Affiliations: Portland, OR
Dr. Nichols is Assistant Professor of Medicine, Dr. Maziarz is Associate Professor of Medicine, Microbiology and Immunology, and Dr. Haupt is Professor, Department of Medicine at Oregon Health Sciences University.
Correspondence to: Dane Nichols, MD, Division of Pulmonary and Critical Care Medicine, Oregon Health Sciences University, UHS8Q, 3181 SW Sam Jackson Park Rd, Portland, OR 97221
Respiratory failure in the context of bone marrow
transplant (BMT) is distressingly common and is associated with
low long-term survival rates among patients requiring mechanical
ventilation. Although disparate processes may lead to invasive
ventilatory support, when multiorgan system dysfunction supervenes
survival may be impossible. The bleak experience in this population has
lead some authors to suggest early withdrawal of support in the
presence of key organ dysfunction or, in the extreme view, withholding
the option of mechanical ventilation altogether.1–2
In this issue of CHEST (see page 1012), Dr. Shorr and
colleagues provide important lessons for those of us who routinely
manage the critical care aspects of patients undergoing hematopoietic
stem cell transplant. First, not all transplanted patients with
respiratory failure can be expected to succumb. Second, in the select,
low-risk population of autologous stem cell transplant, respiratory
failure requiring mechanical ventilation is still encountered
relatively frequently (10.7%). Third, the survival rate in the subset
of patients with respiratory failure posttransplant falls within a
range that we see for many other diseases that are routinely managed in
ICUs, such as sepsis-related ARDS and anoxic
Recent case reports and observational studies support the conclusion
that the prognosis for many of these patients is significantly improved
over that reported in the 1980s and early 1990s. Survival rates for
intubated bone marrow and stem cell transplant patients may now be as
high as 16 to 19% overall.1,6–7 Within certain groups,
such as those with engraftment syndrome, survival may be considerably
higher. Lee et al8 reported 68% survival in patients
described as having ARDS following engraftment of autologous cells when
treated with high-dose steroids.
What factors are responsible for this trend toward improved survival?
New molecular diagnostics and more effective prophylaxis leading to
lower rates of viral and fungal pneumonias may have contributed to
better outcomes. Reduced toxicity of conditioning regimens culminating
in less severe expression of the idiopathic pneumonia syndrome may be
playing a role as well.7,9 While difficult to quantitate,
it is also quite likely that experience with this group of patients has
played a role in increasing our ability to recognize and effectively
manage potentially lethal complications.
It is, therefore, incumbent on critical care practitioners to
aggressively support these patients in the early phase of respiratory
failure, to diligently attempt to establish an etiology, and to avoid
contributing to the climate of pessimism that frequently surrounds
their ICU care. This can best be achieved by collaboratively managing
patients with the transplant team. It is important to recall
that many of these patients do not progress to severe respiratory
failure and often can be managed by way of noninvasive mask
ventilation. Treatable diagnoses such as capillary leak associated with
engraftment, congestive heart failure, and pulmonary hemorrhage also
represent opportunities for therapeutic intervention.
Despite the improved outlook, ≥ 80% of patients undergoing
hematopoietic stem cell transplant who develop respiratory failure will
not survive their ICU or hospital stay. In those who progress,
it is our obligation to limit treatment … in those patients who
are overwhelmed by their disease. The appropriate timing for
withholding or withdrawal of support is probably the most contentious
issue surrounding the care of these patients. Our ability to
prognosticate early in the course of an ICU stay is hampered by the
poor performance of standard scoring systems such as the
simplified acute physiology score II, Mortality Probability
Model II, and acute physiology and chronic health evaluation
II.5Although they are more accurate (and are difficult to
apply), other models specific to oncology patients have, at present,
unacceptably wide confidence intervals.6,10
Guidelines proposed by Rubenfeld and Crawford1 provide a
reasonable approach for the ICU practitioner who is dealing with this
high-risk population. A period of mechanical ventilation should
generally be offered to all patients who are requesting this level of
support. This is continued until progressive organ dysfunction and/or
ongoing vasopressor requirements lead to the conclusion that further
aggressive treatment would be futile. Certain combinations of organ
involvement in particular carry a very low probability of survival and
may allow for the early limitation of care. In the Seattle experience,
where the majority of patients were allogeneic marrow recipients, no
patient survived with the combination of lung injury, hepatic and renal
failure, and hypotension necessitating vasopressor use.1
With the availability of such data, it is instructive to review
what impact antecedent knowledge and awareness of these outcome
measures has on physician performance and patient preference. Paz and
colleagues11 developed an educational program in their
institution that stressed the dismal survival rates for intubated BMT
recipients. ICU admission rates were not significantly changed over
time despite general acceptance of the data. What proved even more
striking was the self-reported timing of discussions regarding ICU
care. Approximately one third of the attending oncologists did not
raise the issue of efficacy of intensive care in the pretransplant or
immediate posttransplant period.
Studies in other groups, such as the elderly, suggest that conveying
this kind of information is important in guiding patient
preferences.12However, the same may not hold true in the
BMT population. Reports dealing with informed consent in BMT patients
suggest that the survival imperative and trust in their physician carry
more weight in the decision-making process than risk and outcome
data.13–14 Clearly, during this time of emotional turmoil
and physical stress, asking patients and their families to provide
direction for their future care in a disinterested, objective manner is
So what should we consider for the future? (1) Patients must be
encouraged to generate advanced directives. (2) Informed consent should
include intubation outcome data based on current evidence. (3)
Proactive collaboration between the transplant team and ICU service
must be established early in the ICU stay. (4) Physicians need to stay
abreast of the changing spectrum of respiratory disease in this
population, and they must move quickly to identify processes for which
treatment can be expected to lead to reasonable survival rates.
Finally, when uncertainty exists over the direction that care should
take, decisions should be guided by probability estimates that are
undertaken in a climate of understanding and compassion.
Become a CHEST member and receive a FREE subscription as a benefit of membership.
Individuals can purchase this article on ScienceDirect.
Individuals can purchase a subscription to the journal.
Individuals can purchase a subscription to the journal or buy individual articles.
Learn more about membership or Purchase a Full Subscription.
Institutional access is now available through ScienceDirect and can be purchased at myelsevier.com.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 8
Customize your page view by dragging & repositioning the boxes below.
Enter your username and email address. We'll send you a reminder to the email address on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.