The principal role of the clinician for
patients with advanced pancreatic cancer
is to manage progressive symptoms and
relieve pain and suffering.
Pierre-Auguste Renoir. Roses, 1890.
Palliative Care in Pancreatic Cancer
Frank J. Brescia, MD, MA, FACP
Background: Pancreatic cancer is a formidable health problem, representing the 10th most common malignancy
in the United States and the 4th most common cause of all cancer deaths. The overall 5-year survival rate is 4%,
making this disease a model tumor in which to address the specialized care issues of palliative medicine.
Methods: General considerations in both medical decision-making and symptom management are reviewed.
Treatment of patients with locally unresectable, recurrent, or metastatic disease is individualized, based on
considerations that include patient age, patient wishes, family influence, insurance constraints, and geographic
Results: Success in managing progressive symptoms is needed to palliate patients with advanced pancreatic
cancer. Common problems include biliary obstruction, depression, pain, intestinal obstruction, and fatigue.
Conclusions: Relief of pain and suffering associated with critical illness is required in managing patients with
cancer. Pancreatic cancer is a model illness that mandates this need.
pathology and biology of the disease, as well as improved
diagnostic imaging and staging studies, the overall 5-year
Pancreatic cancer is a formidable health problem with
survival rate remains 4% for all stages and races.
Although this tumor represents
Adenocarcinoma of the pancreas comprises 90% to
only 2% of new cancer diagnoses in both men and women
95% of all malignant tumors of the exocrine pancreas. It is
and is the 10th most common malignancy in the United
one of the most lethal malignancies, and its geographic loca-
States, it is the fourth most common cause of all cancer
tion within the body makes imaging studies and biopsy pro-
deaths. Despite advances in the understanding of the
cedures more difficult compared with other tumors. There
are no clear-cut high-risk populations to follow, even if effec-
tive screening procedures were available. More problemat-
From the Hollings Cancer Center at the Medical University of South
Carolina, Charleston, South Carolina.
ic is the reality that presenting symptoms are vague, diverse,
Submitted July 8, 2003; accepted December 15, 2003.
and long-evolving before medical attention is sought.
Address reprint requests to Frank J. Brescia, MD, MA, FACP, 96 Jonathan
The clinical presentation is often dramatic, with “pain-
Lucas Street, Hollings Cancer Center, Medical University of South Car-
less” obstructive jaundice. There is often a history of mild
olina, CSB 903, Charleston, SC 29425. E-mail: firstname.lastname@example.org
but progressive discomfort or pain in the mid-abdomen,
No significant relationship exists between the author and the compa-
occasionally with radiation to the back, and usually noted
nies/organizations whose products or services may be referenced in
worse at the end of the day.
January/February 2004, Vol. 11, No. 1
Cancer Control 39
Ten percent of patients have a new onset of diabetes.
patients may be good candidates for a clinical trial. (3)
Others describe fatigue, anorexia, nonspecific gastroin-
Physicians feel the need to instill a level of hope. (4)
testinal symptoms, weight loss, and depressed mood — all
Financial incentives may affect behavior of some treat-
of which can go unnoticed until there is an obvious need
ment teams. (5) Some patients distrust all conventional
to seek medical care. Patients may require symptom relief
medical care, including hospice, and choose an alternative
before any treatment interventions can begin, and some
or complimentary medicine approach.
patients move rapidly to a state where the options are
Success in managing the complex, multiple, progres-
aimed solely at comfort. Multiple complaints, poor per-
sive signs and symptoms in pancreatic cancer is important
formance status, and comorbid illness make definitive
for all patients at all stages of disease, but this becomes the
surgery a less likely option. Interestingly, long-standing
overriding imperative for those facing advanced illness.
symptoms, weight loss, and anemia are not negative pre-
Even with treatment, most gastrointestinal cancers still fail
dictors of survival by univariate analysis, if the patient is
to show complete tumor regression, and despite obvious
able to undergo a resection.2 In a series of 13,560 patients
responses, meaningful survival benefit may be modest.
with pancreatic cancer, a major predictor of survival was
The oncology literature now recognizes the importance of
the ability to undergo a complete resection, whereby sur-
symptom control measures and supportive services that
vival rates were 2-fold greater compared with palliative
diminish the unacceptable side effects of beneficial anti-
bypass procedures.3 Yeo et al4 reported that residual dis-
cancer therapies such as antiemetics, erythropoietin,
ease, manifested by positive vs negative margins, translates
colony-stimulating growth factors, and antibiotics.7 How-
into a 5-year survival rate of 8% vs 26%. This disease is a
ever, at times, efforts to manage symptom progression are
model tumor to address the specialized care issues of sup-
addressed by escalating intensity or prolonging treatment
portive and palliative medicine.
regimens rather than by comforting patients whose
tumors are truly failing to respond.8
The clinical ability to effectively relieve a patient’s dis-
tress begins with an understanding of the natural history
of the tumor as well as the biological behavior or tempo
of progression. This knowledge allows the treatment team
Although unresectability represents the poorest prognos-
to anticipate and manage the changing picture of the clin-
tic factor, even patients who are able to have a Whipple
Consideration of the patient’s age,
procedure with clear margins often relapse, making this
comorbid problems, functional status, mental capacity,
one of the most biologically aggressive cancers. Manage-
preferences, and previous responses to adequate clinical
ment of the minority of patients intended for curative
trials all help to clarify the potential outcome and the
resection usually causes little controversy. However, those
aggressiveness of further therapy. The following consider-
with locally unresectable, recurrent, or metastatic disease
ations may be useful in the decision-making process for
generate additional issues that go beyond the pure med-
patients with malignancies:
ical picture: patient age, patient wishes, family influence,
insurance constraints, and geographic practice variations.
• Establish with reasonable certainty the extent of
There is an increasing mandate in oncology to
disease and whether the cancer is beyond curative
improve outcomes, to limit adverse events, to control
therapeutic intervention. Is the tumor resectable?
escalating costs, and to improve patient satisfaction.
These are meaningful goals that drive a treatment
• Evaluate the etiology and severity of symptoms and
approach for an illness that has a 5-year mortality rate of
relate them to the pace of progression and extent of
95%. Half of patients may be too ill for entrance into a
disease. Inexperienced physicians may be less critical
clinical trial, and the remaining half may not qualify
in their clinical assessment, sacrificing useful options
because of age, nonmeasurable disease, serious comorbid
that may produce meaningful tumor responses and
illness, or a second malignancy.5 In may cases, because of
reduce intensity of symptoms.
patient and/or family insistence, patients with poor per-
formance are given chemotherapy outside of a clinical
• Determine if the patient’s signs and symptoms are
study, sometimes with a minimally toxic regimen. Many
compatible with the natural history of pancreatic
patients are never considered for hospice programs or are
cancer. It is wise to rule out treatable noncancer caus-
referred late for end-of-life care. This occurs for a variety
es, such as pulmonary embolus, and to recognize and
of practical reasons: (1) Some patients and/or their fami-
manage problems related to the treatment itself.
ly members refuse to accept a noncurative, nonaggressive
Pancreatic cancer is generally a disease of older indi-
approach. One study showed that 53% of cancer patients
viduals — the mean age of onset is 65 years, with 80%
were willing to accept intensive chemotherapy and
of patients 60 years or older. Many of these patients
endure toxicity for a 1% cure rate. More than 40% would
have concurrent morbidities and use several medica-
do the same to extend their lives by 3 months.6 (2) Some
tions, which may confusion to the clinical picture.
40 Cancer Control
January/February 2004, Vol. 11, No. 1
• What potential limitations of function are imposed on
varied from 0% to 60%, with a median and mode of 5%.
the patients due to the disease, diagnostic interven-
Almost one fifth of physicians chose a threshold point of
tion, or treatment? Early pancreatic cancer is not easy
20% or higher, while 4% of those surveyed were willing to
to diagnose. Computed tomographic or magnetic res-
have a cut-off point (for example, no further active therapy
onance imaging scans are often not definitive, so fur-
at 50%).12 This lack of consensus shows a wide inter-indi-
ther diagnostic studies that include fine-needle aspi-
vidual difference among clinicians to define futile inter-
ration and endoscopic retrograde cannulation of the
ventions and outcome measures. However, the variations
pancreatic duct (ERCP) may be indicated.
in response may reflect only the differences among clini-
cians rather than differences in specific clinical scenarios.
• What information needs to be communicated to the
Patients usually wish to avoid an experimental or
patient, family, and other members of the treatment
risky procedure if the probable outcome of a standard
team? How much is clearly understood? Are there
treatment is in their perceived range of benefit. As expect-
elements of denial or unrealistic expectations driving
ed, patients will be risk-takers for a novel approach when
treatment decisions that may be inappropriate?
the known outcome of standard care is not in their favor.13
More problematic for patients as well as their physicians is
• What are the management goals for this patient?
the ability to access overwhelming information, some of
These will change over time. Is further chemothera-
which is based on media hype of “promising” therapies.14
py futile? The meaning of futility may need to be
It would be helpful if the practice of asking patients to
defined for specific objectives (eg, survival, function,
choose between the options of only palliative care or anti-
pain, satisfaction, and quality of life).
cancer therapy was discontinued in order to make both
patient’s goals and expectations become of para-
simultaneous supportive care and chemotherapy part of
“best care.” Pilot demonstration programs to address the
feasibility of such a model of care have been reported.15
• What other options and recommendations can be
offered if the patient refuses the clinician’s plan of
Predicting Survival and Prognosis
Most oncologists understand the distinction between
The ability to predict with accuracy the life expectancy of
therapy for curative results and therapy for end-of-life care.
an individual with cancer is difficult, even among experi-
The challenge is determining whether further palliative
enced oncologists.16-18 Oncologists have an obligation to be
treatment remains valid as disease progresses. There is no
honest while still cultivating their implied promise of deliv-
benefit to continue anticancer treatment that fails to pro-
ering hope to their patients. In the case of Arato v Avedon
long life, improve functional limitations, or reduce symp-
(Cal. 1993, 858 P.2d 598), a patient’s family sued the attend-
toms of illness. At this point, management measures should
ing oncologist for violating California’s Informed Consent
focus on pain and symptom distress. The medical oncolo-
Doctrine. They claimed the patient was not told that 95%
gist should know not only when interventions are indicat-
of people with his diagnosis and stage of pancreatic cancer
ed, but also when they are not.9 Some have suggested that
would die within 5 years. The family argued that if the
we need to redefine anticancer medications to include
patient knew the facts, he would have elected to die “at
those modalities that enhance patient well-being.10
peace” without experiment therapy. The physician coun-
Indeed, the use of gemcitabine in this disease is more valu-
tered that the patient was anxious and thus a detailed dis-
able for palliating symptoms than extending survival!
closure of such poor prognostic information was inappro-
For patients with advanced cancer, the manner and
priate to share with him. The Supreme Court of California
process of delivering medical care, as well as the possible
agreed with the physician and claimed “…we believe it
outcomes, are intimately woven and connected. On the
unwise to require as a matter of law that a particular species
other hand, physicians may experience feelings of suffer-
of information be disclosed.” In a survey of oncologists on
ing, uselessness, loneliness, disappointment, and failure
this matter, all believed that giving hope was a necessary
when treating dying patients.11 Clinicians have a height-
goal of their profession.19 Kodish and Post20 reported that
ened, uncomfortable awareness of their own human
patients expect clinicians to disclose information after
frailty, vulnerability, and mortality. Factors that exacerbate
some interpretation of the data, but they do not want “cal-
this physician distress include inadequate training, fear of
lous disclosure of grim diagnosis and prognosis.” Patients
dying, poor communication skills, conflicts in goals of
need to appreciate that the whole range of possibilities for
care, unrealistic expectations, and uncertainty of treat-
their particular cancer may not follow a set course and in
fact may be better or worse than the probabilities show.
Patients and physicians often disagree about qualita-
This obligation of physicians to nourish their patients’
tive and quantitative thresholds for futility.12
hope — to see some good to their future — remains a
responses to futility cut-off points (ie, a chance of success)
challenge. Lamont and Christakis21 have shown that physi-
January/February 2004, Vol. 11, No. 1
Cancer Control 41
cians usually make optimistic errors in foreseeing patient
rate, with the majority successfully placed during the first
life expectancy, and they make larger conscious optimistic
procedure. However, plastic endoprostheses require chang-
errors in the actual disclosure to patients as to their sur-
ing at 3 to 6 months because of occlusion and a return of
vival time. Christakis22 further reports that “it is important
jaundice, fever, and discomfort. It is unclear whether antibi-
to realize that physicians often believe that prognostica-
otics or bile salts help stents to remain patent.28 The place-
tion is itself intrinsically and fundamentally a lie…any
ment of a metal stent in a patient with repeated bouts of
statement about it — but especially one that might be con-
cholangitis is also controversial. The stent can become
strued as definitive — is seen as necessarily mendacious.”
blocked by tumor invasion without prolonging survival.29
Physicians regard this as a “considered use” of information
However, the relief offered by stent placement may
rather than a “deliberate distortion” of information and
enhance the patient’s overall quality of life, even in a popu-
thus see no conflict between presenting an optimistic pic-
lation of individuals whose survival is limited.
ture and their need to be truthful.
Symptom Control and Palliation
It has long been known that depression is more common in
Despite clinicians’ mandate to comfort their patients, the
patients with pancreatic cancer than in those with other
current practice of medicine to relieve suffering is “more
malignancies. A well-described study of 139 patients who
a hope than standard” for most patients.23 Forty percent
were admitted for possible colon or pancreatic cancer
will die with unrelieved pain,24 and 50% of hospitalized
reported that 76% of patients with pancreatic cancer had
terminally ill patients have pain that is ignored by their
depressive symptoms prior to surgery compared to 17% of
caregivers.25 Physicians and nurses tend to minimize their
patients with colon cancer.30 A small National Cancer Insti-
assessment of pain, especially when patients describe sig-
tute study also revealed major depressive symptoms in half
nificant pain. 26
of the patients with pancreatic cancer compared with none
Treatment planning for patients with pancreatic can-
in patients with gastric cancer.31 A large literature of retro-
cer usually includes measures that aim at simultaneous
spective reviews has tried to connect the issue of unre-
tumor response and symptom management. Because cura-
lieved pain and misdiagnosis to depressive symptoms in
tive outcomes or long-term remissions are not likely to
Foley33 noted that pain was a presenting
occur in a disease so biologically aggressive, the challenge
symptom in 80% of patients and occurred sometime in
to integrate hope, treat the cancer aggressively, and simul-
their illness in 90% of patients. Because of the widely held
taneously face reality becomes a frustrating and difficult
belief that pain and depression are common with pancreat-
experience for the clinician.
ic cancer, Kelsen et al34 prospectively evaluated these symp-
toms in 130 newly diagnosed patients: 83 patients prior to
a surgical procedure and 47 before their first chemothera-
py treatment. All patients had excellent performance status
and were being treated at a major tertiary cancer center.
Since 70% to 85% of patients have tumors involving the
The Beck Depression Inventory (BDI) and Beck Hopeless-
pancreatic head, the development of jaundice is a com-
ness Scale (BHS) were utilized, as well as other validated
mon initial presentation. Most patients have a previous
tools, to measure pain and symptom distress. Only 29% of
history of unsuspecting vague, nonspecific abdominal dis-
these patients complained of moderate to severe pain. The
comfort that predates the jaundice. Biliary obstruction
patients receiving chemotherapy reported more pain than
may occur later in the illness due to growth of an unre-
did preoperative patients. BDI scores were ≥15 in 38%, sug-
sected primary tumor, recurrent tumor, enlarged regional
gesting high levels of depressive symptoms. There was a
nodes, or biliary stent occlusion.
Ninety percent of
strong correlation between increasing pain and depressive
patients will have jaundice at some time in their illness,
symptoms among those with pain. However, the authors
with associated symptoms of malaise, pruritus, loss of
concluded that moderate or severe pain and symptoms of
appetite, fever, and abdominal discomfort. The optimal
depression were less prevalent than originally thought. This
strategy for treatment may not be initially obvious due to
study examined patients earlier in their disease where
the patient’s age, life expectancy, and generally poor well-
severity of depression and pain would be less prevalent.
being, or the physician’s experience and expertise. Biliary
The patients receiving chemotherapy who were more like-
bypass surgery has long been utilized for patients with
ly to be depressed may have seen themselves as less likely
unresectable disease or, in cases where other options are
to be cured. It is difficult to be sure since the number of
unavailable, for relieving disturbing symptoms and per-
cases in this treatment arm was small.
haps prolonging patient survival.27
Some data suggest that patients with a prior history of
Endoscopic placement of Teflon stents, introduced in
depression have a worse survival when cancer occurs
the 1980s, is now routinely performed with a 90% success
than would be expected on the basis of their cancer diag-
42 Cancer Control
January/February 2004, Vol. 11, No. 1
nosis alone.35 If prior depression affects life expectancy
with a “rescue” dose calculated at approximately 15% of
among patients with cancer, effective intervention should
the 24-hour baseline dose. Oral doses may need to be
be studied to substantiate survival outcomes. Brief psy-
given every hour for relief, and the severity of the pain will
chotherapy (ie, fewer than six sessions with a psychiatrist)
determine the dose, route, and frequency of the analgesic
and cognitive therapy appear to be beneficial for patients
intervention. Finding the correct opioid may be empirical,
in a palliative setting by addressing depressive symptoms,
ie, trial and error. Common reasons for inadequate pain
anxiety, and adjustment of patients to their illness (D.
control include making errors in dosing, failing to escalate
Schuyler, MD, personal communication, 2003). Interest-
total and breakthrough dose, not addressing side effects,
ingly, there is evidence that depressive symptoms may
and not using alternative opioids and adjuvant analgesics
abate in patients with pancreatic cancer when the malig-
(eg, antidepressant, anticonvulsants, corticosteroids).
nancy has been surgically excised.36
Patients with rapidly advancing illness present special
problems, and progressive cancer is a major reason for
increasing opioid dosage.
The final opioid dose required for relief is the dose
that works with an acceptable side effect profile. The dos-
In a prospective study of 1,107 patients admitted to a pal-
ing requirements necessary to deliver adequate pain relief
liative setting, approximately 44% of those with pancreat-
vary widely among patients. For example, at our center,
ic cancer had severe pain.37
Again, the prevalence of
nearly one third of patients with pancreatic cancer require
depressive disorders of all types were found to be higher
a 48-hour dosing schedule with transdermal fentanyl
in cancer patients with severe pain, raising an inference of
patches. The development of tolerance is not an impor-
causation. This link between pain and depression, along
tant issue when dose escalation is required.
with anxiety, underscores the problem of undertreatment
quickly become tolerant to the side effects of respiratory
for pain as the most common opioid abuse issue in the
depression, nausea, and sedation, but do not become tol-
care of the dying.38
erant of constipation.
Pain is the aspect of cancer that is most worrisome to
Percutaneous celiac plexus blockage can be a benefi-
both patients and their families. Half of respondents to
cial adjunctive interventional technique in individuals
public surveys about pain believed physicians cannot
whose pain is poorly controlled with opioids and who are
make a difference and this fear translated to 20% claiming
bothered by escalating adverse effects. In most cases, anal-
they would avoid seeking cancer treatment.39 The paradox
gesic responses are high (>50%) with a wide range of dura-
of cancer pain is the following: it is the most feared symp-
tion.41 No controlled trials have been conducted in which
tom, the most connected and interwoven to other cancer
conventional pain management is compared to neurolytic
symptoms (insomnia, fatigue, nausea, constipation), and yet
intervention. Complications are usually mild, but in rare
the most treatable of cancer complaints. Oral analgesics
cases the procedure can cause significant complications,
provide relief to 90% of patients with cancer.40
including pneumothorax, paraplegia, and ischemic gan-
Pain syndromes with pancreatic cancer can occur
grene of the bowel. In some centers, the procedure is now
due to the proximity of the organ to a number of other
recommended as the first approach to pain, followed by
critical structures: the duodenum, liver, stomach, jejunum,
the titration and escalation of opioid analgesia.
and transverse colon. The pancreas itself is innervated by
Intraspinal drug delivery (intrathecal or epidural) also
nerve networks that interact with both the parasympa-
can be effective in selected patients with intolerable
thetic and sympathetic systems. Pain may be felt at multi-
abdominal cancer pain.42 Smith and colleagues43 recently
ple and distant sites. Discomfort arising from the body of
reported the value of an implantable drug delivery system
the pancreas appears as midepigastric discomfort, while
in a randomized clinical trial, which also reported a sur-
pain coming from the tail is often localized in the left epi-
vival benefit in patients with refractory cancer pain.
gastrium and left intercostal space. Obstructive symptoms
Radiotherapy in pancreatic cancer is principally used
are cramps, poorly localized with a crescendo-decrescendo
as a palliative modality. There are no controlled random-
quality, while destruction of pancreatic tissue itself causes
ized trials to evaluate the impact on pain with other pain
further inflammation and discomfort. Pain can be referred
therapies, such as celiac plexus block or the use of opioids
to somatic structures without tumor infiltration of somatic
alone. However, in patients with advanced local disease,
nerves. The pain is progressive, and its character, quality,
radiation is often considered to manage pain.
and temporal nature worsen as the illness progresses. The
liver is a common site of metastasis, and pain can arise due
to nociceptive sensitive areas located within the liver cap-
sule and biliary tract. Pain can be referred to the right
shoulder or neck.
Despite the fact that less than 5% of patients present with
Most patients with chronic malignant pain will require
duodenal obstruction until late in the course of illness
an opioid regimen consisting of around-the-clock dosing,
(often a preterminal event), obstruction can manifest any-
January/February 2004, Vol. 11, No. 1
Cancer Control 43
where in the gastrointestinal tract. A thorough history
reported observing fatigue, and oncologists described 76%
and physical examination is essential to assess etiology
of their patients with this complaint.49 The pathogenesis
and define treatment options. Many patients have pro-
of cancer-related fatigue is unknown and may represent a
gressing tumor with associated poor performance, immo-
final common pathway for multiple possible mechanisms.
bility, and dehydration, and they are taking opioids for
Several possible contributing factors can be identified in
pain. The differential diagnosis must include concerns of
patients with pancreatic cancer: depression, pain, opioid
obvious progression of the cancer, decreased bowel
use, anemia, chemotherapy with or without radiation,
motility, ileus, opioid-induced nausea, and constipation.
insomnia, dehydration, and cachexia. The abnormal pro-
Nasogastric suctioning and fluid replacement often
duction and distribution of cytokines may be other possi-
relieves the situation, at least temporarily. Surgical inter-
ble mechanisms, as they have been implicated in the
vention is usually not considered because these patients
chronic fatigue syndrome.50 Treatment is often difficult
are poor surgical risks and often have widespread intra-
and focused on obvious factors that are correctable: pain,
abdominal disease with multiple points of obstruction.
anemia, insomnia, depression, and dehydration.
Medical management can be difficult, and treatment deci-
sions depend on the age of the patient, past treatments,
and closeness to death.
Pharmacologic agents can
include dexamethasone, haloperidol, and octreotide,
which have been helpful to some with intestinal obstruc-
The dying process — the final phase of any illness — can
tion, nausea, and increased intestinal secretions.44-46
be overwhelming for both patient and family. Difficulties
Aggressive nutritional support with total parenteral nutri-
with this process can be accentuated by clinicians’ confu-
tion is reserved for patients whose survival and quality of
sion, indifference, negligence, and frank abandonment.
life might be enhanced by active anticancer therapy. No
Observing the last hours of 200 consecutive terminal
data are available to support its use in patients with
patients in a quiet hospice setting revealed that nearly 40%
advanced disease. The decision to continue active sup-
had a specific problem that required attention, but the
port may be influenced by patient or family pressure, but
majority (91%) were thought to die “peacefully,” without
most often this is not the case, and care at the end of the
signs of bleeding, hemoptysis, dyspnea, pain, restlessness,
life can be maintained at home even without intravenous
or regurgitation.51 Less than 10% of these patients were
fluids for those obviously near death.
overtly confused, with nearly one third conscious until
chemotherapy offer little help, especially since in most
death and only 1% unresponsive for more than 48 hours
cases, disease has progressed despite utilizing these
prior to death.
modalities weeks or months before.
It is rare for symptoms to be so unmanageable near
Expandable metal stents may be helpful in desperate
death that total sedation is required to ease a patient’s suf-
nonsurgical candidates. Advanced pancreatic cancer of
fering. If sedation is necessary, several issues need to be
the pancreatic head can commonly obstruct the gastric
addressed: (1) have a collaborative medical consultation,
outlet. Approximately 90% of patients with gastroduode-
(2) communicate clearly with the family to understand the
nal stents improve clinically, and oral intake can resume
patient’s wishes, (3) ensure that death is imminent, (4)
quickly.47 Endoscopic placement allows the stent to reach
ascertain that symptoms are progressive, severe, and
the obstruction directly with relief to the proximal
refractory to treatment measures, (5) have clear goals and
Complications include perforation, bleeding,
a plan (eg, no cardiopulmonary resuscitation), (6) establish
stent migration malposition, and failure due to tumor over-
good communication with non-physicians, particularly the
growth and blockage.
nurses involved in care, and (7) maintain good documen-
tation of the above recommendations.
There is always a moral imperative for medical care
that provides comfort. Pancreatic cancer is a model illness
that mandates this imperative. Physicians need to develop
Fatigue is the most common symptom in patients with
a better appreciation of the nature of their patients’ dis-
cancer.48 Cancer-related fatigue is a subjective experience
tress beyond the physical nature of their symptoms. They
that has a detrimental influence on the patient’s quality of
must also recognize that the severity and chronicity of any
life, diminishing physical, emotional, work, and social rela-
specific complaint are interwoven with their patients’ per-
tionships. The pervasive extensions of profound malaise
ception of their life quality and meaning. The patient’s
can alter patients’ ability to retain information and to con-
preparedness to die depends on what remains for that per-
tinue with their treatment programs, thus lessening their
son to do and with the physical and mental capacity of
success of achieving an antitumor response. The magni-
that person to accomplish it. The compassionate and com-
tude of the problem of fatigue has been underestimated,
petent manner in which care is delivered at the end of life
and there are no validated animal models to study, as there
helps the sickest of our patients and their families to tran-
are with pain. Ninety per cent of patients’ loved ones
44 Cancer Control
January/February 2004, Vol. 11, No. 1
29. Lammer J, Hausegger KA, Fluckiger F, et al. Common bile duct
obstruction due to malignancy:
treatment with plastic versus
1. Jemal A, Murray T, Samuels A, et al. Cancer statistics, 2003. CA Can-
metal stents. Radiology. 1996;201:167-172.
cer J Clin. 2003;53:5-26.
30. Fras I, Litin EM, Pearson JS. Comparison of psychiatric symptoms
2. Compton CC, Mulvihill SJ. Prognostic factors in pancreatic carci-
in carcinoma of the pancreas with those in some other intraab-
noma. Surg Oncol Clin N Am. 1997;6:533-554.
dominal neoplasms. Am J Psychiatry. 1967;123:1553-1562.
3. Bramhall SR, Allum WH, Jones AG, et al. Treatment and survival in
31. Joffe RT, Rubinow DR, Denicoff KD, et al. Depression and carcino-
13,560 patients with pancreatic cancer, and incidence of the dis-
ma of the pancreas. Gen Hosp Psychiatry. 1986;8:241-245.
ease, in the West Midlands: an epidemiology study. Br J Surg.
32. Massie MJ, Holland JC. The cancer patient with pain: psychiatric
complications and their management.
Med Clin North Am.
4. Yeo CJ, Cameron JL, Lillemoe KD, et al. Pancreaticoduodenectomy
for cancer of the head of the pancreas: 201 patients. Ann Surg.
33. Foley KM. Pain assessment and cancer pain syndromes. In: Doyle
D, Hanks GWC, MacDonald N, eds. Oxford Textbook of Palliative
5. Winn RJ. Obstacles to the accrual of patients to clinical trials in the
Medicine. 2nd ed. Oxford, New York: Oxford University Press;
community setting. Semin Oncol. 1994(4 suppl 7);21:112-117.
6. Slevins ML. Making decisions about palliative chemotherapy. In:
34. Kelsen DP, Portenoy RK, Thaler HT, et al. Pain and depression in
ASCO Educational Book. 1992.
patients with newly diagnosed pancreas cancer. J Clin Oncol.
7. Braverman AS. Medical oncology in the 1990s. Lancet. 1991;337:
35. Angelino AF,Treisman GJ. Major depression and demoralization in
8. Klastersky J.
what for? Curr Opin Oncol.
cancer patients: diagnostic and treatment considerations. Support
Care Cancer. 2001;9:344-349.
9. Markman M.
treatment of malignancies poorly
36. Shakin EJ, Holland J. Depression and pancreatic cancer. J Pain
responsive to currently available antineoplastic therapy. Semin
Symptom Manage. 1988;3:194-198.
Oncol. 1994;21(4 suppl 7):1-2.
37. Brescia FJ, Portenoy RK, Ryan M, et al. Pain, opioid use, and survival
10. Bailes JS. Cost aspects of palliative cancer care.
in hospitalized patients with advanced cancer.
J Clin Oncol. 1992;
1995;22(2 suppl 3):64-66.
38. Solomon MZ, O’Donnell L, Jennings B, et al. Decisions near the end
11. Cherny N, Coyle N, Foley KM. Guidelines in the care of the dying
of life: professional views on life-sustaining treatments. Am J Pub-
cancer patient. Hematol Oncol Clin North Am. 1996;10:261-286.
lic Health. 1993;83:14-23.
12. McCrary SV, Swanson JW,Youngner SJ, et al. Physicians’ quantitative
39. Levin DN, Cleeland CS, Dar R. Public attitudes toward cancer pain.
assessments of medical futility. J Clin Ethics. 1994;5:100-105.
13. Sox HC Jr, Blatt MA, Higgins MC, et al, eds. Medical Decision-Mak-
40. Saunders C. Appropriate treatment, appropriate death. In: Saun-
ing. Boston, Mass: Butterworth-Heinemann; 1988.
ders DC, ed. The Management of Terminal Malignant Disease.
14. Daugherty CK. The “cure” for cancer: can the media report the
London, Baltimore, Md: Edward Arnold; 1984.
hope without the hype? J Clin Oncol. 2002;20:3761-3764.
41. Kawamata M, Ishitani K, Ishikawa K, et al. Comparison between
15. Meyers FJ, Linder J. Simultaneous care: disease treatment and pal-
celiac plexus block and morphine treatment on quality of life in
liative care throughout illness. J Clin Oncol. 2003;21:1412-1415.
patients with pancreatic cancer pain. Pain. 1996;64:597-602.
16. Pearlman RA. Inaccurate predictions of life expectancy. Dilemmas
42. Seamans DP, Wong GY, Wilson JL. Interventional pain therapy for
and opportunities. Arch Intern Med. 1988;148:2537-2538.
intractable abdominal cancer pain. J Clin Oncol. 2000;18:1598-
17. Addington-Hall JM, MacDonald LD, Anderson HR. Can the Spitzer
Quality of Life Index help to reduce prognostic uncertainty in ter-
43. Smith TJ, Staats PS, Deer T, et al. Randomized clinical trial of an
minal care? Br J Cancer. 1990;62:695-699.
implantable drug delivery system compared with comprehensive
18. Schonwetter RS,Teasdale TA, Storey P, et al. Estimation of survival
medical management for refractory cancer pain: impact on pain,
time in terminal cancer patients: an impedance to hospice admis-
drug-related toxicity, and survival. J Clin Oncol. 2002;20:4040-
sions? Hosp J. 1990;6:65-79.
19. DelVecchio Good MJ, Good BJ, Schaffer C, et al. American oncology
44. Davis MP, Nouneh C.
Modern management of cancer-related
and the discourse on hope. Cult Med Psychiatry. 1990;14:59-79.
intestinal obstruction. Curr Oncol Rep. 2000;2:343-350.
20. Kodish E, Post SG. Oncology and hope. J Clin Oncol. 1995;13: 1817.
45. Philip J, Lickiss N, Grant PT, et al. Corticosteroids in the manage-
21. Lamont EB, Christakis NA. Some elements of prognosis in terminal
ment of bowel obstruction on a gynecological oncology unit.
cancer. Oncology (Huntington). 1999;13:1165-1180.
Gynecol Oncol. 1999;74:68-73.
22. Christakis NA. Death Foretold: Prophecy and Prognosis in Med-
46. Mangili G, Franchi M, Mariani A et al. Octreotide in the manage-
ical Care. Chicago, Ill: University of Chicago Press; 1999.
ment of bowel obstruction in terminal ovarian cancer. Gynecol
23. Emanuel EJ, Emanuel LL. The promise of a good death. Lancet.
47. Baron TH. Expandable metal stents for the treatment of cancerous
24. Field MJ, Cassel CK, eds. Approaching Death: Improving Care at
obstruction of the gastrointestinal tract. N Engl J Med. 2001;
the End of Life. Committee on Care at the End of Life, Division of
Health Care Services, Institute of Medicine.
48. Miaskowski C, Portenoy RK. Update on the assessment and man-
National Academy Press; 1997.
agement of cancer-related fatigue. Principles and Practice of Sup-
25. A controlled trial to improve care for seriously ill hospitalized
portive Oncology Updates. 1998;1:1-10.
patients. The study to understand prognoses and preferences for
49. Simon AM, Zittoun R. Fatigue in cancer patients. Curr Opin Oncol.
outcomes and risks of treatments (SUPPORT). The SUPPORT Prin-
cipal Investigators. JAMA. 1995;274:1591-1598.
50. Dunlop RJ, Campbell CW. Cytokines and advanced cancer. J Pain
26. Grossman SA, Sheidler VR, Swedeen K, et al. Correlation of patient
Symptom Manage. 2000;20:214-232.
and caregiver ratings of cancer pain. J Pain Symptom Manage.
51. Lichter I, Hunt E. The last 48 hours of life. J Palliat Care. 1990;6:
27. el-Kamar FG, Grossbard ML, Kozuch PS.
cancer: emerging strategies in chemotherapy and palliative care.
28. Tarnasky PR, Cotton PB. Randomized trial of prevention of biliary
stent occlusion by ursodeoxycholic acid plus norfloxacin. Gas-
trointest Endosc. 1995;42:103-104.
January/February 2004, Vol. 11, No. 1
Cancer Control 45