You are on page 1of 13

The n e w e ng l a n d j o u r na l of m e dic i n e

Review Article

Dan L. Longo, M.D., Editor

Necrotizing Soft-Tissue Infections
Dennis L. Stevens, Ph.D., M.D., and Amy E. Bryant, Ph.D.​​

N
ecrotizing fasciitis is a surgical diagnosis characterized by From the Veterans Affairs Medical Cen­
friability of the superficial fascia, dishwater-gray exudate, and a notable ter, Boise, ID; and the University of Wash­
ington School of Medicine, Seattle. Ad­
absence of pus. This and other necrotizing soft-tissue infections have dress reprint requests to Dr. Bryant at the
multiple causes, risk factors, anatomical locations, and pathogenic mechanisms, Veterans Affairs Medical Center, Infec­
but all such infections result in widespread tissue destruction, which may extend tious Diseases Section, 500 W. Fort St.
(Mail Stop 151), Boise, ID 83702, or at
from the epidermis to the deep musculature. ­amy​.­bryant@​­va​.­gov.
Necrotizing infections can occur after major traumatic injuries, as well as after
N Engl J Med 2017;377:2253-65.
minor breaches of the skin or mucosa (e.g., tears, abrasions, lacerations, or insect DOI: 10.1056/NEJMra1600673
bites), varicella infection, nonpenetrating soft-tissue injuries (e.g., muscle strain or Copyright © 2017 Massachusetts Medical Society.
contusion), or routine obstetrical and gynecologic procedures; they can also occur
in postsurgical and immunocompromised patients (Table 1). Although necrotizing
infections have common clinical features, various entities have been defined, such
as progressive bacterial synergistic gangrene, synergistic necrotizing cellulitis,
streptococcal gangrene, gas gangrene (clostridial myonecrosis), and nonclostridial
anaerobic cellulitis. Subtle differences may distinguish one entity from another,
but the clinical approaches to diagnosis and treatment are similar.
In this review, we describe the clinical and laboratory features of necrotizing
fasciitis and other necrotizing soft-tissue infections. We also discuss diagnostic
pitfalls and recommended treatment approaches, as well as the effect of delays in
surgical intervention on mortality. (Details about pathogenic mechanisms are
provided in the Supplementary Appendix, available with the full text of this article
at NEJM.org.)

Epidemiol o gic Fe at ur e s a nd C ause s of Necro t i zing
Fa scii t is
Types I and II Infections
Necrotizing fasciitis types I and II are distinguished largely by whether the cause is
polymicrobial (type I) or monomicrobial (type II).1 The annual incidence of necrotiz-
ing fasciitis ranges from 15.5 cases per 100,000 population in Thailand2 to 0.3 to
5 cases per 100,000 in other regions.3-5 The relative incidence of type I and type II
infections varies considerably. In some studies, type II infections have accounted
for 55 to 87% of all cases of necrotizing fasciitis,6,7 whereas type I infections have
been more prevalent, by a factor of approximately 2, in other studies,8,9 and in
some studies, the incidence of the two types of infection has been similar.10-12
Necrotizing fasciitis type I is a polymicrobial infection involving aerobic and
anaerobic organisms. It is usually seen in the elderly or in those with underlying
illnesses. Predisposing factors include diabetic or decubitus ulcers, hemorrhoids,
rectal fissures, episiotomies, and colonic or urologic surgery or gynecologic pro-
cedures. Type I infection is often associated with gas in the tissue and thus is

n engl j med 377;23 nejm.org  December 7, 2017 2253
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e

Table 1. Factors Conferring a Predisposition to Specific Necrotizing Soft-Tissue Infections.*

Predisposing Factor Clinical Syndrome Etiologic Agent
Major penetrating trauma: crush or deeply penetrating Gas gangrene Clostridium perfringens, C. histolyticum,
wound or C. novyi
Minor penetrating trauma NF type II
Freshwater laceration Aeromonas hydrophila
Saltwater laceration Vibrio vulnificus
Minor nonpenetrating trauma: muscle strain, sprain, NF type II or streptococcal myonecrosis Streptococcus pyogenes
or contusion
Mucosal breach: mucosal tear (rectal, vaginal, urethral); NF type I Mixed aerobic and anaerobic organisms
gastrointestinal, genitourinary or gynecologic
surgery
Skin breach
Varicella lesions NF type II or streptococcal myonecrosis S. pyogenes
Insect bites NF type II or streptococcal myonecrosis S. pyogenes
Injection drugs Gas gangrene C. perfringens, C. histolyticum, C. novyi,
or C. sordellii
Immunocompromised state
Diabetes with peripheral vascular disease NF type I Mixed aerobic and anaerobic organisms
Cirrhosis and ingestion of raw oysters NF type II V. vulnificus
Neutropenia Gas gangrene C. septicum
In women: pregnancy, childbirth, abortion (spon­ NF type II, streptococcal myonecrosis, S. pyogenes, C. perfringens, or C. sordellii
taneous or medically induced), gynecologic or clostridial myonecrosis
­procedures or surgery
Occult factors: colonic lesions, including carcinoma Spontaneous gas gangrene C. septicum

* Gas gangrene is also known as clostridial myonecrosis.

difficult to distinguish from gas gangrene. Non- as progressive bacterial synergistic gangrene or
clostridial anaerobic cellulitis and synergistic large phagedenic ulcer15 can follow surgery in-
necrotizing cellulitis are type I variants. Both volving colostomy sites or wire sutures. Though
occur in patients with diabetes and typically in- large ulcerations often develop, the process does
volve the feet, with rapid extension into the leg. not involve the fascia.
Though nonnecrotizing cellulitis is common in Necrotizing fasciitis type II is a monomicro-
patients with diabetes, necrotizing fasciitis should bial infection (Table 1). Among gram-positive
be considered in patients with systemic manifes- organisms, group A streptococcus remains the
tations of sepsis, such as tachycardia, leukocyto- most common pathogen,2,4,7,11,16 followed by meth-
sis, acidosis, or marked hyperglycemia. icillin-resistant Staphylococcus aureus (MRSA).17,18
Bacterial penetration into the fascial compart- Unlike type I infections, type II infections may
ments of the head and neck may result in Lud- occur in any age group and in persons without
wig’s angina (i.e., infection of the submandib­ any underlying illness.
ular fascial spaces) or Lemierre’s syndrome Other pathogens include Aeromonas hydrophila
(thrombophlebitis of the jugular vein), with or and Vibrio vulnificus. Some experts have proposed
without severe sepsis.13,14 Breach of the gastro- that infections with these microbes, and possi-
intestinal or urethral mucosa may result in bly clostridial species, be classified as necrotiz-
Fournier’s gangrene, which begins abruptly with ing fasciitis type III. Monomicrobial necrotizing
severe pain and may spread rapidly from the fasciitis due to gram-negative pathogens (bacte-
perineal region to the anterior abdominal wall, roides and Escherichia coli) has also been report-
the gluteal muscles, and in males, the genitalia. ed,6,19-21 though these infections are typically
Finally, an indolent polymicrobial infection known seen in immunocompromised, diabetic, obese,

2254 n engl j med 377;23 nejm.org  December 7, 2017

The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
Necrotizing Soft-Tissue Infections

or postoperative patients or in those with pre- pain escalation sufficiently severe to require
existing, chronic organ dysfunction and are not ketorolac or narcotics) may be present, and such
typically classified as necrotizing fasciitis type II. pain prompts patients to seek urgent medical
Resolution of these nomenclature issues requires care. Malaise, myalgias, diarrhea, and anorexia
a consensus among international infectious dis- may also be present in the first 24 hours. Since
ease physicians, surgeons, and intensive care cutaneous manifestations are absent initially,
practitioners. the infection is often misdiagnosed or the cor-
The classic clinical and histologic features of rect diagnosis is delayed,30 and as a result, the
necrotizing group A streptococcal and clostridial mortality exceeds 70%.28 By the time ecchymoses
infections are mediated by potent bacterial exo- and bullae develop, tissue destruction is exten-
toxins and by the host response. For details, see sive, and systemic toxicity and organ failure are
the Supplementary Appendix. evident. Emergency surgery, including extensive
débridement or multiple amputations, is often
Invasive Group A Streptococcal Soft-Tissue required to ensure survival and necessitates pro-
Infections longed hospitalization.26,30-32 Erroneous diagnoses
A 2005 report estimated that more than 18 mil- include severe muscle strain and deep-vein throm-
lion people worldwide have invasive Streptococcus bophlebitis; because of the associated gastro­
pyogenes diseases,22 including postpartum infec- intestinal manifestations, food poisoning may
tions.23 In developed countries, the annual inci- also be diagnosed in error. Although seeding of
dence of invasive infections has remained steady the deep tissues probably occurs through tran-
at 3 to 5 cases per 100,000 population,3,24,25 with sient bacteremia from the nasopharynx, reports
an average mortality of 29%.24,26 Mortality is rarely document coexisting or antecedent symp-
higher among patients in whom streptococcal tomatic pharyngitis (unpublished data). This
toxic shock syndrome or septic shock develops might be expected, given the incidence of inva-
(38% and 45%, respectively).24,25 sive disease (18 million cases) as compared with
Two distinct clinical presentations have been pharyngitis (>600 million cases).22 In children,
described: infection with a defined portal of invasive streptococcal infections have been asso-
bacterial entry and infection that arises sponta- ciated with varicella zoster33-35 and influenza vi-
neously in the deep tissue, without an overt rus infections,36 as well as streptococcal pharyn-
wound or lesion (Fig. 1). S. pyogenes gains entry gitis,37 though relatively few cases of necrotizing
to the deep tissues through superficial cutane- fasciitis have been reported.
ous lesions (chickenpox vesicles, insect bites, or The toxicity of group A streptococcal necrotiz-
lacerations), after breaches of skin or mucosal ing fasciitis is severe and more fulminant than
integrity (due to drug injections, surgical inci- that described by Meleney in 1924.38 Ecchymoses
sions, or childbirth), or after penetrating trauma. and bullae develop more rapidly (in 2 to 3 days)
The initial lesion may appear to be only mildly and deep-muscle involvement is more common in
erythematous, but over a period of 24 to 72 hours, contemporary cases. The mortality is also higher.
inflammation becomes extensive, the skin turns Using only “bear claw fasciotomy” and irrigation
dusky and then purplish, and bullae appear. with Dakin’s solution (hypochlorous acid) as
Bacteremia is frequently present, and metastatic treatment, Meleney reported a mortality of 20%,
infections may occur. Very rapidly, the skin be- as compared with a mortality of 30 to 80% in
comes frankly gangrenous and undergoes exten- the current era.27 Given the involvement of epider-
sive sloughing. The patient is now perilously ill, mis, dermis, subcutaneous tissue, fascia, and mus-
with a high temperature and extreme prostration. cle, “necrotizing soft-tissue infection” appears
At this stage, mortality is high, even with appro- to be a more accurate term than “necrotizing
priate treatment.24,27 fasciitis” to describe the contemporary disease.
In approximately 50% of patients with group
A streptococcal necrotizing fasciitis or myone- Nonsteroidal Antiinflammatory Drugs
crosis, infection initiates deep in the soft tissues, and Group A Streptococcal Infection
without a portal of entry, often at sites of non- In the 1980s, an association between the use of
penetrating trauma (muscle strain or bruise).26,28,29 nonsteroidal antiinflammatory drugs (NSAIDs)
Initially, only fever and crescendo pain (rapid and the development of group A streptococcal

n engl j med 377;23 nejm.org  December 7, 2017 2255
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e

A Defined Portal of Entry

Erythema and Ecchymoses Bullae
Skin or Bullae
mucosal breach swelling
Dermal
papilla Large
bulla
Local tissue
Capillaries EPIDERMIS
damage
Exotoxins

Damage to
blood supply
Organisms Vascular
or spores occlusion

DERMIS Platelet–leukocyte Necrosis in multiple
aggregation tissue layers
SUBCUTANEOUS
FAT

Blood
vessels
Damage to
MUSCLE Deep fascia deep tissue

BONE

Exotoxins cause local tissue Deeper tissues become infected.
Organisms or spores are Erythema and swelling
damage. Platelet–leukocyte Larger venules and arterioles are
introduced into soft tissue. become widespread. Bullae
aggregates occlude capillaries occluded. Necrosis affects all
Exotoxins are released. and ecchymoses develop.
and damage vascular endothelium. tissue layers.

B No Defined Portal of Entry
Ecchymoses

Bulla

EPIDERMIS

DERMIS Vimentin Gradual upward
progression of
Myogenic infection
Activated
progenitor
myogenic
progenitors Group A Platelet–leukocyte
streptococci aggregation

Leukocytes
Arterial
Deep fascia Venous occlusion
occlusion

Streptococci
Exotoxins
Deep-tissue
injury

A nonpenetrating deep-tissue injury In susceptible hosts with Arteries become occluded, causing
Exotoxins are released. Venous
stimulates a repair response. There transient bacteremia, organisms necrosis in deep tissue that spreads
occlusion leads to necrosis in
is an influx of leukocytes and activation are trafficked to injury site in a to upper tissue layers. Bullae and
deep tissue.
of myogenic progenitor cells. vimentin-mediated process. ecchymoses later develop.

2256 n engl j med 377;23 nejm.org December 7, 2017

The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
Necrotizing Soft-Tissue Infections

Figure 1 (facing page). Evolution of Necrotizing Fasciitis
increased trafficking of circulating group A
or Myonecrosis. streptococci to strain-injured muscles in mice.49
Panel A shows the evolution of infections with a defined
portal of entry. A breach in the skin or mucosal integrity Necrotizing Clostridial Infections
facilitates the introduction of organisms or spores into Gas gangrene (clostridial myonecrosis) is an
soft tissue, resulting in monomicrobial infection (with acute invasion of healthy living tissue that oc-
group A streptococcus, Aeromonas hydrophila, or Vibrio
vulnificus) or polymicrobial infection (with both aerobes
curs spontaneously or as a result of traumatic
and anaerobes). Penetrating trauma sufficient to inter­ injury. Recurrent gas gangrene, occurring sev-
rupt the blood supply favors the development of gas eral decades after the primary infection, has
gangrene (clostridial myonecrosis). Bacteria proliferate also been described.50
and release exotoxins, which cause local tissue damage Deeply penetrating injuries that compromise
and impair inflammatory responses. Toxin-induced
platelet–leukocyte aggregates occlude capillaries and
the blood supply create an anaerobic environ-
damage the vascular endothelium, resulting in fluid ment that is ideal for spore germination and
leakage, tissue swelling, and erythema. Erythema and bacterial proliferation.51 Such trauma accounts
swelling become more widespread, bullae and ecchy­ for approximately 70% of cases of gas gangrene.
moses develop, and deeper tissues become infected. Other predisposing conditions are bowel and
Exotoxin production leads to occlusion of larger venules
and arterioles, with subsequent ischemic necrosis of
biliary tract surgery, intramuscular epinephrine
all tissue layers, from the dermis to the deep muscula­ injection, retained placenta, prolonged rupture
ture. Panel B shows the evolution of group A strepto­ of the membranes, and intrauterine fetal death.
coccal infection with no defined portal of entry (cryp­ Clostridium perfringens causes approximately 80%
togenic infection). A deep, nonpenetrating tissue injury of such infections; other pathogens include
such as muscle strain, sprain, or hematoma stimulates
a repair response, including an influx of leukocytes and
C. septicum, C. novyi, and C. histolyticum.
activation and proliferation of myogenic progenitor cells. Data regarding contamination versus active
In a susceptible host with a transient bacteremia (pos­ infection of traumatic wounds come from stud-
sibly arising from asymptomatic pharyngeal carriage of ies performed during World Wars I and II. In
Streptococcus pyogenes), the injury leads to trafficking 1915, Fleming documented that 60.4% of war
of the organisms to the injured site. Increased vimen­
tin expression on activated myogenic progenitor cells
wounds were contaminated with clostridia.52 Yet
and infiltrated macrophages serves as a ligand for MacLennan53 found that active infection (gas
group A streptococcus binding. Bacterial proliferation gangrene or “anaerobic cellulitis”) occurred in
with local production of exotoxins (e.g., streptolysin O fewer than 10 patients per 1000 wounded.54 In
and streptococcal pyrogenic exotoxin A) ensues. Ab­ 1941, Qvist55 suggested that anaerobic cellulitis
sorption of streptolysin O may stimulate intravascular
accumulation of platelet–leukocyte aggregates, first in
required only débridement of tissue that was
postcapillary venules and then in arterioles and larger damaged by trauma itself, whereas in gas gan-
vessels, resulting in vascular occlusion. Ischemic de­ grene, amputation was necessary to control
struction of the deep soft tissues ensues. Cutaneous rapid invasion of healthy tissue and thus ensure
manifestations of necrotizing infection (ecchymoses survival — a premise that guides clinical prac-
and bullae) develop later in the course of infection.
tice today.
Spontaneous (nontraumatic) gas gangrene is
necrotizing fasciitis was proposed. Proponents commonly caused by C. septicum, which is more
recognized that NSAIDs can suppress critical aerotolerant than other clostridial pathogens.56
neutrophil functions39 and augment the produc- Most infections occur in patients with gastroin-
tion of tumor necrosis factor α, a key mediator testinal portals of entry such as adenocarcino-
of septic shock.39 Other people argued that ma56 or in those with congenital or cyclic neu-
NSAIDs merely mask the signs and symptoms tropenia.57 C. sordellii infections can affect women
of developing infection, delaying diagnosis and after natural childbirth, as well as after abortion
treatment. Numerous clinical and epidemiologic or other gynecologic procedures. Such infections
studies have investigated, but not resolved, this can also develop in men, women, and children
issue.40-46 Experimental evidence is limited, after traumatic injuries and surgical procedures
though two studies clearly showed that non­ or illicit-drug injection.58 Common sites include
selective NSAIDs (e.g., ketorolac and ibuprofen) the skin, muscle, uterus, and perineum. Systemic
accelerated the disease course and worsened signs include an absence of fever, profound hypo-
outcomes.47,48 In addition, ketorolac significantly tension, diffuse capillary leak, hemoconcentration

n engl j med 377;23 nejm.org  December 7, 2017 2257
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e

Table 2. Pitfalls in the Diagnosis of Necrotizing Soft-Tissue Infection.*

Pitfall Explanation
Absence of fever Fever is often absent in patients with necrotizing soft-tissue infections because of NSAIDs that
are self-administered or prescribed in the emergency department or in postsurgical set­
tings. Fever is also absent in patients with necrotizing infection due to C. sordellii.
Absence of cutaneous manifestations Patients with spontaneous or cryptogenic necrotizing infections (i.e., infections without an
­obvious bacterial portal of entry) that begin in the deep soft tissues often do not have
­cutaneous signs of infection until late in the course of the disease.
Attributing severe pain to injury Severe pain is a key finding in patients with necrotizing infections. However, when such infec­
or procedure tions develop after surgery or parturition, pain may be erroneously attributed to the proce­
dure itself. Similarly, perineal pain may be attributed to hemorrhoids, epididymitis, or vaginal
or rectal trauma. Severe pain associated with spontaneous or cryptogenic infections is often
wrongly attributed to muscle strain or venous thrombosis. If pain is out of proportion to
the suspected cause or requires opioids or ketorolac for management, a developing necro­
tizing infection should be considered. Pain may be absent because of the use of narcotics
or NSAIDs or because of neuropathy in patients with diabetes.
Nonspecific imaging tests In patients with necrotizing infections, radiographs may show only edema, with no evidence of
gas in the deep tissue. Since this finding is consistent with noninfectious causes (e.g., soft-
tissue injury and postsurgical and postpartum conditions), it may confound the diagnosis.
Attributing systemic manifestations Nausea, vomiting, and diarrhea may be early manifestations of toxemia from group A strepto­
to other causes coccal infection, though they are often wrongly attributed to food poisoning or viral illness.

* NSAIDs denotes nonsteroidal antiinflammatory drugs.

(hematocrit, 50 to 80%), and a marked leuke- coccal infection (i.e., infection with no portal of
moid reaction (white-cell count, 50,000 to entry), the process begins deep in the tissues.
150,000 per cubic millimeter). Mortality is 70 to Crescendo pain is the most important clinical
100%, and death occurs within 2 to 4 days after clue for such infection, and its onset typically
hospital admission. occurs well before shock or organ dysfunction is
manifested. However, crescendo pain may be
absent or attenuated in patients who are receiv-
Di agnosis of Necro t i zing
Fa scii t is a nd O ther Necro t i zing ing analgesic agents, including NSAIDS; in pa-
Infec t ions tients who have undergone surgery, childbirth,
or trauma, the pain may be incorrectly attributed
Diagnostic Pitfalls
to normal postoperative soreness, typical post-
Early diagnosis of necrotizing infections may be partum discomfort, or the trauma itself, respec-
confounded by numerous factors (Table 2), as tively, rather than to acute infection. Pain may
discussed below. Physicians must be aware of also be absent in patients with altered mental
these potential pitfalls because delays in diagno- status or those with diabetes-related neuropathy.
sis and treatment have dire consequences. A basic In these instances, the absence of a strong clini-
diagnostic algorithm is shown in Figure 2. cal clue delays the correct diagnosis and appro-
priate treatments. Thus, all patients presenting
Clinical Findings with a sudden onset of severe pain in an extrem-
Classic manifestations of necrotizing fasciitis in- ity, with or without an obvious portal of bacte-
clude soft-tissue edema (in 75% of cases), ery- rial entry or the presence of fever, should be
thema (72%), severe pain (72%), tenderness evaluated for severe soft-tissue infection on an
(68%), fever (60%), and skin bullae or necrosis emergency basis.
(38%).8 In a recent case–control study, factors
that differentiated necrotizing fasciitis from cel- Imaging Tests
lulitis were recent surgery, pain out of propor- Radiographs, computed tomographic (CT) scans,
tion to clinical signs, hypotension, skin necro- or magnetic resonance imaging (MRI) studies
sis, and hemorrhagic bullae.59 will show soft-tissue swelling in patients with
In patients with cryptogenic group A strepto- group A streptococcal infection and will show

2258 n engl j med 377;23 nejm.org  December 7, 2017

The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
Necrotizing Soft-Tissue Infections

Clinical Symptoms
Swelling, erythema, pain

Systemic Signs
Tachycardia, >120 beats/min; hypotension;
No systemic signs
elevated CK level; CRP, >15 mg/dl;
LRINEC score, >6

Radiographic Evidence

Gas in the tissue No gas in the tissue

Immediate Surgery
Inspection and débridement Inspection and débridement
Specimen collection Specimen collection

Gram’s Staining
and Culture
Gram-positive Mixed aerobes Gram-positive Gram-negative
rods and anaerobes cocci rods

Differential Diagnosis
Clostridial myonecrosis NF Type I NF Type II NF Type II Cellulitis
(aggressive) Nonclostridial crepitant Streptococcus pyogenes Aeromonas Erysipelas
Spontaneous — cellulitis Staphylococcus aureus Freshwater Cutaneous abscess
Clostridium septicum Synergistic necrotizing or group A strepto- Vibrio
Traumatic — cellulitis coccal myonecrosis Saltwater
C. perfringens Progressive bacterial
Gynecologic — synergistic gangrene
C. sordellii
Clostridial anaerobic
cellulitis (indolent)

Figure 2. Algorithm for the Diagnosis of Necrotizing Infections.
In the algorithm, early clinical signs and symptoms and available results of laboratory tests and imaging studies are used to establish
the diagnosis and cause of a diverse array of skin and soft-tissue infections. CK denotes creatine kinase, CRP C-reactive protein, LRINEC
Laboratory Risk Indicator for Necrotizing Fasciitis, and NF necrotizing fasciitis.

gas in the tissues of patients with gas gangrene A study of enhanced CT in patients with docu-
or necrotizing fasciitis type I. Imaging evidence mented necrotizing fasciitis as compared with
of gas in the tissues, or the presence of crepitus, those who had other musculoskeletal infections
should prompt immediate surgical consultation. suggested that the absence of fascial enhance-
A finding of swelling alone may not be useful in ment was specific for necrotizing fasciitis.60
patients who have had a traumatic injury or have
undergone surgery or childbirth, since swelling Tissue Biopsy, Histologic Tests, and Gram’s
cannot be used to distinguish between infection, Staining
trauma, and inflammation. MRI may show thick- Gram’s staining of surgically obtained material
ening and hyperintensity of intermuscular fascia is crucial for determining the cause of infection
on T2-weighted images, findings that are sensitive and guiding empirical treatment. Percutaneous
but not entirely specific for necrotizing fasciitis. biopsy and examination of a frozen section has

n engl j med 377;23 nejm.org  December 7, 2017 2259
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e

been proposed to aid in the diagnosis of necro- T r e atmen t
tizing infection.61,62 However, this technique is
subject to sampling error and is not a good Surgical Intervention
substitute for open surgical inspection and biop- For patients with aggressive soft-tissue infection
sy. Group A streptococcal necrotizing infection or those with mild infection plus evidence of sys-
is characterized histologically by the destruction temic toxicity, prompt surgical exploration is ex-
of muscle tissue, a paucity of infiltrating phago- tremely important8,32,70 for three reasons: to deter-
cytes, and large numbers of gram-positive cocci mine the extent of infection, to assess the need
at the site (Fig. 3). The histologic findings are for débridement or amputation, and to obtain
similar for gas gangrene, though with more evi- specimens for Gram’s staining and culture. When
dence of edema, gas formation, or both. infection is near the vital structures of the neck,
surgical intervention may be necessary to pre-
Surrogate Markers for Early Diagnosis vent airway obstruction. Reinspection of the sur-
of Necrotizing Fasciitis gical site within 24 hours after surgery is recom-
A C-reactive protein level of more than 200 mg mended.8,70 Inspection and débridement should
per liter,63 a modestly increased white-cell count be continued every 1 to 2 days until necrotic
with a marked left shift,64 and an elevated serum tissue is no longer present.8,32,62,70-72 Negative-
creatinine level in the absence of hypotension pressure devices have shown promise in facili-
are suggestive of severe group A streptococcal tating closure and healing of these complex
infection. Marked leukemoid reactions (50 to wounds in small series of patients.73-75
150,000 white cells per cubic millimeter) and There is universal agreement that early surgi-
profound hemoconcentration are characteristic cal débridement is crucial in managing these
of C. sordellii infection. A white-cell count of complex cases. But how early is early? Pinpoint-
more than 15,400 per cubic millimeter plus a ing the critical time for surgical intervention on
serum sodium level of less than 135 mmol per the basis of published data is problematic, since
liter distinguishes necrotizing fasciitis in gen- the starting point for measuring the time to
eral from nonnecrotizing soft-tissue infections, surgery varies among studies, particularly retro-
with a negative predictive value of 99% but a spective analyses, with some studies using the
positive predictive value of only 26%.64 Elevated time from establishment of a definitive diagno-
levels of serum creatine phosphokinase or serum
aspartate aminotransferase suggest deep infec-
tion involving muscle or fascia (as opposed to
cellulitis).
The Laboratory Risk Indicator for Necrotizing
Fasciitis (LRINEC) scoring system uses the total
white-cell count and hemoglobin, sodium, glu-
cose, creatinine, and C-reactive protein levels to
distinguish between mild soft-tissue infections
and necrotizing fasciitis.65,66 For adults with
LRINEC scores of 5.8 or higher (on a scale of
0 to 13, with higher scores indicating a greater
likelihood of necrotizing soft-tissue infection),
the positive predictive value for necrotizing fas-
ciitis ranged from 57 to 92% in three stud- Figure 3. Histopathological Features of Group A Strepto-
ies,4,65,67 with negative predictive values of 86% coccal Necrotizing Fasciitis and Myonecrosis.
and 96% in two studies.65,68 The disparities may Routine hematoxylin and eosin staining of a muscle
specimen from a patient who died from cryptogenic
be attributable, in part, to the fact that the
group A streptococcal infection shows the classic fea­
specificity of the LRINEC score is greatest for tures of this infection: widespread tissue destruction,
severe disease.9 In a study involving children lack of a tissue inflammatory response, and large num­
with necrotizing fasciitis, the median LRINEC bers of bacteria in the tissues.
score was only 3.7.69

2260 n engl j med 377;23 nejm.org December 7, 2017

The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
Necrotizing Soft-Tissue Infections

sis, some using the time from initial recognition in group A streptococcus have increased, to 15%
of the infection, and others using the time from in the United States82 and to 95.5% in China.83
hospital admission. Studies at tertiary care hos- Treatment failures have been reported in cases
pitals typically report the shortest times to sur- of clindamycin resistance.84 In a study of experi-
gery, probably because the diagnosis was made mental myonecrosis due to group A streptococcus
elsewhere, before admission to the study hos- with resistance to erythromycin and clindamy-
pital. We agree with Bandyopadhyay and col- cin, tedizolid, a second-generation oxazolidinone
leagues71 that the definition of the time to sur- antibiotic, was highly efficacious and superior to
gery should be standardized. linezolid.85
Nevertheless, survival is significantly increased
among patients taken to surgery within 24 hours Other Necrotizing Fasciitis Type II Infections
after admission as compared with those in Current guidelines recommend that A. hydrophila
whom surgery is delayed for more than 24 infections be treated with doxycycline plus either
hours.8,76 Survival is further increased with ear- ciprofloxacin or ceftriaxone.70 A combination of
lier surgical intervention (e.g., within 6 hours),77,78 doxycycline plus either ceftriaxone or cefotaxime
supporting the notion that the earlier surgery is is recommended for V. vulnificus infections.70 For
performed, the better the outcome. MRSA infections, vancomycin, linezolid, dapto-
mycin, or ceftaroline is likely to be effective,
Pharmacologic Treatment though such treatment has not been adequately
Polymicrobial Necrotizing Infections studied.
For mixed aerobic and anaerobic infections of the
head and neck, abdomen, perineum, or gyneco- Traumatic or Spontaneous Gas Gangrene
logic organs, definitive treatment should be Treatment with penicillin plus clindamycin for
based on Gram’s staining, culture, and sensitiv- 10 to 14 days is recommended for traumatic or
ity information. Because of antibiotic resistance spontaneous gas gangrene.70 The recommenda-
in gram-positive microbes and among the Entero- tion of penicillin is based on in vitro sensitivity
bacteriaceae,79 broader antibiotic coverage may data. The recommendation of clindamycin is
be necessary, particularly if the patient has re- based on data showing that it is more effective
cently been hospitalized or treated with antibiot- than penicillin in animal models of gas gan-
ics. Treatment should also be guided by local grene caused by C. perfringens86; clinical trials of
antibiograms, since the emergence of resistance clindamycin have not been performed.
is geographically determined and specific.80 The
Infectious Diseases Society of America (IDSA) Care of Critically Ill Patients
publishes guidelines for the treatment of skin Guidelines for the care of critically ill patients
and soft-tissue infections.70 The current guide- have recently been published.87 However, prob-
lines recommend vancomycin or linezolid plus lems specifically associated with necrotizing
one of the following therapies: piperacillin–tazo- infections are of concern, as noted below.
bactam, a carbapenem, or ceftriaxone–metroni-
dazole. Additional information on specific and Capillary Leak Syndrome
alternative treatments is also provided in the Circulating bacterial toxins and host mediators
IDSA guidelines. Studies are under way to evalu- cause diffuse endothelial damage. Intravenous
ate ceftazidime–avibactam for highly resistant fluid requirements may be extremely high (10 to
gram-negative microbes.81 12 liters of normal saline per day). However,
profound hypoalbuminemia (0.5 to 1 g per deci-
Group A Streptococcal Infections liter) is also common, and replacement with
Treatment with clindamycin in combination colloid (albumin) may therefore be necessary to
with penicillin for 10 to 14 days is recommended maintain oncostatic pressure.
for group A streptococcal infection.70 Clindamy-
cin monotherapy should be considered only after Intravascular Hemolysis
antibiotic susceptibility has been determined, Bacterial hemolysins cause striking and rapid
since both constitutive and inducible resistance reductions in the hematocrit in the absence of

n engl j med 377;23 nejm.org  December 7, 2017 2261
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e

disseminated intravascular coagulopathy. Thus, ciitis is based on its ability to neutralize extra-
the hematocrit may be a better indicator of cellular toxins that mediate pathogenesis. Clini-
the need for transfusion than the hemoglobin cal studies suggesting that there are benefits to
level. IVIG have had serious limitations, including
differences in surgical intervention or clindamy-
Cardiomyopathy cin use between the group that received IVIG
Global hypokinesia, as indicated by echocar- and the group that did not,95 lack of power due
diography and cardiac output, is seen in some to the small sample size,96 low mortality in the
patients with streptococcal toxic shock syn- group that did not receive IVIG,97 and differ-
drome.88 Among survivors, this cardiomyopathy ences in the incidence of necrotizing fasciitis
is reversible, fully resolving in 3 to 24 months between the two study groups.98 Furthermore,
after infection. Some patients have survived with both the quantity and quality of neutralizing
the use of cardiac-assist devices. Management is antitoxin antibodies vary from batch to batch of
difficult, since use of vasopressors increases IVIG. In view of these limitations and the lack
afterload, resulting in decreased peripheral per- of data from definitive double-blind, controlled
fusion and reduced cardiac output. Symmetric studies, the IDSA does not recommend IVIG for
gangrene resulting in loss of one to four ex- necrotizing group A streptococcal infections.70
tremities has been described. Careful monitor- Other investigators are in agreement.72 In a well-
ing and maintenance of mean arterial pressure controlled 2017 study involving 4127 patients
so that it does not exceed 65 mm Hg in patients with necrotizing fasciitis and streptococcal toxic
with this infection seem prudent, though no shock syndrome in 130 hospitals in the United
clinical studies have been performed to support States, IVIG had no effect on mortality or length
this recommendation. of hospital stay.99 Thus, though IVIG has its ad-
vocates, a consensus supporting its use has not
Adjunctive Measures been reached.
Hyperbaric Oxygen
A review of 57 studies performed between 1997 Other Measures
and 2003 concluded that hyperbaric oxygen is A phase 2 trial of a new inhibitor of bacterial
not useful for the treatment of necrotizing fasci- superantigens showed no significant benefit
itis,89 a finding that is similar to the results of with respect to survival, number of surgical
other studies.8,90,91 In contrast, a significant sur- débridements, or serum cytokine levels.100
vival benefit of hyperbaric oxygen in necrotizing
fasciitis was documented in recent studies from Sum m a r y
the United States and Australia.11,92 Other studies
have also suggested a beneficial role of hyper- Necrotizing soft-tissue infections share many
baric oxygen in the treatment of gas gan- clinical and pathological features, and all such
grene,89,91 though experimental studies showed infections result in extensive tissue destruction.
no benefit.93 Recently, a study has been initiated No single clinical laboratory test or group of
to evaluate the effect of hyperbaric oxygen on tests can adequately replace surgical inspection
inflammatory and vasoactive biomarkers in nec- for diagnosis of these infections. Early diagno-
rotizing infections.94 Meanwhile, its benefits re- sis, prompt surgical intervention, and appropri-
main controversial. Surgical débridement, which ate antibiotic treatment are essential to reduce
is essential for the treatment of necrotizing fas- mortality and improve outcomes.
ciitis, should not be delayed in order to pursue
hyperbaric oxygen treatment. Dr. Stevens reports receiving grant support from Motif Biosci-
ences; and Dr. Bryant, receiving grant support from Merck. No
Intravenous Immune Globulin other potential conflict of interest relevant to this article was
reported.
The rationale for using intravenous immune Disclosure forms provided by the authors are available with
globulin (IVIG) in patients with necrotizing fas- the full text of this article at NEJM.org.

2262 n engl j med 377;23 nejm.org  December 7, 2017

The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
Necrotizing Soft-Tissue Infections

References
1. Pasternack MS, Swartz MN. Cellulitis, itis and septic embolization. Nihon Naika tions associated with a toxic shock–like
necrotizing fasciitis, and subcutaneous Gakkai Zasshi 2016;​105:​99-104. (In Japa- syndrome and scarlet fever toxin A. N Engl
tissue infections. In:​Bennett JE, Dolin R, nese.) J Med 1989;​321:​1-7.
Blaser MJ, eds. Mandell, Douglas, and 14. Tawa A, Larmet R, Malledant Y, Se- 27. Wong CJ, Stevens DL. Serious group A
Bennett’s Principles and practice of infec- guin P. Severe sepsis associated with streptococcal infections. Med Clin North
tious diseases. 8th ed. Philadelphia:​Else- Lemierre’s syndrome: a rare but life- Am 2013;​97:​721-736, xi-xii.
vier, 2015:​1194-215. threatening disease. Case Rep Crit Care 28. Adams EM, Gudmundsson S, Yocum
2. Khamnuan P, Chongruksut W, Jear- 2016;​2016:​1264283. DE, Haselby RC, Craig WA, Sundstrom
wattanakanok K, Patumanond J, Tantra- 15. Luckett WH. VII. Large phagedenic WR. Streptococcal myositis. Arch Intern
worasin A. Necrotizing fasciitis: epide- ulcer of the abdomen. Ann Surg 1909;​50:​ Med 1985;​145:​1020-3.
miology and clinical predictors for 605-8. 29. Nuwayhid ZB, Aronoff DM, Mulla ZD.
amputation. Int J Gen Med 2015;​8:​195- 16. Arif N, Yousfi S, Vinnard C. Deaths Blunt trauma as a risk factor for group A
202. from necrotizing fasciitis in the United streptococcal necrotizing fasciitis. Ann
3. Naseer U, Steinbakk M, Blystad H, States, 2003-2013. Epidemiol Infect 2016;​ Epidemiol 2007;​17:​878-81.
Caugant DA. Epidemiology of invasive 144:​1338-44. 30. Bisno AL, Stevens DL. Streptococcal
group A streptococcal infections in Nor- 17. Cheng NC, Wang JT, Chang SC, Tai infections of skin and soft tissues. N Engl
way 2010-2014: a retrospective cohort HC, Tang YB. Necrotizing fasciitis caused J Med 1996;​334:​240-5.
study. Eur J Clin Microbiol Infect Dis by Staphylococcus aureus: the emergence of 31. Schurr M, Engelhardt S, Helgerson R.
2016;​35:​1639-48. methicillin-resistant strains. Ann Plast Limb salvage for streptococcal gangrene
4. Glass GE, Sheil F, Ruston JC, Butler Surg 2011;​67:​632-6. of the extremity. Am J Surg 1998;​ 175:​
PE. Necrotising soft tissue infection in a 18. Miller LG, Perdreau-Remington F, 213-7.
UK metropolitan population. Ann R Coll Rieg G, et al. Necrotizing fasciitis caused 32. Anaya DA, Dellinger EP. Necrotizing
Surg Engl 2015;​97:​46-51. by community-associated methicillin- soft-tissue infection: diagnosis and man-
5. Bocking N, Matsumoto CL, Loewen resistant Staphylococcus aureus in Los Ange- agement. Clin Infect Dis 2007;​44:​705-10.
K, et al. High incidence of invasive group les. N Engl J Med 2005;​352:​1445-53. 33. Brogan TV, Nizet V, Waldhausen JHT,
A streptococcal infections in remote in- 19. Tsai YH, Shen SH, Yang TY, Chen PH, Rubens CE, Clarke WR. Group A strepto-
digenous communities in northwestern Huang KC, Lee MS. Monomicrobial necro- coccal necrotizing fasciitis complicating
Ontario, Canada. Open Forum Infect Dis tizing fasciitis caused by Aeromonas hydroph- primary varicella: a series of fourteen pa-
2016;​4(1):​ofw243. ila and Klebsiella pneumoniae. Med Princ tients. Pediatr Infect Dis J 1995;​14:​588-94.
6. Jabbour G, El-Menyar A, Peralta R, et al. Pract 2015;​24:​416-23. 34. Outbreak of invasive group A strepto-
Pattern and predictors of mortality in 20. Yahav D, Duskin-Bitan H, Eliakim- coccus associated with varicella in a child-
necrotizing fasciitis patients in a single Raz N, et al. Monomicrobial necrotizing care center — Boston, Massachusetts, 1997.
tertiary hospital. World J Emerg Surg fasciitis in a single center: the emergence MMWR Morb Mortal Wkly Rep 1997;​46:​
2016;​11:​40. of Gram-negative bacteria as a common 944-8.
7. Nordqvist G, Walldén A, Brorson H, pathogen. Int J Infect Dis 2014;​28:​13-6. 35. Laupland KB, Davies HD, Low DE,
Tham J. Ten years of treating necrotizing 21. Lee CY, Kuo LT, Peng KT, Hsu WH, Schwartz B, Green K, McGeer A. Invasive
fasciitis. Infect Dis (Lond) 2015;​47:​319- Huang TW, Chou YC. Prognostic factors group A streptococcal disease in children
25. and monomicrobial necrotizing fasciitis: and association with varicella-zoster virus
8. McHenry CR, Piotrowski JJ, Petrinic gram-positive versus gram-negative patho- infection. Pediatrics 2000;​105:​E60.
D, Malangoni MA. Determinants of mor- gens. BMC Infect Dis 2011;​11:​5. 36. Zakikhany K, Degail MA, Lamagni T,
tality for necrotizing soft-tissue infections. 22. Sims SA, Colquhoun S, Wyber R, Cara- et al. Increase in invasive Streptococcus pyo-
Ann Surg 1995;​221:​558-63. petis JR. Global disease burden of group genes and Streptococcus pneumoniae infections
9. van Stigt SF, de Vries J, Bijker JB, et al. A streptococcus. In:​Ferretti JJ, Stevens DL, in England, December 2010 to January
Review of 58 patients with necrotizing fas- Fischetti VA, eds. Streptococcus pyogenes: 2011. Euro Surveill 2011;​16:​19785.
ciitis in the Netherlands. World J Emerg basic biology to clinical manifestations. 37. Zachariadou L, Stathi A, Tassios PT,
Surg 2016;​11:​21. Oklahoma City:​University of Oklahoma Pangalis A, Legakis NJ, Papaparaskevas J.
10. Wong CH, Chang HC, Pasupathy S, Health Sciences Center, 2016:​19785-8. Differences in the epidemiology between
Khin LW, Tan JL, Low CO. Necrotizing 23. Hamilton SM, Stevens DL, Bryant AE. paediatric and adult invasive Streptococcus
fasciitis: clinical presentation, microbiol- Pregnancy-related group a streptococcal pyogenes infections. Epidemiol Infect 2014;​
ogy, and determinants of mortality. J Bone infections: temporal relationships between 142:​512-9.
Joint Surg Am 2003;​85-A:​1454-60. bacterial acquisition, infection onset, clin- 38. Meleney FL. Hemolytic streptococcus
11. Devaney B, Frawley G, Frawley L, ical findings, and outcome. Clin Infect gangrene. Arch Surg 1924;​9:​317-64.
Pilcher DV. Necrotising soft tissue infec- Dis 2013;​57:​870-6. 39. Stevens DL. Could nonsteroidal anti-
tions: the effect of hyperbaric oxygen on 24. Nelson GE, Pondo T, Toews KA, et al. inflammatory drugs (NSAIDs) enhance
mortality. Anaesth Intensive Care 2015;​ Epidemiology of invasive group A strepto- the progression of bacterial infections to
43:​685-92. coccal infections in the United States, toxic shock syndrome? Clin Infect Dis
12. Kulasegaran S, Cribb B, Vandal AC, 2005-2012. Clin Infect Dis 2016;​63:​478- 1995;​21:​977-80.
McBride S, Holland D, MacCormick AD. 86. 40. Aronoff DM, Bloch KC. Assessing the
Necrotizing fasciitis: 11-year retrospec- 25. Lepoutre A, Doloy A, Bidet P, et al. relationship between the use of nonste-
tive case review in South Auckland. ANZ Epidemiology of invasive Streptococcus pyo- roidal antiinflammatory drugs and necro-
J Surg 2016;​86:​826-30. genes infections in France in 2007. J Clin tizing fasciitis caused by group A strepto-
13. Suzuki K, Hayashi Y, Otsuka H, et al. Microbiol 2011;​49:​4094-100. coccus. Medicine (Baltimore) 2003;​ 82:​
Case report: a case of Lemierre’s syn- 26. Stevens DL, Tanner MH, Winship J, 225-35.
drome associated with necrotizing fasci- et al. Severe group A streptococcal infec- 41. Lesko SM. The safety of ibuprofen

n engl j med 377;23 nejm.org  December 7, 2017 2263
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
The n e w e ng l a n d j o u r na l of m e dic i n e

suspension in children. Int J Clin Pract cer patients: a 12-year experience. Cancer patients with suspected necrotizing soft
Suppl 2003;​135:​50-3. 1991;​67:​1928-42. tissue infections: a case-control study.
42. Lamagni TL, Neal S, Keshishian C, et 57. Stevens DL, Musher DM, Watson DA, J Pediatr Surg 2016;​51:​1022-5.
al. Severe Streptococcus pyogenes infections, et al. Spontaneous, nontraumatic gangrene 70. Stevens DL, Bisno AL, Chambers HF,
United Kingdom, 2003-2004. Emerg In- due to Clostridium septicum. Rev Infect Dis et al. Practice guidelines for the diagnosis
fect Dis 2008;​14:​202-9. 1990;​12:​286-96. and management of skin and soft tissue
43. Dubos F, Hue V, Grandbastien B, Cat- 58. Aldape MJ, Bryant AE, Stevens DL. infections: 2014 update by the Infectious
teau B, Martinot A. Bacterial skin infec- Clostridium sordellii infection: epidemiology, Diseases Society of America. Clin Infect
tions in children hospitalized with vari- clinical findings, and current perspec- Dis 2014;​59(2):​e10-e52.
cella: a possible negative impact of tives on diagnosis and treatment. Clin 71. Bandyopadhyay D, Jacobs JV, Panchab-
non-steroidal anti-inflammatory drugs? Infect Dis 2006;​43:​1436-46. hai TS. What’s new in emergencies, trauma
Acta Derm Venereol 2008;​88:​26-30. 59. Alayed KA, Tan C, Daneman N. Red and shock? The tortuous path in the man-
44. Mikaeloff Y, Kezouh A, Suissa S. Non- flags for necrotizing fasciitis: a case con- agement of necrotizing fasciitis: is early
steroidal anti-inflammatory drug use and trol study. Int J Infect Dis 2015;​36:​15-20. surgical intervention critical? J Emerg
the risk of severe skin and soft tissue 60. Carbonetti F, Cremona A, Carusi V, Trauma Shock 2016;​9:​1-2.
complications in patients with varicella or et al. The role of contrast enhanced com- 72. Hakkarainen TW, Kopari NM, Pham
zoster disease. Br J Clin Pharmacol 2008;​ puted tomography in the diagnosis of TN, Evans HL. Necrotizing soft tissue in-
65:​203-9. necrotizing fasciitis and comparison with fections: review and current concepts in
45. Leroy S, Marc E, Bavoux F, et al. Hos- the Laboratory Risk Indicator for Necro- treatment, systems of care, and out-
pitalization for severe bacterial infections tizing Fasciitis (LRINEC). Radiol Med comes. Curr Probl Surg 2014;​51:​344-62.
in children after exposure to NSAIDs: 2016;​121:​106-21. 73. Crew JR, Varilla R, Allandale Rocas T
a prospective adverse drug reaction report- 61. Stamenkovic I, Lew PD. Early recogni- III, Abdul Rani S, Debabov D. Treatment
ing study. Clin Drug Investig 2010;​30:​179- tion of potentially fatal necrotizing fas- of acute necrotizing fasciitis using nega-
85. ciitis — the use of frozen-section biopsy. tive pressure wound therapy and adjunc-
46. Das DK, Baker MG, Venugopal K. Risk N Engl J Med 1984;​310:​1689-93. tive NeutroPhase irrigation under the foam.
factors, microbiological findings and out- 62. Majeski J, Majeski E. Necrotizing fas- Wounds 2013;​25:​272-7.
comes of necrotizing fasciitis in New Zea- ciitis: improved survival with early recog- 74. Baharestani MM. Negative pressure
land: a retrospective chart review. BMC nition by tissue biopsy and aggressive wound therapy in the adjunctive manage-
Infect Dis 2012;​12:​348. surgical treatment. South Med J 1997;​90:​ ment of necrotizing fasciitis: examining
47. Hamilton SM, Bayer CR, Stevens DL, 1065-8. clinical outcomes. Ostomy Wound Man-
Bryant AE. Effects of selective and non- 63. Chelsom J, Halstensen A, Haga T, age 2008;​54:​44-50.
selective nonsteroidal anti-inflammatory Høiby EA. Necrotising fasciitis due to 75. Lee JY, Jung H, Kwon H, Jung SN. Ex-
drugs on antibiotic efficacy of experimen- group A streptococci in western Norway: tended negative pressure wound therapy-
tal group A streptococcal myonecrosis. incidence and clinical features. Lancet assisted dermatotraction for the closure
J Infect Dis 2014;​209:​1429-35. 1994;​344:​1111-5. of large open fasciotomy wounds in nec-
48. Weng TC, Chen CC, Toh HS, Tang HJ. 64. Wall DB, Klein SR, Black S, de Virgilio rotizing fasciitis patients. World J Emerg
Ibuprofen worsens Streptococcus pyogenes C. A simple model to help distinguish Surg 2014;​9:​29.
soft tissue infections in mice. J Microbiol necrotizing fasciitis from nonnecrotizing 76. Freischlag JA, Ajalat G, Busuttil RW.
Immunol Infect 2011;​44:​418-23. soft tissue infection. J Am Coll Surg 2000;​ Treatment of necrotizing soft tissue in-
49. Hamilton SM, Bayer CR, Stevens DL, 191:​227-31. fections: the need for a new approach. Am
Lieber RL, Bryant AE. Muscle injury, vi- 65. Wong CH, Khin LW, Heng KS, Tan J Surg 1985;​149:​751-5.
mentin expression, and nonsteroidal anti- KC, Low CO. The LRINEC (Laboratory 77. Bucca K, Spencer R, Orford N, Catti-
inflammatory drugs predispose to cryptic Risk Indicator for Necrotizing Fasciitis) gan C, Athan E, McDonald A. Early diag-
group A streptococcal necrotizing infec- score: a tool for distinguishing necrotiz- nosis and treatment of necrotizing fasci-
tion. J Infect Dis 2008;​198:​1692-8. ing fasciitis from other soft tissue infec- itis can improve survival: an observational
50. Stevens DL, Laposky LL, McDonald P, tions. Crit Care Med 2004;​32:​1535-41. intensive care unit cohort study. ANZ J
Harris I. Spontaneous gas gangrene at a 66. Bechar J, Sepehripour S, Hardwicke J, Surg 2013;​83:​365-70.
site of remote injury — localization due to Filobbos G. Laboratory Risk Indicator for 78. Hadeed GJ, Smith J, O’Keeffe T, et al.
circulating antitoxin. West J Med 1988;​ Necrotising Fasciitis (LRINEC) score for Early surgical intervention and its impact
148:​204-5. the assessment of early necrotising fasci- on patients presenting with necrotizing
51. Stevens DL. Clostridial myonecrosis itis: a systematic review of the literature. soft tissue infections: a single academic
and other clostridial diseases. In:​Bennett Ann R Coll Surg Engl 2017;​99:​341-6. center experience. J Emerg Trauma Shock
JC, Plum F, eds. Cecil textbook of medi- 67. Burner E, Henderson SO, Burke G, 2016;​9:​22-7.
cine. 20th ed. Philadelphia:​W.B. Saunders, Nakashioya J, Hoffman JR. Inadequate 79. Hawser S, Hoban DJ, Badal RE, et al.
1996:​2090-3. sensitivity of laboratory risk indicator to Epidemiology and antimicrobial suscepti-
52. Fleming A. On the bacteriology of rule out necrotizing fasciitis in the emer- bility of Gram-negative aerobic bacteria
septic wounds. Lancet 1915;​186:​638-43. gency department. West J Emerg Med causing intra-abdominal infections dur-
53. MacLennan JD. The histotoxic clos- 2016;​17:​333-6. ing 2010-2011. J Chemother 2015;​27:​67-
tridial infections of man. Bacteriol Rev 68. Holland MJ. Application of the Labo- 73.
1962;​26:​177-276. ratory Risk Indicator in Necrotising Fas- 80. Satlin MJ, Chen L, Patel G, et al. Multi-
54. Stewart JC. Anaerobic cellulitis. J Am ciitis (LRINEC) score to patients in a center clinical and molecular epidemiolog­
Med Assoc 1905;​45:​528-35. tropical tertiary referral centre. Anaesth ical analysis of bacteremia due to carba­
55. Qvist G. Anaerobic cellulitis and gas Intensive Care 2009;​37:​588-92. penem-resistant Enterobacteriaceae (CRE)
gangrene. Br Med J 1941;​2:​217-21. 69. Putnam LR, Richards MK, Sandvall in the CRE epicenter of the United States.
56. Bodey GP, Rodriguez S, Fainstein V, BK, Hopper RA, Waldhausen JH, Harting Antimicrob Agents Chemother 2017;​61(4):​
Elting LS. Clostridial bacteremia in can- MT. Laboratory evaluation for pediatric e02349-16.

2264 n engl j med 377;23 nejm.org  December 7, 2017

The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.
Necrotizing Soft-Tissue Infections

81. Krapp F, Grant JL, Sutton SH, Ozer 87. Rhodes A, Evans LE, Alhazzani W, et al. immune response and effects of hyper-
EA, Barr VO. Treating complicated carba­ Surviving Sepsis Campaign: international baric oxygenation — the protocol of the
penem-resistant Enterobacteriaceae infec- guidelines for management of sepsis and prospective cohort BIONEC study. BMJ
tions with ceftazidime/avibactam: a retro- septic shock: 2016. Intensive Care Med Open 2015;​5(5):​e006995.
spective study with molecular strain 2017;​43:​304-77. 95. Kaul R, McGeer A, Norrby-Teglund A,
characterisation. Int J Antimicrob Agents 88. Stevens DL, Shelly MP, Stiller R, Villa­ et al. Intravenous immunoglobulin ther-
2017;​49:​770-3. senor-S A, Bryant AE. Acute reversible apy for streptococcal toxic shock syndrome
82. DeMuri GP, Sterkel AK, Kubica PA, cardiomyopathy in patients with strepto- — a comparative observational study.
Duster MN, Reed KD, Wald ER. Macrolide coccal toxic shock syndrome. In:​Proceed- Clin Infect Dis 1999;​28:​800-7.
and clindamycin resistance in group A ings of the XVIIth Lancefield Interna- 96. Darenberg J, Ihendyane N, Sjölin J,
streptococci isolated from children with tional Symposium on Streptococci and et al. Intravenous immunoglobulin G ther-
pharyngitis. Pediatr Infect Dis J 2017;​36:​ Streptococcal Diseases, Athens, June 22– apy in streptococcal toxic shock syndrome:
342-4. 26, 2008. a European randomized, double-blind,
83. Peng XM, Yang P, Liu S, et al. The ge- 89. Wang C, Schwaitzberg S, Berliner E, placebo-controlled trial. Clin Infect Dis
netic features of drug resistance to group Zarin DA, Lau J. Hyperbaric oxygen for 2003;​37:​333-40.
A streptococcus and macrolides antibiot- treating wounds: a systematic review of 97. Shah SS, Hall M, Srivastava R, Subra-
ics among pediatric patients in Beijing the literature. Arch Surg 2003;​138:​272-9. mony A, Levin JE. Intravenous immuno-
2012. Zhonghua Yu Fang Yi Xue Za Zhi 90. Jallali N, Withey S, Butler PE. Hyper- globulin in children with streptococcal
2013;​47:​1040-4. (In Chinese.) baric oxygen as adjuvant therapy in the toxic shock syndrome. Clin Infect Dis
84. Lewis JS II, Lepak AJ, Thompson GR management of necrotizing fasciitis. Am 2009;​49:​1369-76.
III, et al. Failure of clindamycin to eradi- J Surg 2005;​189:​462-6. 98. Linnér A, Darenberg J, Sjölin J, Hen-
cate infection with beta-hemolytic strep- 91. Willy C, Rieger H, Vogt D. Hyperbaric riques-Normark B, Norrby-Teglund A. Clin-
tococci inducibly resistant to clindamycin oxygen therapy for necrotizing soft tissue ical efficacy of polyspecific intravenous
in an animal model and in human infec- infections: contra. Chirurg 2012;​83:​960- immunoglobulin therapy in patients with
tions. Antimicrob Agents Chemother 2014;​ 72. (In German.) streptococcal toxic shock syndrome: a com-
58:​1327-31. 92. Shaw JJ, Psoinos C, Emhoff TA, Shah parative observational study. Clin Infect
85. Bryant AE, Aldape MJ, Bayer CR, McIn- SA, Santry HP. Not just full of hot air: Dis 2014;​59:​851-7.
doo E, Stevens DL. Efficacy of tedizolid in hyperbaric oxygen therapy increases sur- 99. Kadri SS, Swihart BJ, Bonne SL, et al.
experimental myonecrosis caused by eryth- vival in cases of necrotizing soft tissue Impact of intravenous immunoglobulin
romycin/clindamycin-sensitive and -resis- infections. Surg Infect (Larchmt) 2014;​15:​ on survival in necrotizing fasciitis with
tant group A streptococcus. In:​Proceed- 328-35. vasopressor-dependent shock: a propen-
ings and Abstracts of 2016 IDWeek, New 93. Stevens DL, Bryant AE, Adams K, sity score-matched analysis from 130 US
Orleans, October 26–30, 2016. Mader JT. Evaluation of hyperbaric oxy- hospitals. Clin Infect Dis 2017;​ 64:​
877-
86. Stevens DL, Maier KA, Laine BM, gen therapy for treatment of experimental 85.
Mitten JE. Comparison of clindamycin, Clostridium perfringens infection. Clin Infect 100. Bulger EM, Maier RV, Sperry J, et al.
rifampin, tetracycline, metronidazole, and Dis 1993;​17:​231-7. A novel drug for treatment of necrotizing
penicillin for efficacy in prevention of ex- 94. Hansen MB, Simonsen U, Garred P, soft-tissue infections: a randomized clini-
perimental gas gangrene due to Clostridi- Hyldegaard O. Biomarkers of necrotising cal trial. JAMA Surg 2014;​149:​528-36.
um perfringens. J Infect Dis 1987;​155:​220-8. soft tissue infections: aspects of the innate Copyright © 2017 Massachusetts Medical Society.

images in clinical medicine
The Journal welcomes consideration of new submissions for Images in Clinical
Medicine. Instructions for authors and procedures for submissions can be found
on the Journal’s website at NEJM.org. At the discretion of the editor, images that
are accepted for publication may appear in the print version of the Journal,
the electronic version, or both.

n engl j med 377;23 nejm.org  December 7, 2017 2265
The New England Journal of Medicine
Downloaded from nejm.org by R LOPEZ MATA on December 6, 2017. For personal use only. No other uses without permission.
Copyright © 2017 Massachusetts Medical Society. All rights reserved.