Handbook of Cancer Chemotherapy (Lippincott Williams & Wilkins Handbook Series), 8th Ed.

3. Systematic Assessment of the Patient With Cancer and Consequences of Treatment

Roland T. Skeel


A. Pathologic diagnosis is critical

Although it might seem obvious that the diagnosis of cancer must be firmly established before chemotherapy or any other treatment is administered, the critical nature of an accurate diagnosis warrants a reminder. As a rule, there must be cytologic or histologic evidence of neoplastic cells together with a clinical picture consistent with the diagnosis of the cancer under consideration. It is rarely acceptable to initiate treatment based solely on clinical exam, radiologic evidence, and nontissue laboratory evidence, such as tumor markers. Commonly, patients present to their physician with a complaint such as a cough, bleeding, pain, or a lump; through a logical sequence of evaluation, the presence of cancer is revealed on a cytologic or histologic specimen. Less frequently, lesions are discovered fortuitously during routine examination, evaluation of an unrelated disorder, or systematic screening for cancer. With some types of cancer, pathologists can establish the diagnosis based on small amounts of material obtained from needle biopsies, aspirations, or tissue scrapings. Other cancers require larger pieces of tissue for special staining, immunohistologic evaluation, flow cytometry, examination by electron microscopy, or more sophisticated studies such as evaluation for genetic deletions, amplifications, or other mutations.

It is often helpful to confer with the pathologist before obtaining a specimen to determine what kind and size of specimen is adequate to establish the complete diagnosis. When a tissue diagnosis of cancer is made by the pathologist, it is incumbent on the clinician to review the material with the pathologist. This practice is good medicine. It also allows the clinician to tell the patient that he or she has actually seen the cancer and to avoid administering chemotherapy without a firm pathologic diagnosis. In addition, the pathologist often gives a better consultation—not just a tissue diagnosis—when the clinician shows a personal interest.

B. Pathologic and clinical diagnosis must be consistent

Once the tissue diagnosis is established, the clinician must be certain that the pathologic diagnosis is consistent with the clinical findings. If the two are not consistent, a search must be made for additional information, clinical or pathologic, that allows the clinician to make a unified diagnosis. A pathologic diagnosis, like a clinical diagnosis, is also an opinion with varying levels of certainty. The first part of the pathologic diagnosis—and usually the easier part—is an opinion about whether the tissue examined is neoplastic. Because most pathologists rarely render a diagnosis of cancer unless the degree of certainty is high, a positive diagnosis of cancer is generally reliable. The clinician must be more cautious if the diagnosis rendered states that the tissue is “highly suggestive of” or “consistent with the diagnosis of” cancer. Absence of definitively diagnosed cancer in a specimen does not mean that cancer is not present, however; it means only that it could not be diagnosed on the tissue obtained, and clinical circumstances must establish if additional tissue sampling is necessary. A second part of the pathologist's diagnosis is an opinion about the type of cancer and the tissue of origin. This determination is not necessary in all circumstances but is usually helpful in selecting the most appropriate therapy and making a determination of prognosis.

C. Treatment without a pathologic diagnosis

There are rare circumstances in which treatment is undertaken before a pathologic diagnosis is established. Such circumstances are clearly exceptions, however, and involve less than 1% of all patients with cancer. Therapy is begun without a pathologic diagnosis only when the following conditions are met:

1. The clinical features strongly suggest the diagnosis of cancer, and the likelihood of a benign diagnosis is remote.

2. Withholding prompt treatment or carrying out the procedures required to establish the diagnosis would greatly increase a patient's morbidity or risk of mortality.

Two examples of such circumstances are (1) a primary tumor of the midbrain and (2) superior vena cava syndrome from a large mediastinal mass with no accessible supraclavicular nodes and no endobronchial disease found on bronchoscopy in the occasional patient in whom the risk of bleeding from mediastinoscopic biopsy is deemed greater than the risk of administering radiotherapy for a disease of uncertain nature.


Once the diagnosis of cancer is firmly established, it is important to determine the anatomic extent or stage of the disease. The steps taken for staging vary considerably among cancers because of the differing natural histories of the tumors.

A. Staging system criteria

For most cancers, a system of staging has been established based on the following factors:

1. Natural history and mode of spread of the cancer

2. Prognostic import of the staging parameters used

3. Value of the criteria used for decisions about therapy.

B. Staging and therapy decisions

In the past, surgery and radiotherapy were used to treat patients with cancer in early stages, and chemotherapy was used when surgery and radiotherapy were no longer effective or when the disease was in an advanced stage at presentation. In such circumstances, chemotherapy was only palliative (except for gestational choriocarcinoma), and in the absence of exquisitely sensitive tumors or strikingly potent drugs, the likelihood of increasing the survival was low. As knowledge has increased about the genetic determinants of cancer growth, tumor cell kinetics, and the development of resistance, the value of early intervention with chemotherapy has been transposed from animal models to human cancers. To plan this intervention and evaluate its effectiveness, careful staging has become increasingly important. Only when the exact extent of disease has been established can the most rational plan of treatment for the individual patient be devised, whether it is surgery, radiotherapy, chemotherapy, or molecular targeted therapy alone or in combination.

Although no single staging system is universally used for all cancers, the system developed jointly by the American Joint Committee on Cancer and the TNM Committee of the International Union Against Cancer is most widely used for staging solid tumors. It is based on the status of the primary tumor (T), regional lymph nodes (N), and distant metastasis (M). For some cancers, tumor grade (G) is also taken into account. The stage of the tumor is based on a condensation of the total possible TNM and G categories to create stage groupings, usually stages 0, I, II, III, and IV, which are relatively homogeneous with respect to prognosis. When relevant to the specific cancers whose chemotherapy is discussed in Section II of this handbook, the staging system or systems most commonly used for that cancer are discussed.


The performance status refers to the level of activity of which a patient is capable. It is a measure independent from the anatomic extent or histologic characteristics of the cancer and of how much the cancer or comorbid conditions have affected the patient, and a prognostic indicator of how well the patient is likely to respond to treatment.

A. Types of performance status scales

Two performance status scales are in wide use:

bull The Karnofsky Performance Status Scale (Table 3.1) has 10 levels of activity. It has the advantage of allowing discrimination over a wide scale but the disadvantages of being difficult to remember and perhaps of making discriminations that are not clinically useful.



bull The Eastern Cooperative Oncology Group (ECOG)/World Health Organization (WHO)/Zubrod Performance Status Scale (Table 3.2) has the advantages of being easy to remember and making discriminations that are clinically useful.

According to the criteria of each scale, patients who are fully active or have mild symptoms respond more frequently to treatment and survive longer than patients who are less active or have severe symptoms. A clear designation of the performance status distribution of patients in therapeutic clinical trials is thus critical in determining the comparability and generalizability of trials and the effectiveness of the treatments used.

B. Use of performance status for choosing treatment

In the individualization of therapy, the performance status is often a useful parameter to help the clinician decide whether the patient will benefit from treatment or will be made worse. For example, unless there is some reason to expect a dramatic response of a cancer to chemotherapy, treatment may be withheld from patients with an ECOG Performance Status Scale score of 3 or 4 because responses to therapy are infrequent and toxic effects of the treatment are likely to be great.

C. Quality of life

A related but partially independent measure of performance status can be determined based on patients' own perceptions of their quality of life (QOL). QOL evaluations have been shown to be independent predictors of tumor response and survival in some cancers, and they are important components in a comprehensive assessment of response to therapy. For some cancers, improvement in QOL measures early in the course of treatment is the most reliable predictor of survival time.


Response to therapy may be measured by survival (with or without disease), objective change in tumor size or in tumor product (e.g., immunoglobulin in myeloma), and subjective change.

A. Survival

One goal of cancer therapy is to allow patients to live as long and with the same QOL as they would have if they did not have the cancer. If this goal is achieved, it can be said that the patient is cured of the cancer (though biologically, the cancer may still be present). From a practical standpoint, we do not wait to see if patients live a normal lifespan before saying that a given treatment is capable of achieving a cure, but we follow a cohort of patients to see if their survival within a given timespan is different from that in a comparable cohort without the cancer. For the evaluation of response to adjuvant therapy(additional treatment after surgery or radiotherapy that is given to treat potential nonmeasurable, micrometastatic disease), survival analysis (rather than tumor response) must be used as the definitive objective measure of antineoplastic effect. With neoadjuvant therapy(chemotherapy or biologic therapy given as initial treatment before surgery or radiotherapy), tumor response, and resectability are also partial determinants of effectiveness.

B. Definitions

The overall survival rate is used to describe the percentage of people in a cohort who are alive for a specified period of time after diagnosis or initiation of a given treatment. The median survival time is the time after either diagnosis or treatment at which half of the patients with a given disease are still alive. Disease-free survival, the length of time after treatment for a specific disease during which a patient survives with no sign of the disease, is often a useful comparator in clinical studies of adjuvant therapy, as return of disease most often represents loss of curability. Progression-free survival (PFS) is the length of time during and after treatment in which a patient is living with a disease that does not get worse. It is used primarily in studies of metastatic or unresectable disease.

C. Other considerations

It is, of course, possible that a patient may be cured of the cancer that was treated but dies early owing to complications associated with the treatment, including second cancers. Even with complications (unless they are acute ones such as bleeding or infection), survival of patients who have been cured of the cancer is likely to be longer than if the treatment had not been given, though shorter than if the patient had never had the cancer.

If cure is not possible, the reduced goal is to allow the patient to live longer than if the therapy under consideration were not given. It is important for physicians to know if, and with what likelihood, any given treatment will result in a longer life. Such information helps physicians to choose whether to recommend treatment and the patient to decide whether to undertake the recommended treatment program.

It is important to learn from the patient what his or her goals of therapy are and to have a frank discussion about whether those goals are realistic. This can avoid unnecessary surprises and anger at some later time, which can occur when the patient has set a goal that is not realistic and the physician has not discussed what may or may not reasonably be expected as a consequence of therapy.

D. Objective response

Although survival is important to the individual patient, it is determined not only by the initial treatment undertaken but also by biologic determinants of the patient’s individual cancer and subsequent treatment; thus, survival does not give an early measurement of a given treatment effectiveness. Tumor regression, on the other hand, when measurable, frequently occurs early in the course of effective treatment and is therefore a readily used determinant of treatment benefit. Tumor regression can be determined by a decrease in size of a tumor or the reduction of tumor products.

1. Tumor size. When tumor size is measured, responses are usually classified by the Response Evaluation Criteria in Solid Tumors (RECIST) methodology first published in 2000 and revised in 2008 (RECIST 1.1), reported by Eisenhauer et al. in the European Journal of Cancer in 2009, and available online at http://www.eortc.be/recist/documents/RECISTGuidelines.pdf.

a. Baseline lesions are characterized as “measurable” or “non-measurable.” To be measurable, non-lymph node lesions must be 20 mm or more in longest diameter and measurable by calipers using conventional techniques, or 10 mm or more in longest diameter using computed tomography (CT). On CT scan, lymph nodes must be ≥ 15 mm for target lesions or 10 to 15 mm in short axis for nontarget lesions. Smaller lesions and truly nonmeasurable lesions are designated nonmeasurable. To assess response, all measurable lesions up to a maximum of two per organ and five in total are designated as “target” lesions and measured at baseline. Except for lymph nodes, only the longest diameter of each lesion is measured. The sum of the longest diameters of all target lesions is designated the “baseline sum longest diameter.”

There are a variety of lesions in cancer that cannot be measured. These include blastic and sclerotic metastatic lesions to the bone, effusions, lymphangitic disease of the lung or skin, and lesions that have necrotic or cystic centers. Bone lesions are measurable only if they include an identifiable soft-tissue component, which constitutes the measureable lesion.

b. Response categories are based on measurement of target lesions.

(1) Complete response (CR) is the disappearance of all target lesions. If lymph nodes are included in target lesions, each node must achieve a short axis 10 mm.

(2) Partial response (PR)is a decrease of atleast 30% in sum of the longest diameters of target lesions, using as reference the baseline sum of the longest diameters.

(3) Stable disease (SD) is when there is neither sufficient shrinkage to qualify for partial response nor sufficient increase to qualify for progressive disease.

(4) Progressive disease (PD) is an increase of 20% or more in the sum of the longest diameters of target lesions, taking as reference the smallest sum's longest diameter recorded since the treatment started or the appearance of one or more new lesions. The sum must also demonstrate an absolute increase of at least 5 mm. While the fluorode-oxyglucose (FDG)-positron emission tomography (PET) scan cannot be used to determine measureable disease, if there is negative FDG-PET at baseline with a positive FDG-PET at follow-up, this is sign of PD, based on a new lesion.

(5) Inevaluable for response is a category used where there is early death by reason of malignancy or toxicity, tumor assessments were not repeated, etc

c. Time to progression based on response criteria is an additional indicator that is often used, similar to PFS. It takes into account the fact that from the patient's perspective, CR, PR, and SD may be meaningless distinctions so long as the tumor is not causing symptoms or impairment of function. It also takes into account that some agents result in disease stability for a substantial period, despite failure to produce measurable disease shrinkage. This is particularly true for biologic targeted agents where it has been shown that time to progression for some cancers is substantially prolonged, despite no measurable reduction in tumor size.

Time to progression can also be used as an indicator of disease status when there was no measurable disease at the outset of therapy or when the therapeutic modalities were not comparable. For example, if one wanted to compare the results of surgery alone with those of chemotherapy alone, time to progression from the onset of treatment would allow a valid comparison of the effectiveness of the treatments, whereas the traditional tumor response criteria would not. Time to progression thus places each of the agents or modalities on an even basis.

d. Survival curves. If survival curves of patient populations having different categories of response are compared, those patients with a CR frequently survive longer than those with a lesser response. If a sizable number of CRs occur with a treatment regimen, the survival rate of patients treated with that regimen is likely to be significantly greater than that of patients who are untreated. When the number of complete responders in a population rises to about 50%, the possibility of cure for a small number of patients begins to appear. With increasing percentages of complete responders, the frequency of cures is likely to increase correspondingly.

Although patients who have partial response to a treatment usually survive longer than those who have SD or progression, it is often not easy to demonstrate that the overall survival of the treated population is better than that of a comparable untreated group. In part, this difficulty may be due to a phenomenon of small numbers. If only 15% to 20% of a population respond to therapy, the median survival rate may not change at all, and the numbers may not be high enough to demonstrate a significant difference in survival duration of the longest surviving 5% to 10% of patients (the “tail” of the curves) of the treated and untreated populations. It is also possible that the patients who achieve a PR to therapy are those who have less aggressive disease at the outset of treatment and thus will survive longer than the nonresponders, regardless of therapy. These caveats notwithstanding, most clinicians and patients welcome even a PR as a sign that offers hope for longer survival and improved QOL.

2. Tumor products. For many cancers, objective tumor size changes are difficult or impossible to document. For some of these neoplasms, tumor products (hormones, antigens, antibodies) may be measurable and may provide a good, objective way to evaluate tumor response. Two examples of such markers that closely reflect tumor cell mass are the abnormal immunoglobulins (M proteins) produced in multiple myeloma and the human chorionic gonadotropin produced in choriocarcinoma and testicular cancer. Other markers such as prostate-specific antigen or carcinoembryonic antigen are not quite as reliable but are nonetheless helpful measures of response of the tumor to therapy. In some cancers, reduction in the number of circulating tumor cells is also an indicator of response to therapy.

3. Evaluable disease. Other objective changes may occur but are not easily quantifiable. When these changes are not easily measurable, they may be termed evaluable. For example, neurologic changes secondary to primary brain tumors cannot be measured with a caliper, but they can be evaluated using neurologic testing. An arbitrary system of grading the degree of severity of neurologic deficit can be devised to permit surrogate evaluation of tumor response. Evaluable disease is not a category of the RECIST criteria.

4. Performance status changes may also be used as a measure of objective change; although in some respects, the performance status is as representative of subjective aspects as it is of the objective status of the disease.

E. Subjective change and QOL considerations

A subjective change is one that is perceived by the patient but not necessarily by the physician or others around the patient. Subjective improvement and an acceptable QOL are often of far greater importance to the patient than objective improvement: If the cancer shrinks, but the patient feels worse than before treatment, he or she is not likely to believe that the treatment was worthwhile. It is not valid to look at subjective change in isolation, however, because temporary worsening in the perceived state of well-being may be necessary to achieve subsequent long-term improvement.

This point is particularly well illustrated by the combined modality treatment in which chemotherapy is used to treat micrometastases after surgical removal of the macroscopic tumor. In such a circumstance, the patient is likely to feel entirely well after the primary surgical procedure, but the side effects of chemotherapy increase the symptoms and make the patient feel subjectively worse for the period of treatment. The patient should be encouraged to continue treatment, however, because if the chemotherapy treatment of the micrometastases is successful, he or she will be cured of the cancer and can be expected to have a normal or near-normal life expectancy rather than dying from recurrent disease. Most patients agree that the temporary subjective worsening is not only tolerable but well worth the price if cure of the cancer is a distinct possibility. This judgment depends on the severity and duration of symptoms, functional impairment, and perceptions of illness during the acute phase of the treatment; the expected benefit (increased likelihood of survival) anticipated as a result of the treatment; and the potential long-term adverse consequences of the treatment.

In contrast, when chemotherapy is given with a palliative intent, patients (and less often physicians) may be unwilling to tolerate significant side effects or subjective worsening from treatment. Fortunately, subjective improvement often accompanies objective improvement, so those patients in whom there is measurable improvement of the cancer also feel better. The degree of subjective worsening that each patient is willing to tolerate varies, and the patient and physician together must discuss and evaluate whether the chemotherapy treatment program is worth continuing. Such discussions should include a clear presentation of the scientific facts that include objective survival and tumor response data together with whatever QOL information has been documented for the treatment proposed. Moreover, the expressed goals and desires and the social, economic, psychological, and spiritual situations of the patient and his or her family must be sensitively considered.

A word of caution about discussions of response and survival is important. Patients can more easily understand the notion of response rates than survival probabilities. For example, a 50:50 chance of the cancer shrinking helps them to understand the goals and expectations of therapy and does not lead to undue anxiety over time. On the other hand, understanding and dealing with median or expected survival estimates is more problematic intellectually and even more difficult emotionally. It is therefore usually best to give the patient a range of expected survival rather than a discrete number. For example, the physician can say, “Some patients may have progression of their disease and possibly die within 6 months, but others may go on feeling fairly well and functioning well for 2 or more years.” This helps the patient and family not to focus on a single number (“They said I only had 13 months to live”) and to avoid some of the feeling of impending doom.


A. Factors affecting toxicity

One of the characteristics that distinguishes cancer chemotherapeutic agents from most other drugs is the frequency and severity of anticipated side effects at usual therapeutic doses. Because of the severity of the side effects, it is critical to monitor the patient carefully for adverse reactions so that therapy can be modified before the toxicity becomes life-threatening. Most toxicity varies according to the following factors:

1. Specific agent

2. Dose

3. Schedule of administration, including infusion rate and frequency of dose

4. Route of administration

5. Predisposing factors in the patient, including genetic variants,* that may be known and predictive for toxicity or unknown and resulting in unexpected toxic effects


* An example is homozygosity for the UGT1A1*28 allele, a variation of a uridine diphosphate glucuronosyltransferase gene and its corresponding enzyme (UGT1A1), which is responsible for glucuronidation of bilirubin and involved in deactivation of Sn-38, a toxic active metabolite of irinotecan.

B. Clinical testing of new drugs for toxicity

Before the introduction of any agent into wide clinical use, the agent must undergo testing in carefully controlled clinical trials. The first set of clinical trials are called phase I trials. They are carried out with the express purpose of determining toxicity in humans and establishing the maximum tolerated dose; although with antineoplastic agents, they are done only in patients who might benefit from the drug. Such trials are undertaken only after extensive tests in animals have been completed. Much human toxicity is predicted by animal studies, but because of significant species differences, initial doses used in human studies are several times lower than doses at which toxicity is first seen in animals. Phase I trials are carried out using several schedules, and the dose is escalated in successive groups of patients once the toxicity of the prior dose has been established.

At the completion of phase I trials, there is usually a great deal of information about the spectrum and anticipated severity of acute drug effects (toxicity). However, because patients in phase I trials often do not live long enough to undergo many months of treatment, chronic or cumulative effects may not be discovered. Discovery of these toxicities may occur only after widespread use of the drug in phase II trials (to establish the spectrum of effectiveness of the drug), in phase III trials (to compare the new drug or combination with standard therapy), or from postmarketing reports (when even larger numbers and less rigorously selected patients are treated).

C. Common acute toxicities

Some toxicities are relatively common among cancer chemothera-peutic agents. Common acute toxicities include the following:

1. Myelosuppression with leukopenia, thrombocytopenia, and anemia

2. Nausea, vomiting, and other gastrointestinal effects

3. Mucous membrane ulceration and cutaneous effects, including alopecia

4. Infusion reactions.

Some of these toxicities occur because of the cytotoxic effects of chemotherapy on rapidly dividing normal cells of the bone marrow and epithelium (e.g., mucous membranes, skin, and hair follicles) incidental to the mechanism of action of the drugs; others such as nausea and vomiting or infusion reactions are not related to the antineoplastic mechanism of action.

D. Selective toxicities

Other toxicities are less common and are specific to individual drugs or classes of drugs. Examples of drugs and their related toxicities include the following:

1. Anthracyclines and anthracenediones: irreversible cardiomyopathy

2. Asparaginase: anaphylaxis (allergic reaction), pancreatitis

3. Bleomycin: pulmonary fibrosis

4. Cisplatin: renal toxicity, neurotoxicity

5. Epidermal growth factor receptor inhibitors: acneiform rash

6. Fludarabine, cladribine, pentostatin, and temozolomide: prolonged suppression of cellular immunity with heightened risk for opportunistic infection

7. Ifosfamide and cyclophosphamide: hemorrhagic cystitis

8. Ifosfamide: central nervous system toxicity

9. Mitomycin: hemolytic-uremic syndrome and other endothelial cell injury phenomena

10. Monoclonal antibodies (e.g., rituximab, trastuzumab): hypersensitivity reactions

11. Paclitaxel: neurotoxicity, acute hypersensitivity reactions

12. Procarbazine: food and drug interactions

13. Trastuzumab: reversible cardiomyopathy

14. Vascular endothelial growth factor inhibitors: gastrointestinal perforation, impaired wound healing

15. Vinca alkaloids: neurotoxicity.

E. Recognition and evaluation of toxicity

Anyone who administers chemotherapeutic agents must be familiar with the expected and the unusual toxicities of the agent the patient is receiving, be prepared to avert severe toxicity when possible, and be able to manage toxic complications when they cannot be avoided. The specific toxicities of commonly used individual chemotherapeutic agents are detailed in Chapter 33.

For the purpose of reporting toxicity in a uniform manner, criteria are often established to grade the severity of the toxicity. For many years, a simplified set of criteria was used by several National Cancer Institute (NCI)–supported clinical trial groups for the most common toxic manifestations. Although this document was helpful, it was, in many respects, incomplete. To address this issue, a new set of more comprehensive toxicity criteria, the Common Toxicity Criteria, was developed in 1999. A revised version of these criteria (Common Terminology Criteria for Adverse Events v3.0 [CTCAE]) was published in 2003 and updated again in 2009 (CTCAE v4.0) and is available online at http://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm. A host of other helpful information can be obtained online at http://ctep.cancer.gov/. All new clinical trials approved by the NCI Cancer Therapy Evaluation Program use these new toxicity criteria. Such standardization is important in the evaluation of the toxicity of cancer treatment.

F. Acute toxicity management

Prevention and treatment of bone marrow suppression can be partially achieved using filgrastim, sargramostim, epoetin alfa, and oprelvekin. Treatment of its infectious, bleeding, and anemia consequences is discussed in Chapters 27 and 28. Management of nausea and vomiting, mucositis, and alopecia as well as diarrhea, nutrition problems, and drug extravasation are discussed in Chapter 26. Other acute toxicities are discussed with the individual drugs in Chapter 33. Long-term medical problems are a special issue and are highlighted in the subsequent section.


A. Late organ toxicities

Late organ toxicities may be minimized by limiting doses when thresholds are known. In most instances, however, individual patient effects cannot be predicted. Treatment is primarily symptomatic.

1. Cardiac toxicity (e.g., congestive cardiomyopathy) is most commonly associated with high total doses of the anthracyclines (doxorubicin, daunorubicin, epirubicin). In addition, high-dose cyclophosphamide as used in transplantation regimens may contribute to congestive cardiomyopathy. When mediastinal irradiation is combined with these chemotherapeutic agents, cardiac toxicity may occur at lower doses. Although evaluation of ventricular ejection fraction with echocardiography or nuclear radiography studies has been useful for acutely monitoring the effects of these agents on the cardiac ejection fraction, studies have reported late onset of congestive heart failure during pregnancy or after the initiation of vigorous exercise programs in adults who were previously treated for cancer as children or young adults. The cardiac reserve in these previously treated cancer patients may be marginal. It is probable that there are some changes that take place even at low doses, and it is only because of the great reserve in cardiac function that effects are not measurable until higher doses have been used. Mediastinal irradiation also accelerates atherogenesis and may lead to premature symptomatic coronary artery disease.

Because of the large number of women with breast cancer who are treated with doxorubicin as part of an adjuvant chemotherapy regimen, this group is of special concern and warrants ongoing clinical follow-up, although adverse cardiac effects do not appear to increase with time, up to 13 years. Trastuzumab (and perhaps other targeted agents) is also associated with cardiac toxicity, but the mechanism of toxicity is different from that of the anthracyclines and is usually reversible.

2. Pulmonary toxicity has been classically associated with high doses of bleomycin (more than 400 U). However, a number of other agents have been associated with pulmonary fibrosis (e.g., alkylating agents, methotrexate, nitrosoureas). Premature respiratory insufficiency, especially with exertion, may become evident with aging.

3. Nephrotoxicity is a potential toxicity of several agents (e.g., cisplatin, methotrexate, nitrosoureas). These agents can be associated with both acute and chronic toxicities. Other nephrotoxic agents such as amphotericin or aminoglycosides may exacerbate the problem. Even usually benign agents such as the bisphosphonates or allopurinol may be a problem. Rarely, some patients may require hemodialysis as a result of chronic toxicity.

4. Neurotoxicity has been particularly associated with the vinca alkaloids, cisplatin, oxaliplatin, epipodophyllotoxins, taxanes, bortezomib, and ixabepilone. Peripheral neuropathy can cause considerable sensory and motor disability. Autonomic dysfunction may produce debilitating postural hypotension. Whole-brain irradiation, with or without chemotherapy, can be a cause of progressive dementia and dysfunction in some long-term survivors. This is particularly a problem for patients with primary brain tumors and for some patients with small-cell lung cancer who have received prophylactic therapy. Survivors of childhood leukemia have developed a variety of neuropsychological abnormalities related to central nervous system prophylaxis that included whole-brain irradiation.

It has become evident over the years that some patients (up to one in five) who have received adjuvant chemotherapy for carcinoma of the breast also have measurable cognitive deficits such as difficulties with memory or concentration. This appears to be greater for women who have received high-dose chemotherapy than for those women who have received standard-dose chemotherapy; in both groups, the incidence is higher than in control groups. It is not uncommon for patients to refer to the effects of chemotherapy with complaints about memory being worse than it was, not being able to calculate numbers in their head, or just having “chemo-brain.” Rarely patients may have severe, debilitating, idiosyncratic cognitive impairment or even fatal central nervous system damage subsequent to chemotherapy.

5. Hematologic and immunologic impairment is usually acute and temporally related to the cancer treatment (e.g., chemotherapy or radiation therapy). In some instances, however, there can be persistent cytopenias, as with alkylating agents. Immunologic impairment is a long-term problem for patients with Hodgkin lymphoma, which may be due to the underlying disease as well as to the treatments that are used. Fludarabine, cladribine, and pentostatin, with or without rituximab, cause profound suppression of cluster of differentiation 4 (CD4) and CD8 lymphocytes and render treated patients susceptible to opportunistic infections for many months after treatment has been discontinued. Temozolomide causes CD4 lymphopenia and also carries a risk of opportunistic infection. Complete immunologic reconstitution may take 2 years after these therapies or marrow-ablative therapy requiring stem cell reconstitution. Patients who have undergone splenectomy are also at risk of overwhelming bacterial infections.

B. Second malignancies

1. Acute myelogenous leukemia and myelodysplasia may occur secondary to combined modality treatment (e.g., radiation therapy and chemotherapy in Hodgkin lymphoma), prolonged therapy with alkylating agents or nitrosoureas, or other chemotherapy. In general, this form of treatment-related acute leukemia arises in the setting of myelodysplasia and is refractory even to intensive treatment. Treatment with the epipodophyllotoxins also has been associated with the development of acute nonlymphocytic leukemia. This may be the result of a specific gene rearrangement between chromosome 9 and chromosome 11 that creates a new cancer-causing oncogene: ALL-1/AF-9. The peak time of occurrence of secondary acute leukemia in patients with Hodgkin lymphoma is 5 to 7 years after treatment, with an actuarial risk of 6% to 12% by 15 years. Thus, a slowly developing anemia in a survivor of Hodgkin lymphoma should alert the clinician to the possibility of a secondary myelodysplasia or leukemia.

Fortunately, the risk of secondary leukemias in women treated with standard adjuvant therapy for breast cancer (e.g., cyclophosphamide and doxorubicin) is only modestly higher (excess absolute risk of 2 to 5 per 100,000 person years) than that in the general population.

2. Solid tumors and other malignancies are seen with increased frequency in survivors who have been treated with chemotherapy or radiation therapy. Non-Hodgkin lymphomas have been reported as a late complication in patients treated for Hodgkin lymphoma or multiple myeloma. Patients treated with long-term cyclophosphamide are at risk of bladder cancer. Patients who have received mantle irradiation for Hodgkin lymphoma have an increased risk of breast cancer, thyroid cancer, osteosarcoma, bronchogenic carcinoma, colon cancer, and mesothelioma. In these cases, the second neoplasm is usually in the irradiated field. In general, the risk of solid tumors begins to increase during the second decade of survival after Hodgkin lymphoma. As a result, young women who have received mantle irradiation for Hodgkin lymphoma should be screened more carefully for breast cancer, starting at an age earlier than what is advised in standard screening recommendations.

C. Other sequelae

1. Endocrine problems may result from cancer treatment. Patients receiving radiation therapy to the head and neck region may develop subclinical or clinical hypothyroidism. This is a particular risk in patients receiving mantle irradiation for Hodgkin lymphoma. Biennial assessment of thyroid-stimulating hormone should be undertaken in these patients. Thyroid replacement therapy should be given if the thyroid-stimulating hormone level rises in order to decrease the risk of thyroid cancer. Short stature may be a result of pituitary irradiation and growth hormone deficiency.

2. Premature menopause may occur in women who have received certain chemotherapeutic agents (e.g., alkylating agents, procarbazine) or abdominal and pelvic irradiation. The risk is age-related, with women older than 30 years at the time of treatment having the greatest risk of treatment-induced amenorrhea and menopause. Early hormone replacement therapy should be considered in such women, if not otherwise contraindicated, to reduce the risk of accelerated osteoporosis and premature heart disease from estrogen deficiency.

3. Gonadal failure or dysfunction can lead to infertility in both male and female cancer survivors during their peak reproductive years. Azoospermia is common, but the condition may improve over time after the completion of therapy. Retroperitoneal lymph node dissection in testicular cancer may produce infertility due to retrograde ejaculation. Psychological counseling should be provided to these patients to help them adjust to these long-term sequelae of therapy. Cryopreservation of sperm before treatment should be considered in men. For women, there are limited means available to preserve ova or protect against ovarian failure associated with treatment. Abdominal irradiation in young girls can lead to future pregnancy loss due to decreased uterine capacity.

4. The musculoskeletal system can be affected byradiation therapy, especially in children and young adults. Radiation may injure the growth plates of long bones and lead to muscle atrophy. Short stature may be a result of direct injury to bone. Aromatase inhibitors increase bone loss and can contribute to pathologic osteoporotic fractures.

5. Psychological and social concerns can be severe as patients who have had cancer often carry an ongoing sense of vulnerability and frequent worry of the cancer returning. Changes in body image and sexual function can lead to difficulty with marriage and other relationships. Survivors may also suffer from discrimination on the job and find it difficult or impossible to get insurance, despite having been cured from their cancer.


The author is indebted to Dr. Patricia A. Ganz, who contributed to previous editions of this chapter. Most of the section on the late consequences of cancer treatment in this revision of the handbook represents Dr. Ganz's work.

Selected Readings

Cancer Therapy Evaluation Program. Common terminology criteria for adverse events (CTCAE) v4.0, 2009. Retrieved from http://ctep.cancer.gov/protocolDevelopment/electronic_applications/ctc.htm

Centers for Disease Control and Prevention. A national action planfor cancer survivorship: advancing public health strategies. Atlanta: Centers for Disease Control and Prevention; 2004.

Curtis RE, Boice J-D Jr, Stovall M, et al. Risk of leukemia after chemotherapy and radiation treatment for breast cancer. N Engl J Med. 1992;326:1745–1751.

Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: Revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45:228–247. Retrieved from http://www.eortc.be/recist/documents/RECISTGuidelines.pdf

Ganz PA. A teachable moment for oncologists: cancer survivors, 10 million strong and growing! J Clin Oncol. 2005;23:5458–5460.

Ganz PA Late effects of cancer in adult survivors: what are they and what is the oncologist's role in follow-up and prevention? Alexandria: American Society of Clinical Oncology; 2005: 724–730.

Ganz PA, Hussey MA, Moinpour CM, et al. Chemotherapy in breast cancer survivors treated on Southwest Oncology Group protocol S8897. J Clin Oncol. 2008;26:1223–1230.

Goldhirsch A, Gelber PD, Simes RJ, et al. Costs and benefits of adjuvant therapy in breast cancer: a quality-adjusted survival analysis. J Clin Oncol. 1989;7:36–44.

Greene FL, Page DL, Fleming ID, et al. (Eds.) AJCC cancer staging manual. 6th ed. New York: Springer, 2002.

Hewitt M, Greenfield S, Stovall S. (Eds.). From cancer patient to cancer survivor: lost in transition. Washington: The National Academies Press; 2005.

Howard RA, Gilbert ES, Chen BE, et al. Leukemia following breast cancer: an international population based study of 376,825 women. Breast Cancer Res Treat. 2007;105:359–368.

Hudson MM, Mertens AC, Yasui Y, et al. Health status of adults who are long-term childhood cancer survivors: a report from the Childhood Cancer Survivor Study. JAMA. 2003;290:1583–1592.

Loescher LJ, Welch-McCaffrey D, Leigh SA, et al. Surviving adult cancers. Part 1: physiologic effects. Ann Intern Med. 1989;111:411–432.

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.

Nieman CL, Kazer R, Brannigan RE, et al. Cancer survivors and infertility: a review of a new problem and novel answers. J Support Oncol. 2006;4:171–178.

Oeffinger KC, Mertens AC, Sklar CA, et al. Prevalence and severityofchronic diseases in adult survivors of childhood cancer: a report from the Childhood Cancer Survivor Study. J Clin Oncol. 2005;23:3s (abstr 9).

Oken MM, Creech RH, Tormey DC, et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 1982;5:649–655.

Pedersen-Bjergaard J, Sigsgaard TC, 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.

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.

Schagen SB, van Dam FS, Muller MJ, et al. Cognitive deficits after postoperative adjuvant chemotherapy for breast carcinoma. Cancer. 1999;85:640–650.

Tallman MS, Gray R, Bennett JM, et al. Leukemogenic potential of adjuvant chemotherapy for early-stage breast cancer: the Eastern Cooperative Oncology Group experience. J clin Oncol. 1995;13:1557–1563.

van Leeuwen FE, Klokman JW, Hagenbeek A, et al. Second cancer risk following Hodgkins disease: a 20-year follow-up study. J Clin Oncol. 1994;12:312–325.