory CLL, and 2′-deoxycoformycin (DCF) and 2-chlorodeoxyadenosine (CdA) for HCL. Although comparison with the SEER database demonstrated an increased incidence of secondary malignancies for fludarabine and CdA compared with that of a normal population, the values were consistent with the increase already associated with these diseases. Consequently, these cytotoxic/immunosuppressive agents do not appear to increase the risk of secondary malignancies in CLL and HCL.196

Further evidence supporting the contention that long-term immunosuppression contributes to neoplastic induction is found in the experience of inadvertent engraftment of human tumors in donor kidneys. In one case, immunosuppression led to the development of a tumor of donor origin, but tumor rejection occurred rapidly after cytotoxic therapy ceased. Immunosuppression is not an entirely satisfactory explanation for the high incidence of lymphomas in transplant patients because long-term alkylating-agent therapy leads to nonlymphocytic leukemia in patients with multiple myeloma or ovarian carcinoma. Continued investigations into the role of immune surveillance in carcinogenesis are necessary to define the mechanisms responsible for the development of neoplasms in immunosuppressed patients. The complex interaction of various factors (such as the interleukins and interferons) is being defined. How antineoplastic drugs interact with these factors must be defined before the impact of antineoplastic agents on immune surveillance is known.

CONCLUSION

Both clinical and laboratory studies have implicated alkylating agents and epipodophyllotoxins as potent carcinogens. Strong evidence exists for carcinogenicity in laboratory systems for the antitumor antibiotics and procarbazine; the clinical evidence suggests less of a risk. Antimetabolites as a group are much less hazardous, likely because of fewer interactions with DNA. Newer agents such as the topoisomerase I inhibitors and the taxanes have not been used for a sufficient duration to allow estimation of any carcinogenic risk. Long-term immunosuppression with azathioprine has led to an increased incidence of lymphoid malignancies, perhaps by an entirely different mechanism than those producing mutagenic effects. The combined use of chemotherapy and radiotherapy definitely increases the risk of tumor induction. All of this, however, must be interpreted in the context of the need to successfully treat a potentially lethal primary cancer.

The available data suggest that certain guidelines should be followed in the design, use, and follow-up of chemotherapy (and radiation therapy) for patients with potentially curable diseases. A careful surveillance must be conducted for secondary neoplasms during long-term follow-up of these patients. An attempt should be made to establish the quantitative risk of neoplasia for any regimen that proves curative, and efforts should be made to limit the use of the more highly carcinogenic agents. On the basis of present information, caution is required when using alkylating agents or epipodophyllotoxins. Careful prospective and retrospective studies should be aimed at establishing whether a total-dose threshold exists for carcinogenicity of suspected carcinogens in humans and whether modification of the schedule of administration affects this risk. Pharmacogenetic factors may soon play a role in determining the most appropriate therapy to reduce the risk of secondary malignancies. Finally, further attention should be directed to the development of new agents that do not have mutagenic or cytotoxic actions, but that exert regulatory actions on cell growth and differentiation.

REFERENCES

1. Haddow A. Mode of action of chemical carcinogens. Br Med Bull 1947;4:331–342.

2. Lerner HJ. Acute myelogenous leukemia in patients receiving chlorambucil as long-term adjuvant chemotherapy for stage II breast cancer. Cancer Treat Rep 1978;62:1135–1138.

3. Fisher B, Rockette H, Fisher ER, et al. Leukemia in breast cancer patients following adjuvant chemotherapy or postoperative radiation: the NSABP experience. J Clin Oncol 1985;3:1640–1658.

4. Boice JD, Greene MH, Killen JY, et al. Leukemia after adjuvant chemotherapy with semustine (methyl-CCNU). N Engl J Med 1986;314:119–120.

5. Kapadia SB, Krause JR, Ellis LD, et al. Induced acute non-lymphocytic leukemia following long-term chemotherapy. Cancer 1980;45:1315–1321.

6. Rizzo SC, Ricevuti G, Gamba G, et al. Multimodal treatment in operable breast cancer. BMJ 1981:283–437.

7. Miller A. Carcinogenesis by chemicals: an overview— G.H.A. Clowes Memorial Lecture. Cancer Res 1970;30:559–576.

8. Farber E. Carcinogenesis—cellular evolution as a unifying thread: presidential address. Cancer Res 1973;33:2537–2550.

9. Miller EC. Some current perspectives on chemical carcinogenesis in humans and experimental animals: presidential address. Cancer Res 1978;38:1479–1496.

10. Pitot HC. The molecular biology of carcinogenesis. Cancer 1993;72:962–970.

11. Keller R. Suppression of natural antitumor defense mechanisms by phorbol esters. Nature 1979;282:729–731.

12. Price CC, Gaucher GM, Koneru P, et al. Mechanism of action of alkylating agents. Ann N Y Acad Sci 1969;163:593–600.

13. Singer B. Sites in nucleic acids reacting with alkylating agents of differing carcinogenicity or mutagenicity. J Toxicol Environ Health 1977;2:1279–1295.

14. Alban, JM, Wild, C, Rollinson, S, et al. Polymorphism in glutathione S-transferase P1 is associated with susceptibility to chemotherapy-induced leukemia. PNAS 2001;98:11592–11597.

15. Haase, D, Binder, Bunger, J et al. Increased risk for therapy-associated hematologic malignancies in patients with carcinoma of the breast and combined homozygous gene deletions of glutathione transferases M1 and T1. Leukemia Res 2002;226: 249–254.

16. Kelly KM, Perentis JP. Polymorphisms of drug metabolizing enzymes and markers of genotoxicity to identify patients with Hodgkin's lymphoma at risk of treatment-related complications. Ann Oncol 2002;13(Suppl1):34–39.

17. Hagan CR, Rubin CM. Mobile genetic element and genotoxic cancer therapy: potential clinical implications. Am J PharmacoGenomics 2002;2:25–35.

18. Zarbl H, Sukumar S, Arthur AV, et al. Direct mutagenesis of ha-ras-1 oncogenes by N-nitroso-N-methylurea during initiation of mammary carcinogenesis in rats. Nature 1985;315:382–385.

19. Nath J, Krichna G. Fundamental and applied genetic toxicology. In: Craig E, Stitzel RE, eds. Modern Pharmacology with Vlinical Applications. Boston: Little, Brown and Company, 1997:69–77.

20. McCann J, Ames BN. A simple method of detecting environmental carcinogens as mutagens. Ann N Y Acad Sci 1976;271:5–13.

21. McCann J, Choi E, Yamasaki E, et al. Detection of carcinogens as mutagens in the Salmonella/microsome test: assay of 300 chemicals. Proc Natl Acad Sci U S A 1975;72:5135–5139.

22. McCann J, Ames BN. Detection of carcinogens as mutagens in the Salmonella/microsome test: assay of 300 chemicals [discussion]. Proc Natl Acad Sci U S A 1975;73:950–954.

23. Seino Y, Nagao M, Yahagi T, et al. Mutagenicity of several classes of antitumor agents to Salmonella typhimurium TA98, TA100, and TA92. Cancer Res 1978;38:2148–2156.

24. Brundrett RB, Colvin M, White EH. Comparison of mutagenicity, antitumor activity, and chemical properties of selected nitrosoureas and nitrosoamides. Cancer Res 1979;39:1328–333.

25. Genther CS, Schoeny RS, Loper JC. Mutagenic studies of folic acid antagonists. Antimicrob Agents Chemother 1977;12:84–92.

26. Benedict WF, Baker MS, Haroun L, et al. Mutagenicity of cancer chemotherapeutic agents in the Salmonella/ microsome test. Cancer Res 1977;37:2209–2213.

27. Rinkus SJ, Legator MS. Chemical characterization of 465 known or suspected carcinogens and their correlation with mutagenic activity in the Salmonella typhimurium system. Cancer Res 1979;39:3289–3318.

28. Albertini RJ, Castle KL, Borcherding WR. T-cell cloning to detect the mutation 6-thioguanine-resistant lymphocytes present in human peripheral blood. Proc Natl Acad Sci U S A 1982;79: 6617–6621.

29. Hirota H, Kubota M, Hashimoto H, et al. Analysis of hprt gene mutation following anti-cancer treatment in pediatric patients with acute leukemia. Mutat Res 1993;319:113–120.

30. Hirota H, Kirota M, Adachi A, et al. Somatic mutation at T-cell antigen receptor and glycophorin A loci in pediatric leukemia patients following chemotherapy: comparison with HPRT locus mutation. Mutat Res 1994;315:95–103.

31. Musalis JC, Monteforte JA, Winegar RA. Transgenic animal models for detection of in vivo mutations. Annu Rev Pharmacol Toxicol 1995;35:145–164.

32. Kato H. Spontaneous and induced sister chromatid exchanges as revealed by the BUdr-labeling method. Int Rev Cytol 1977;49: 55–97.

33. Perry P, Evans HJ. Cytological detection of mutagen-carcinogen exposure by sister chromatid exchange. Nature 1977;258: 121–125.

34. Guerrero PR, Rounds DE, Hall TC. Bioassay procedure for the detection of mutagenic metabolites in human urine with the use of sister chromatid exchange analysis. J Natl Cancer Inst 1979;62:805–809.

35. Banerjee A, Benedict WF. Production of sister chromatid exchanges by various cancer chemotherapeutic agents. Cancer Res 1979;39: 797–799.

36. Lambert B, Ringborg U, Harper E, et al. Sister chromatid exchanges in lymphocyte cultures of patients receiving chemotherapy for malignant disorders. Cancer Treat Rep 1979; 62:1413–1419.

37. Lambert B, Ringborn U, Linblad A, et al. The effects of DTIC, melphalan, actinomycin D and CCNU on the frequency of sister chromatid exchanges in peripheral lymphocytes of melanoma patients. In: Jones, Salmon SE, eds. Adjuvant Therapy of Cancer II. New York: Grune & Stratton, 1979:55–62.

38. Ohtsuru M, Ishi Y, Takai S, et al. Sister chromatid exchanges in lymphoctyes of cancer patients receiving mitomycin C treatment. Cancer Res 1980;40:477–480.

39. Heidelberger C. Chemical oncogenesis in culture. Cancer Res 1973;18:317–366.

40. Heidelberger C. Chemical carcinogenesis. Cancer 1977;40: 430–433.

41. Benedict WF, Banerjee A, Gardner A. Induction of morphological transformation in mouse C3H/10T1/2 clone 8 cells and chromosomal damage in hamster A(T1)C1-3 cells by cancer chemotherapeutic agents. Cancer Res 1977;37:2202–2208.

42. Jones PA, Benedict WF, Baker MS. Oncogenic transformation of C3H/10T1/2 clone 8 mouse embryo cells by halogenated pyrimidine nucleosides. Cancer Res 1976;36:101–107.

43. Gabrielson EW, Harris CC. Use of cultured human tissues and cells in carcinogenesis research. J Cancer Res Clin Oncol 1985;110:1–10.

44. Schamhl D, Habs M. Experimental carcinogenesis of antitumor drugs. Cancer Treat Rev 1978;5:175–184.

45. Weisburger JH, Griswald DP, Prejean JD, et al. The carcinogenic properties of some of the principal drugs used in clinical cancer chemotherapy: recent results. Cancer Res 1975;52:1–17.

46. Weisburger EK. Bioassay program for carcinogenic hazards of cancer chemotherapeutic agents. Cancer 1977;40:1935–1949.

47. Solcia E, Ballerini L, Bellini O, et al. Mammary tumors induced in rats by Adriamycin and daunomycin. Cancer Res 1978;38: 1444–1446.

48. Sieber SM, Adamson RH. Toxicity of antineoplastic agents in man: chromosomal aberrations, antifertility effects, congenital malformations and carcinogenic potential. Cancer Res 1978;22:57–155.

49. Harris C. Chemical and physical carcinogenesis: advances and perspectives for the 1990s. Cancer Res 1991;151:5023S–5044S.

50. Sugamiua H, Weston A, Caporaso NE. Biochemical and molecular epidemiology of cancer. Biomed Environ Sci 1991;4:73–92.

51. Harris RA, Yang A, Stein RC, et al. Cluster analysis of an extensive human breast cancer cell line protein expression map database. Proteomics 2002;2:2122–2223.

52. Pucci-Minafra I, Fontana S, Cancemi P, et al. Proteomic patterns of cultured breast cancer cells and epithelial mammary cells. Ann N Y Acad Sci 2002;963:122–139.

53. Wu W, Tang X, Hu W, et al. Identification and validation of metastasis associated proteins in head and neck cancer cell lines by two-dimensional electrophoresis and mass spectrometry. Clin Exp Metastasis 2002;19:319–326.

54. Smith MA, McCaffrey RP, Karp JE. The secondary leukemias: challenges and research directions. J Natl Cancer Inst 1996;88: 407–418.

55. Bokemeyer C, Schmoll HJ. Treatment of testicular cancer and the development of secondary malignancies. J Clin Oncol 1995;13: 283–292.

56. Van Leeuwen F. Second cancers. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology. 6th Ed. Philadelphia: Lippincott, 2001:2939–2964.

57. Travis, LB. Therapy associated solid tumors. Acta Oncologica 2002;41:323–333.

58. Makuch R, Simon R. Recommendations for the analysis of the effect of treatment on the development of second malignancies. Cancer 1979;44:250–253.

59. Gleevec [package insert]. Novartis Pharmaceuticals Corp; East Hanover, NJ 2005.

60. Hughes RG, Colquhoun M, Alloub M, et al. Cervical intraepithelial neoplasia in patients with breast cancer: a cytological and colposcopic study. Br J Cancer 1993;67:1082–1085.

61. Islam MQ, Kopf I, Levan A, et al. Cytogenetic findings in 111 ovarian cancer patients: therapy-related chromosome aberrations and heterochromatic variants. Cancer Genet Cytogenet 1993;65:35–46.

62. De Wit PE, de Vaan GA, de Boo TM, et al. Prevalence of naevocytic naevi after chemotherapy for childhood cancer. Med Pediatr Oncol 1990;18:336–338.

63. Baird EA, McHenry PM, MacKie RM. Effect of maintenance chemotherapy in childhood on numbers of melanocytc naevi. Br J Med 1992;305:799–801.

64. Li FP, Abramson DH, Tarone RE, et al. Hereditary retinoblastoma, lipoma and second primary cancer. J Natl Cancer Inst 1997;89:83–84.

65. Tucker MA, D'Angi GJ, Boice JD, et al. Bone sarcomas linked to radiotherapy and chemotherapy in children. N Engl J Med 1987;317:588–593.

66. Tucker MA, Meadows AT, Boice JD, et al. Leukemia after therapy with alkylating agents for childhood cancer. J Natl Cancer Inst 1987;78:459–464.

67. Neglia JP, Friedman DL, Yasui Y, et al. Second malignant neoplasms in five-year survivors of childhood cancer: childhood cancer survivor study. J Natl Cancer Inst 2001;93:618–629.

68. Winick NJ, McKenna RW, Shuster JJ, et al. Secondary acute myeloid leukemia in children with acute lymphoblastic leukemia treated with etoposide. J Clin Oncol 1993;11:209–217.

69. Pui CH, Ribeiro RC, Hancock ML, et al. Acute myeloid leukemia in children treated with epipodophyllotoxins for acute lymphoblastic leukemia. N Engl J Med 1991;325:1682–1687.

70. Kreissman SG, Gelber RD, Cohen HJ, et al. Incidence of secondary acute myelogenous leukemia after treatment of childhood acute lymphoblastic leukemia. Cancer 1992;70: 2208–2213.

71. Kimball Dalton VM, Gelber RD, Li F, et al. Second malignancies in patients treated for childhood acute lymphoblastic leukemia. J Clin Oncol 1998;16:2848–2853.

72. Felix CA. Secondary leukemias induced by topoisomerase- targeted drugs. Biochim Biophys Acta 1998;1400:233–255.

73. Smith MC, Rubinstein L, Ungerleider RS. Therapy-related acute myeloid leukemia following treatment with epipodophyllotoxins: estimating the risk. Med Pediatr Oncol 1994;23:86–98.

74. Smith MA, Rubinstein L, Anderson JR, et al. Secondary leukemia or myelodysplastic syndrome after treatment with epipodophyllotoxins. J Clin Oncol 1999;17:569–577.

75. Kushner BH, Cheung NKV, Kramer K, et al. Neuroblastoma and treatment related myelodysplasia/leukemia: the MSKCC experience. J Clin Oncol 1998;16:3880–3889.

76. Green DM, Zevon MA, Reese PA, et al. Second malignant tumors following treatment during childhood and adolescence for cancer. Med Pediatr Oncol 1994;22:1–10.

77. Jankovic M, Fraschini D, Amici A, et al. Outcome after cessation of therapy in childhood acute lymphoblastic leukaemia. Eur J Cancer 1993;29A:1839–1843.

78. Morland BJ, Radford M. Cutaneous squamous cell carcinoma following treatment for acute lymphoblastic leukaemia. Med Pediatr Oncol 1993;21:150–152.

79. Dunst J, Ahrens S, Paulussen M, et al. Second malignancies after treatment for Ewing's sarcoma: a report of the CESS- studies. Int J Radiat Oncol Biol Phys 1998;42:379–384.

80. Fuchs, B, Valenzuela, RG Petersen, IA, et al. Ewings sarcoma and the development of secondary malignancies. Clin Orthop Relat Res 2003;415:82–89.

81. Aung L, Gorlick RG, Shi W, et al. Second malignant neoplasms in long-term survivors of osteosarcoma: Memorial Sloan-Kettering Cancer center experience. Cancer 2002;95:1728–1734.

82. Wolden SL, Lamborn KR, Cleary SF, et al. Second cancers following pediatric Hodgkin's disease. J Clin Oncol 1998;16: 536–544.

83. Bhatia S, Yasui Y, Robison LL, et al. High risk of subsequent neoplasms continue with extended follow-up of childhood Hodgkin's disease: report from the Late Effects Study Group. J Clin Oncol 2003;21:4386–4394.

84. Travis LB, Hill DA, Dores GM, et al. Breast cancer following radiotherapy and chemotherapy among young women with Hodgkin disease. JAMA 2003;290:465–470.

85. Horwich A, Swerdlow AJ. Second primary breast cancer after Hodgkin's disease. Br J Cancer 2004;90:294–298.

86. Schellong G, Potter R, Bramswig J, et al. High cure rates and reduced long-term toxicity in pediatric Hodgkin's disease: the German-Austrian multicenter trial DAL-HD-90-The German-Austrian Pediatric Hodgkin's Disease Study Group. J Clin Oncol 1999;17:3736–3744.

87. Wong FL, Boice JD, Abramson DH, et al. Cancer incidence after retinoblastoma:radiation dose and sarcoma risk. JAMA 1997; 278:1262–1267.

88. Breslow NE, Takashima JR, Whitton JA, et al. Second malignant neoplasms following treatment for Wilms' tumor: a report from the National Wilms tumor Study Group. J Clin Oncol 1995;13: 1851–1859.

89. Bagley CM, Young RC, Canellos GP, et al. Treatment of ovarian carcinoma: possibilities for progress. N Engl J Med 1972;287: 856–862.

90. Einhorn N. Acute leukemia after chemotherapy (melphalan). Cancer 1978;41:444–447.

91. Sotrel G, Jafari K, Lash AF, et al. Acute leukemia in advanced ovarian carcinoma after treatment with alkylating agents. Obstet Gynecol 1976;47:67S–71S.

92. Morrison J, Yon JL. Acute leukemia following chlorambucil therapy of advanced ovarian and fallopian tube carcinoma. Gynecol Oncol 1978;6:115–120.

93. Casciato DA, Scott JL. Acute leukemia following prolonged cytotoxic agent therapy. Medicine 1979;53:32–47.

94. Reimer PR, Hoover R, Fraumeni J, et al. Acute leukemia after alkylating-agent therapy of ovarian cancer. N Engl J Med 1977;297:177–181.

95. Greene MH, Harris EL, Gershenson DM, et al. Melphalan may be a more potent leukemogen than cyclophosphamide. Ann Intern Med 1986;105:360–367.

96. Kaldor JM, Day NE, Pettersson F, et al. Leukemia following chemotherapy for ovarian cancer. N Engl J Med 1990;322:1–6.

97. Travis LB, Curtis RE, Boice JD, et al. Second malignant neoplasms among long-term survivors of ovarian cancer. Cancer Res 1996;56:1564–1570.

98. Sprance HE, Hempling RE, Piver MS. Leukemia following cisplatin-based chemotherapy for ovarian carcinoma. Eur J Gynecol Oncol 1992;13:131–137.

99. Reed E, Evans MK. Acute leukemia following cisplatin-based chemotherapy in a patient with ovarian cancer. J Natl Cancer Inst 1990;82:431–432.

100. Carbone PP, Bauer M, Baud P, et al. Chemotherapy of disseminated breast cancer: current status and prospects. Cancer 1977; 39:2916–2922.

101. Chan PYM, Sadoff L, Winkley JH. Second malignancies following first breast cancer in prolonged thio-TEPA adjuvant chemotherapy. In: Salmon SE, Jones SE, eds. Adjuvant Therapy of Cancer. Amsterdam: North-Holland, 1977:597–607.

102. Fisher B, Glass A, Redmond C, et al. L-phenylalanine mustard (L-PAM) in the management of primary breast cancer: an update of earlier findings and a comparison with those utilizing L-PAM plus 5-fluorouracil (5-FU). Cancer 1977;39:2883–2903.

103. Valagussa P, Tancini G, Bonadonna G. Second malignancies after CMF for resectable breast cancer. J Clin Oncol 1987;5:1138–1142.

104. Valagussa P, Moliterni A, Terenziani M, et al. Second malignancies following CMF-based adjuvant chemotherapy in resectable breast cancer. Ann Oncol 1994;5:803–808.

105. Arriagada R, Rutqvist LE. Adjuvant chemotherapy in early breast cancer and incidence of new primary malignancies. Lancet 1991;338:535–538.

106. Ben Yehuda, D, Krichevsky, S, Caspi, O, et al. Microsatellite instability and p53 mutations in therapy-related leukemia suggest mutator phenotype. Blood 1996;88:4296–4303.

107. Pederson-Bjergaard J, Siqsgaard T, Nielsen D, et al. Acute monocytic or myelomonocytic leukemia with balanced chromosome translocations to band. 11q23 after therapy with 4 epidoxorubicin and cisplatin or cyclophosphamide for breast cancer. J Clin Oncol 1992;10:1444–1451.

108. Diamandidou E, Buzdar AU, Smith TL, et al. Treatment-related leukemia in breast cancer patients treated with fluorouracil-doxorubicin-cyclophosphamide combination adjuvant chemotherapy: the University of Texas M.D. Anderson Cancer Center experience. J Clin Oncol 1996;14:2722–2730.

109. Chaplain G, Milan C, Sgro et al. Increased risk of acute leukemia after adjuvant chemotherapy for breast cancer: a population based study. J Clin Oncol 2000;18:2836–2842.

110. Curtis RE, Boice JD, Moloney WC, et al. Leukemia following chemotherapy for breast cancer. Cancer Res 1990;50:2741–2746.

111. Curtis RE, Boice JD, Stovall M, et al. Risk of leukemia after chemotherapy and radiation treatment for breast cancer. N Engl J Med 1990;326:1745–1751.

112. Henne T, Schmahl D. Occurrence of second primary malignancies in man—a second look. Cancer Treat Rev 1985;12:77–94.

113. Neugut AI, Murray T, Santos J, et al. Increased risk of lung cancer after bresat cancer radiation therapy in cigarette smokers. Cancer 1994;73:1615–1620.

114. Yap J, Chuba PJ, Thomas R, et al. Sarcoma as a second malignancy after treatment for breast cancer. Int J Rad Oncol Biol Phys 2002;52:1231–1237.

115. Brady MS, Garfein CF, Petrek JA, et al. Post-treatment sarcoma in breast cancer patients. Ann Surg Oncol 1994;1:66–72.

116. Rutqvist LE. Long-term toxicity of tamoxifen. Cancer Res 1993;127:257–266.

117. Satyaswaroop PG, Zaino RJ, Marbel R. Estrogen-like effects of tamoxifen on human endometrial carcinoma transplanted into nude mice. Cancer Res 1984;44:4006–4010.

118. Fornander T, Rutqvist LE, Cedermark B, et al. Adjuvant tamoxifen in early breast cancer: occurrence of new primary cancers. Lancet 1989;1:117–120.

119. Andersson M, Storm HH, Mouridsen HT. Incidence of new primary cancers after adjuvant tamoxifen therapy and radiotherapy for early breast cancer. J Natl Cancer Inst 1991;83:1013–1017.

120. Ribeiro G, Swindell R. The Christie Hospital adjuvant tamoxifen trial. Monogr Natl Cancer Inst 1992;11:121–125.

121. Fisher B, Costantino JP, Redmond CK, et al. Endometrial cancer in tamoxifen-treated breast cancer patients: findings from the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-14. J Natl Cancer Inst 1994;86:527–537.

122. Braithwaite RS, Cchlebowski RT, Lau J, et al. Meta-analysis of vascular and neoplastic events associated with Tamoxifen. Journal of General Internal Medicine 2003;18:37–47.

123. Seoud MA, Johnson J, Weed JC. Gynecological tumors in tamoxifen-treated women with breast cancer [review]. Obstet Gynecol 1993;82:165–169.

124. Fornander T, Hellstrom AC, Moberger B. Descriptive clinicopathologic study of 17 patients with endometrial cancer during or after adjuvant tamoxifen in early breast cancer. J Natl Cancer Inst 1993;85:1850–1855.

125. Magriples U, Naftolin F, Schwartz PE, et al. High-grade endometrial carcinoma in tamoxifen-treated breast cancer patients. J Clin Oncol 1993;11:485–490.

126. Katase K, Sugiyama Y, Hasumi K, et al. The incidence of subsequent endometrial carcinoma with tamoxifen use in patients with primary breast cancer. Cancer 1998;82:1698–1703.

127. Rutqvist LE, Johansson H, Signomklao T, et al. Adjuvant tamoxifen therapy for early stage breast cancer and second primary malignancies. J Natl Cancer Inst 1995;87:645–651.

128. Fisher B, Costantino J, Redmond C, et al. A randomized clinical trial evaluating tamoxifen in the treatment of patients with node negative breast cancer. N Engl J Med 1989;320:479–484.

129. Stewart HJ. The Scottish trial of adjuvant tamoxifen in node-negative breast cancer: Scottish Cancer Trials Breast Group. Monogr Natl Cancer Inst 1992;11:117–120.

130. Rutqvist LE, Cedermark B, Glas U, et al. Contralateral primary tumors in breast cancer patients in a randomized trial of adjuvant tamoxifen therapy. J Natl Cancer Inst 1991;83: 1299–1306.

131. Rubagotti A, Perrotta A, Casella C, et al. Risk of new primaries after chemotherapy and/or tamoxifen treatment for early breast cancer. Ann Oncol 1996;7:239–244.

132. Ragaz J, Coldman A. Survival impact of adjuvant tamoxifen on competing causes of mortality in breast cancer survivors. J Clin Oncol 1998;16:2018–2024.

133. Friedman MA, Trimble EL, Abrams JS. Tamoxifen: trials, tribulations, and tradeoffs. J Natl Cancer Inst 1994;86:478–479.

134. Rosner F, Grunwald H. Multiple myeloma terminating in acute leukemia. Am J Med 1974;57:927–939.

135. Kyle RA, Pierre RV, Bayard ED. Multiple myeloma and acute leukemia associated with alkylating agents. Arch Intern Med 1975;135:185–192.

136. Bergsagel DE, Bailey AJ, Langley GR, et al. The chemotherapy of plasma-cell myeloma and the incidence of acute leukemia. N Engl J Med 1979;301:743–748.

137. Sieber SM. Cancer chemotherapeutic agents and carcinogenesis. Cancer Chemother Rep 1975;59:915–918.

138. Moertel CG, Hagedorn AB. Leukemia or lymphoma and coexistent primary malignant lesions: a review of the literature and study of 120 cases. Blood 1957;12:788.

139. Tucker MA, Coleman CN, Cox RS, et al. Risk of second cancers after treatment for Hodgkin's disease. N Engl J Med 1988;318: 76–81.

140. Krikorian JG, Burke JS, Rosenberg SA, et al. Occurrence of non-Hodgkin's lymphoma after therapy for Hodgkin's disease. N Engl J Med 1979;300:452–458.

141. Canellos GP, Arseneau JC, De Vita VT, et al. Second malignancies complicating Hodgkin's disease in remission. Lancet 1975;1:947–949.

142. Arseneau JC, Canellos GP, Johnson R, et al. Risk of new cancers in patients with Hodgkin's disease. Cancer 1977;40: 1912–1916.

143. Rodriguez MA, Fuller LM, Zimmerman SO, et al. Hodgkin's disease: study of treatment intensities and incidences of second malignancies. Ann Oncol 1993;4:125–131.

144. Van Leeuwen FE, Chorus AM, van den Belt-Dusebout AW, et al. Leukemia risk following Hodgkin's disease: relation to cumulative dose of alkylating agents, treatment with teniposide combinations, number of episodes of chemotherapy, and bone marrow damage. J Clin Oncol 1994;12:1063–1073.

145. Brusamolini E, Anselm AP, Klersy C, et al. The risk of acute leukemia in patients treated for Hodgkin's disease is significantly higher after combined modality programs than after chemotherapy alone and is correlated with the extent of radiotherapy and type and duration of chemotherapy: a case-control study. Haematologica 1998;83:812–823.

146. Travis LB, Curtis RE, Stovall M, et al. Risk of leukemia following treatment for non-Hodgkin's lymphoma. J Natl Cancer Inst 1994;86:1450–1457.

147. Hancock SL, Tucker MA, Hoppe RT. Breast cancer after treatment of Hodgkin's disease. J Natl Cancer Inst 1993;85:25–31.

148. Sont JK, van Stiphout WA, Noordijk EM, et al. Increased risk of second cancers in managing Hodgkin's disease: the 20-year Leiden experience. Ann Hematol 1992;65:213–218.

149. Henry-Amar M. Second cancers after treatment of Hodgkin's disease: experience at the International Database on Hodgkin's disease (IDHD). Bull Cancer 1992;79:389–391.

150. Abrahamsen JF, Andersen A, Hannisdal E, et al. Second malignancies after treatment of Hodgkin's disease: the influence of treatment, follow-up time, and age. J Clin Oncol 1993;11: 255–261.

151. Boice JD. Second cancer after Hodgkin's disease—the price of success? J Natl Cancer Inst 1993;85:4–5.

152. Swerdlow AJ, Douglas AM, Vaughn Hudson G, et al. Risk of second primary cancers after Hodgkin's disease by type of treatment. Br J Med 1992;304:1137–1143.

153. Slanina J, Heinemann F, Henne K, et al. Second malignancies after the therapy of Hodgkin's disease: the Freiburg collective 1940 to 1991. Strahlenther Onkol 1999;175:154–161.

154. Josting A, Wiedenmann S, Franklin J, et al. Secondary myeloid leukemia and myelodysplastic syndromes in patients treated for Hodgkin's disease: a report from the German Hodgkin's lymphoma study group. J Clin Oncol 2003;21:3440–3446.

155. Duggan, D, Petroni, G, Johnson, J, et al. Randomized comparison of ABVD and MOPP/ABV hybrid for the treatment of advanced Hodgkin's disease: report of an intergroup trial. J Clin Oncol 2003;21:607–614.

156. Travis, LB, Gospodarowicz, M, Curtis, RE, et al. Lung cancer following chemotherapy and radiotherapy for Hodgkin's disease. J Natl Cancer Inst 2002;94:182–192.

157. Salloum E, Doria R, Schubert W, et al. Second solid tumors in patients with Hodgkin's disease cured after radiation or chemotherapy plus adjuvant low-dose radiation. J Clin Oncol 1996;14:2435–2443.

158. Blayney DW, Longo DL, Young RC, et al. Decreasing risk of leukemia with prolonged follow-up after chemotherapy and radiotherapy for Hodgkin's disease. N Engl J Med 1987;316: 710–714.

159. Meadows AT, Baum E, Fossati-Bellani F, et al. Second malignant neoplasms in children: an update from the Late Effects Study Group. J Clin Oncol 1985;3:532–538.

160. Glanzmann C, Veraguth A, Lutolf UM. Incidence of second solid cancer in patients after treatment of Hodgkin's disease. Strahlenther Onkol 1994;170:140–146.

161. Andre M, Mounier N, Leleu X, et al. Second cancers and late toxicities after treatment of aggressive non-Hodgkin lymphoma with the ACVBP regimen: a GELA chort study on 2837 patients. Blood 2004;103:122–128.

162. Armitage JO, Carbone PP, Connors JM, et al. Treatment realated myelodysplasia and acute leukemia in non-Hodgkin's lymphoma patients. J Clin Oncol 2004;21:897–906.

163. Finazzi G, Ruggeri M, Rodeghiero F et al. Second malignancies in patients with essential thrombocythaemia treated with busulphan and hydroxyurea: long-term follow-up of a randomized clinical trial. Br J Haematol 2002;116:923–924.

164. Van Leeuwen FE, Stiggelbout AM, van den Belt AW, et al. Second cancer risk following testicular cancer. J Clin Oncol 1993;11: 415–424.

165. Fossa SD, Langmark F, Aass N, et al. Second non-germ cell malignancies after radiotherapy of testicular cancer with or without chemotherapy. Br J Cancer 1990;61:639–643.

166. Wanderas EJ, Fossa SD, Tretli S. Risk of subsequent non–germ cell cancer after treatment of germ cell cancer in 2006 Norwegian male patients. Eur J Cancer 1997;33:253–262.

167. Bokemeyer C, Schmoll HJ. Secondary neoplasms following treatment of malignant germ cell tumors. J Clin Oncol 1993; 11:1703–1709.

168. Bachaud JM, Berthier F, Souile M, et al. Second non-germ cell malignancies in patients treated for state I–II testicular seminoma. Radiother Oncol 1999;50:191–197.

169. Nichols CR, Hoffman R, Einhorn LJ, et al. Hematologic malignancies associated with primary mediastinal germ cell tumors. Ann Intern Med 1985;102:603–609.

170. Williams SD, Birch R, Einhorn LH, et al. Treatment of disseminated germ-cell tumors with cisplatin, bleomycin and either vinblastine or etoposide. N Engl J Med 1987;316:1435–1440.

171. Nichols CR, Breeden ES, Loehrer PJ, et al. Secondary leukemia associated with a conventional dose of etoposide: review of serial germ cell tumor protocols. J Natl Cancer Inst 1993;85:36–40.

172. Bajorin DF, Motzer RJ, Rodriguez E, et al. Acute nonlymphocytic leukemia in germ cell tumor patients treated with etoposide-containing chemotherapy. J Natl Cancer Inst 1993;85:60–62.

173. Kollmannsberger C, Beyer J, Droz JP, et al. Secondary leukemia following high cumulative doses of etoposide in patients treated for advanced germ cell tumors. J Clin Oncol 1998;16:3386–3391.

174. Smith, MA, Rubinstein, L, Anderson, JR, et al. Secondary leukemia or myelodysplastic syndrome after treatment with epipodophyllotxins. J Clin Oncol 1999;17:569–77.

175. Travis LB, Curtis RE, Storm H, et al. Risk of second malignant neoplasms among long-term survivors of testicular cancer. J Natl Cancer Inst 1997;89:1429–1439.

176. Oddou S, Vey N, Viens P, et al. Second neoplasms following high-dose chemotherapy and autologous transplantation for malignant lymphomas. Leuk Lymphoma 1998;31:187–194.

177. Curtis RE, Rowlings PA, Deeg HJ, et al. Solid cancers after bone marrow transplantation. N Engl J Med 1997;336:897–904.

178. Sobecks RM, Le Beau MM, Anastasi J, et al. Myelodysplasia and acute leukemia following high-dose chemotherapy and autologous bone marrow or peripheral blood stem cell transplantation. Bone Marrow Transplant 1999;23:1161–1165.

179. Stone RM, Neuberg D, Soiffer R, et al. Myelodysplastic syndrome as a late complication following autologous bone marrow transplantation for non-Hodgkin's lymphoma. J Clin Oncol 1994;12:2535–2542.

180. Kushner BH, Heller G, Cheung N, et al. High risk of leukemia after short-term dose-intensive chemotherapy in young patients with solid tumors. J Clin Oncol 1998;16:3016–3020.

181. Pedersen-Bjergaard J, Andersen KM, Christiansen DH. Therapy-related acute myeloid leukemia and myelodysplasia after high-dose chemotherapy and autologogous stem cell transplant. Blood 2000;95:3273–3279.

182. Metayer C, Curtis RE, Vose J, et al. Myelodysplastic syndrome and acute myeloid leukemia after autotransplantation for lymphoma: a multicenter case-control study. Blood 2003;101: 1015–1023.

183. Kroger N, Zander AR, Martinelli G, et al. Low incidence of secondary myelodysplasia and acute myeloid leukemia after high-dose chemotherapy as adjuvant therapy for breast cancer patients: a study by the Solid Tumors Working Party of the European Group for Blood and Marrow Transplantation. Ann Oncol 2003;14:554–558.

184. Rodenhuis S, Bontenbal M, Beex LVAM, et al. High-dose chemotherapy with hematopoietic stem-cell rescue for high-risk breast cancer. N Engl J Med 2003;349:7–16.

185. Tallman MS, Gray R, Robert NJ, et al. Conventional adjuvant chemotherapy with or without high-dose chemotherapy and autologous stem-cell transplantation in high-risk breast cancer. N Engl J Med 2003;349:17–26.

186. Baker KS, DeFor TE, Burns LJ, et al. New malignancies after blood or marrow stem-cell transplantation in children and adults: incidence and risk factors. J Clin Oncol 2003;21: 1352–1358.

187. Bhatia S, Louie AD, Bhatia R, et al. Solid cancers after bone marrow transplantation. J Clin Oncolo 2001;19:464–471.

188. Bielack SS, Rerin JS, Dickerhoff R, et al. Osteosarcoma after allogeneic bone marrow transplantation. A report of four cases from the Cooperative Osteosarcoma Study Group (COSS). Bone Marrow Transplantation 2003;31:353–359.

189. Siu LL, Moore MJ. Evidence based guidelines: use of mesna in the prevention ofifosfamide induced urotoxicity. Support Care Cancer 1998;6:144–152.

190. Inagaki T, Ebisuno S. Cyclophosphamide induced urinary bladder and renal pelvic tumor: a case report. Nippon Hinyokika Gakkai Zasshi 1998;89:674–677.

191. Travis LB, Curtis RE, Glimelius B, et al. Bladder and kidney cancer following cyclophosphamide therapy for non- Hodgkin's lymphoma. J Natl Cancer Inst 1995;87:524–530.

192. Roberts MM, Bell R. Acute leukemia after immunosuppressive therapy. Lancet 1976;2:768–770.

193. Penn I. Second malignant neoplasms associated with immunosuppressive medication. Cancer 1976;37:1024–1032.

194. Steinberg AD, Plotz PH, Wolff SM, et al. Cytotoxic drugs in treatment of nonmalignant disease. Ann Intern Med 1972;76: 619–642.

195. Penn I. Occurrence of cancer in immune deficiencies. Cancer 1974;34:858–866.

196. Cheson BD, Vena DA, Barrett J, et al. Second malignancies as a consequence of nucleoside analog therapy for chronic lymphoid leukemias. J Clin Oncol 1999;17:2454–2460.