Necrotising perineal infection

Introduction: Necrotizing perineal infection or Fournier's gangrene is an uncommon but lethal complication of ischiorectal fossa abscesses. It is associated with a high mortality, especially in diabetics and immunocompromised individuals. Attempts have been made to study factors which could serve as prognostic indicators. The role of faecal diversion has not been clearly determined. Materials and methods:The medical records of 8 patients who presented with necrotizing perineal infection as a complication of ischiorectal fossa abscesses were reviewed. Various parameters were studied to see if any of them could serve as predictors of outcome. Mean surface area of involvement was calculated using modified burns assessment criteria. To study the effect of colostomy on the general condition of the patient the physiological and biochemical parameters before and after the procedure were compared . Statistical analysis was done using the unpaired and paired 't' tests. Results: The mean age of the patients was 50.6 ± 10.3 years. Five patients were diabetic, of whom four died; all the non-diabetics survived. The mean surface area of involvement was 5.1 ± 0.75%, among the survivors, and 9.6 ± 3.4% among the non-survivors. Colostomy was performed in four patients one of whom died. While in one patient the colostomy was created along with the initial radical debridement, in three other patients it was formed on days three, five and five, respectively. There was a significant improvement in their general status and biological parameters. All patients with testicular involvement died. Conclusion: Evidence of systemic sepsis at presentation, extent of tissue and testicular involvement, a low haematocrit, a high blood urea and creatinine and a low serum albumin, were associated with a higher mortality. Prompt recognition of the condition, urgent radical surgical debridement and the use of appropriate antibiotics are the mainstays of management. Formation of a diverting colostomy appears to favour survival.

Keywords: abscess, Fournier's gangrene, ischiorectal fossa, necrotising perineal infection

Necrotizing perineal infection or Fournier's gangrene is an uncommon complication of ischiorectal fossa abscesses. Though this condition was first described as idiopathic scrotal gangrene, a cause can be found in about 95% of the cases.^{1, 2}

The commonest source of sepsis is the anorectum.³ The condition is more prevalent in the older age group with the age of incidence being greater in those over 50 years of age.^4,5 Ischiorectal abscesses, perineal abscesses and inter-sphincteric abscesses have all been described as sources of the infection^6-8. Fournier's gangrene is typically described as a synergistic necrotizing fasciitis of the peri-rectal, perineal and genital areas which is characterised by obliterative endarteritis of the subcutaneous arteries resulting in gangrene of the subcutaneous tissue and the overlying skin. This tissue damage may extend to the penis, anterior abdominal wall, buttocks or thighs.^5,9 Tissue ischaemia lowers the local oxygen tension favouring the proliferation of anaerobic organisms. Anaerobic metabolism results in the accumulation of hydrogen and nitrogen, which gives the characteristic crepitus on examination.¹⁰ The mortality rate is valuable and reported to be between 7-75%.² Mortality is much higher amongst patients who have diabetes. In one series, the mortality was 80% amongst diabetics while amongst nondiabetics it was 30-35%.¹⁰ Mortality is also higher when the source of infection is the anorectum rather than the urethra or skin¹¹although there have been attempts to study factors that may determine prognosis and survival, there are still no reliable prognostic indicators. Extent of involvement, number of debridements, age of the patient, presence of systemic sepsis, physiological state at time of presentation, source of infection, bacteriological load in the wound, have all been studied as determinants of outcome.^4,9,12,13 Diversion of urine by a suprapubic cystostomy and of faeces by a colostomy are controversial issues with no consensus concerning their use.^3,4,13

The medical records of eight patients, who presented with necrotizing perineal infection as a complication of an ischiorectal fossa abscess, were reviewed. Their main presenting features and the presence of diabetes mellitus or other co-morbid conditions were noted. The extent of involvement was calculated by using a modified burns assessment criteria.^4,9 The penis, scrotum and perineum each constituted 1% of the surface area and each ischiorectal fossa was 2.5%. Involvement of the abdominal wall or thigh was calculated by using the 'rule of 9'. The general condition of the patient was noted with special emphasis on the mental status, temperature and blood pressure. The patients were described as having systemic sepsis if their temperature was greater than 38°C (100.4° F), heart rate greater than 90 beats per minute, respiratory rate greater than 20 breaths per minute and the white blood cell count (WBC) greater than 12,000 per cubic mm, along with a positive blood culture.

Special note was made of any change in these parameters in the survivors after a colostomy was fashioned. The bacteriological profile of the wounds was noted. Both aerobic and anaerobic cultures of wound swabs were carried out. All patients underwent optimum radical debridements on an urgent basis. The antibiotics administered, the need for blood, platelet and albumin/fresh frozen plasma transfusions were noted. The extent of involvement, the presence of diabetes mellitus, laboratory parameters such as haemoglobin, urea, serum creatinine, WBC and serum albumin, were studied to see if any could serve as predictors of outcome. Statistical analyses were undertaken using the unpaired and paired 't' tests.

Eight male patients presented with perineal necrotizing infection as a complication of an ischiorectal fossa abscess. The mean age of the patients was 50.6 ± 10.3 (SD) years (range: 38-62 years). The mean age of the survivors was 45.2 ± 11.3 years while the mean age of the non-survivors was 56 ± 5.6 years. The ischiorectal abscess was right-sided in seven cases and left-sided in one case. Five out of the eight patients were diabetics. Local perineal pain and fever were present in all patients. Four patients had been treated previously in other hospitals, prior to being referred. Three patients had undergone incision and drainage of the ischiorectal fossa abscesses, while one had received only antibiotics. Five patients had evidence of systemic sepsis at presentation. Local examination revealed bronze discolouration, erythema, blackening and bullous formation of the skin. Crepitus of the scrotum was present in all patients who subsequently showed scrotal involvement during surgery. In three patients, skin discolouration and induration in the ischiorectal fossa prompted an aspiration, which revealed pus.

Using a modified burn assessment criteria the extent of involvement was calculated. The surface area involved varied from 4.5% to 11.5% with a mean of 7.1 ± 3.1%. The mean surface area of involvement amongst the survivors was 5.1 ±0.75% in the non-survivors was 9.6 ± 3.4% (p< 0.05).

All the patients had a WBC >12,000cmm at presentation. The liver function tests were abnormal in two patients, both of whom did not survive. The coagulation profile was abnormal in one patient diagnosed as having disseminated intravascular coagulation; this patient died soon after admission. The nonsurvivors had significantly lower levels of haemoglobin and albumin and higher levels of blood urea and serum creatinine (Table 1). Bacteriological profile of the wounds revealed Escherichia (E.Coli) alone in two patients and a mixed bacterial flora in six patients (E.Coli and Klebsiella or Pseudomonas). Culture for anaerobes was done in only six patients and was positive in five out of these six patients.

Table 1: Blood parameters in survivors and non-surviving patients with necrotising fasciitis

	Mean ± SD	Survivors Mean ± SD	Non-survivors Mean ± SD	P Value (<)
Haemoglobin (g/dl)	9.2 ± 1.6	10.6± 0.4	7.8 ± 1.0	0.005
Serum Albumin (g/dl)	2.3 ± 0.8	3.0 ± 0.8	1.6 ± 0.4	0.013
Blood Urea Nitrogen (mg/dl)	46.8± 0.9	34.5 ± 10.1	59.0 ± 13.9	0.015
Serum Creatinine (mg/dl)	2.5 ± 0.9	1.9 ± 0.7	3.1 ± 0.5	0.04

Five patients underwent radical surgical debridement with excision of all necrotic material and tissue of doubtful viability. Two patients underwent incision and drainage of the ischiorectal fossa abscesses initially. Subsequent examinations showed spreading fasciitis for which they underwent radical debridement on the 4th and 5th day after admission, respectively. One patient underwent deroofing and debridement of an already burst abscess and on detection of spread into the perineum and scrotum, underwent radical debridement on the 5th day.

Orchidectomy was performed for gangrenous testes in three patients. Two patients underwent unilateral orchidectomy on the same side as the ischiorectal abscess while one patient underwent bilateral orchidectomy.

All patients received a third generation cephalosporin with an aminoglycoside and metronidazole, with dose adjustment according to renal status. Six patients received blood transfusions. Four patients received blood due to serious blood loss during debridement. One patient received a blood transfusion following serious bleeding from his wound, which required re-operation. Complications occurred in two patients; one patient succumbed to bronchopneumonia on day 17 and one patient had severe bleeding from the wound. Among the survivors the mean hospital stay was 27.5 ± 5.2 days. Colostomy was performed on four patients. In one patient, it was performed along with the initial radical debridement. In three other patients colostomy was performed on days 3, 5 and 5, respectively. There was a documented change in the general status of the patients with a fall in temperature, rise in blood pressure, decrease in purulent discharge from the wound and an improvement in their biochemical parameters measured two days after the procedure (Table 2).

Table 2 Changes in blood parameters before and following creation of a colostomy in patients with necrotising faciitis

	Pre-colostomy	Post-colostomy	P Value (<)
WBC Count (per cubic mm)	20,600 ± 3401	8500 ± 577	0.05
Blood Urea Nitrogen (mg/dl)	54.2 ± 15.0	20.2 ± 8.6	0.05
Serum Creatinine (mg/dl)	2.1 ± 0.7	0.7 ± 0.3	0.05

Four patients died, all of whom were diabetics. Three patients died from multi-organ failure soon after admission while one patient died of pneumonia 17 days after admission.

Ischiorectal fossa abscesses are rarely life threatening though there have been previous reports of the condition resulting in a fatal outcome.^6,14 The mechanism of spread from the anorectal area to the perineum and genitalia and, subsequent further spread, remains unclear. It has been suggested that the attachment of the Colles' fascia to the undersurface of the urogenital diaphragm is fenestrated. Thus, infection in the perirectal area penetrates the urogenital diaphragm through these fenestrae.^2,6 Once the fascia is infected, spread along its attachments to Buck's fascia, Scarpa's fascia and into the thigh is inevitable.

It has been noted that patients with a colorectal source of infection have a longer hospital stay, require more operative procedures and have a more complex bacteriological spectrum. Diabetic patients more commonly have extensive rather than localised infection.¹³ In the diabetic, the condition has a fulminant and fatal course because of the inherent neutrophil dysfunction with decreased phagocytic and intra-cellular bactericidal activity.¹⁵

As this necrotising condition is fatal, there have been attempts to identify factors that may serve as prognostic indicators. Extent of infection has been one of the variables said to determine prognosis. Clayton et al (1990) found no survival advantage in patients with localised compared with extensive infection.¹³ In two other series the authors found that those with less than 5% surface area of involvement, calculated using the modified burns assessment criteria described in this article, had a higher likelihood of survival.^4,9 In our series the mean surface area of involvement amongst the survivors was 5.1± 0.75% while amongst the non- survivors was 9.6 ± 3.4%.

In this series, non-survivors had significantly lower levels of haemoglobin and serum albumin and higher levels of blood urea and serum creatinine. This has been documented previously.⁴ Anaemia is probably a result of intravascular thrombosis, coupled with decreased production due to sepsis.¹⁶ The importance of assessing the physiological state of the patient, rather than studying individual variables has been stressed.^4,9 A severity score, similar to one used in acute pancreatitis, has been proposed by Laor et al.⁴ This included temperature, pulse rate, respiratory rate, serum sodium, potassium, and creatinine levels, haematocrit estimation, serum bicarbonate and WBC count; these variables were used to calculate a score. The authors found that there was a 75% probability of death with a score greater than 9.

Although there is no consensus on the extent of initial debridement, we chose to radically debride the tissues once we suspected spreading necrosis. Underlying fasciitis facilitates the separation of the skin and the subcutaneous tissue from the fascia.¹⁷ Difficulty is experienced in defining viable from doubtful areas. It is imperative that the zone of erythema and induration invariably present at the edges of definitely necrotic areas is meticulously and frequently monitored. Any evidence of necrosis should prompt an urgent repeat debridement. Adequate surgical debridement results in prompt improvement in the patient's condition.¹⁸ If the initial debridement is adequate, later debridements involve only limited excision,⁷ especially at the edges of the wound and the slough that frequently develops once demarcation occurs.

The testis is usually spared from the infective process because its blood supply arises in the abdomen from the aorta. Testicular necrosis and gangrene results from testicular artery thrombosis following retroperitoneal spread of infection.³ Exploratory laparotomy and debridement of the retroperitoneal tissues, in the presence of testicular necrosis, is recommended.^3,10 All three patients in our series who underwent orchidectomy for testicular gangrene died. Two of them died soon after admission while one patient died of bronchopneumonia 17 days later. In retrospect, this patient could probably have been saved with a timely retroperitoneal exploration, as the retroperitoneal sepsis must have served as a nidus of infection.

Although split thickness skin grafts and flaps have been described in the reconstruction of the denuded area, none of the survivors in our series needed these procedures. It has been found that the remaining scrotal skin is capable of effective regeneration and that wounds in these areas can easily heal by secondary intention. ^6,8,10,19

Faecal diversion has been recommended in cases of necrotising perineal infection, especially where the source of infection is from the colon-rectum.^6,14,20 Some authors have recommended diversion only when: the sphincter is infected, there is demonstrable rectal perforation, if the rectal wound is large or if the patient is immunocompromised.³ Some authors suggest that a colostomy is virtually never needed, even if massive necrosis is present.⁷ In our series, four colostomies were performed. Among the patients in whom colostomies were formed, one patient died of bronchopneumonia. Diversion was performed in these patients because of the persistence of systemic sepsis in spite of optimal radical debridement.

As a result of our experience with these fulminant complications of ischiorectal fossa abscesses, we have adopted the following policy in their management; all diabetics and older patients with ischiorectal fossa abscesses are admitted; a prompt and detailed examination of the affected ischiorectal fossa, perineum and scrotum is carried out; emergency drainage and adequate deroofing of the abscess cavity is carried out by an experienced surgeon.10 In the post-operative period, these patients are administered antibiotics and undergo at least two repeat examinations of the wound. Clean wounds with healthy skin edges are reexamined on an outpatient basis. The presence of spreading perineal necrosis as a result of spread from an ischiorectal or perianal abscess should prompt emergency radical debridement with the use of a diverting colostomy. Colostomy is a relatively simple procedure, which should be performed along with the initial debridement in patients who have an anorectal source of infection, especially when they show evidence of systemic sepsis and in whom the infection is extensive. It should also be performed in patients who do not show any reversal of their septic state and in whom the wound shows evidence of persistent infection. This can be performed under general anaesthesia during subsequent debridements or even under local anaesthesia. Other than individual parameters, the physiological state of the patient is the most important factor in determining prognosis. Other determinants of prognosis are the extent of involvement and the presence of testicular necrosis.

Correspondence to : Mr. K. Moorthy, Specialist Registrar, Department of Surgery, Hinchingbrooke Hospital Huntingdon PE18 8NT, U.K. Email- moorthyk@hotmail.com

ORIGINAL ARTICLES

Necrotising perineal infection: a fatal outcome of ischiorectal fossa abscesses