In the early days of medical oncology, the term palliative chemotherapy would have been considered by many to be an oxymoron at best, a cruel jest at worst. Chemotherapy was widely seen, and not completely without justice, as a toxic assault upon a dying patient in the attempt, usually vain, to prolong his miserable life by a few extra weeks. This perception was coeval with the science itself. The first effective chemotherapy for any malignancy was reported in 1948 by Sidney Farber et al. They treated 16 children with acute leukemia. Only five entered remission, and the longest of these remissions was a paltry 6 weeks. Most of the children he treated experienced toxicity without benefit. Yet, that experience in treating acute leukemia in children proved to be the springboard from which modern chemotherapy has jumped. Now, three fourths of children with leukemia can be cured of their disease and go on to enjoy a normal life.

Although that progress in the management of childhood leukemia is certainly admirable, it does not address the semantic challenge of the term “palliative chemotherapy.” Clearly, modern chemotherapy of childhood leukemia is not simply palliative; its goal is cure. It does not conform to the World Health Organization definition of palliative care: “the active total care of patients whose disease is not responsive to curative treatment.” Of course, under this definition, any chemotherapy in which cure is not the goal or expected outcome would be considered palliative chemotherapy. Such a definition is clearly too broad for our purposes. It would include almost all chemotherapy now used to treat cancer, since only a minority of cancer cases is expected to be cured by chemotherapy.

Confounding the problem of defining “palliative chemo-therapy” is the fact that the Medicare Hospice Benefit pays for drugs “used primarily for the relief of pain and symptom control related to the individual’s terminal illness.” With this in mind, some medical oncologists have argued that any chemotherapy that is not curative in intent is, by default, palliative chemotherapy, and should be a covered service for hospice patients. Hospices have understandably resisted this interpretation, both because it is contrary to the prima facie understanding of hospice care, and because the high cost of most chemotherapy would quickly bankrupt any hospice attempting to pay for it with the fixed per diem reimbursement it currently receives from Medicare, about $107 per day.

In the first edition of this book, the term “chemotherapy” seemed less problematic than the term “palliative,” but this is no longer the case. In 1909, when Paul Ehrlich concocted the first effective drug therapy for syphilis, an inorganic compound based on arsenic, the term “chemotherapy” was used to distinguish synthesized medications from those derived from plant or animal products. Now that most pharmacotherapy is synthesized, the term “chemotherapy” has come to mean the drug therapy of cancer in particular, in contradistinction to treatment of cancer by surgery or radiation therapy. But even here, the term is unclear. For example, tamoxifen is a synthesized chemical used to treat cancer, but most oncologists would refer to it as “hormone therapy” rather than chemotherapy. Strontium-89 chloride is a simple two-atom chemical used to treat cancer, but most oncologists would refer to it as a variant of radiation therapy, rather than chemotherapy. Some treatments straddle categories. The monoclonal antibody rituximab, for example, is an immunotherapy, but it is a chimeric molecule, part murine, and part human, produced in a laboratory in a cell suspension culture derived from a third mammal, the hamster. Is it not also a chemotherapy? When a radioactive element is affixed to a monoclonal antibody, as in the case of yttrium-90 radiolabeled ibritumomab tiuxetan (Zevalin), does it thereby become radiation therapy?

Clearly, as the science of treating cancer advances, the lexicography will scurry along after it, but never quite keep up with it. For the purpose of this discussion, palliative chemotherapy refers to the administration of medications that kill or suppress growth of cancer cells, when used to ameliorate the symptoms caused by an incurable malignancy. Although they may reasonably be called forms of chemotherapy, this chapter will not discuss the palliative use of hormonal agonists or antagonists, nor will it discuss the palliative use of injectable radionuclides except in the context of radioimmunoconjugates.

Chemotherapy can be used as adjuvant, curative, or palliative treatment, depending on the type and stage of malignancy. When used in the latter two circumstances, its effect is usually described by response rate. Response rate primarily refers to objective measurements of changes in cancer size, sites of metastatic disease, and measurable serologic or other markers of tumor activity.

Adjuvant chemotherapy refers to systemic treatment administered after all gross evidence of disease has been controlled by surgery or radiation therapy, but when the possibility of cancer recurrence is high. Adjuvant chemotherapy is given with the intent of completely eradicating any undetectable cancer cells that may be present. The benefit-to-toxicity ratio of adjuvant chemotherapy is extremely important. Most medical oncologists do not suggest very toxic or potentially lethal adjuvant chemotherapy if the likelihood of cancer recurrence is low and the percentage of patients likely to benefit is also low (i.e., residual cancer cells may not be killed with the treatment). In this instance, the treatment may harm more patients than it helps. Yet the individual with the cancer can have a very different perspective regarding this issue. Patients are often willing to undergo adjuvant therapies of prolonged duration and endure significant impairment in their quality of life (QOL) for a relatively small increase in the likelihood of cure (1). An additional benefit of adjuvant chemotherapy is the positive psychological effect of “doing something” to decrease the risk of cancer recurrence. Medical oncologists are wise, however, to keep in mind the distinction between “doing something” and “doing something beneficial” (2).

The goal of curative chemotherapy is to eradicate all cancer completely and to prevent its recurrence. Diffuse large cell lymphoma, testicular cancer, and Hodgkin’s disease are just three of several different kinds of malignancies for which chemotherapy has been shown to provide the possibility of cure in the nonadjuvant setting. Chemotherapy should be recommended in most circumstances for highly chemotherapy-responsive cancers. Maintenance chemotherapy (that is, treatment continued for prolonged or indefinite periods of time after complete response is obtained) has been shown to be of no benefit for a number of chemotherapy-curable or highly chemosensitive cancers (testicular cancer, small cell lung cancer, ovarian cancer, lymphomas, and Hodgkin’s disease).

Chemotherapy used in conjunction with another anticancer therapy can be palliative if it decreases treatment-related morbidity, even if it does not improve the chance of cure. For example, anal or laryngeal carcinoma can often be locally controlled or cured with aggressive surgery. This success, however, comes with significant morbidity—the need for colostomy or loss of voice. Organ function can be maintained for patients with these and other malignancies by combined chemotherapy and radiation therapy. It is anticipated that more malignancies will be treated with these organ-sparing treatments in the future.

Chemotherapy as a single modality may also be given with no expectation of curing the cancer. In such a situation, the goal of treatment is symptom reduction, life prolongation, or the combination of the two. Whereas, most patients will accept potentially curative therapy almost regardless of the toxicity it carries, the emotional response to noncurative therapy is quite different. The patient quite rightly asks what price in side effects he must pay for life prolongation, and how much life prolongation is he likely to get for this price? All the more so, when the therapeutic goal is the more modest one of symptom reduction without life prolongation, the patient asks if the same symptom control might be achieved with less toxic therapy. The good medical oncologist asks himself the same questions.

Not surprisingly, those noncurative chemotherapies which have a high response rate and which yield a clinically significant prolongation of life also reduce cancer-related symptoms, and improve the overall QOL. The reason is self evident: the symptoms were due to the cancer, and reducing the cancer reduces the symptoms. Examples of this are initial treatment of metastatic breast cancer or extensive stage small cell lung cancer.

It is not the intent of this chapter in a textbook of palliative care and supportive oncology to discuss curative therapies, adjuvant therapies, or even dramatically life-prolonging palliative therapies. These mainstays of medical oncology are better discussed in standard textbooks of medical oncology and in oncology journals. Rather, this chapter focuses on the more problematic situation. The question concerning poorly responsive cancers addressed here are twofold:

The tacit assumption of this approach is that chemotherapy may have an anticancer effect too small to greatly prolong survival, but still have enough anticancer activity and low-enough toxicity to improve QOL.

After recovering from the initial shock of a cancer diagnosis, patients appear to be quite willing to accept chemotherapy’s side effects in an attempt to control the malignancy. In a study of 100 patients with cancer, Slevin reported the differences between patients’ attitudes toward either mild or intensive chemotherapy (3). This information was compared to a similar survey of doctors, nurses, and a control group matched to the patients for age, sex, and occupation. Study participants were asked to consider various probabilities of cure, prolongation of life, and relief of symptoms necessary for them to accept toxicities associated with mild and aggressive treatments. Patients with cancer were more likely than the other groups to accept intensive and toxic treatments for an extremely small probability of cure, prolongation of life, or symptom relief. For example, 53% of patients with cancer would undergo a very toxic chemotherapeutic regimens for a 1% chance of cure, and 42% would accept the same regimen for a 3-month prolongation of life. In contrast, only 20% of the matched control subjects would accept the treatment for a 1% chance of cure, and the same percentage would accept it for a 3-month prolongation of life. The respective endorsement by medical oncologists was only 10% for each question. The control group was also less likely to accept the treatment-related side effects, and wanted more potential benefits for any particular risk. Medical oncologists were more likely to accept aggressive treatments than general practitioners. Oncology nurses’ scores fell between those of the generalists and the control group. Radiation oncologists were the least likely of any group to accept treatment. After 3 months of actual treatment, the same questionnaire was readministered to patients. Their responses were essentially unchanged. This important study indicates that individuals’ attitudes change dramatically when they actually receive the diagnosis of cancer; when the situation is real, rather than hypothetical, more risks are taken. Patients younger than 40 years of age are even more likely to undergo chemotherapy for extremely little predicted chance of symptomatic or longevity benefit (4). These studies bring to mind the Siberian gulag-inspired observation of Alexander Solzhenitsyn in A Day in the Life of Ivan Denisovich: “A warm man can never understand a cold one.”

The likelihood of palliative chemotherapy benefit varies among cancer types. Malignancies with moderate to high sensitivity to
chemotherapy are listed in Table 50.1. Anticipated chemotherapy response rates are in the 30 to 80% range. The primary goal of chemotherapy for these incurable, locally advanced or metastatic cancers is to achieve a partial or complete tumor response and to prolong progression-free and overall survival.

Metastatic or locally advanced cancers that have low expected response rates to chemotherapy are listed in Table 50.2. Partial responses to chemotherapy of less than 25%, usually of short duration, have been reported for these malignancies. Complete responses are extremely rare. The oncologist can recommend chemotherapy of limited proven benefit, investigational treatments, or nonchemotherapeutic palliative supportive care as reasonable alternatives for patients with these malignancies. Most cancer specialists have anecdotal experience of an occasional patient’s unexpected excellent response in these chemotherapy-resistant cancers. One is cautioned not to base treatment of the average patient on the anecdotal (and at times inexplicable or coincidental) outcome of an individual patient.

Some chemotherapeutic agents have received U.S. Food and Drug Administration (FDA) approval primarily because they have improved QOL. A number of treatment studies report decreased pain, or other symptomatic improvement, induced by chemotherapy, despite little or no objective tumor response. Hypotheses to explain this phenomenon include alteration in the local milieu of the neoplasm and inhibition of production of circulating cytokines responsible for many adverse symptoms associated with malignancies (5, 6). Changes in growth factor production could result in tumor growth inhibition, as opposed to tumor shrinkage. Decreased inflammatory responses induced by many chemotherapeutic drugs or antikinins have led to their widespread use in rheumatologic diseases. Similar local environment alterations may occur at tumor sites with a decrease in tumor-associated symptoms, but without necessarily influencing tumor size.

The physician has considerable influence in the patient’s decision about starting or continuing chemotherapy. Although most of adult patients with cancer in one report (N = 439) wanted “all” information to be given, only 69% wanted to participate in the therapeutic decision-making. Twenty-five percent wanted the physician to make all of the decisions. This latter group consisted of predominantly older and sicker men (7). Although patients tend to accept their physicians’ advice, this is not universal. A better-informed and educated patient, or a patient who believes that his physician is ambivalent about a recommendation, is more likely to decide independently (8).

Oncologists should not feel obligated to offer chemotherapy to every patient with malignancy. A group of patients for whom the predicted toxicity of treatment should preclude consideration of chemotherapy can be defined (see the section Prognostic Factors). Other patients may fall into a gray zone in which the benefit to toxicity ratio of treatment is less clear.

The practice of giving subtherapeutic doses of chemotherapy primarily to prevent patients from believing their cases are hopeless is strongly discouraged. This is more likely to reflect the physician’s discomfort in dealing with a terminally ill patient than it is to address the patient’s psychological needs. A more compassionate and an ethical approach is to review the medical situation with the patient and family, and to provide them with an informed assessment of possible treatment choices. Usually, this will include nonchemotherapy palliative care. This is especially true near the end of life. Patient and family desires for open communication and participation in decision making, which may be tempered by cultural differences, must be considered in these discussions.

Investigational studies of newer chemotherapy or palliative treatments are considerations when no proven beneficial anticancer treatment exists. The patient and family should understand that the benefit of investigational treatment is unproved; it may cause more harm than good. Patients and families may mistakenly believe that “no chemotherapy” means “no need for the physician to follow up and treat the patient” or “nothing more can be done.” Unfortunately, some medical oncologists share that mistaken belief. Palliative care should be offered as a positive intervention, rather than just the absence of chemotherapy. Future visits should be scheduled with the physician who is supervising the patient’s care. Visits should be at intervals that meet the physical and psychological needs of the patient and family. Even when it seems likely that the patient will be too sick to keep any more appointments, they should be scheduled nonetheless. Such scheduling provides a message of ongoing concern for the welfare of the patient and caregivers.

Patients and families often request statistics regarding the likelihood of tumor response, expected duration of response, possible treatment-related toxicities, and average survival prolongation that may occur with chemotherapy. This information may have significant impact on the decision to accept a recommended trial of chemotherapy. In addition to giving an honest reply, the physician should respond to these inquiries by trying to learn if there are specific factors leading to the question. What questions and decisions need to be addressed in the patient’s or family’s lives? Is there a contemplated retirement, a reunion, a special upcoming event, or significant unfinished business that must be completed? Are there fears about dying, or spiritual issues that need interventions from the physician or other supportive caregivers?

Patients and families often misconstrue a survival estimate on the basis of published or personal experiences as a precise number to live or die by. For example, a physician’s statement that the median survival is 6 months is frequently misunderstood by the patient and family to mean that the patient has 6 months to live. Similarly, an expected “20% response rate to treatment” for a published cohort of patients may be misconstrued by the patient to mean that he specifically has a 20% chance of responding to the same treatment. The patient must be informed that statistics refer to groups of patients and not individuals. Oftentimes, patients or their families insist on a numeric answer concerning therapeutic efficacy or expected longevity. They may honestly be informed with such phrases as, “I expect the time would more likely be measured in weeks than in months,” or “There is a reported response rate of 60%, which lasts on average for 4 months.” This is an opportunity to discuss possible scenarios and alternatives to the question, “What if a response does not occur?” Although these replies are not very precise—indeed, it is very difficult for a reply to these questions to be both honest and precise—they at least give the patients information to correct gross misconceptions about their prognoses. It is possible to have these difficult discussions and still impart hope for comfort, maintenance of functionality for as long as possible, and finally, a death without uncontrolled pain or loss of dignity.

Patients often fear becoming a burden to their loved ones. It is unwise to deny the possibility that the patient will become a burden, for it is true. But the patient may be reminded that many we have loved most in our lives, such as children and parents, have been burdens to us from time to time. Every person, at one time or another in his life is useless. But no person is ever worthless. The family needs the opportunity to show its love by bearing this burden, and the patient wrongs them not to allow this opportunity. There is truth in the proverb “Even more than the calf wants to suck, the cow wants to suckle.” Also important is the reassurance that the physician will continue to care for the patient and family, even if chemotherapy is no longer used.

When patients and their families are provided with the best possible prognostic information, decisions frequently differ from those made when less information is imparted. Patients with metastatic colon and lung cancer who believed that they had a significant chance of dying within 6 months were less likely to choose a possible life-extending therapy over comfort care (9). When informed that the likelihood of 6-month survival was 90, 75, 50, 25, or 10%, the likelihood that the patient would choose a life-extending therapy was decreased to 51, 29, 29, 31, and 21%, respectively. In this study, patients, as compared to their physicians, more commonly overestimated their chance of surviving 6 months. For example, whereas 50% of physicians believed that a patient group would have a 6-month survival, only 12.5% of patient members of that group thought that they would survive less than 6 months. Physician estimates of prognosis were more accurate than the patient’s own estimates, as evidenced by the fact that only 45% of the patient group was alive at 6 months. If chemotherapy is chosen by mutual agreement between the patient and physician, their attitude should be that all that is necessary will be done to assure that the particular patient will benefit from the treatment, even if the response rate is reported to be relatively low. This optimistic approach after a realistic discussion is appropriate and of great psychological benefit to the patient and the family, as well as the physician and other members of the health care team.

A composite of patient and disease-related factors will help determine if chemotherapy is a reasonable treatment option. Probably the single most important factor (other than tumor type) that determines possible benefit-to-toxicity ratio of chemotherapy is the performance status (PS) or activity level of the patient. Two measures used to quantify this are the Eastern Cooperative Oncology Group (ECOG) and Karnofsky scales (Table 50.3). Because the ECOG scale offers the rater only five choices, although the Karnofsky scale offers ten (ignoring the bizarre inclusion of 0% = dead on a performance scale), it is not surprising that there is less interobserver variability with use of the ECOG scale (10). Moreover, at least in the case of lung cancer, the ECOG scale has a somewhat better prognostic capability (11). Regardless of the cause of debility (malignancy or other comorbid medical illness), a severely weakened patient with a restricted PS (ECOG 3 or 4; Karnofsky<50%) is more likely to experience excessive toxicities from chemotherapy than a beneficial response. It is worth emphasizing, however, that the data correlating benefit of chemotherapy and PS derive mostly from studies of cytotoxic chemotherapy. As more targeted (and therefore less toxic) therapies are developed, the range of PS values for which chemotherapy will prove reasonable will doubtless expand.

The sites of metastatic spread may allow better predictions of the potential initial response to chemotherapy as well as overall prognosis. For example, breast cancer metastatic to the skin, bone, and lymph nodes is usually more responsive to chemotherapy than are liver or pulmonary lymphangitic metastases. Chemotherapy “sanctuary sites,” such as the brain and other central nervous system structures, are usually least responsive to systemic cytotoxic treatments.

Numerous other prognostic features exist for specific malignancies. Data are increasingly available to support genetic aberrations, biochemical markers, and even psychometric assessments as predictive of short- and long-term prognosis for malignancies, as well as the likelihood of response to chemotherapy (12). Patients whose cancers are refractory to ongoing chemotherapy are less likely to respond to second or subsequent chemotherapy treatments. Similarly, cancers that have recurred within 6 months of cessation of chemotherapy are less likely to respond to retreatment with the same chemotherapy than are those that recur after a longer disease-free period. Absent routinely curative therapies, however, prognostication must remain, at best, predictive rather than prophetic (13).

Low PS is clearly a risk factor for poor outcome of cancer chemotherapy, and PS is at least somewhat related to age.
However, the impact of age per se on the outcome of chemotherapy is far less clear.

We do know that elderly patients are less likely to be offered chemotherapy than younger patients. For example, a report based on the Surveillance, Epidemiology, and End Results (SEER)-Medicare database showed that women over 80 years old were less likely to receive adjuvant chemotherapy for stage II ovarian cancer than women aged 65 to 69. Survival was better in women receiving chemotherapy. Controlling for other factors, mortality was not associated with increasing age (14). Similarly, a study of women over 50 years old who were eligible for adjuvant chemotherapy of breast cancer by consensus guidelines showed that increasing age was associated strongly with a decreasing likelihood of receiving a recommendation in favor of chemotherapy. This study was conducted at a tertiary referral center, the kind of facility seldom criticized as being overly timid in its treatment. Patient age correlated with treatment recommendation even when the data were adjusted for comorbidity scores. Moreover, this reduced recommendation of chemotherapy was not a reflection of patient preference; the authors found that patient acceptance of chemotherapy did not vary with age (15).