Prognostic factors in Fournier gangrene

Summary

Background

Fournier gangrene is a necrotizing fasciitis, arising in the genital and perineal area. This entity is still associated with a high mortality rate despite improvements in antibiotic and surgical treatment.

Methods

This is a retrospective study of all the patients diagnosed and surgically treated for Fournier gangrene at General University Hospital Ramon y Cajal between 1988 and 2008. Possible prognostic factors that could have any influence on the evolution of Fournier gangrene were analyzed.

Results

Seventy patients were analyzed, 62 males (88.6%) and 8 females (11.4%) with a mean age of 57.9 ± 13.5 years. Most frequent clinical manifestations were perineal pain (82.9%) and fever (60%). Physical examination revealed edema (91.4%), erythema (88.6%) and perineal skin necrosis (60%). All the patients underwent surgical debridement of necrotic tissue. In 54.3% reoperations were necessary for new surgical debridements. Medical complications rate was 27.1% and mortality one 22.9%. Ethylism, coexistence of neoplasms, presence of skin necrosis, myonecrosis, abdominal wall affection, number of debrided areas, reoperations, concentration of creatinine in serum>1.4 mg/dL, and hemoglobin <10 g/dL, and platelet count <150 × 10⁹/L in whole blood are associated with higher mortality rates.

Conclusion

Identification of prognostic factors may help to determine high-risk patients in order to establish an optimal treatment, according to severity of the infection and general status.

Keywords

Fournier gangrene;outcome;prognostic factor

1. Introduction

Fournier gangrene is a form of necrotizing fasciitis, arising in the genital and perineal area, that may extend through fascial layer to the groin, thigh and even abdominal wall. Infection leads to thrombosis of subcutaneous blood vessels, which is the main cause of the overlaying skin necrosis. Fournier gangrene begins with local tenderness, edema, and erythema of the scrotal and perianal skin. This progresses to necrosis of the scrotal fascia. The scrotum enlarges to several times its normal diameter. If the process continues beyond the penile-scrotal region to the abdomen or the upper legs, the normal picture of necrotizing fasciitis can be seen. Necrotizing fasciitis is defined as a progressive, rapidly spreading, inflammatory infection located in the deep fascia, with secondary necrosis of the subcutaneous tissue.^{1 ;  2}

Diverse microorganisms (Gram-positive and Gram-negative bacteria, anaerobic bacteria, and fungi) are involved in the infection. Actually, this entity is still associated with a high mortality rate despite the improvement of antibiotic and surgical treatment.^{1 ;  3}

Diverse factors have been established as poor outcome predictors, such as advanced age, comorbidities (diabetes mellitus, alcoholism, cachexia), delayed diagnosis and treatment, extension of the gangrene developing a necrotizing fasciitis, reoperations and some laboratory data at diagnosis (low hematocrit and albumin concentration, and high creatinine and alkaline phosphatase).^{2 ;  3} The aim of this study is to analyze possible prognostic factors that could influence on the evolution of Fournier gangrene, reviewing our experience in the management of this entity during the last 20 years.

2. Patients and methods

This is a retrospective study of all the patients diagnosed of Fournier gangrene and undergoing surgical treatment at University Hospital Ramon y Cajal, Madrid, Spain from 1988 to 2008. Age, gender, personal history, clinical manifestations and exploratory findings, laboratory data, microbiological cultures, surgical procedure, infection origin, depth affection, reoperations, debrided anatomical areas, postoperative complications, mortality, hospital stay, and follow-up were investigated.

Inclusion criteria for the diagnosis of Fournier gangrene were the appearance of local tenderness, edema, and erythema arising initially in the genital or perianal area, independently of the posterior progression of the infection to the abdomen or upper legs, developing a necrotizing fasciitis. Exclusion criteria were the development of these signs and symptoms in the genital and perianal area as continuation of a primary infectious process located in another site.

Statistic analysis was performed with SPSS 14.0 for Windows. Quantitative variables following a normal distribution were defined by mean, standard deviation, and range. In non-Gaussian variables, median was used instead of mean. Qualitative variables were defined by number of cases and percentage. Quantitative and qualitative variables were compared with the Student t method (or Mann-Whitney method when quantitative variables did not follow Gaussian distribution). Qualitative variables were compared with Chi-square method. A p-value < 0.05 was considered significant.

3. Results

Seventy patients were analyzed, 62 males (88.6%) and 8 females (11.4%) with a mean age of 57.9 ± 13.5 years (range 22–84 years). Relevant personal history data are summarized in Table 1.

Table 1. Personal history.

	Global	Survivors	Nonsurvivors	p
	n (%)	n (%)	n (%)
Diabetes mellitus	39 (55.7)	34 (63)	5 (40)	NS
Hematological disorders	10 (14.3)	6 (11.1)	4 (25)	NS
Hepatopathy	8 (11.4)	4 (7.4)	4 (25)	NS
Ethilism	6 (8.6)	2 (3.7)	4 (25)	0.045
Neoplasms	6 (8.6)	2 (3.7)	4 (25)	0.045
Recent surgery	4 (5.7)	4 (7.4)	0	NS
Morbid obesity	4 (5.7)	2 (7.4)	2 (12.5)	NS

NS = nonsignificant.

Diabetes mellitus coexisted in 39 patients (55.7%). Haematological disorders were AIDS in five patients (7.1%), chronic myeloid leukemia in one (1.4%), and pancytopenia secondary to myelodysplastic syndrome in one. Eight patients (11.4%) presented hepatic disorders, five (7.1%) of them secondary to hepatitis C and three (4.3%) to alcoholism. Two patients (2.8%) presented prostatic neoplasms; in the other four with previous diagnosis of cancer, these were located in larynx, lung, right colon and bladder. Two subjects (2.8%) presented a recent surgical antecedent: prostatic adenomectomy and transurethral resection of a bladder tumor.

The most frequent clinical manifestations were perineal pain (n = 58, 82.9%) and fever (n = 42, 60%). Median evolution time of the symptoms up to medical examination was 7 days (range 2–30 days). Physical examination revealed edema (n = 64, 91.4%), erythema (n = 62, 8.6%) and perineal skin necrosis (n = 42, 60%); regional crepitation was discovered in 22 patients (31.4%).

Laboratory data are described in Table 2. The origin of the gangrene was located in the perineal area in 64 patients (91.4%) and in the urological tract in 6 (8.6%). Depth of infection was limited to skin and subcutaneous tissue in 40 patients (57.1%), in 24 patients (34.3%) the fascia was overcome, developing a necrotizing fasciitis, and in six (8.6%) muscle necrosis (myonecrosis) was evident. All the patients underwent surgical debridements of necrotic tissue. In eight patients (11.4%) fecal diversion had proceeded, in six because of sphincterian myonecrosis, and two cases due to iatrogenic rectal perforation. Debrided areas are explained in Table 3. In 38 patients (54.3%) reoperations were necessary for new surgical debridements; in four patients (5.6%) two reoperations were performed, one patient underwent three and another one required six. Indications for new surgical debridement were bad aspect of the wound with purulent segregation or necrotic tissue, progression to necrotizing fasciitis and septic status. In five patients with septic status and wound and perilesional tissues with apparently good evolution, a pelvic computed tomography scan was performed to evaluate deep tissue affection, revealing a pelvic abscess in two cases.

Table 2. Laboratory data.

	Global	Survivors	Nonsurvivors	p
White blood cells ( /mm3)	16428 ± 8634	17478 ± 8880	10915 ± 4734	NS
Hemoglobin (g/dl)	11.9 ± 2.4	12.5 ± 2.2	9.2 ± 1.5	0.01
Platelets ( /μl)	2277330 ± 149156	271667 ± 146553	95933 ± 41321	0.007
Creatinine (mg/dl)	1.21 ± 0.45	1.1 ± 0.3	2.1 ± 0.7	0.001
Urea (mg/dl)	61.4 ± 17.5	63.7 ± 19.2	52 ± 11.4	NS
BUN (mg/100 ml)	26.6 ± 7.8	27.4 ± 9.4	24.2 ± 8.1	NS
GOT (U/L)	16.7 ± 11.6	10 ± 1	30 ± 4.2	NS
GPT (U/L)	11.7 ± 7.4	7.5 ± 2.1	20 ± 5.4	NS
Bilirubin (mg/dl)	0.7 ± 0.5	0.5 ± 0.4	0.8 ± 0.2	NS
Alkaline phosphatase (U/L)	217.5 ± 236.9	285 ± 202.1	150 ± 240.2	NS
Glucose (mg/dl)	273.3 ± 124.7	279.4 ± 128.9	217.5 ± 78.5	NS
Prothrombin time (seconds)	17.3 ± 9.2	19 ± 10.4	12.1 ± 6	NS
Cephalin time (seconds)	39.8 ± 12	38.8 ± 14.8	41.5 ± 7.8	NS
Fibrinogen (mg/dl)	848.8 ± 251.1	883.2 ± 210.2	688.2 ± 412.3	NS

Results are defined in mean ± standard deviation.

NS = nonsignificant.

Table 3. Debrided areas.

	Global	Survivors	Nonsurvivors	p
	n (%)	n (%)	n (%)
Perianal	52 (74.3)	38 (70.4)	14 (87.5)	NS
Scrotum	46 (65.7)	32 (59.3)	14 (87.5)	NS
Perineum	42 (60)	30 (55.6)	12 (75)	NS
Groin	28 (40)	20 (37)	8 (50)	NS
Thigh	14 (20)	10 (18.5)	4 (25)	NS
Abdominal wall	10 (14.3)	4 (7.4)	6 (37.5)	0.033
Lumbar region	2 (2.9)	2 (3.7)	0	NS

Number of debrided regions: median 3 (range 1-6).

NS = nonsignificant.

After surgery, wide-spectrum antibiotic treatment was started in all the patients. The drugs employed have changed over the years; in the first decade of the studied period, amoxicillin/clavulanic acid was mostly used (1 g/8 h), but in the last decade imipenem and meropenem (1 g/8 h) were widely employed. Patients presenting allergies to beta-lactamic agents were mostly treated with gentamycin 240 mg/24 h and metronidazole (500 mg/8 h).

Medical complications were seen in 19 patients (27.1%): 14 nosocomial pneumonias, three central line catheter infections, and two urinary infections; all were satisfactorily managed with antibiotic treatment. Mortality rate was 22.9% (16 patients). All the deceased cases were caused by septic shock and multiorganic failure, related to an uncontrolled extension of the necrotizing fasciitis despite correct surgical debridements. Dividing the sample into 2 groups depending on the date of diagnosis (1988–1998 and 1999–2008), there were no significant differences among mortality (23.1% vs. 22.8%).

During the follow-up, 5 patients (7.1%) were diagnosed with anorectal tumors at 1, 2, 4, 6, and 8 months after Fournier gangrene.

Evaluating prognostic factors, it was observed that alcoholism, coexistence of neoplasms, presence of skin necrosis, myonecrosis, abdominal wall affection, number of debrided areas, reoperations, serum creatinin, hemoglobin levels, and platelet count were associated with higher mortality. Cut points were established for significant laboratory data in creatinine > 1.4 mg/dL, hemoglobin < 10 g/dL, and platelets < 150 × 10⁹/L. The results of the statistical analysis are in Table 4.

Table 4. Factors associated to mortality.

	Odds ratio (95%CI)	p
Ethylism	8.7 (1.6 – 112)	0.045
Neoplasms	8.7 (1.6 – 112)	0.045
Skin necrosis	1.4 (1.06 – 1.95)	0.05
Mionecrosis	9.6 (1.12 – 98.7)	0.04
Abdominal wall affection	7.5 (1.2 – 57.2)	0.033
Number of debrided areas > 4	8.7 (1.7 – 102.3)	0.043
Need for reoperations	1.7 (1.2 – 2.5)	0.004
Creatinine > 1.4 mg/dL	2 (1.3 – 5.3)	0.009
Hemoglobin < 10 g/dL	9.6 (1.3 – 73)	0.008
Platelets < 150 × 109/L	6 ((1.4 – 44.9)	0.018

4. Discussion

Fournier gangrene is a surgical emergency with a high mortality rate despite adequate treatment.³Although antimicrobial drugs have undergone important advances in recent decades, with the appearance of new and stronger agents with wide spectrum, the main treatment of Fournier gangrene remains surgical debridement. In our series we have observed that the mortality rate was similar among the patients diagnosed in the first decade to those diagnosed in the second, despite the use of newer drugs in the last 10 years of the study. In some recent publications, mortality associated with this pathology is estimated around 20% to 35%.^{2 ;  4} Similar to other entities, better results are related to an early diagnosis and adequate surgical treatment.^{1; 3 ;  4} Unfortunately, early diagnosis does not depend on physicians in most cases, thus the patients consult usually in advanced phases, leading these to rapid progression of the infection and development of a necrotizing fasciitis. Median evolution time up to medical evaluation in our series was 7 days. Notwithstanding, identification of prognostic factors could have some influence on survival, helping the surgeon to optimize the treatment to the personal situation of each patient.

Older patients and those with comorbidities, such as diabetes mellitus, alcoholism or cachexia, are considered to present worse prognosis. Similarly, patients with a longer period up to the treatment, with wider extension of the infection or those requiring a colostomy are associated with worse outcome. A low socioeconomical level has also been described as a poor prognostic factor.^{2 ;  6}

In our series we could not demonstrate advanced age to be associated with higher mortality, the mean age of our patients being 57.9 years, similar to those described in other recent studies.³ Diabetes mellitus is the most frequent comorbidity among our patients (55.7%), but its association with higher mortality could not be demonstrated, as described by other authors.⁷ In the same way, glucose levels at diagnosis in our patients were not associated with higher mortality.

Referring to alcoholism and coexistence of neoplasms, we found an association with higher mortality rate (odds ratio [OR] 8.7, 95% confidence interval [95%CI] 1.6-112) in both groups, as described by other authors.⁸

The extension of the infection is one of the most important prognostic factors in Fournier gangrene.^{4 ;  9} In our study, we have divided the affected surface into areas, as explained in Table 3. When 4 or more regions are affected, survival is significantly worse (OR 8.7, 95%CI 1.7–102.3). Reoperations are also associated with the number of affected areas (p = 0.011). The extension of the infection to the abdominal wall has been a bad prognostic factor in our study, independently of the number of affected areas (OR 7.5, 95%CI 1.2–57.2). Myonecrosis was also associated with higher mortality (OR 9.6, 95%CI 1.12–98.7), as described in the literature. ⁵ Moreover, we have observed that patients with skin necrosis at the time of diagnosis presented worse outcome (OR 1.4 95%CI 1.06 – 1.95).

Those patients undergoing reoperations presented a higher mortality risk, because these are usually necessary when infections progress and develop necrotizing fasciitis despite initial debridements, revealing extremely aggressive features.¹⁰ Other authors have reported that there is no relationship between the number of reoperations and mortality rate,¹¹ in contrast to what occurred in our patients.

Referring to laboratory data, low hematocrit and albumin concentration, and increased creatinine and alkaline phosphatase have been associated with a higher mortality rate. In our series we have established cut points; when creatinine values are over 1.4 mg/dL, mortality is twice (95%CI 1.3–5.3), hemoglobin levels lower than 10 g/dL present a risk 9.6-fold higher (95%CI 1.3–73) and when platelets count is <150 × 10⁹/L the risk of decease is 6-fold higher (95%CI 1.4–44.9). Increased creatinine levels reflect renal dysfunction, probably related to septic shock and may be the initial stage of a multiorgan failure. Reduced hemoglobin levels show the worsening of the general status, also related with sepsis originated from Fournier gangrene; on the other hand it facilitates the appearance of postoperative complications. Platelet count has not been described as a bad prognostic factor, but it is probably associated with the reduction of hemoglobin levels, also related with the initial septic status.^{2; 12 ;  13}

In the literature, the best prognostic tool to evaluate the severity of Fournier gangrene is the Fournier Gangrene Severity Index Score, which assigns a numerical score describing the acuity of the disease. This index describes patients' vital signs (temperature, heart and respiratory rates) and metabolic parameters (sodium, potassium, creatinine, and bicarbonate concentrations, hematocrit, white blood cell count) and computes a score relating to the severity of the disease at that time.⁴ One of the main limitations of our study is that we could not apply this Index, because our study is a long-term retrospective series and unfortunately some data were not available in the clinical history, mainly nonrecorded patient vital signs and some analytical parameters.

Based on the results of this study, we suggest that when facing patients with risk factors, a more aggressive treatment (surgical and antibiotic) should be started from the beginning and a careful follow-up (frequent clinical examination of the surgical site, blood analysis and eventual imaging tests when considered necessary) during the postoperative period must be performed, in order to achieve early detection of any possible complication.

5. Conclusion

Fournier gangrene is a surgical emergency, associated with a high mortality rate despite adequate surgical debridement and wide spectrum antibiotherapy. Identification of prognostic factors may help to determine patients with higher mortality risk and to establish an optimal treatment, according to severity of the infection and general status. We have established alcoholism, coexistence of neoplasms, presence of skin necrosis, mionecrosis, gangrene extension, abdominal wall affection, reoperations and laboratory data (creatinine > 1.4 mg/dL, hemoglobin < 10 g/dL, platelets < 150 × 10⁹/L) as prognostic factors associated with higher mortality.

References

1. 1 J. Arias Diaz, X. Guirao Garriga; Infecciones graves de los tejidos blandos; Infecciones Quirúrgicas. Guías Clínicas De La Asociación Española De Cirujanos, Aran, Madrid (2006), pp. 295–315
2. 2 H.J. Jeong, S.C. Park, I.Y. Seo, J.S. Rim; Prognostic factors in Fournier gangrene; Int J Urol, 12 (2005), pp. 1041–1044
3. 3 J. Medina Polo, A. Tejido Sánchez, F. de la Rosa Kehrmann, N. Felip Santamaría, M. Blanco Álvarez, O. Leiva Galvis; Gangrena de Fournier: estudio de los factores pronósticos en 90 pacientes; Actas Urol Esp, 32 (2008), pp. 1024–1030
4. 4 A. Tuncel, O. Aydin, U. Tekdogan, V. Nalcacioglu, Y. Capar, A. Atan; Fourniers gangrene: Three years of experience with 20 patients and validity of the Fourniers gangrene severity index score; Eur Urol, 50 (2006), pp. 838–843
5. 5 J.M. Enriquez, S. Moreno, M. Devesa, V. Morales, A. Platas, E. Vicente; Fourniers syndrome of urogenital and anorectal origin. A retrospective, comparative study; Dis Colon Rectum, 30 (1987), pp. 33–37
6. 6 N. Eke; Fourniers gangrene: a review of 1726 cases; Br J Surg, 87 (2000), pp. 718–728
7. 7 A.A. Nisbet, I.M. Thompson; Impact of diabetes mellitus on the presentation and outcomes of Fourniers gangrene; Urology, 60 (2002), pp. 775–779
8. 8 M. Basoglu, I. Ozbey, S.S. Atamanalp, et al.; Management of Fourniers gangrene: review of 45 cases; Surg Today, 37 (2007), pp. 558–563
9. 9 L.S. Palmer, H.I. Winter, B.M. Tolia, R.E. Reid, E. Laor; The limited impact of involved surface area and surgical debridement on survival in Fourniers gangrene; Br J Urol, 76 (1995), pp. 208–212
10. 10 H.R. Unalp, E. Kamer, H. Derici, et al.; Fourniers gangrene: evaluation of 68 patients and analysis of prognostic variables; J Postgraduate Med, 54 (2008), pp. 102–105
11. 11 E. Laor, L.S. Palmer, B.M. Tolia, R.E. Reid, H.I. Winter; Outcome prediction in patients with Fourniers gangrene; J Urol, 154 (1995), pp. 89–92
12. 12 J.N. Olsofka, E.H. Carrillo, D.A. Spain, H.C. Polk Jr.; The continuing challenge of Fourniers gangrene in the 1990s; Am Surg, 65 (1999), pp. 1156–1159
13. 13 K. Moorthy, P.P. Rao, A.N. Supe; Necrotising perineal infection: a fatal outcome of ischiorectal fossa abscesses; J R Coll Surg Edinb, 45 (2000), pp. 281–284