Necrotizing soft-tissue infecctions

LETTERS TO THE EDITOR
Necrotizing soft-tissue infections
(SUPPLEMENT 4 TO VOLUME 74, AUGUST 2007)
This letter concerns an article in a Cleveland Clinic
Journal of Medicine supplement (Infections in
Hospitalized Patients: Urgent Challenges, Evolving
Management) distributed to only a portion of the Journal’s
regular readership, owing to the terms of the grant supporting
the supplement. The supplement is available free to all
online at www.ccjm.org/toc/hospitalinfections.htm.
TO THE EDITOR: I read with great interest your excellent
supplement, Infections in Hospitalized Patients:
Urgent Challenges, Evolving Management (August
2007). I was particularly interested in the section
on necrotizing fasciitis in the article by Merlino and
Malangoni on complicated skin and soft-tissue
infections.1 The authors state that amputation may
be necessary in up to one-third of patients with
necrotizing soft-tissue infections involving the
extremities, citing a retrospective patient series at a
university center reported by McHenry et al in
1995.2 However, McHenry et al reported an amputation
rate of nearly 50% if the infection involved
an extremity, in addition to an overall mortality
rate of 29% in the series.2
Additionally, Drs. Merlino and Malangoni make
no mention of the role of adjunctive hyperbaric oxygen
therapy for necrotizing soft-tissue infections.
This may represent a bias on their part, as the 1995
report by McHenry et al (of which Dr. Malangoni
was a coauthor) states: “No patient in our series was
treated with hyperbaric oxygen. To date, the efficacy
of hyperbaric oxygen in treatment of [necrotizing
soft-tissue infections] has not been established by
controlled studies, and delays in debridement have
been reported with its use.”2
I believe it is important to include this potentially
useful adjunctive therapy in discussions of necrotizing
soft-tissue infections, in light of several recent
papers that support the use of hyperbaric oxygen—in
terms of lower mortality, lower rates of morbidity
(amputation), earlier wound closure, and no delays
to debridement—in these difficult cases.3-6
PAUL CIANCI, MD
Professor of Internal Medicine (Emeritus),
University of California, Davis
Medical Director, Department of Hyperbaric Medicine
Saint Francis Memorial Hospital, San Francisco, CA
John Muir Medical Center, Walnut Creek, CA
Doctors Medical Center, San Pablo, CA
IN REPLY: We thank Dr. Cianci for his thoughtful comments.
Our review on the treatment of complicated
soft-tissue infections1 was not intended to be a dissertation
on the management of necrotizing fasciitis.
We agree that any comprehensive discussion of this
entity should include a review of the role and possible
benefits of hyperbaric oxygen.
The literature on the benefits of hyperbaric oxygen
for necrotizing soft-tissue infections is inconsistent.
The Undersea and Hyperbaric Medicine Society makes
the following statement on its use for these infections:
“The primary treatments for necrotizing soft tissue
infection are surgical excision of infected tissue and
administration of appropriate antibiotics. In selected
cases [emphasis added], addition of hyperbaric oxygen
therapy may be both lifesaving and cost effective.”7
Unfortunately, no prospective controlled trials
have examined hyperbaric oxygen therapy for this
indication. The concern about retrospective studies
in this population is that sicker patients are likely to
be excluded from treatment with hyperbaric oxygen
because they cannot be sustained or attended to in a
hyperbaric chamber. These exclusions would bias the
treatment effect to favor hyperbaric oxygen when in
fact there might be no benefit. In addition, the use of
hyperbaric oxygen is associated with complications,
including barotrauma, oxygen toxicity, and
decompression illness, that may be detrimental to the
many patients with necrotizing soft-tissue infections
who are critically ill.
The literature on the risk of extremity amputation
for these infections has been relatively consistent over
time, and this risk is significantly influenced by delays
in presentation and debridement.8 Primary therapy
remains fluid resuscitation, broad-spectrum antibiotics,
and rapid and aggressive surgical debridement. Any
delay in treatment—especially to transfer a patient
with an active infection to another facility solely to
institute hyperbaric oxygen therapy—is ill-advised and
may result in loss of life.
JAMES I. MERLINO, MD
MARK A. MALANGONI, MD
MetroHealth Medical Center
Case Western Reserve University
School of Medicine
Cleveland, OH
■ REFERENCES
1. Merlino JI, Malangoni MA. Complicated skin and soft-tissue
infections: diagnostic approach and empiric treatment
options. Cleve Clin J Med 2007; 74(suppl 4):S21–S28.
2. McHenry CR, Piotrowski JJ, Petrinic D, Malangoni MA.
768 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 74 • NUMBER 11 NOVEMBER 2007
Determinants of mortality for necrotizing soft-tissue infections.
Ann Surg 1995; 221:558–565.
3. Escobar SJ, Slade JB Jr, Hunt TK, Cianci P. Adjuvant hyperbaric
oxygen therapy (HBO2) for treatment of necrotizing fasciitis
reduces mortality and amputation rate. Undersea Hyperb
Med 2005; 32:437–443.
4. Hollabaugh RS Jr, Dmochowski RR, Hickerson WL, Cox CE. Fournier’s
gangrene: therapeutic impact of hyperbaric oxygen. Plast
Reconstr Surg 1998; 101:94–100.
5. Wilkinson D, Doolette D. Hyperbaric oxygen treatment and survival
from necrotizing soft tissue infection. Arch Surg 2004;
139:1339–1345.
6. Elliott DC, Kufera JA, Myers RAM. Necrotizing soft tissue infections.
Ann Surg 1996; 224:672–683.
7. Indications/HBO. Undersea and Hyperbaric Medicine Society
web site. http://www.uhms.org/Indications/LayIndic.htm.
Accessed October 2, 2007.
8. Anaya DA, Dellinger EP. Necrotizing soft-tissue infection: diagnosis
and management. Clin Infect Dis 2007; 44: 705–710.
CT screening for lung cancer
(JUNE 2007)
The CCJM publishes editorials, such as the one from
Dr. Yankelevitz in our June 2007 issue, to provide our
readers with an alternative view of a topic reviewed in
our journal. They do not undergo our usual peer
review process; we expect an “opinion piece,” and
count on the authors to present their perspective in a
forthright, accurate manner. Drs. Bach and Yankelevitz
obviously disagree about defining symptoms due to risk
factors for lung cancer and those due to early lung
cancer. A detailed discussion can be found in letters to
JAMA 2007; 298:513–516.
TO THE EDITOR: My colleagues and I recently published an
evaluation of computed tomographic (CT) screening,
in which we found that screening did not reduce lung
cancer mortality rates.1 In an editorial in the Cleveland
Clinic Journal of Medicine,2 Yankelevitz characterizes our
study differently, and in doing so makes several inaccurate
claims.
First, as we stated in our manuscript, all eligible
subjects who were screened in our three cohorts and
who could be followed through public death records
(99% of enrolled subjects) were included in our analyses
of lung cancer mortality rates. Yankelevitz thinks
we should have excluded from our study several of the
patients who died, because they could have harbored
symptoms of lung cancer at study entry. As our study
enrolled older patients with heavy smoking histories,
nonspecific pulmonary symptoms were likely common,
and so my colleagues and I disagree with Yankelevitz.
We believe we assessed the cohort most relevant to
our study question. Either way, it was inaccurate for
Yankelevitz to suggest that he was presenting new
information about our study in his editorial. He was
just stating his opinion.
Second, as we stated in our manuscript, we compared
the observed (“O”) and expected (“E”) number
of deaths from lung cancer using the equation (O –
E)2/E, which produces a chi-squared statistic with 1
degree of freedom. So, our primary result of 38.8
expected deaths vs 38 observed resulted in the equation
(38 – 38.8)2/38.8, a chi-squared statistic of 0.02,
and a two-sided P value of .90. Yankelevitz declares
that removing from our analysis five patients who
died would create the appearance of a statistically
significant reduction in the number of lung cancer
deaths. Yet removing five deaths changes the calculation
to (33 – 38.8)2/38.8, which has a chi-squared
statistic of 0.87 and a P value of .35. So here too
Yankelevitz was inaccurate when he suggested that
our findings were sensitive to small changes in the
number of observed events. One only achieves a P
value less than .05 after removing 12, not 5, of the
38 deceased patients from the analysis.
Lastly, Yankelevitz asserts that patients who have
any symptoms suggestive of lung cancer are actually
ineligible for enrollment in a screening program. Yet
in Yankelevitz’s own multicenter International Early
Lung Cancer Action Program (I-ELCAP) study of
CT screening, which he claims in his editorial
showed that screening is beneficial, there was no
exclusion of subjects for symptoms of lung cancer.
Instead, the protocol specified a common regimen of
screening but allowed each participating institution to
specify its criteria for enrollment.3 The Danish Lung
Cancer Screening trial, begun in 2004, has no exclusion
for patients based on symptoms (http://clinicaltrials.
gov), and neither does the Dutch-Belgian randomized
lung cancer screening trial (the “NELSON”
trial).4 The National Lung Screening Trial in the
United States (the Lung Screening Study) excludes
subjects who had recent profound weight loss (> 15
pounds) or hemoptysis, but not other symptoms that
could be due to lung cancer.5 So, here again, it was
inaccurate for Yankelevitz to assert as scientific dogma
a set of standards for eligibility that neither he, nor
his colleagues in I-ELCAP, nor the large consortia of
scientists evaluating screening in the United States
and Europe apply.
PETER B. BACH, MD
Memorial-Sloan Kettering
Cancer Center
New York
CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 74 • NUMBER 11 NOVEMBER 2007 769
■ REFERENCES
1. Bach PB, Jett JR, Pastorino U, Tockman MS, Swensen SJ, Begg CB.
Computed tomography screening and lung cancer outcomes.
JAMA 2007; 297:953–961.
2. Yankelevitz DF. Perhaps it is time for a change in policy on
lung cancer screening. Cleve Clin J Med 2007; 74:438–440.
3. Henschke CI, Yankelevitz DF, Libby DM, Pasmantier MW, Smith JP,
Miettinen OS. Survival of patients with stage I lung cancer
detected on CT screening. N Engl J Med 2006; 355:1763–1771.
4. van Iersel CA, de Koning HJ, Draisma G, et al. Risk-based selection
from the general population in a screening trial: selection
criteria, recruitment and power for the Dutch-Belgian randomised
lung cancer multi-slice CT screening trial (NELSON).
Int J Cancer 2007; 120:868–874.
5. Gohagan JK, Marcus PM, Fagerstrom RM, et al. Final results of the
Lung Screening Study, a randomized feasibility study of spiral
CT versus chest X-ray screening for lung cancer. Lung Cancer
2005; 47:9–15.
IN REPLY: There really should be no confusion about what
screening for a cancer is about. On the National
Cancer Institute Web site’s “Dictionary of Cancer
Terms,”1 screening is defined as “checking for disease
when there are no symptoms [emphasis added],” and the
National Cancer Institute expert advisory panel, the
Physician Data Query, gives the following definition:
“Screening is a means of detecting disease early in
asymptomatic people [emphasis added].”2 The American
Cancer Society defines screening as “the search for disease,
such as cancer, in people without symptoms
[emphasis added].”3 The recently published American
College of Chest Physicians guidelines on screening for
lung cancer, of which Bach is the lead author, states
that the objective was to “review the evidence for and
against screening for lung cancer with low-dose CT
and offer recommendations regarding its usefulness for
asymptomatic patients [emphasis added].”4
In my editorial, I referred to an article by Bach et
al5 that stated that the subjects of their study were
“individuals with a smoking history and no prior history
or symptoms suggestive of lung cancer.” In his letter
he changes the description of the cohort to one that
includes participants who had “nonspecific symptoms,”
rejects the notion that being asymptomatic is a requirement
for screening, and implies that the International
Early Lung Cancer Action Program (I-ELCAP) has no
such criteria for ensuring absence of symptoms.
That is incorrect. The I-ELCAP protocol clearly
states in its section on indications for screening that,
“symptomatic persons are ineligible for enrollment.”6
When Bach notes that each participating institution
specified its criteria for enrollment, this is without context.
The protocol makes it obvious that the indications
we are referring to are age and smoking history,
not the absence of symptoms of lung cancer.6 Regarding
the National Lung Screening Trial, the only large study
in the United States outside of the I-ELCAP, on its
Web site under “eligibility” it states, “no present symptoms
suggestive of current lung cancer, including: unexplained
weight loss of over 15 pounds within the past
12 months or unexplained hemoptysis [emphasis
added],”7 with the implication that other symptoms are
also considered.
The reason I noted that only 5 deaths needed to be
excluded from the analysis of Bach et al to result in a statistically
significant reduction in mortality by CT screening
for lung cancer in contrast to the 12 deaths said to
be required by Bach is as follows. In Bach et al,5 it is
reported that the mathematical model employed actually
predicts 48.3 deaths, not 38.8, and the data show 39
observed deaths, not 38. But Bach et al lower the number
of deaths the model predicts by arbitrarily excluding
those predicted in year 1 of CT screening, while in an
earlier publication that describes the validation of the
model,8 Bach includes deaths in year 1. After such a
major change it seems to me that Bach’s claim that a
validated model was employed is unjustified. Part of the
rationale given for the exclusion of year 1 was that even
though “CT screening would have appeared to reduce
lung cancer mortality by 20% [with inclusion of year 1
deaths] . . . this reduction would not have been statistically
significant.”5 However, the exclusion of 5 participants
with symptoms of lung cancer on entry and who
died of their lung cancer causes the number of observed
deaths to be only 34, and results in a significant difference
between the observed and expected deaths. Thus,
with the exclusion of 5 participants, not 12, Bach et al
would show a benefit associated with CT screening.
DAVID F. YANKELEVITZ, MD
New York-Presbyterian Hospital
New York
■ REFERENCES
1. National Cancer Institute. Dictionary of Cancer Terms.
www.cancer.gov/dictionary. Accessed 10/10/07.
2. Cancer Screening Overview. National Cancer Institute Physician
Data Query (PDQ) 1. www.cancer.gov/cancertopics/pdq/screening/
overview. Accessed 9/15/07.
3. American Cancer Society. Glossary Search.
www.cancer.org/docroot/GRY/GRY_0.asp. Accessed 10/03/07.
4. Bach P, Silvestri G, Hanger M, Jett J. Screening for lung cancer.
Chest 2007; 132:69S–77S.
5. Bach PB, Jett JR, Pastorino U, Tockman MS, Swensen SJ, Begg CB.
Computed tomography screening and lung cancer outcomes.
JAMA 2007; 297:953–961.
6. I-ELCAP enrollment and screening protocol.
http://ielcap.org/professionals/docs/ielcap.pdf. Accessed 10/03/07.
7. American College of Radiology Imaging Network. ACRIN #6654
Contemporary Screening for the Detection of Lung Cancer.
http://acrin.org/pdf_file2.html?file=protocol_docs/A6654partial_
summary.pdf. Accessed 8/7/07.
8. Bach PB, Elkin EB, Pastorino U, et al. Benchmarking lung cancer
mortality rates in current and former smokers. Chest 2004;
126:1742–1749.
LETTERS TO THE EDITOR
770 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 74 • NUMBER 11 NOVEMBER 2007