Special Feature
Intensive Care for Cancer Patients
By Stephen W. Crawford, MD
For almost two decades, critical care physicians have debated the merits of intensive care for patients with cancer. It is clear that the mortality rates associated with critical illness in this patient population are very high. Some feel that the prognosis usually is so grim that cancer patients should not be admitted to the ICU. Others favor a limited period of life-support and withdrawal if improvement does not occur rapidly.
At the same time, the treatment of cancer continues to evolve. New therapies, such as hematopoietic stem cell transplantation and biological response modifiers, among others, offer the possibility of cure of malignancies and other fatal diseases. It seems that the evolving treatments of cancer permit us to hope that the prognosis for the next patient that we see will be different. Should we provide these patients with heroic levels of care in the hope that life-support (including mechanical ventilation) will allow them to survive treatment and return to normal function? What guidelines should we use when faced with cancer patients in our ICU?
ICU for Cancer Patients
The published data regarding outcomes of cancer patients in the ICU are daunting. In 1979, Snow and colleagues reviewed the experience of 180 cancer patients with respiratory failure at M. D. Anderson Hospital.1 Only 26% of these patients were extubated; only half of these left the hospital, and six-month survival was only 7%. Hauser et al reported that more than half of the patients with cancer in the ICU at the University of Miami died.2 The mortality rate was 75% among those with respiratory failure. (See Table 1.) Patients with lung cancer requiring mechanical ventilation faired even worse in a study of Ewer and associates,3 and only four of 46 patients survived. Similar experiences are reported with other types of cancers. Peters and colleagues from the Mayo Clinic noted in 1988 that hospital mortality was 82% among 116 patients with hematological malignancies who developed respiratory failure.4 It was distressing that the authors found no significant clinical differences on admission to the ICU between survivors and non-survivors. Thus, predicting who would survive was difficult.
Table 1
Common Reasons for ICU Admission in Cancer Patients
· Respiratory failure
· Shock
· Cardiac arrhythmia
· Oral stomatitis
· Altered mental status
· Seizure
· Post-operative care
· Bleeding
The most impassioned plea for a change in attitude toward cancer patients in the ICU came from Schuster and Marion in 1983.5 They found an 80% hospital mortality among 77 patients with hematological malignancies admitted to the ICU. Most of these patients were admitted with either shock or respiratory failure. Only four of 52 requiring mechanical ventilation survived, and none of these required prolonged support. Refractory hypotension, not hypoxemia, was the terminal event in most patients. Based on these findings, the investigators urged that life-support be provided on a limited basis to such patients. They advocated withdrawal of support if there is not a significant improvement in the clinical status within several days of admission to the ICU.
A financial incentive to limit the extent of critical care for cancer patients was provided in a 1993 study of Schapira and associates.6 These investigators at the H. Lee Moffitt Cancer Center in Florida examined the costs of critical care for cancer patients. The results were staggering. For the minority of patients who survived to hospital discharge, the cost of a year of life gained at home was astronomical: an estimated $95,000 (for solid tumor) to $450,000 (for hematological malignancy).
ICU after Transplantation
Treatment for many types of cancer now includes rescue from marrow ablative chemotherapy with marrow or stem cell transplantation (now called "hematopoietic stem cell transplantation," or HSCT). In fact, the proliferation of centers and oncologists using transplantation makes it increasingly likely that most critical care physicians will encounter the procedure. This proliferation is, in part, driven by the availability of pheresis units to assist with peripheral stem cell collection and transplantation. Have these technological advances in oncology altered the incidence or outcome of critical illness in cancer patients? In short, the answer is probably "no."
The incidence of critical illness after HSCT varies depending on the types of cancers treated and techniques used. The rates vary from 7% to 40% of patients, being lowest at centers using autologous stem cell sources predominantly. Most centers report ICU admission rates of 15-20%. The majority of ICU admissions after HSCT are for respiratory failure, followed in frequency by shock.7-9 There is hope that the use of peripheral stem cells as a transplant source will decrease the incidence of complications, as this source is associated with shortened period of neutropenia and fewer infections. However, initial reports do not suggest a significant decrease in critical complications associated with the peripheral stem cells.
In all reports from transplant centers, the mortality rates associated with respiratory failure after HSCT is high. In fact, the rates are consistently higher than those reported with conventional cancer treatments. Large studies from the Fred Hutchinson Cancer Research Center in Seattle and from the University of Minnesota reported that more than 90% of patients with respiratory failure die in the hospital.10-12 Six-month survival is 3% at both centers. It appears that similar rates affect children as adults after HSCT.13,14
The patients at risk for respiratory failure after HSCT have identifiable characteristics. The risk factors are older age, malignancy in relapse, and mismatched allogeneic donor source. These also are risks for hepatic and renal complications, as well as graft-vs.-host disease. Not surprisingly, respiratory failure is often accompanied (and preceded) by multiple organ dysfunction.
These risk factors have not been closely associated with outcome once respiratory failure develops. While older patients tend to do worse than younger ones, most patients who die will develop a combination of organ failures within four days of onset of respiratory failure. Respiratory failure accompanied by hypotension or hepatic and renal insufficiency was associated with no survivors out of an estimated 398 patients at the Fred Hutchinson Cancer Research Center.15 Thus, it can be determined with a very high degree of certainty among the majority of HSCT patients within a few days of life-support that survival is unprecedented.
The incidence of critical illness and the overall prognosis appear similar for all cancer patients, regardless of whether the treatment included HSCT. It is unclear whether the prognostic factors for mortality among those with respiratory failure apply to treatments other than HSCT. At present, I see no reason to believe that the same characteristics would be found in all cancer patients. Schuster and Marion found that most cancer patients died with refractory hypotension.5 This finding supports the hypothesis that multi-organ failure or hypotension predicts death among ventilated cancer patients. Complicated computer-modeling programs are currently being developed to assist caregivers in predicting survival specifically among patients with cancer.16 (See Table 2.)
Table 2
Prognostic Factors for Mortality
· Respiratory failure
· Prolonged neutropenia
· Allogeneic HSCT
· Recurrent or progressive cancer
· Cardiac arrest
Goals of ICU
We have recognized for the last 10 years that approximately 15-20% of HSCT recipients develop critical illness prompting admission to the ICU. The mortality rates for these patients have consistently been reported to be more than 80%. Among those patients with respiratory failure, only 5-6% survive to hospital discharge. Are these outcomes so dismal that none of these patients should be admitted to the ICU? The answer lies in part in the goal of the interventions.
The goals of care in the setting of malignancy should be common to those of medicine in general. The goals of medicine in Western culture were expressed early in medical literature. These goals as stated in the Hippo-cratic Corpus are: 1) attenuation of disease, 2) relief of suffering, and 3) restraint from treatment of hopelessly ill patients (lest physicians be thought of as charlatans).17
Throughout most of history, there was little that medicine could offer to alter the outcome of disease. The principal endeavors of medicine were to provide diag-noses, prognoses, and supportive care. However, modern advances in anesthesia, surgical technique, antisepsis, antibiotics, and vaccines have changed the capabil-ities of medicine. Consequently, modern society has altered its expectations. Because medical technology now permits us to alter disease processes, Western society has imposed additional goals on medical providers. Attenuation of disease and relief of suffering, while still important endeavors, are now joined (or supplanted) by the goals of maintaining or restoring function and prolonging life.
These notions that medicine can "save life" and "cure disease" are recent in human history and a reflection of scientific advances. Immunology and transplant biology represent novel extensions of our newly discovered capacities to influence the outcome of disease. Thus, while Hippocratic goals are still relevant, it is harder to apply the last of the Hippocratic injunctions, namely, refrain from treating the hopeless case. The question today is: what case is hopeless? For example, should the promise of cure with successful transplantation impart hope to otherwise desperate cases?
Allocation of ICU Resources
The issue of admitting cancer patients to the ICU is part of the large dilemma of how to allocate scarce ICU resources. The American Thoracic Society Bioethics Task Force recently addressed this topic.18 The Task Force believed that ICU admission should be allocated on a "first come, first served" basis among patients who would benefit from intensive care. This "benefit" was viewed in the context of the mission of the ICU. This mission was three-fold: preserve life, provide specialized rehabilitative care, and/or provide compassionate and attentive care to the dying. In this light, admission of cancer patients to the ICU until a diagnosis and prognosis can be established seems appropriate.
The Dilemma of Cancer Patients
The cancer patient in the ICU presents problems to the critical care staff beyond those framed by the question of whether to admit the patient to the unit. I believe that many of us face complex emotional dilemmas in caring for these patients. (See Table 3.)
First, oncology requires chronic care. Many intensivists are oriented to acute conditions and feel uncomfortable with long-term therapy. Second, appropriate care of the cancer patient requires a multidisciplinary approach. Coordination of critical care with oncology and often surgical care expertise is mandatory. The drugs, treatments, and complications often are unfamiliar not only to the medical staff, but also to the nursing staff. This unfamiliarity can be disconcerting to the critical care physician and nurse accustomed to "being in control."
The daily integration of care from various disciplines and the delineation of responsibilities can be tedious and trying. However, I believe that integration of both medical and nursing care by staff familiar with both critical care and oncology is crucial to the management of these patients. Third, many critical care procedures appear to be risky in the patients who are often thrombocytopenic. Fortunately, recent reports suggest that the complication rates for invasive procedures such as central line placement and bronchoscopy are not increased provided the operator is experienced.19,20
Table 3
The Challenges of Cancer Patients in the ICU
· Chronic problems
· Multiple medical issues require multiple disciplines
· Unfamiliar treatments and complications
· Coagulation defects common
· Immunosuppression
· Long-term prognosis often unclear
· Young age of patients
· Identification with patient's plight
The last dilemma faced by many of us in caring for the critically ill cancer patient is facing our own mortality. Many of these patients are young, often our own age. They may not be victims of self-abusive behavior. We often feel helpless to prevent the same fate from befalling us. Cancer can be the great equalizer. We may feel even more helpless to alter what seems to be an inevitable fortune for the patient. The danger lies in our saying that the patient does "belong" in the ICU, when the reality may be that we do not want to have to "deal" with the emotions.
Conclusion
Patients with cancer have a very high mortality rate when critically ill. However, in many cases, the prognosis can be definitively established within a short period of life-support. Recovery from the critical illness holds the real chance for long-term survival. The ICU does offer a potential benefit to these patients. Our task is to realistically assess the prognosis and provide or withdraw support because of the data, and not our emotions.
References
1. Snow RM, et al. JAMA 1979;241:2039-2042.
2. Hauser MJ, et al. Arch Intern Med 1982;142: 527-529.
3. Ewer MS, et al. JAMA 1986;256:3364-3366.
4. Peters SG, et al. Chest 1988;94:99-102.
5. Schuster DP, Marion JM. Am J Med 1983;75:402-408.
6. Schapira DV, et al. JAMA 1993;269:783-786.
7. Afessa B, et al. Mayo Clin Proc 1992;67:117-122.
8. Torrecilla C, et al. Intensive Care Med 1988;14: 393-398.
9. Paz HL, et al. Chest 1993;104:527-531.
10. Crawford SW, et al. Am Rev Respir Dis 1988;137: 682-687.
11. Crawford SW, Petersen FB. Am Rev Respir Dis 1992; 145:510-514.
12. Faber-Langendoen K, et al. Bone Marrow Transplantation 1993;12:501-507.
13. Nichols DG, et al. Crit Care Med 1994;22:1485-1491.
14. Bojko T, et al. Crit Care Med 1995;23:755-759.
15. Rubenfeld GD, Crawford SW. Ann Intern Med 1996; 125:625-633.
16. Groeger, JS et al. J Clin Oncol 1998;16:761-770.
17. Amundsen D. Hastings Center Rep 1978;8:23-24.
18. American Thoracic Society. Am J Respir Crit Care Med 1997;156:1282-1301.
19. Barrera R, et al. Cancer 1996;78:2025-2030.
20. Weiss SM, et al. Chest 1993;104:1025-1028.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content