Summary of Key Points
The best palliative outcomes are obtained when palliative care is involved early in advanced lung cancer.
Integration of palliative care early in the advanced cancer patient has important health utilization and economic outcomes besides patient-related outcomes.
Supportive oncology and palliative care are not the same. Supportive oncology involves therapies to treat or minimize anticancer therapy toxicity.
There is in general a misunderstanding of palliative care as hospice care. This misunderstanding is a barrier to referral.
Patient-related outcome measures predict relevant outcomes such as survival, tolerance to chemotherapy, and performance status.
Hope in advanced lung cancer can be maintained without the promise of therapies that are unlikely to alter the cause of cancer.
There are multiple symptom and quality-of-life tools with lung cancer–specific items that can be used to assess pain, nonpain symptoms, and quality of life in lung cancer.
Pain, fatigue, dyspnea, cough, and anorexia can be successfully treated using guidelines established for specific symptom management.
Definition of Palliative Care
The World Health Organization (WHO) defines palliative care as the “holistic care of patients with advanced progressive illness. Managing pain and other symptoms, and providing psychosocial and spiritual support are paramount. The goal of palliative care is to achieve the best quality of life for patients and their families. Many aspects of palliative care also apply earlier in the course of illness, in conjunction with curative, disease-modifying or rehabilitating treatments.” This definition establishes three important points: (1) palliative care is patient-oriented rather than disease-oriented; (2) palliative care is interdisciplinary, using multiple medical and nonmedical specialties to achieve the best quality of life for the patient and family; and (3) palliative care is complementary to disease-related care. However, this definition does not provide clinical guidelines for successfully implementing early integration of palliative care. Specific guidelines are necessary for defining when and how palliative care needs to be integrated into clinical pathways for successful implementation.
Many palliative care programs have adopted multiple names, including supportive care, to facilitate early integration. According to National Council for Hospice and Specialist Palliative Care Services, supportive care helps the patient and family cope with cancer and treatment, from before diagnosis through the process of diagnosis, treatment, cure, continuing illness or death, and (family) bereavement. Supportive care helps patients maximize the benefits of cancer treatment and live as well as possible with the side effects of the disease and its treatment. Supportive care is of equal priority with diagnosis and treatment. Supportive care originally focused on side effects from anticancer therapies, such as neutropenic fever and nausea, related to chemotherapy agents. Unlike palliative care, supportive care does not have a recognized subspecialty status. Like palliative care, supportive care is multidisciplinary. Many trials have used the term best supportive care, but this term is not defined in trial documents. In one review, best supportive care was largely limited to biomedical support, such as transfusions, antibiotics, and antiemetics, and did not include advance directives, communication, and psychosocial or spiritual care or support.
End-of-life care has been largely associated with hospice care. In the United States, because of the hospice Medicare benefit, end-of-life care is assumed to be the 6 months, or less, prior to the patient’s death. Definitions of end-of-life care have broadened considerably regarding disease trajectory, when end-of-life care should be introduced, and the clinical conditions on which it focuses. As noted previously, there are multiple of terms and definitions for end-of-life care and palliative care. The fact that programs frequently use several descriptors leads to confusion. The reason palliative care programs have adopted the term supportive care is largely due to the fact that physicians and patients find the term more acceptable than palliative care.
The first inpatient palliative care unit in North America was established by Dr. Balfour Mount at the Royal Victoria Hospital, Montreal, Canada in 1975. The establishment of this unit was just 8 years after St. Christopher’s Hospice was opened in London, England by Cicely Saunders. The term “palliative” was adopted at that time to describe the function of the unit and the purpose of the program. The main reason for opening this inpatient palliative care unit within an acute care hospital was because of disturbing deficiencies in the care of patients with incurable illnesses. Dr. Mount recognized the misalignment between the goals for treating incurable and terminal illnesses, and the main goals of the acute care hospital, which were to investigate, diagnose, cure, and prolong life. In contrast, the three main goals of cancer treatment were to cure, prolong life, and palliate. Prior to the establishment of the palliative care unit, most, if not all, effort and resources were being applied toward curing or prolonging life. The palliative needs of terminally ill patients and their families, including medical, emotional, and spiritual needs, were generally neglected in an acute care hospital.
Philosophical Differences Between Oncology and Palliative Care
A shift to a scientific-based, disease-oriented medical model, which occurred in the latter half of the 19th century, resulted in major advances in health care. Louis Pasteur’s germ theory led to Joseph Lister’s management of surgical wounds and the remarkable reduction in postoperative mortality and morbidity due to infections. Florence Nightingale used statistics and collected mortality data to prove that sanitary conditions reduce mortality, and the randomized clinical trial based on disease models and evidence-based therapeutics became the language of medicine. However, the adverse outcome of this shift was to objectify patients and identify them with their disease (e.g., cancer patients) and to depersonalize the healing process through reductionism. Quantification of outcomes became a priority when treating large populations.
The curative and the disease-modifying models include the inherent assumptions that analytical, rational, and clinical encounters are the basis for scientific inquiry. The object of analysis is the disease rather than the patient and their experience of illness. Symptoms are valued as clues to a diagnosis instead of a disease worthy of treatment. Cure is contingent on effective diagnosis and treatment. Treatment is largely empiric, uniformly applied, rather than person-centered, and based on stringently controlled trials of persons with similar disease states and measurable disease-related objective outcomes that are viewed as most important (survival, disease-free survival, or progression-free survival). Laboratory and radiographic data are trusted more than the patient’s self-report, and patient-related outcomes are of secondary importance to disease-related outcomes. In fact, most physicians are unfamiliar with patient-related outcomes and outcome measures. Few oncologists routinely use patient-related outcome measures in practice and largely depend on laboratory and radiographic disease states to guide treatment and clinical decisions. The disease-oriented and curative model tends to ignore phenomena that cannot be scientifically explained. Patients are perceived as component parts, and treatment is delivered by subspecialists. Physicians tend to think in terms of molecules, cells, organ systems, and genomes, particularly those with critical cancer-related (so-called addicted) pathways. They place secondary importance on relational perspectives. Oncology within the cancer center is largely hierarchical and physician centered. Multidisciplinary tumor boards are almost entirely comprised of physicians from various subspecialties and rarely involve nonmedical and noncancer medical specialties such as nursing, palliative care, social work, and rehabilitation services. Discussions within tumor boards are almost entirely centered on managing the disease; treatment recommendations are largely biomedical in nature and limited to radiotherapy, surgery, and chemotherapy, or a combination of antitumor therapies. Death in a curative model is seen as defeat. There are no so-called good deaths in oncology. The most commonly heard excuses are that treatment failed, or, even worse, the patient failed the treatment, or nothing more can be done. The curative model fosters the involvement of palliative care in a transitional approach, in which palliative care is considered only when anticancer therapy is exhausted, rather than as simultaneous care. As a result, in the absence of palliative care support, physicians do what they have been trained to do; they give chemotherapy or targeted agents even in a terminally ill patient with little-to-no expected benefit. Reviews published regarding management of nonsmall cell lung cancer (NSCLC) have sometimes explicitly used an either-or approach when comparing chemotherapy with palliative care. In the curative model, palliative care is considered end-of-life care, and referral is made once treatment is no longer effective.
In the palliative-care approach, treatment centers around the patient and family, with the goals of providing relief of pain and other symptoms, reducing psychosocial and spiritual stressors, restoring function, including social life and the role within the family, and improving quality of life. Palliative care supports patient values, concerns and personal choices. Treatment of pain and other symptoms is a legitimate outcome and goal of medicine. Diagnosis is not a predetermined goal, but is pursued if compatible with the patient’s personal goals. Treatment is individualized based on evidence from guidelines. Palliative care, like hospice, seeks neither to hasten nor delay death and sees death as a natural part of living. The structure of palliative care services is nonhierarchical, with multiple medical and nonmedical specialists, including physicians, forming an interdisciplinary team that recognizes different roles and responsibilities. Encounters are usually lengthy. To achieve the most favorable outcome, palliative care is best practiced within multiple encounters over a prolonged period of time and is best used early in the course of illness, rather than as a crisis intervention or as a death management service at the end of life.
There are three basic integrated palliative care models. The first involves the oncologist playing both roles in cancer care. This requires extensive time and expanded expertise, and, in general, it is impractical, because of time constraints and the need for secondary training. Oncologists should be skilled in basic palliative care. In areas where there is a lack of palliative care services, this is the default model. The second approach is the so-called cafeteria model, in which the oncologist forms an interdisciplinary team by arranging multiple consultations with specialists in radiotherapy, surgery, palliative care, social work, psychology, and spiritual care. The oncologist makes up the hub of the wheel and multiple specialists serve as the spokes of the wheel. This approach is time-consuming and expensive for patients who have to make multiple visits to many consultants. A breakdown in communication is also more likely to occur, and knowing which specialist to contact for which symptom, stressor, family concern, or financial matter may be problematic for the patient. The third model involves a simultaneous consult with a palliative care specialist and the oncologist, early in the course of cancer treatment. Communication and rapport can be established with both specialists, and the oncologist is free to focus on cancer management. This is a time-efficient practice that improves communication between two services. In this model, the palliative care specialist may potentially act as a so-called treatment broker. As the patient and oncologist gain trust in the palliative care specialist, the patient may feel free to use the palliative care specialist as a sounding board to clarify preferences and goals of care. Palliative care includes several components, and interdisciplinary palliative care teams have many distinct characteristics (see Boxes 58.1 and 58.2 ).
Association or affiliation with hospice services
Home palliative care services
Grand Rounds and other teaching modalities
Provision for teaching internal medicine residents, oncology fellows, other fellows (gynecology, oncology, radiation, pain management)
Prospective observational and interventional studies
Dedicated time, fellows, personnel, and resources
Integration into oncology as supportive care trials
Continuity: reconciliation of service lines
Assessment of symptoms: expert use of patient-related outcome measures
Treatment of cancer complications and associated symptoms (expertise in pharmacologic and nonpharmacologic management of symptoms)
Treatment goals and prognosis
Assess values and preferences and understanding regarding disease stage, course, outlook, and goals of care
End-of-life care and decisions, preference for site of care
Active cancer treatment plus palliative care to active palliative care
Active palliative care to hospice
Facilitate family meetings
Understand family systems
Management of family distress, dysfunction, and family grief
Management of distress, depression, demoralization
Management of anticipatory grief, complicated grief, and depression in grief
Recognize existential suffering
Able to take a spiritual history
Understand diverse religious practices
Referral to pulmonary and nonpulmonary rehabilitation, physical therapy, and occupational therapy
Treatment of toxicity and complications related to anticancer therapy
Care of the actively dying
Research in supportive and palliative care interventions, service structures, and complex research designs
Why Palliative Care is Needed Early in the Management of Advanced Lung Cancer
Individuals presenting with lung cancer are usually highly symptomatic. Most patients with advanced disease have at least four distressing symptoms. Their average pain severity is 6 to 7, which is considered moderate, based on a numerical rating scale with 0 being no pain and 10 being severe pain. The most frequently reported symptoms are anorexia, cough, dyspnea, fatigue, pain, and insomnia. Individuals with lung cancer experience a greater prevalence of depression and anxiety than individuals with other cancers and also have more frequent and prolonged fatigue compared with patients with other cancers. Breathlessness occurs in 70% of patients, is associated with a substantially greater symptom burden, and can adversely influence the experience of caring for a loved one. Symptom burden may, in fact, be greater than patients indicate. That is why it is important to use a standardized symptom questionnaire. Studies have shown that, when presented with a symptom checklist, patients report three to four times more symptoms than they volunteer.
Even individuals with relatively early-stage lung cancer (stages I–IIIB) have substantial symptoms. When using the Lung Cancer Symptom Scale at disease presentation, patients reported an average of 11 symptoms, and, even when successfully treated, they reported an average of 6 to 7 symptoms still present a year after treatment. As mentioned previously, individuals with early-stage lung cancer experience lack of energy, worry, dyspnea, cough, and insomnia. The Karnofsky performance status often remains impaired at 36 to 52 weeks after treatment, but symptom distress gradually declines by 1 year. Symptom distress is the primary reason for clinical encounters in the outpatient setting and the primary reason for unscheduled admissions. Despite treatments for NSCLC, there is gradual deterioration in physical function, activities of daily living, and cognitive function and an increasing need for social support.
Physicians frequently believe that tumor reduction is a surrogate for improved patient-related outcomes. However, based upon a systematic review of quality of life associated with standard chemotherapy for advanced NSCLC, there are no major changes in the global quality of life for patients on standard chemotherapy regimens. There is only a modest correlation (0.35) between objective tumor size and symptoms. Improved patient-related outcomes with treatment last a much shorter time (3.8 months on average) than duration of objective tumor response (6.4 months). Objective tumor response and patient-related outcomes contribute unique, independent information. One cannot be substituted for the other. Unfortunately, most oncologists are unfamiliar with standard symptom questionnaires, and most are not aware of clinically meaningful differences in patient-related outcome measures. However, having patients complete quality-of-life and symptom questionnaires gives patients a sense of better continuity with their physician and the feeling that their treating oncologist has considered their daily activities and emotional state. Patients think that such questionnaires are helpful and not burdensome. Longitudinal assessment of symptoms is also important because treatment reduces certain symptoms but increases others.
At least 40% of outpatients with cancer are undertreated for pain, as measured by the Pain Management Index. This reflects inadequate analgesic choices based on pain severity, but does not reflect pain response. Despite the fact that greater than 60% of patients have substantial pain, there is no change in the Pain Management Index score with follow-up. Oncologists self-rate their ability to manage pain as high (7 to 10, with 10 being the best management), but rate their colleagues’ ability to control pain lower (3 to 10). Oncologists perceive lack of assessment, time constraints, and patient reluctance to complain as barriers to pain management. Most oncologists know the WHO analgesic stepladder and prescribe opioids around-the-clock for chronic pain. However, most (greater than 60%) fail to correctly answer questions regarding opioid management (dose, schedule, conversion and rotation ratios, and titration) when tested through clinical scenarios. This has been documented in several studies. Symptom burden increases in intensity and number of symptoms as cancer progresses and then plateaus in the last month of survival. Eighty percent have increasing fatigue, dyspnea, and anorexia, and most have chest pain in the last 3 months of life. Left poorly managed, these individuals experience a painful death, and families experience complicated grief or depression in bereavement. Integrating palliative care into outpatient oncology practice early in the course of cancer reduces symptom burden, improves quality of life, and provides the care and support patients and families want (see Box 58.3 ).
Be treated as individuals
Be valued for their skills and knowledge
Exercise real choice about treatment and services
Receive detailed, high-quality information about the disease, disease trajectory, prognosis, and goals of care
Know options and alternatives, including access to support groups, self-help services, and complementary therapy
Know they will undergo interventions that they have been informed of and have agreed to
Have excellent face-to-face communication
Know that services are well coordinated and are of high quality
Know that physical symptoms will be assessed and managed to their satisfaction and within the physician’s current expertise and knowledge
Have services available that can provide support and advice about financial concerns, including employment
Have access to spiritual care and be supported spiritually
Die in the place of choice
Be assured that the family will be supported throughout their illness and into bereavement
Honest communication is important but can be marred when a patient and physician maintain false hope in anticancer therapy. Good communication includes discussions about alternative therapies, prognosis, the goals of therapy, advance directives, and end-of-life care. Giving bad news requires some special skills, and few oncologists have had that type of communication training. In the US Cancer Care Outcomes Research and Surveillance (Can CORS) study, only half of patients with advanced cancer had discussions about hospice care, and yet greater than 70% died within 6 months. Of the more than 4000 physicians caring for Can CORS patients, most would not initiate discussions about prognosis, advance directives, resuscitation, and hospice care, even if death was expected within 6 months and despite national guidelines. Most physicians stated that they would postpone these discussions until the patient was highly symptomatic or failed disease-modifying therapy. Some will not conduct these discussions unless initiated by the family or patient. Medical oncologists explain the disease course in 53% of consultations and discuss the absence of cure in 84%. However after consultation, most patients with advanced lung cancer still believe there is some chance of cure with chemotherapy. Even after being fully informed using decision aids, one-third of patients with incurable cancer feel that their cancer has some chance of being cured with chemotherapy. Patients do want to be fully informed about the stage at diagnosis and the prognosis, but comprehension often lags behind. Within consultations, oncologists address symptoms in 35% of patient encounters and discuss prognosis in 39%. Patients may perceive prognosis differently than incurability, which has no time line. Even when prognosis is discussed, physicians tend to be overly optimistic. Documentation of discussions about prognosis appears in less than 40% of medical charts. However, documented discussions about prognosis are also associated with documented discussions about options regarding ongoing anticancer therapy (odds ratio 5.8), and documented do-not-resuscitate (DNR) orders (odds ratio, 2.2). Individuals with advanced cancer who overestimate their prognosis or who are given an overly optimistic estimate of survival are more likely to choose therapies for which the burden outweighs the benefit, are less likely to discuss preferences with surrogates, and are less likely to obtain information that would improve the quality of end-of-life care.
The picture may be a bit different from the patient’s perspective. In a study involving 276 patients from four major medical centers, 40% of patients rated communication with their oncologist about the potential for cure of their lung cancer low, and 80% gave low ratings about communication regarding resuscitation, life-sustaining treatments, and preparation of advance directives. Over half of patients reported that communication with their oncologist was inadequate. These findings may be related to a patient’s perception and inability to comprehend the seriousness of the situation, which is only realized as cancer progresses.
A large number of patients choose palliative chemotherapy at the end-of-life, for little to no benefit, because it helps them maintain a sense of control. Adverse effects are less of a concern for patients who value quantity of life. In this situation, alternatives are frequently not discussed. There are few, if any, decision aids to help patients make choices when considering palliative chemotherapy. Patients who have a priority of a longer life rather than better quality of life and who have had a previous response to anticancer treatment are likely to choose aggressive treatment. Patients experience cognitive dissonance when oncologists attempt to discuss end-of-life care and palliative chemotherapy within the same visit. Few patients with advanced cancer have completed advance directives, and less than one-quarter of patients want to discuss advance directives with their oncologists. Sixty percent of oncologists prefer not to discuss advance directives and end-of-life care, including resuscitation and hospice, until anticancer treatments are exhausted.
A collusion of hope surrounding anticancer therapy is maintained by continuing aggressive anticancer therapies despite little or no benefit. Although once important, patients seem to disregard quality of life and give precedence to quantity of life in choosing salvage chemotherapy. Physicians are inclined to continue anticancer therapy, despite lack of benefit, to maintain hope rather than provide supportive and palliative care. Paradoxically, this occurs despite the fact that there is little survival benefit with chemotherapy given within 1 to 3 months of death. In fact, chemotherapy may shorten survival, and early palliative care and hospice care may prolong survival.
As a result, the average time from the last chemotherapy to death is 50 to 60 days, and the time from the last targeted therapy to death is 40 to 50 days. Fourteen to eighteen percent of individuals receive chemotherapy or targeted therapy within 30 days of dying. The most common targeted agents used at the end of life are erlotinib and bevacizumab. Patients with lung cancer had greater odds (2.6) of being on a targeted agent within the last 30 days of life than individuals with other advanced cancers. In one study at a large cancer center, the median time from palliative care consult to death was 1.4 months (interquartile 0.5 to 4.2 months) but the median length of time from the first encounter with an oncologist to consultation about palliative care was 20 months (interquartile 6 to 45). Therefore, there were multiple missed opportunities to include palliative care as part of patient care earlier in the course of disease. Half of patients with lung cancer are within 2 months of death before end-of-life care and hospice are mentioned. The average stable performance score of patients with lung cancer, as measured by the Palliative Prognostic Index, a modification of the Karnofsky performance score, is 8 to 9 months. Once the Palliative Prognostic Index has dropped to 30 or less, the average survival is 0.38 months. At this stage, very few patients (less than 5%) will have improvement in their performance score. Consulting palliative care services after the patient has become bedridden, or has a poor performance score, provides little time to manage symptoms and primarily requires crisis intervention. Offering additional anticancer therapy when a performance score improves is unlikely to occur and, if discussed with the patient, will lead to a false sense of hope and will delay advanced care planning and hospice referral. Even patients with technically treatable small cell lung cancer and an Eastern Cooperative Oncology Group performance score of 3 or 4 do poorly. Only 20% of them will finish the standard four cycles. With a performance score of 4, the median survival is 7 days, and, with a performance score of 3, it is 64 days.
The end results of poor communication and a collusion of hope in anticancer therapy are aggressive care at the end of life, chemotherapy in the last 14 days of life, intensive care unit admissions in the last 30 days of life, and acute hospital-based care in the last 30 days of life. About half of patients with advanced lung cancer will have aggressive therapy at the end of life. The detrimental effect of aggressive care at the end of life is not just economic. It also creates a greater risk of caregiver depression and complicated grief. In retrospect, many patients and families regret their choices of aggressive therapy near the end of life.
In general, patients feel that advance directives should be discussed earlier than physicians do, and the majority of families wished palliative care had been involved earlier in the course of cancer treatment. In one study, end-of-life care discussions took place a median of 33 days before death. In hospitalized patients with advanced cancer, palliative consults usually occur late within the hospital stay, when death is imminent, after a substantial hospital stay, or after admission to an intensive care unit. Consultations often involve the transfer of care.
Most patients in large cancer centers do have a DNR directive in place at the time of death. The average time of signing a DNR directive is less than 3 days before death, and one-third are signed by surrogates. As few as 5% of individuals with advanced cancer who die in the hospital have signed outpatient DNR directives.
Benefits of Early Integration of Palliative Care Into Oncology
Symptom Management, Prognostic Information, and Hope
One of the benefits of integrating palliative care into outpatient oncology is decreased symptom burden. The use of symptom assessment questionnaires will uncover more bothersome symptoms, which can then be managed by the palliative care team. Individuals referred to palliative care early in the course of treatment are more likely to perceive and retain accurate information about prognosis and are less likely to receive aggressive chemotherapy at the end of life. Prognostic discussions do not dampen hope. They empower individuals by providing realistic expectations to help them make informed choices about their medical care. Early palliative care is associated with longer intervals between the last chemotherapy and death and increased hospice enrollment more than 7 days before death (60% compared with 33%). Individuals who have less than 30 days of exposure to palliative care are more likely to receive chemotherapy within 30 days of death. Hope is more influenced by a caring relationship between patient and physician than by prognostic disclosure. The great majority of caregivers feel that avoiding discussions about prognosis is an inappropriate way of maintaining hope.
Communication, Quality of Life, and Patient-Related Outcomes
Palliative care programs use quality-of-life questionnaires that have been shown to improve communication. Providing palliative care during anticancer therapy improves quality of life. Patient-centered communication takes time and involves sensitivity. Physicians must be able to respond to emotive verbal and nonverbal cues. Oncologists respond to approximately 20% to 30% of emotional cues and are quite responsive to informational cues. Within family meetings, physicians often do most of the talking (71%) compared with family members (29%). One of the core competencies required of physicians completing a fellowship in palliative medicine is to be able to conduct a family meeting. Therefore, when compared with oncology trainees, palliative medicine specialists are more likely to be better equipped to effectively communicate within the family conference. Emotional and psychosocial issues assume greater importance at the end of life. Communication training, which is part of a palliative medicine fellowship, can improve physicians’ attitudes and their responses to emotions, as well as the satisfaction of the patients’ families. Time constraints are a major issue for cancer specialists because cancer treatment and related issues need to be addressed. Paradoxically, despite the intensive involvement oncologists have with their patients, patients prefer to have conversations and communications about end-of-life care with physicians other than their oncologist. They want their oncologist to remain “optimistic” and focused on treating their cancer.
General rehabilitation, exercise with strength training, and aerobics and pulmonary rehabilitation are often neglected in the various phases of treating patients with lung cancer. Many individuals with lung cancer have chronic obstructive lung disease (COPD), and evidence indicates pulmonary rehabilitation in chronic lung disease improves quality of life, dyspnea, and fatigue and empowers patients to be involved in their own therapy. Although moderate physical activity reduces fatigue and improves symptoms, function and quality of life three-fourths of patients with lung cancer in the United States do not meet physical activity guidelines, and 51% do not participate in moderate activities. For some individuals with cancer who have undergone potentially curative therapy, physical exercise can actually reduce recurrences and all-cause mortality. Physical exercise is safe and feasible and may even potentially benefit individuals with advanced incurable cancer. There is an open trial that is investigating the benefits of 2 months of a physical exercise intervention on fatigue and quality of life in patients with unresectable lung cancer. Rehabilitation and physical exercise are most often discussed in palliative care, interdisciplinary team meetings rather than in oncology tumor boards. Therefore, patients who are seen by the palliative care team are more likely to be considered for pulmonary rehabilitation and exercise training.
Research and complementary therapies of integration of palliative care into cancer treatment have resulted in supportive and complementary therapeutic trials during standard chemotherapy. In one trial, patients using American ginseng had improved cancer-related fatigue, particularly during chemotherapy. In another trial, patients using omega-3 fatty acids during chemotherapy for NSCLC experienced improved muscle mass compared with standard treatment.
Early integration of palliative care into cancer care not only reduces aggressive care at the end of life, but also has economic advantages, without adversely influencing survival. Nearly 40% of Medicare dollars are spent in the last months of life. Even when palliative care is used as crisis intervention, transfer of appropriate patients from inpatient acute care to inpatient palliative care reduces costs by 66%. Palliative care inpatient consult teams reduce daily inpatient costs by US $239. The authors of two randomized trials and a cohort study found that care provided in inpatient palliative care units reduced costs by 38% to 50% compared with care provided in regular hospital wards. Likewise, using in-home palliative care teams reduces readmissions to the hospital and emergency department compared with standard outpatient cancer care. The average costs per day for in-home outpatient care was $95, which was substantially less than the usual care ($213). Financial comparisons of acute care hospital services have been possible using the Centers for Medicare & Medicaid Services case mix index (CMI) and All Patient Refined-Diagnosis Related Group (APR-DRG) data. Based on CMI and APR-DRG data, Cleveland Clinic’s Inpatient Palliative Medicine acute care unit’s total mean charges per admission are $7800 lower than at peer institutions without palliative care inpatient units, despite an equivalent severity of illness, longer length of stay, and higher mortality. The lower charges are due primarily to lower laboratory and pharmaceutical charges. A systematic review of costs and cost-effectiveness of palliative care has been published, but the quality of the studies varied, studies used different methods, and some studies were small. Cohort studies were at risk of bias through potentially unobserved confounding variables. However, all palliative care service structures (inpatient, home, and outpatient palliative care) were found to have economic advantages, largely due to reduced health-care utilization (readmissions and referral to hospice) and direct-care cost savings. For example, earlier palliative care enrollment reduced acute care days.
Early integration of palliative care into cancer care does not shorten survival, and may actually prolong survival. Early palliative care also leads to earlier hospice transition, which has a short-term survival advantage for patients with lung cancer. The survival advantage of palliative care needs to be confirmed by other studies because this information was based on a post hoc analysis in the Boston Study, and there has been much speculation about the mechanism used to calculate the survival advantage. Quality of life and mood are associated with survival; reduced quality of life and depression are associated with shortened survival; and improved quality of life is associated with improved survival. Interventions such as palliative care improve quality of life and reduce depression, and may have a biologic effect related to prolonged survival. It is interesting to note that, in the Boston Study, mood was improved without increased use of antidepressants. Survival benefits also may be related to reduced aggressive care at the end of life. Lastly, greater social support fostered by the palliative care team and the integration of the family into the plan of care may improve survival.
Studies Documenting Benefits to Integrating Palliative Care Into Cancer Care
The feasibility of integrating palliative care into oncology has been demonstrated in multiple clinical trials. In general, patients participating in these trials have experienced improved symptoms, quality of life, patient satisfaction, less aggressive care at the end of life, and no decrease in survival. It is important to note that the greatest benefit occurred when palliative care was integrated early in the course of advanced cancer.
Several studies involving different cancers have evaluated benefits to home nursing and symptom-support visits during chemotherapy and after treatment. Certain symptoms, including depression and dyspnea, improved. In addition, chemotherapy toxicity was lower than in the control group, patient satisfaction improved, and there were fewer emergency room visits and hospitalizations.
The benefits of inpatient palliative care consultations are relatively small unless palliative care teams assume direct care of patients. Without the palliative care team’s involvement in direct care, recommendations for medical management are carried out in less than a quarter of patients. Also, although mild improvements in anxiety and dyspnea were noted, compared with care given without palliative care consultation, there was no improvement in depression or pain.
Acute inpatient palliative care units with an interdisciplinary palliative care team have some advantages. Individuals directly managed by a palliative care team in an inpatient unit have fewer intensive care admissions, longer median hospice stays, and a greater number of completed advance directives. There were no detrimental effects on survival.
Benefits to early integration of palliative care into cancer care were found in two large studies. The Educate, Nurture, Before Life Ends (ENABLE) study involved education, problem-solving, symptom management, advanced-care planning, and monthly telephone follow-up. A full palliative care interdisciplinary team was not involved in care. The primary outcomes of mood and quality of life improved. There were no changes in symptom intensity or health-care utilization. There was a trend toward better survival with the intervention (14 vs. 8.5 months) ( p = 0.14). The second trial, conducted by Temel et al involved individuals with newly diagnosed advanced lung cancer. Patients were randomly assigned to either usual care or integrated palliative care. The interdisciplinary team saw patients as outpatients within 8 weeks of diagnosis. Individuals were repeatedly seen when they returned to the clinic, or at least monthly. Management was guided by the National Consensus Project for Quality Palliative Care. The primary criteria were mood and quality of life at 12 weeks, using patient-related outcome measures (Functional Assessment of Cancer Therapy-Lung and Hospital Anxiety and Depression Scale). Other criteria were aggressive care at the end of life and survival. The results were improved mood without an increase in antidepressant use, improved quality of life as demonstrated by a clinically meaningful change in the quality-of-life scale, earlier referral to hospice, and significant improvement in survival. The study was not powered for survival. As a result of several prospective studies, multiple national and international organizations have recommended early integration of palliative care into cancer treatment.
Barriers to Integrating Palliative Care Into Cancer Care
The collusion of hope between patient and physician, associated with aggressive anticancer therapy, leads to a belief that survival benefit increases with each salvage therapy, and delays or prevents palliative care referrals. If discussions about prognosis and end-of-life care are put off, or not introduced, until crisis episodes, it is likely that palliative care will not be used until needed for crisis intervention, or not at all. Continuing to offer therapies that are unlikely to offer benefit, and then attempting to introduce discussions regarding end-of-life care, can confuse patients. Under these circumstances, end-of-life care discussions are not likely to be well received, or may be put off by patients in favor of continuing anticancer therapy. Continuing aggressive care may be perceived by patients as the only reasonable choice and may be framed in terms of wanting to live, implying a survival advantage to ongoing aggressive therapy. These patients will have a false “either or” dichotomy between cancer therapy and palliative care. Physicians delay referral because they don’t want to destroy the patient’s hope by recommending a palliative care specialist. There is a false impression that palliative care services are dependent on disease outcomes and prognosis, rather than symptom burden, regardless of stage or expected disease outcomes. To overcome these barriers, the term supportive care has been adopted because it is more acceptable to patients and physicians and does not imply end-of-life care.
Palliative care is under resourced. Although most patients with advanced cancer have a substantial symptom burden, and nearly 40% die from cancer, only 1% of the National Institutes of Health budget is devoted to palliative care. Although palliative care services are available in more than 90% of National Cancer Institute (NCI) designated facilities and more than 75% of community cancer centers, and despite the fact that the great majority (over 80%) of cancer specialists rate palliative care as important to cancer care, less than 20% of cancer programs are likely to devote resources toward integrating palliative care into cancer treatment. The average number of full-time equivalent physicians per cancer program is only two, and they are largely overworked.
Other reasons for late referrals to palliative care include inequitable access to services and lack of standardized criteria for referral. In addition, physicians often lack an understanding of the extent and availability of services. They also lack education and information about palliative care. These factors can lead to delayed referral or no referral. Although standardized symptom assessment and measured patient-related outcomes are key to understanding symptom burden and patients’ needs, most oncologists do not routinely use symptom assessment tools and have had little palliative care experience during training. Because most patients do not report the full extent of the symptoms they are experiencing, the oncologist will remain unaware of the needs of patients.
A review of 12 textbooks published by multiple medical specialties found that hematology-oncology textbooks ranked 10th in the palliative care content. There has been improvement over the decade since that review was written, but it is interesting that, among reviews in managing advanced lung cancer published in major journals, palliative care often is either not included or is added as the last paragraph; outcomes of therapy are usually described in terms of progression-free survival and overall survival. Although there are exceptions, these data seem to indicate that oncologists have little exposure to information on palliative care in the materials they are reading to remain current in their specialty.
Some palliative care services, such as hospice, will not follow-up on patients who are being treated with chemotherapy. Integration of palliative care into cancer treatment may be difficult if palliative care specialists do not have an understanding of the natural history of cancer, treatment by stage and histology, common chemotherapy side effects, and new developments in anticancer therapies, including new targeted agents and their toxicity. As a result, palliative care specialists may refer patients to hospice programs prematurely, because they mistake treatment toxicities for progressive cancer. To overcome this barrier, palliative care specialists need to have a fundamental knowledge of oncology, a basic understanding of new developments, and close communication with the oncologists.
Public exposure to information on palliative care is minimal, but public exposure to information on cancer is usually high, because of major news stories about sensational new discoveries and individual reports of dramatic successes based on N-of-1 experiences. Because public fundraising is largely motivated by cure rather than care (e.g., Race for the Cure campaigns), funding care may be seen as less important than funding cure.
Health-care policy involving changing from pay for services to value-based reimbursement may be responsible for either improving or diminishing palliative care services. Value-based reimbursement is based upon cost-effectiveness and requires quality indicators. Palliative care quality indicators are different from those of oncology, and consensus about quality indicators is not universal. Palliative care quality indicators related to cancer care are underdeveloped. Based on value-based reimbursement determined by health-care policy, inpatient units with high mortality may be viewed as unfavorable. In addition, if direct costs alone, rather than direct costs plus indirect cost savings, are not taken into account within the administrative matrix, then inpatient palliative care units may be seen as losing propositions. Compared with general inpatient wards, the type of patient admitted to palliative care units may have higher or lower costs; but deaths are certainly higher than in the general wards. Patients on palliative care inpatient units have greater symptom severity, more serious psychosocial problems, and higher complexity of care. The All Patient Refined-Disease Related Index, although useful, may not adequately reflect the complexity of care, severity of illness, or risk of mortality seen on inpatient palliative care units. Part of the art of demonstrating case mixed severity is to use the appropriate word codes for symptoms and diseases to demonstrate for administrators and policymakers the type of patient treated on inpatient palliative care wards.
As palliative care expands to chronic nonmalignant illnesses and moves upstream to comanage cancer patients undergoing disease modifying therapy, there will be negative consequences based on availability of services. There is not enough funding, enough training programs, or enough time to train the number of palliative care specialists needed to meet the demand. It is unlikely that the medical system can take another layer of specialized care for seriously ill patients on top of expensive and already complex health care. Bundling payments will discourage the practice of multiple consultations. If palliative care specialists are able to assume all of the tasks of palliation, other specialists will begin to believe that basic symptom assessment, management and psychosocial care are not part of their responsibility.
To address these barriers, programs in all medical specialties should include training in basic palliative care skills, with treatment centered on patient goals and values, and basic symptom assessment and management. Core competency would include symptom assessment using standardized instruments, basic management of pain and nonpain symptoms, and screening for distress and psychosocial and spiritual concerns. Oncologists should discuss prognosis, goals of care, suffering, and advance directives and assess the patient’s understanding of these issues. To give providers an opportunity to assess the value of palliative care in improving the quality of oncologic care, the American Society of Clinical Oncology, through a grant from the Agency for Healthcare Research and Quality, in collaboration with the American Academy of Hospice and Palliative Medicine, is developing and disseminating a primary palliative care curriculum based on best evidence. The aim is to enhance oncologists’ understanding of basic palliative care, while palliative care specialists concentrate on more complex or refractory symptoms and problems.
As lung cancer and its symptoms progress, the preference for information also progresses. Although the preference for information regarding diagnosis, prognosis, and treatment options does not change, the preference for information about palliative care and end-of-life decisions (if not integrated early into cancer care) does change. Preferences may be for either more information or less information. As a result, it is important that oncologists update patients on disease status and prognosis and ask patients if they want information about palliative care, advance directives, and end-of-life care.
There are no universal definitions of end-of-life care, regardless of whether it is any length of time or a disease state or the patient’s preference. Unfortunately, the WHO definition of palliative care is used as a master definition for end-of-life care. Although the term end-of-life care is commonly used, there is no consensus about the components of its definition. It tends to imply either a time frame of survival or boundaries between cure and care that can be detrimental to integrating palliative care into cancer care (if palliative care is used synonymously with end-of-life care).
Assessing Symptoms and Quality of Life in Lung Cancer
Symptom control remains one of the most important practices in cancer care, palliative care, and end-of-life care and should be done continuously throughout the course of cancer treatment and into survival. System assessment is one of the major limitations to symptom management. Fatigue, dyspnea, pain, anorexia, and cachexia are the most common physical symptoms related to lung cancer. As symptom burden increases, global health and survival decreases. For patients with lung cancer, symptom burden, when measured by the number of symptoms and severity, inversely correlates with quality of life and prognosis. Symptom burden increases up to the last 4 weeks of life and then plateaus.
Over the last decade, more than 50 tools have been developed to measure quality of life in individuals with lung cancer. Commonly used quality-of-life measurement tools include the Functional Assessment of Cancer Therapy-Lung (FACT-L) and its modification (NCCN-FACT-17), Lung Cancer Symptom Scale (LCSS), and the European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Lung Cancer (EORTC-LC-13). The FACT-L and the EORTC-LC-13 are generic, quality-of-life instruments with specific items related to lung cancer symptoms attached as a disease-specific module.
The FACT-L consists of 41 self-assessment questions, of which 34 pertain to five domains. Combining the physical, functional, and lung cancer modules creates the Trial Outcome Index, a tool that is sensitive to change over time and has provided clinically meaningful, patient-related outcomes in clinical trials. Symptoms (dyspnea, difficulty breathing, cough, chest tightness, appetite and weight loss, and cognitive function) are assessed using a categorical scale with a time frame of 1 week. A change of 2 points on the 7-item symptom scale is clinically significant. The questionnaire is relatively insensitive to treatment-related symptoms.
The LCSS consists of two scales, one rated by physicians and one rated by patients. The patient questions include six symptoms and three summary questions that are marked on a visual analog scale (a 100-mm horizontal line). The mean of the six symptom items is the average symptom burden. Physician items involve six main lung cancer symptoms. A change of 10 mm in the patient’s scale is clinically meaningful. The LCSS does not measure social or affect quality-of-life domains, which is a drawback.
The EORTC-LC-13 is a 13-item, lung cancer symptom module that includes cough, hemoptysis, dyspnea (3 items), sore mouth or tongue, dysphagia, hair loss, tingling hands, tingling feet, pain (3 items), and pain medications. All questions are framed in a 4-point categoric scale and a 7-point numerical scale, with the time frame of 1 week. This quality-of-life scale is clinically valid and useful. It is also sensitive to treatment-related symptoms.
There are problems with measuring quality of life. There can be intraobserver and interobserver errors, missing data, fatigued patients, particularly for long questionnaires, and attrition, which can favorably influence the outcomes, because patients who fail to complete questionnaires are usually sicker or have dropped out of the study. In the case of attrition, statistical adjustments need to be made to prevent bias. Comparison across studies is difficult because of differences in patient mix. Visual analog scales are more difficult for patients to understand and complete. A shift response may occur as patients recalibrate the severity of their symptoms or change their priorities and feelings about the relative importance of quality-of-life domains over time. More severe symptoms will tend to lower the mean over time. Lastly, the quality of dying cannot be measured by quality-of-life questionnaires. What patients and their loved ones want at the end of life has been identified (see Box 58.4 ).
Peace with God or a supreme being
Legacy (being of some help to others)
Ability to pray and/or meditate
Ability to make funeral arrangements
Absence of burden to others
Feeling that one’s life is complete
Closure on relationships (being able to reconcile and say goodbye)
Patients should complete quality-of-life questionnaires during the initial consultation and whenever there is a reevaluation of cancer response during treatment. In between, a symptom scale may be helpful. The Edmonton Symptom Assessment System (ESAS) is a nine-question scale with an additional question for patient-specific symptoms not covered in the nine questions. It gauges symptom burden by the number and severity of symptoms using a numerical rating scale (0 = no symptoms, 10 = severe symptoms). Completion rate is high, and it can be completed daily in the inpatient setting. The ESAS may be a good way to screen for distress. Distress is a multifactor, unpleasant, emotional experience of a psychosocial and spiritual nature that may interfere with the ability to effectively cope with cancer. Pain and fatigue are the major contributors to distress. It is recommended that all patients with cancer be assessed for distress. It is most often recommended that a so-called 11-point distress thermometer be used to screen patients and that a triage system be used to manage patients with distress.
Managing Lung Cancer Symptoms
Symptoms are present both at diagnosis and with recurrent cancer. Symptoms recur, or new symptoms develop, with relapse after definitive therapy, or with progression on maintenance therapy, and are usually associated with weight loss. At least half of individuals with lung cancer make emergency department visits for intolerable symptoms sometime during the course of their cancer. Part of the reason for high emergency department utilization is that disease progression—not symptoms—is used frequently as an indicator of relapse. As mentioned earlier, symptom progression, on average, occurs before disease progression. An alternative to detection of relapse would be symptom monitoring, using self-assessed symptom forms that are completed by outpatients on a weekly basis. A categoric, self-assessment scale, which measures the severity of weight loss, fatigue, pain, cough, and breathlessness on a weekly basis, has been developed and used in a feasibility study. Sensitivity, specificity, and positive predictive value of these sentinel symptoms and weight loss were 86%, 93%, and 86%, respectively. Relapses were detected 6 weeks earlier on average than would have occurred with planned imaging. Use of this assessment tool may reduce the number of imaging procedures (which usually have a low yield in asymptomatic individuals) and allow for earlier intervention before symptoms become severe and require emergency department visits or emergency inpatient admissions.
Fatigue is almost universally experienced by individuals with advanced cancer and is the main symptom that detrimentally influences quality of life. Fatigue is a distressing persistent sense of tiredness or exhaustion that interferes with daily activities. Unlike normal physical fatigue, it occurs with normal activity and can occur without any physical activity. It is pervasive and lasts beyond the normal expected recovery time. Descriptors for fatigue, such as tiredness, induration, lack of vigor, and asthenia, may not exactly describe every patient’s experience. Fatigue can be screened through a numerical rating scale (0 = no fatigue, 10 = severe fatigue), with fatigue ratings greater than 4 being clinically significant. People with cancer who have fatigue should be screened for depression. Cancer-related fatigue is not associated with anhedonia, hopelessness, or worthlessness. Insomnia and pain should be treated. Patients with obstructive sleep apnea will also experience fatigue and may benefit from continuous positive airway pressure while sleeping. Anemia, hypothyroidism, and hypogonadism will contribute to the severity of fatigue, and treatment for these conditions may improve fatigue. Comorbidities, such as heart failure and COPD, should be assessed and treatment maximized.
In one randomized double-blind trial, cancer-related fatigue was improved with corticosteroids. However, the long-term use of corticosteroids is associated with side effects, including osteoporosis, myopathy, insomnia, thromboembolism, and psychotomimetic side effects. Dexamethasone (4 mg in the morning and at noon) was used in this trial. Doses should be tapered to the lowest effective dose, or discontinued if no response is observed after a 2-week trial period. In a second randomized, double-blind trial, American ginseng was effective in reducing cancer-related fatigue. The side effects were similar to those with placebo, and ginseng had few drug interactions. Psychostimulants were reported to be effective in prospective, single-arm studies, but these positive results have not been duplicated in randomized trials. Strength and endurance training may help to address fatigue and loss of physical function during chemotherapy, as such training has been shown to increase the 6-minute walk time, stair climbing, strength capacity, and, in patients with dyspnea, perception of shortness of breath during submaximal walking.
Dyspnea can be described as chest tightness, rapid shallow breathing, air hunger, and not getting enough air, and each descriptor reflects a different pathophysiology. Chest tightness is associated with coronary artery disease. Rapid shallow breathing is the result of a mismatch between respiratory drive and lung capacity. Air hunger is associated with increasing retention of carbon dioxide. Not getting enough air is associated with hyperinflation. Most patients with cancer have dyspnea at the end of life.
In advanced cancer, particularly lung cancer, there are usually a multitude of reasons for dyspnea. These reasons include loss of lung volume from treatment and tumor size, pleural effusions, cardiac tamponade, COPD, coronary artery disease, thromboembolism, pneumonia and wasting, anxiety and depression, and uncontrolled pain. Wasting is associated with quantitative and qualitative changes in the diaphragm muscle, which leads to reduced tidal volume with exertion and inspiratory capacity.
Dyspnea can be screened by the numerical scale within the Edmonton Symptom Assessment System. The Cancer Dyspnea Scale and the Dyspnea Numeric Scale, which assess dyspnea interference with activities, may also be used.
Depending on the underlying etiology, treatment for cancer-related dyspnea may include surgery, chemotherapy, radiotherapy, thoracentesis, pericardiocentesis, the use of drainage tubes, bronchoscopy with laser or stenting, brachytherapy, corticosteroids and antibiotics, and transfusions. Nonintervention approaches may also reduce dyspnea. Pulmonary rehabilitation and walking aids that activate accessory muscles can improve dyspnea. Oxygen should be used for patients who are hypoxic (oxygen saturation of less than 90%). Blowing air across the face, by way of a room or handheld fan, may reduce dyspnea in those who have normal oxygen levels. Noninvasive ventilation, using bilevel positive airway pressure, reduces dyspnea and avoids intubation. It does not require sedation.
Morphine reduces dyspnea without causing hypercapnia. Initial dose is 2.5 mg to 5 mg every 4 hours, as needed. If patients are opioid tolerant, a 25% increase in the opioid dose may be helpful. Nebulized opioids and furosemide should not be used as standard therapy, because of lack of evidence supporting their effectiveness. Most individuals with chronic dyspnea also have episodes of worsening dyspnea, or air hunger, that should be treated with an opioid, as needed. Benzodiazepines and sedating phenothiazines have been used to treat dyspnea, but there is conflicting evidence. There are negative and positive trials with few high-quality randomized studies. Both benzodiazepines and sedating phenothiazines may be added or substituted for opioids if patients are intolerant of opioids or wish not to go on opioids. Lastly, for patients with refractory dyspnea, palliative sedation, using subcutaneous phenobarbital, or using parenteral haloperidol plus a benzodiazepine (e.g., midazolam or lorazepam), may be necessary to control dyspnea. Palliative sedation should be discussed with patients and families before instituting treatment. The patient should have a DNR order, and, if gastric tube feedings are in place, they should be removed.
Symptomatic cough will occur in at least half of patients with lung cancer. Patients with advanced lung disease have greater dependence on cough for mucus clearance than healthy individuals. In patients with lung cancer, the cough response is prompted by excessive noxious stimulation of afferent sensory fibers through the vagus. There also can be central sensitization of neurons governing the cough reflex. Evaluation of cough should include whether the cough is productive or not, what triggers the cough, the timing of the episodes (daytime or nocturnal), the patient’s medications, (e.g., angiotensin-converting enzyme inhibitors), and the underlying comorbidities. A cough scale can be used to gauge frequency and severity. The Manchester Cough and Lung Cancer Scale uses 10 items to assess cough in lung cancer. It is presently being validated. Chest radiographs and computed tomography (CT) images may demonstrate an obstructive bronchus, pleural or pericardial effusions, atelectasis, pneumonia, bronchopleural or bronchoesophageal fistula, lymphangitic carcinomatosis, superior vena cava obstruction, or treatment-related pneumonitis.
Radiotherapy, laser therapy, brachytherapy, or stenting may relieve an obstructive bronchus. Thoracentesis, chest tube drainage plus pleurodesis, or placement of tunnelled indwelling pleural catheter can reduce cough and dyspnea associated with pleural effusions. Pericardiocentesis may be relieved with a cardiac tamponade. Simple hydration, humidification, and mucolytics may be helpful, particularly with mobilization of secretions. However, there are no well-conducted randomized trials that support the use of mucolytics in the management of cough. There is also sparse evidence for chest wall vibration and manual clearance of secretions. Flutter valve oscillation through a mouthpiece has been used. Positive end-expiratory pressure, via face mask, for 45 minutes per day improves cough and dyspnea.
Cough suppression is desirable when cough is no longer useful as a function or if cough produces pain or is fatiguing. Opioids commonly used as cough suppressants include codeine, dextromethorphan, and morphine. Dextromethorphan controls the intensity of cough better than codeine. Sustained-release morphine reduces cough severity by 40%. There is no evidence of a dose–response relationship when using morphine, or of one opioid being superior for the treatment of cough.
Proton-pump inhibitors can reduce cough caused by gastric reflux, and gabapentin has been shown to reduce cough stemming from central sensitization. Prednisone (30 mg daily for 2 weeks) may reduce cough associated with bronchospasm, or pleural, pericardial, or diaphragmatic irritation caused by the tumor or treatment. Baclofen may also work, if standard therapies have been ineffective. For patients who are imminently dying, treatment with a cholinergic inhibitor, such as glycopyrrolate, may reduce secretions and the so-called death rattle. The recommended glycopyrrolate dose is 0.1 mg to 0.2 mg (IV or subcutaneous) every 6 to 8 hours. It is best given around-the-clock because it prevents, rather than treats, secretions. Glycopyrrolate is a quaternary scopolamine derivative that does not cross into the central nervous system, thus preventing anticholinergic-induced cognitive dysfunction or delirium. Alternatives are inhaled ipratropium or scopolamine ophthalmic solution dropped onto the tongue.
Cachexia and Anorexia
Cachexia is most often recognized as involuntary weight loss. Precachexia is defined as a weight loss of less than 5%, and cachexia is defined as weight loss greater than 5%. However, if the body mass index (BMI) is less than 20, then a greater than 2% body weight loss would be defined as cachexia. Symptoms associated with cachexia include sarcopenia, loss of fat mass, anorexia, fatigue, and elevated inflammatory cytokines and acute phase reactants. Anorexia is a cluster of symptoms that includes bloating, early satiety, taste and smell changes, dysgeusia, and diurnal variations in food intake.
Peripheral and central mechanisms generate cachexia and anorexia. Inflammatory cytokines upregulate the transcription factor NF kappa-B in muscle that in turn, upregulates myostatin, proteasomes, and prostaglandins. Satellite cell proliferation is inhibited and muscle synthesis, through MyoD and mTOR/Akt, is blocked. Mitochondrial function and calcium metabolism) are also adversely affected, leading to reduced oxidative phosphorylation. In addition, there is an increase in reactive oxygen species. Anorexia results from increased neurotransmission of proopiomelanocortin-containing neurons within the arcuate nucleus of the hypothalamus. There is also a simultaneous reduction of neuropeptide Y signals. Neurotransmission through proopiomelanocortin neurons is increased through activation of serotonin receptors (5-HT2C) as well as throughout regulation of interleukin-1.
Inflammatory cytokines (tumor necrosis factor-alpha, interleukin-1, and interleukin-6) increase production of C-reactive protein from the liver and muscle. Hypoalbuminemia develops from loss of the endothelial barrier and extravasation into interstitial spaces. Elevated C-reactive protein is associated with progressive weight loss and is a poor prognostic indicator in lung cancer. Both serum albumin and C-reactive protein are combined in the Glasgow Prognostic Score. Pretreatment Glasgow prognostic score is a useful and important predictor of cancer-specific survival in patients with inoperable NSCLC. It is predictive of the most important aspects of platinum-related toxicity.
Assessment of cachexia is complex. Factors that have been used to follow the course of cachexia and sarcopenia include changes in weight, BMI, anthropometric measurements, bioelectrical impedance, dual-energy x-ray absorptiometry, and measurement of muscle area at the level of the L3 vertebral body. Using CT to measure muscle mass at the L3 vertebral body level has been validated. These measurements can be done routinely to follow the course of cancer. Assessment of anorexia can be made through the Edmonton Symptom Assessment Score. The Functional Assessment of Anorexia Cachexia Therapy scale, which is a 12-item module attached to the Functional Assessment of Cancer Therapy quality-of-life tool, can also be used. There are multiple other nutritional scales, but they do not separate starvation from cachexia. In a study evaluating use of the Mini Nutritional Assessment scale for individuals with lung cancer, the score on the scale correlated with laboratory parameters of inflammation and was independently associated with survival.
The best method for treating cancer-related cachexia is to cure the cancer. If this is not possible, then the goals of therapy are to maintain muscle mass, nutritional intake, and function. Because it is unlikely that a single drug will have a major impact on cancer cachexia and anorexia, treatment should be multimodal. Appetite stimulants include corticosteroids, progesterone, and olanzapine. In a randomized trial, the combination of megestrol acetate and olanzapine was superior to megestrol acetate as a single agent, with substantial improvements in appetite, nausea, weight gain, and quality of life. Cachexia has been treated with single-agent antioxidants, l-carnitine, omega-3 fatty acids, thalidomide, nonsteroidal antiinflammatory drugs (NSAIDs), and megestrol acetate, with little to marginal benefit. Recent trials of drug combinations have proven to be more effective than single agents in increasing lean body mass, decreasing energy expenditures, and improving appetite. Ghrelin analogs and selective androgen receptor modulators are presently in development.
At least half of individuals with advanced cancer will have chronic pain. There are well-described cancer pain syndromes. Approximately 75% of individuals with cancer have pain related to their cancer, and 25% have pain related to treatment or comorbid illnesses. Many individuals with cancer have cancer pain syndromes because of widespread metastases, and these syndromes have been well described. Clinicians can mistake existential suffering for physical pain. Characteristically, somatic pain does not respond to treatment with opioids.
To properly assess pain, clinicians need to know pain intensity, the point from where it radiates, temporal pattern, pain quality, and provocative and relieving factors. The cause, pathophysiology, and pain syndrome can be inferred from the history and confirmed by physical examination and radiographs. Pain intensity is largely used when choosing analgesics and adjusting doses, but interference with activity and function are equally important. Pain intensity may change little, but function, mood, sleep, and vitality may improve with treatment. Another area that needs to be assessed when considering treatment is the side effects associated with analgesics. Pain can be assessed on a numerical rating scale. Mild pain is less than 4 on an 11-point numerical rating scale; moderate pain is 5 to 7; and severe pain is greater than 7. A reduction of baseline pain severity between 30% and 50% is clinically significant. The Brief Pain Inventory is a validated tool and can be used in the initial assessment. Before prescribing opioids for moderate to severe pain, the patient’s substance abuse history, family substance abuse history, and the patient’s history of depression, anxiety disorder, or personality disorder should be obtained.
Classifying pain into somatic, visceral, or neuropathic is an oversimplification of clinical reality. There are features of neuropathic pain in patients who have visceral metastases, as demonstrated in animal models. Central sensitization can occur in all three subclasses of pain.
For mild pain, acetaminophen or NSAIDs can be used as the analgesics of choice, but should not be used for patients with coagulopathy, or heart or kidney failure. NSAIDs are also not recommended for treating older individuals. Prophylaxis with proton-pump inhibitors should be considered. There is evidence that the combination of NSAIDs and acetaminophen may improve analgesia.
For moderate pain, tramadol, tapentadol, codeine, or low doses of a potent opioid, such as morphine, are reasonable choices. Although it had been assumed that codeine is an analgesic—through conversion to morphine via the cytochrome CYP 2D6—newer evidence suggests that there is actually synergy between codeine and morphine. Therefore, codeine may have to be converted to morphine for analgesia. If patients were originally treated with nonopioid analgesics, such as acetaminophen or NSAIDs, these drugs can be added or continued.
For severe pain, potent opioids such as oxycodone, morphine, hydromorphone, or fentanyl may be used as first-line analgesics. Starting doses are as follows: morphine, 5 mg every 4 hours by mouth; oxycodone, 5 mg every 4 hours; hydromorphone, 1 mg every 4 hours; and transdermal fentanyl at 12 μg per hour. If pain is acute or unstable, transdermal fentanyl should not be used. Sustained-release morphine (15 mg every 12 hours) or oxycodone (10 mg every 12 hours) may be used in place of immediate-release preparations. Most patients have breakthrough pain. A rescue dose of one-sixth of the total daily opioid dose should be provided every 1 to 2 hours. Transient flares of pain may be related to activities (incident pain) or spontaneous. End-of-dose failure is considered suboptimal around-the-clock opioid dosing. If this occurs, the around-the-clock dose should be increased. There are considerable differences in dose requirements between patients who will require dose titration of no less than 25% and no more than 100% of the total daily dose. As an alternative, transmucosal, buccal, sublingual, or intranasal fentanyl have been approved for breakthrough pain. These preparations are expensive and should be reserved as second-line treatments for individuals whose breakthrough pain is unresponsive to oral, immediate-release opioids.
Titration of potent opioids is required to control pain in most individuals. To reach steady-state levels, the dose of immediate-release opioids should not be changed more often than every 24 hours. The dose of sustained-release opioids should not be changed more often than every 48 hours, and the dose of transdermal fentanyl should not be changed more often than every 48 to 72 hours. If pain remains a problem, then the rescue dose can be titrated to response, and the around-the-clock dose adjusted once steady-state is reached. Patients with severe liver disease should be treated with morphine or hydromorphone because these opioids are conjugated and glucuronidation is relatively spared. For patients who are in renal failure, methadone or buprenorphine are the opioids of choice. Because of its unique pharmacology, methadone should be used only by designated prescribers.
Adjuvant analgesics can improve pain and reduce opioid requirements. Antiseizure drugs, predominantly gabapentinoids, tricyclic antidepressants, and selective norepinephrine serotonin reuptake inhibitors (duloxetine and venlafaxine), can improve neuropathic pain. Because tricyclic antidepressants have substantial side effects, secondary amine tricyclic antidepressants may be better tolerated. When adjuvant analgesics are used, the number needed to treat to substantially reduce neuropathic pain in an individual patient ranges from three to five. Corticosteroids reduce symptoms and pain from tumor-related compressive neuropathy, brain metastases, and bowel obstruction. Bisphosphonates have been used for bone pain. For patients in whom pain does not respond to a first-line opioid, or in whom dose-limiting toxicity develops (e.g., confusion, myoclonus, hallucinations, nightmares, or severe nausea), switching to or rotating with an alternative opioid will improve pain control and reduce side effects. Because of noncross tolerance, doses that are 50% to 70% of the equianalgesic dose should be used and further dose adjustment should be made, based on clinical context, including potential drug interactions and organ function. Routine use of analgesic tables can be dangerous. As an alternative, switching to spinal opioids may improve the opioid therapeutic index by reducing side effects. Spinal adjuvant analgesics include bupivacaine and clonidine.
Side effects related to opioids can add symptom burden to patients or may be mistaken for progressive cancer. Constipation, if severe, not only produces nausea and vomiting, but can also resemble a bowel obstruction. Because there is no tolerance for constipation, stool softeners and laxatives should be started proactively when opioids are prescribed. Opioid-induced constipation that is unresponsive to laxatives and enemas should be treated with methylnaltrexone. Opioid-induced sedation may respond to a methylphenidate. The anticholinergic side effects of opioids, (dry mouth, urinary retention, and nausea) may require targeted approaches to management. Antiemetics, such as metoclopramide, prochlorperazine, or ondansetron, may be used to manage nausea. For most individuals, nausea is mild and tolerance develops over several days. Most opioids, except for buprenorphine, induce hypogonadotropic hypogonadism, which can result in altered mood, hot flashes, loss of libido and sexual function, and, over a period of time, loss of muscle and bone. When treatment with an opioid is started or doses are adjusted, there can be mild sedation and mild muddled thinking, which usually resolves over several days. Patients should be taking stable doses of opioids for approximately 2 weeks before they consider driving a car. Overt confusion and delirium induced by opioids require opioid rotation or change in route rather than antipsychotics. Nonpharmacologic modalities should be considered concurrently with analgesics. Mind-body approaches and integrative interventions can be helpful, both in improving pain and in reducing the opioid requirements. Correction of impending fractures of long bones; kyphoplasty; and brachial, celiac, or hypogastric blocks can improve pain and improve opioid responses.