0022-5347/95/1541-0089$03.

00/0
THEJOITLNAL OF UROLOC~Y Vol. 154.89-92.July 1995
Copyright 0 1995 by AMERICAN AL hc.
U R O ~ I C ASSOCIATION. hinted in U S A .

OUTCOME PREDICTION IN PATIENTS WITH FOURNIERS GANGRENE

ELIAHU LAOR, LANE S. PALMER, BHUPENDRA M. TOLIA, ROBERTO E. REID AND
HOWARD I. WINTER*
From the Department of Urology, Montefiore Medical CenterlAlbert Einstein College of Medicine, Bronx, New York

ABSTRACT
We treated 30 patients with Fournier's gangrene during a 15-year period. Data were collected
on demographics, medical history, admission signs and symptoms, physical examination, admis-
sion laboratory studies and bacteriology. The timing and degree of surgical debridement as well
as antibiotic therapy were also reviewed. The extent of disease was calculated from body surface
area nomograms. Data were stratified according to the outcomes of death (13 patients) or
survival (17). Patients who survived were significantly younger (53 years old, range 23 to 90)
than those who died (71 years old, range 53 to 83, p = 0.004). Admission laboratory parameters
that were statistically related to outcome included hematocrit, blood urea nitrogen, calcium,
albumin, alkaline phosphatase and cholesterol levels. White blood count, platelets, potassium,
bicarbonate, blood urea nitrogen, total protein, albumin and lactic dehydrogenase levels 1 week
following hospitalization were also associated with outcome. The greater mean extent of body
surface area involved among patients who died was not statistically different from that of those
who lived (7.16 and 4.32%, respectively, p = 0.1).The number of surgical dbbridements did not
seem to influence outcome.
To assess better the physiological profile of the patients in both outcome categories, the acute
physiology and chronic health evaluation I1 severity score was modified to create a Fournier's
gangrene seventy index. The mean Fournier's gangrene severity index for survivors was 6.9 2
0.9 compared to 13.5 5 1.5 for nonsurvivors. Regression analysis demonstrated a strong corre-
lation between Fournier's gangrene seventy index and death rate (correlation coefficient = 0.934,
p = 0.005). Using a Fournier's gangrene severity index threshold value of 9, there was a 75%
probability of death with a score greater than 9, while a score of 9 or less was associated with a
78% probability of survival (p = 0.008).
In conclusion, Fournier's gangrene is an infectious disease affecting an ever aging population
of patients. Deviation from homeostasis is the most important parameter predictive of outcome
and not the extent of disease or performance of surgical d6bridement. The Fournieis gangrene
severity index is a n objective and simple method to q u a n t the ~ extent of metabolic aberration
that may be used to predict outcome. We recommend the use of the Fournier's gangrene severity
index when evaluating therapeutic options and reporting results.
Kcr WORDS: gangrene; genitalia, male;fasciitis; critical care; prognosis
Despite the development and evolution of modern inten- routinely to assess the extent of burn injuries: the penis,
sive care techniques and antibiotic therapy, the mortality scrotum and perineum accounted for 1%surface area each,
rate from Fournier's gangrene remains high at approxi- each ischiorectal fossa 2.5%and so forth (see figurehs
mately 50%.l,2The factors governing survival or death from We created a Fournier's gangrene severity index by modi-
Fournier's gangrene are not clearly identified, measured and fyrng the acute physiology and chronic health evaluation I1
analyzed in the literature. To determine if prognosticatorsof severity score4 to reflect the measurable data that were uni-
outcome exist, we reviewed our experience with this disease formly available for all patients (table 1). Ultimately, 9 pa-
during the last 15 years, comparing measurable pathophys- rameters were measured in every case and the degree of
iologid data between survivorsand nonsurvivors of Fournier's deviation from normal was graded from 0 to 4. The sum of the
gangrene. individual values was tallied to arrive at the Fournier's gan-
grene severity index score. Data were stratified by outcome,
MATERIALS AND METHODS that is whether the patient died or survived. All data were
The medical records of all patients with Fournier's gan- analyzed gression
statistically by Student's t test, paired t test, re-
analysis and Fisher's exact test, with p <0.05 con-
grene treated at the affiliated hospitals of our institution sidered significant.
between 1978 and 1992 were thoroughly reviewed. Data were
collected in detail about demographics, medical history, ad-
mission signs and symptoms, and physical examination. All RESULTS
available laboratory results were collected for all patients. Ofthe 30 patients studied 13 died and 17 survived. The
Treatment information, including the timing and extent of mean age of the entire group was 61 years (range 23 to 90).
surgical debridement as well as antibiotic therapy, was also Patients who survived were significantlyyounger (mean age
reviewed. The extent of disease was calculated for each pa- 53 years, range 23 to 90)than those who died (mean age 71
tient based on modified body surface area nomograms used years, range 53 to 83, p = 0.004). There was no atatjstically
significant differencein race distribution betwen the 2 out-
* Current a d k s s : Kaiser Permanente, Panorama City, California come groups (p >0.05). Patients presented with v&ow =-
Accepted for ublication December 12,1994.
91402. binations of signs and symptoms, including fever grsatSr
89
90 OUTCOME PREDICTION IN PATIENTS WITH FOURNIERS GANGRENE
DISCUSSION
Necrotizing fasciitis of the male genitourinary tract
I[Fournier’s gangrene) was first described in 1883 as an idio-
pathic, fulminant genital gangrene of acute onset in other-
wise healthy young men.5 There are currently approximately
500 cases reported in the literature, with the largest study
reporting on 57 patients.6 Since the initial reports, Fournier‘B
gangrene has come to involve an older population, is less
fulminant and is known to be an infectious process with an
identifiable source in the majority of cases.1.6-8The reported
age of patients with Fournier’s gangrene has progressively
increased in the literature. In a review of 267 patients
treated before 1945,McCrea reported an average age of 40.9
years.9 Jones et a1 reported that the average age increased to
51.3 years between 1945 and 1979.7In 1984 Spirnak et al
reported an average patient age of 54.6 years.’ The average
age of our patients was 61 years, and surviving patients were
significantly younger than those who died (53 versus 71
years, respectively). Clayton et a1 similarly found that pa-
tients who survived were significantly younger than tho=
who died of Fournier’s gangrene (52versus 69 years, respec-
tivelyL6
The clinical presentation of Fournier’s gangrene in our
patients was similar to that described in the literature.7.S
Typically, a prodromal period of genital discomfort or pruri-
tus associated with fever (with or without chills) was followed
by a constellation of clinical signs and symptoms, including
genital erythema, swelling and crepitance suggestive of sub-
cutaneous gas formation, which may extend to involve the
Percentage allocation of penned surface area anterior abdominal wall as far as the axilla, thighs or but-
tocks. Progression of the gangrenous process will lead to
sloughing and drainage of the affected areas, eventually de-
marcating the area between viable and dead tissue.
than 101F,erythema, purulence or wound discharge, fluctu- Previous reports promoted a n aggressive surgical ap-
ance or crepitance. The duration of symptoms before presen- proach, suggesting a positive effect on survival.1,6,l0-11
tation was not statistically significantly different between Spirnak et al associated the higher mortality rate for pa-
patients who did (2.3days) and did not (7.4days) survive. tients who underwent more frequent operations to a greater
Results of bacteriological studies at presentation are pre- extent of the disease.’ This assertion was contradicted by
sented in table 2.Admission laboratory parameters that were Clayton et al, who found that the mortality rate was not
statistically significantly different depending on outcome in affected by the extent of disease.6 In an attempt to resolve
these patients included hematocrit, blood urea nitrogen these contradictory reports, we devised a method of quanti-
(BUN), calcium, albumin, alkaline phosphatase and choles- fjmg the extent of disease. The mean extent of body surface
terol (table 3).The laboratory values 1 week after hospital- area involved in the necrotizing process in our patients was
ization that were associated with outcome included white not statistically different between those who survived and
blood count, platelets, potassium, bicarbonate, BUN, total those who died, suggesting that the extent of disease is not
protein, albumin and lactic dehydrogenase. predictive of outcome. However, the difference in surface
The mean extent of body surface area involved in the areas was nearly 2-fold between the 2 groups and we find it
necrotizing process in patients who lived and died was 4.32% difficult to discount the importance of this parameter. It is
and 7.16%, respectively. These differences were not statisti- conceivable that if this difference were maintained in a larger
cally significant (p = 0.1).The number of surgical dkbride- sample size, statistical significance would be reached. Fur-
ments did not seem to influence patient outcome. A single thermore, the number of surgical debridements did not ap-
surgical dkbridement was performed in 6 patients who died pear to influence patient outcome, corroborating a similar
finding by Clayton et a1.6
and 9 who survived, while 2 to 3 dkbridements were per-
Numerous predisposing factors leading to Fournier’s gan-
formed in 5 and 6,respectively. Three patients in each group
grene have been reported, including perianal disease, ure-
underwent diverting colostomy and 2 in each group required thral stricture, local trauma, diabetes mellitus, malignancies
suprapubic cystotomy. and other immunocompromised states.6-8.11.12 The majority
The mean Fournier’s gangrene severity index for survivors of patients had at least 1 predisposing or concurrent medical
was 6.9 ? 0.9 compared to 13.5 2 1.5 for those who died. or surgical illness. Among the 13 patients who died 10 had at
Regression analysis demonstrated a strong correlation be- least 3 illnesses, while only 4 of 17 survivors had at least 2
tween Fournier’s gangrene seventy index and death rate illnesses. Among the particular disease states only hepatic
(correlation coefficient = 0.934,p = 0.005).With a Fournier’s dysfunction and renal insufficiency had an adverse effect on
gangrene severity index value of 9 as a threshold, there was survival. The number of patients with diabetes mellitus did
a 75% probability of death with a score greater than 9,while not differ significantly with respect to outcome. The plethora
a score of 9 or less was associated with a 78% probability of of conditions that have been associated with Fournier’s gan-
survival (Fisher’s exact test, p = 0.008).Finally, co-morbid grene, as well as our findings, suggests that a generally
medical conditions that were associated with death included debilitated state establishes a milieu conducive to the infec-
renal failure and hepatic dysfunction (table 4).The incidence tious gangrenous process and has a negative impact on sur-
of diabetes mellitus was similarly distributed among patients vival.
in both groups. In an effort to determine other predictors of outcome, we
 OUTCOME PREDICTION IN PATIENTS WITH FOURNIERS GANGRENE 91
TABLE1. Fournier's gangrene severity index
Physiological Variabld'oint High Abnormal Values Normal Low Abnormal Values
Assignment +4 +3 +2 +1 0 +l +2 +3 +4
Temperature (C) More than 41 39-40.9 - 38.5-38.9 36-38.4 32-33.9 3041.9 Less than 29.9
Heart rate More than 180 140-179 110-139 - 70-109 5549 43-54 hSSthM39
Respiratory rate More than 50 35-49 - 25-34 12-24 6-9 - Lesa than 5
Serum sodium (mmo1.n.) More than 180 160-179 155-159 150-154 130-149 120-129 111-119 Less than 110
Serum potassium (mmo1.n.) More than 7 6-6.9 - 5.5-5.9
-
3.5-5.4 2.5-2.9
-
- L ~ S Sthan 2.5
-
Serum creatinine (mg.1100 ml., More than 3.5 23.4 1.5-1.9 0.6-1.4 Less than 0.6
x 2 for acute renal failure)
Hematoerit (%) More than 60 -
-
50-59.9 46-49.9 30-45.9 20-29.9 - Lessthan20
White blood count ( t ~ t a l / m m . ~ More than 40 2049.9 15-19.9 3-14.9 1-2.9 - Less than 1
x 1,000)
Serum bicarbonate (venous, More than 52 41-51.9 - 3240.9 22-31.9 18-21.9 1617.9 Lessthan 15
m0i.n.)

TABLE2. Bacterial organisms cultured from wounds ofpatients TABLE4. Prediaposinglconcumnt medical or surgical illnesses in
with Fournier's gangrene 30 patients with Fournier's gangrene
Organism Total No. No. Condition No. Survived No. Died
No. Pta. Survived Died Renal insufficiency 1 6
E. coli 11 4 I Diabetes mellitus 5 4
Streptococcus species 11 7 4 Septicemia (pas. blood cultures) 2 2
Enterococeus 9 4 5 Peripheral vascular disease 2 2
Coagulase-neg. staphylocoecus 9 5 4 Perireetal abscess 5 0
Bacteroides species 7 3 4 Hepatic dysfunction 1 4
Klebsiella pneumoniae 5 3 2 Atherosclerotic heart disease 1 3
Unspecified gram-neg. rods 5 3 2 Malignancy 0 2
Enterobacter species 4 2 2 Spermatic cord injury 2 1
Proteus species 3 2 1 Pneumonia 0 2
Pseudomonas species 3 1 2 Hypertension 2 3
Acinetobacter species 1 1 0 Urethral involvementhuanipulation 2 1
Citrobacter species 1 0 1
Coagulase-p. staphylocoecus 1 1 0

BUN, and lactate dehydrogenaae levels, and had signifi-

TABLE3. Laboratory parameters associated with outcome cantly higher platelet counts, serum potassium, bicarbonate,
total protein and albumin levels. Clayton et al, noting a
Mean Mean
Parameter NO.SW- No. p Value
higher mortality rate among patients who presented with
ViVd Died "sepsis," sought to identify outcome prognosticators among
those parameters that defined the septic state and found that
Admission labomtory pammetera
only BUN greater than 50 mgJdl. was signiticant.6 In our
Hematomit 40.8 30.6 0.001 series differences in serum albumin, total protein and cho-
BUN 26 69.7 0.01
CalCiUXU 8.6 7.9 0.009 lesterol levels reflected the underlying degree of debilitation
Albumin 3.1 2.5 0.02 and worse outcome.
Alkaline phosphatase 153 263 0.01 These results, taken individually, did not appear to be
Cholesterol 141 109 0.01 practical in the clinical application. We sought a parameter
Labomtory pammetem I week afier hospitalization that was useful clinically and reflected the physiological pro-
White blood count 10.3 18.5 0.003 tile of patients with Fournier's gangrene. The acute physiol-
Platelets 405 160 O.OOO4 ogy and chronic health evaluation classification system was
Potassium 4.5 3.8 0.002 described in 1981 as a method by which to assess quantita-
Bicarbonate 25 18.5 0.m2
tively the severity of an acute disease." The correlation of
BUN 11 53.7 o.oooo1
Total protein 6.1 4.7 o.oooo1 acute physiology and chronic health evaluation II symptom
Alb- 2.1 2.4 0.01 score with outcome has been documented for many septic
Lactic dehydmgenase 216 417 o.oooo1 proces~es.~~-20 Similarly, the mean Fournier's gangrene se-
verity index for survivors was significantly lower than that
for nonsurvivors and strongly correlated with death rates.
evaluated bacteriologicaldata, and the laboratory data avail- When a F o d e r ' s gangrene severity index score of 9 was
able from hospitalization and 1week aRer surgical dbbride- used as a threshold parameter to predict outcome those with
ment. Synergy between anaerobic and aerobic organisms iso- a score greater than 9 had a 75% probability of death, while
lated from wound cultures has been cited as an important a Fournier's gangrene severity index score of 9 or less was
bacteriological principle in Foumier's gangrene.6 In a tabu- associated with a 78% probability of survival.
lation of isolated bacteria in 194 patients, Paty and Smith
found Escherichia coli, Bacteroides and streptococci to be CONCLUSIONS
the most commonly present organisms.' "he microbiolog- Our data indicate that deviation from homeostasis at pre-
ical results from our patients demonstrated that E. coli sentation with Fournier's gangrene is the most important
and Streptococcus species were the most common organ- general parameter that predicts outcome. The extent of dis-
isms; however, Staphylococcus and Enterococcus were ease, number of surgical d6bridements and other faetors
more commonly isolated than Bacteroides. appear to be much less important. The admission Fournier's
Among the admission laboratory parameters, patients who gangrene severity index is an objective and simple method to
survived had significantly higher hematocrit, serum calcium, quantify the extent of metabolic aberration and can be used
albumin and cholesterol levels, and lower BUN and alkaline to predict outcome. Moreover, there may be a role for the
phosphatase levels. At 1 week following surgical dbbride- Fournier's gangrene severity index in costbenefit analysis or
ment, patients who survived had lower white blood counts, resource allocation algorithms, as has been used by the acute
92 OUTCOME PREDICTION IN PATIENTS WITH FOURNIERS GANGRENE
physiology and chronic health evaluation I1 symptom
score.*l-z3 By our philosophy, t h i s is not the primary inten-
tion but, r a t h e r , a m e a n s of evaluating therapeutic options
a n d reporting results.
REFERENCES
1. Spirnak. J. P., Resnick, M. I., Hampel, N. and Persky, L.: Foumier’s
gangrene: report of 20 patients. J. Urol., 131: 289,1984.
2. Flanigan, R. C.. Kursch, E. D., McDougal, W. S. and Persky, L.:
Synergistic gangrene of the scrotum and penis secondary to
colorectal disease. J. Urol., 1 1 9 369,1978.
3. Curreri. P. W. and Luterman, A,: Bums. In: Principles of Sur-
gery, 5th ed. Edited by S. I. Schwartz, G. T. Shires and F. C.
Spencer. New York: McGraw-Hill Book Co., chapt. 7, pp. 285-
306,1989.
4. Knaus, W. A,, Draper, E. A., Wagner, D. P. and Zimmerman,
J. E.: APACHE 11: a seventy of disease classification. Crit.
Care Med., 1 3 818,1985.
5. Foumier, A. J.: Gangrene foudroyante de la verge. Semaine
Med., 3 345,1883.
6. Clayton, M. D.. Fowler, J. E., Jr., Sharifi, R. and Pearl, R. K.:
Causes, presentation and survival of fifty-seven patients with
necrotizing fasciitis of the male genitalia. Surg., Gynec. &
Obst., 170:49,1990.
7. Jones, R. B., Hirschmann, J. V., Brown, G. S. and Tremann,
J. A.: Fournier’s syndrome: necrotizing subcutaneous infection
of the male genitalia. J . Urol., 122: 279, 1979.
8. Paty, R. and Smith, A. D.: Gangrene and Fournier’s gangrene.
Uml. Clin. N. Amer., 1 9 149, 1992.
9. McRea, L. E.: Fulminating gangrene of the penis: report of a
case. Clinics, 4: 796,1945.
10. Baskin, L. S.. Carroll, P. R., Cattolica, E. V. and McAninch,
J. W.: Necrotising soR tissue infections of the perineum and
genitalia: bacteriology, treatment, and risk assessment. Brit.
J. Urol., 65:524, 1990.
11. Enriquez, J. M., Moreno, S., Devesa, M., Morales, V., Platas, A.
and Vicente, E.: Fournier‘s syndrome of urogenital and ano-
rectal origin. A retrospective and comparative study. Dis. Co-
lon Rectum, 30:33, 1987.
12. Dunaif, C. B.: Fournier’s gangrene. Report of a case and review
of the literature. Plast. Reconstr. Surg., 3 3 84,1964.
13. Knaus, W. A,, Zimmerman, J. E., Wagner, D. P., Draper, E. A.
and Lawrence, D. E.: APACHE-acute physiology and chronic
health evaluation: a physiologically based classification sys-
tem. Crit. Care Med., 9 591,1981.
14. Broughan, T. A., Hermann, R. E., Hardesty, I. J. and Paranandi,
L.: Fascia1 closure in the management of infected pancreatic
necrosis. Amer. Surg., 60:309,1994.
15. Roumen, R. M., Schers, T. J., de Boer, H. H. and Goris, R. J.:
Scoring systems for predicting outcome in acute hemorrhagic
necrotizing pancreatitis. Eur. J. Surg., 158 167,1992.
16. Rattner, D. W., Legermate, D. A,, Lee, M. J., Mueller, P. R. and
Warshaw. A. L.: Early surgical debridement of symptomatic
pancreatic necrosis is beneficial irrespective of infection.
h e r . J. Surg., 163. 105,1992.
17. Kreuzer, E.,KZiab, S., Pilz. G. and Werdan, K.: Early prediction
of septic complications aRer cardiac surgery by APACHE II
score. Eur. J. Cardiothorac. Surg., 6 524, 1992.
18. Fugger, R., Zadrobilek, E., Gotzinger, P., Klimann, S., Rogy, M.,
Winkler, S., Andel, H., Mittelbock, M., Roth, E. and Schulz, F.:
Penoperative TNF alpha and IL-6 concentrations correlate
with septic state, organ function, and APACHE I1 scores in
intra-abdominal infection. Eur. J. Surg., 159: 525, 1993.
19. Vespasiano, M. C., Lewandowski, J. R. and Zimmerman, J. J.:
Longitudinal analysis of neutrophil superoxide anion genera-
tion in patients with septic shock. Crit. Care Med., 21: 666,
1993.
20. Hwang, T. L.,Huang, S. L. and Chen, M. F.: The use of indirect
calorimetry in critically ill patients-the relationship of meas-
ured energy expenditure to injury severity score, septic sever-
ity score, and APACHE I1 score. J. Trauma, 34: 247, 1993.
21. Tuchschmidt, J. A. and Mecher, C. E.: Predictors of outcome
from critical illness. Shock and cardiopulmonary resuscitation.
Crit. Care Clin., 1 0 179,1994.
22. Arregui, L. M., Moyes, D. G., Lipman, J. and Fatti, L. P.: Com-
parison of disease severity scoring systems in septic shock.
Crit. Care Med., 1 9 1165,1991.
23. Ridley, S.,Biggam, M. and Stone, P.: A cost-benefit analysis of
intensive therapy. Anaesthesia, 4 8 14, 1993.