key: cord-0703303-igdep94b
authors: Rimailho, Alain; Riou, Bruno; Richard, Christian; Auzepy, Philippe
title: Fulminant Necrotizing Fasciitis and Nonsteroidal Anti-Inflammatory Drugs
date: 1987-01-03
journal: J Infect Dis
DOI: 10.1093/infdis/155.1.143
sha: b321398b45ab3bdeb128fec0970defe33b09aff9
doc_id: 703303
cord_uid: igdep94b

nan

Necrotizing fasciitis is a distinct clinical entity usually caused by Steptococcus pyogenes. Although accurately defined by Wilson [1] in 1952, the disease is often unrecognized. It is potentially fatal, especially when it becomes fulminant. Even though the host factors and the infectious agent responsible for initiating and spreading necrotizing fasciitis have been identified, the pathogenesis of the disease is unclear. After examining all cases of necrotizing fasciitis occurring in the intensive medical care unit Received for publication 2 June 1986, and in revised form 9 July 1986. We thank Marcelline Liliane Sextus for technical assitance. Please address requests for reprints to Dr. A. Rimailho, Service de Reanimation Medicale, 78 Rue du General Leclerc, Le Kemlin-Bicetre, 94275 Cedex, France. of this hospital between 1983 and 1985, we found that five of seven had a fulminant evolution. This fulminant evolution apparently resulted from nonsteroidal antiinflammatory therapy.

We included only those patients who had necrotizing fasciitis that fulfilled strict criteria: a soft-tissue infection involving the superficial fascia and resulting in extensive undermining of surrounding tissues. The infection was considered fulminant when the delay between the first functional and local signs and surgery did not exceed six days (table 1) .

All patients had previously been healthy, and none had a predisposing condition -previous surgery, chronic ve-_~I i (2 c V'.I""""" " «

Necrotizing fasciitis is a relatively rare disease that is often difficult to diagnose during the early stages, when the skin is apparently still normal. However, the outcome of this potentially fatal disease is largely influenced by early recognition and prompt therapy based upon antibiotics and surgery [2] . Stamenkovic and Lew [3] recently suggested that rapid performance of frozen-section soft-tissue biopsy early in the evolution of a suspect lesion may provide a definitive and life-saving diagnosis. In our patients, as in many others, necrotizing fasciitis was not initially suspected; therefore, no biopsy was performed, and treatments were prescribed for conditions such as arthritis or phlebitis. Necrotizing fasciitis usually requires both a specific organism and an underlying debilitating disease [4, 5] . Although group A l3-hemolytic Streptococcus has long been recognized as the agent of this infection, recent reports [6, 7] indicated that necrotizing fasciitis can be caused by a variety of bacteria. In patient 6, the agent was the gramnegative bacillus Serratia marcescens. The hypothesis [8] that production of hernolysins, fibrinolysins, and hyal-nous or arterial ulcer, or debilitating disease -that could have altered their normal defense mechanisms. Three of the five patients reported recent, minor injuries. Initially, a false diagnosis -arthritis or phlebitis -was made on the basis of pain and edema without subcutaneous abnormalities, and the usual nonsteroidal anti-inflammatory drugs were prescribed, but no antibiotics were administered, Each patient's local and general signs reached a critical point, which included serious disorders (especially shock). An obvious diagnosis of necrotizing fasciitis was established, and the patients were admitted into the intensive care unit. All patients except patient 4 (who received penicillin 36 hr before admission) became septicemic with a single organism -a Lancefield's group A l3-hemolytic Streptococcus in most cases -and had positive cultures of the same organism from blisters or areas of necrosis. Treatment in the intensive care unit included iv penicillin (15-25 X 10 6 Ll/day), fluid replacement, cardiac inotropic drugs, and surgery. Three patients died of septic shock within 15-48 hr after admission, despite excision surgery and aggressive medical treatment. Two patients survived after prolonged hospitalization and repeated surgery and grafting.

During this period, two other patients were admitted with necrotizing fasciitis: a 53-year-old diabetic man with a chronic arterial ulcer and an 82-year-old woman with a chronic venous ulcer. Neither one had previously received anti-inflammatory drugs or antibiotics. The disease had a subacute evolution in both patients (15 and 12 days, respectively, from first local signs to overt necrotizing fasciitis), and cultures of blood were negative. Both patients survived after treatment with penicillin and surgery. .,.

;:: The action of bacterial toxins, together with the effect of bacterial kinases, is further enhanced by the general reduction in host defense that occurs during the postoperative period, by preexisting systemic diseases such as diabetes and malnutrition, or by conditions requiring prolonged treatment with steroids [2, 8] . However, all five of our patients were previously healthy, and no underlying disease could be found. They were all treated with nonsteroidal anti-inflammatory drugs just after diagnosis of the disease. A marked acceleration of the evolution was then observed in all patients, although no other drug was given during the early stages. Necrotizing fasciitis is generally considered a fulminating disease, but the delay between first signs and surgical treatment is usually more than six days. In Freeman's study [7] , most patients with necrotizing fasciitis had symptoms of disease for four days to three weeks before presenting at the hospital; the most severe infections were in patients who waited 10 to 14 days before going to the hospital. Almost all the patients had an underlying disease, and none received immunodepressive drugs [7] . In Hammar's study [4] , patients with necrotizing fasciitis had been ill for four to five days before admission to the hospital, and all but two developed gangrenous ulcers that appeared at least one week after the beginning of the disease. Almost all patients presented with other diseases, but none received oral steroids or other immunodepressive drugs. In Leppard's study [9] , patients with necrotizing fasciitis were divided into two groups: In the first group, the eight patients did not receive any antibiotics before admission and the evolution was considered fulminating (but the time between presentation and diagnosis was not mentioned); in the second group, the five patients had been previously treated with antibiotics, and an interval of several weeks passed before the skin became necrotic. No patients in either group received antiinflammatory drugs. In our intensive care unit, we observed that the two non-fulminating cases of necrotizing fasciitis occurred in patients who did not receive non-steroidal anti-inflammatory drugs or antibiotics. Some cases of necrotizing fasciitis have been recently reported [10] in patients given one or several steroidal and non-steroidal antiinflammatory drugs; the delay from triggering event to overt necrotizing fasciitis and referral to hospital was from four to ten days. The limited number of patients we observed does not warrant a statistical study to affirm that non-steroidal anti-inflammatory drugs were directly responsible for a fulminating evolution in necrotizing fasciitis, but the shorter delays in our study than in previous ones is a convincing argument to establish the deleterious role of such drugs.

The mechanism can be deduced from previous studies. Activation of infection in debilitated patients treated with anti-inflammatory drugs has been reported [11] . Obser-145 vations that granulocyte-mediated functions -chemotaxis, phagocytosis, and bactericidal activity -are impaired in vitro by nonsteroidal drugs have been supported by decreased in vitro bactericidal activity and reduced activation of Staphylococcus aureus and group B Streptococcus by granulocytes incubated with phenylbutazone [12] . Inhibition of granulocyte adherence has been demonstrated to be secondary to the anti-inflammatory effect of aspirin [13] .

In a recent case report [14] , a life-threatening case of necrotizing fasciitis occurred in a 23-year-old, previously healthy woman who took six diflunisal tablets over a 24hr period after a minor injury; lymphocyte function and lymphocyte transformation were severely depressed. The authors suggested that these troubles represented an adverse reaction to the drug that contributed to the severe sepsis that developed. This patient had nearly normal levels of serum immunoglobulins, as did three of our five patients. It appears that the possible deleterious effect of nonsteroidal anti-inflammatory drugs results from deep and quick depression of cellular defense.

Our observations suggest that non-steroidal antiinflammatory drugs should not be used for apparently benign inflammatory cutaneous lesions when infection cannot be avoided. ALAIN 

Necrotizing fasciitis

Diagnosis, clinical course and treatment of acute dermal gangrene

Early recognition of potentially fatal necrotizing fasciitis: the use of frozen-section biopsy

Erysipelas and necrotizing fasciitis

Necrotizing fasciitis

Bacteriology of necrotizing fasciitis

Necrotizing fasciitis

Necrotizing fasciitis. Postgrad

The value of bacteriology and serology in the diagnosis of necrotizing fasciitis

Haemolytic streptococcal gangrene and nonsteroidal anti-inflammatory drugs

Activation of latent infection by indomethacin: a report of three cases

Influence of phenylbutazone on leucocyte chemiluminescence and function

Inhibition of Concise Communications granulocyte adherence by ethanol, prednisone, and aspirin, measured with an assay system

Necrotising fasciitis after diflunisal for minor injury

The Plaque-Neutralization Test as a Measure of Prior Exposure to Measles Virus It has been noted that antibody levels measured by HAl decline over time after either measles or measles vaccination [1] . Some persons may lose all detectable HAl antibody. The immune status of such persons has been unclear.In the spring of 1982, a comprehensive statewide clustersample serosurvey of 6th, 10th, and 12th grade schoolchildren in Massachusetts was undertaken [2] . Of the 1,871 participants, 98.1 070 had a history of measles vaccination. Seroprevalence levels, determined by using a standard HAl technique, were 86.8070 and ranged from 84.7% in 10th grade students to a high of 89.5% in 6th grade students. Students who were seronegative by HAl were retested using a sensitive plaque-neutralization test (PNT). Seroprevalence levels rose to 98.6%, with a low of 97.7% in 6th graders and a high of 99.1 % in 10th graders.These results were difficult to interpret because the clinical significance of neutralizing antibody detected by the sensitive PNT was unclear. To evaluate the meaning of plaque-neutralizing (PN) antibody in the absence of HAl antibody, we undertook a study (in the spring of 1983) to correlate the type of immune response to measles vac-Received for publication 2 April 1986, and in revised form 25 July 1986. This work was presented in part at the 25th Interscience Conference on Antimicrobial Agents and Chemotherapy, held 29 September-2 October 1985 in Minneapolis.Informed consent was obtained from the patients or from their parents or guardians, and the guidelines for human experimentation of the U. S. Department of Health and Human Services were followed in the conduct of this research.We thank Dr. Nicholas J. Fiumara, State Epidemiologist in Massachusetts at the time of the study; John Narkunas for coordinating the vaccination study; Mike Baltier and Dr. George Waterman for guidance and support; Anna Hall and Karen Sanderlin for performing the HAl assay and for fractionating the sera; the epidemiologists and nurses of the Massachusetts Department of Public Health for conducting the study; and Connie Keith and staff for preparing the manuscript.Please address requests for reprints to the Technical Information Services (WAO), Center for Prevention Services, Centers for Disease Control, Atlanta, Georgia 30333. cination with the prior levels of PN antibody. We had the following objectives: (1) to determine whether the absence of detectable PN antibodies before vaccination was associated with a primary immune response after vaccination, as determined by the presence of detectable levels of IgM; (2) to determine whether the presence of detectable PN antibodies before vaccination would result in a secondary immune response, as determined by the absence of IgM; and (3) to determine whether the level of PN antibodies before vaccination correlated with the development of a fourfold or greater rise in titers of antibody after vaccination.

All students who had participated in the previous serosurvey when in the 6th or 10th grade and who did not have detectable HAl antibody at a 1:5 dilution were eligible for the study. The greatest efforts were made to enroll those students with the lowest titers of PN antibodies «20). After obtaining informed consent, students received the measles, mumps, and rubella vaccine. Serum was obtained for analysis of HAl and PN antibodies to measles virus rv2.5-3.5 weeks after vaccination. IgM antibodies were measured in sera by using both techniques on a subsam-pIe of participants chosen to include all students with prevaccination titers of PN antibodies~100.HAl antibodies were measured, starting at a 1:5 dilution, with a standard method using Tween ether-treated antigen, sera inactivated at 56 C to remove nonspecific inhibitors, and vervet monkey erythrocytes as indicator cells [3] .PN antibodies were measured, starting at a 1:4 dilution, in Vero cell monolayer cultures by using a low-passage Edmonston strain of measles virus and the conventional method previously published by our group [4] . The titer of PN antibody was defined as the serum dilution that would reduce the number of plaques by 50070. Levels of IgM antibody to measles virus were determined by each of the above assays from the 19S serum fractions obtained by ultracentrifugation of the serum on sucrose density gra-