home

Gas Producing Microbes in Diabetic Patients




Jamal Al-Wakeel, ABIM, FRCP(C); Saleh Rasheed Al-Ballaa, MACP, FRCP(C); Hani Sergani, MB, BS; Hassan Abu-Aisha, FRCP; Sameer Huraib, FACP, FRCP(C); Ahmed Mitwalli, FRCP(C), FACP, FRCP(Ed)

From the Division of Nephrology (Drs. Al-Wakeel, Abu-Aisha , Huraib and Mitwalli) and Infectious Diseases (Drs. Al-Ballaa and Sergani), Department of Medicine, King Khalid University Hospital, Riyadh.

Address reprint requests and correspondence to Dr. Al-Wakeel:
Associate Professor, Department of Medicine (38), King Khalid University Hospital, P.O. Box 2925, Riyadh 11461, Saudi Arabia.

Accepted for publication 22 June 1994.

The ability of microorganisms to produce gas in infected tissue has long been recognized. Although this was originally linked to Clostridium perferingens and other clostridial species,1 subsequent reports indicated that other bacteria including facultative coliforms, staphylococci, streptococci and other anaerobes are capable of causing gas-forming infections.2,3 These infections range from mild disease with no gangrene to extensive disease with widespread tissue destruction that may lead to death. Most gas-forming infections in diabetics occur in association with diabetic foot ulcers.3 Other sites are less frequently involved. We report gas-forming infection due to Escherichia coli in two diabetic patients with involvement of the knee joint in one patient and the vertebral body and epidural space in the other. To the best of our knowledge, involvement of these sites in diabetic patients were not previously described.

Case Reports

Case 1

A 70-year-old Saudi male with a history of noninsulin-dependent diabetes mellitus for six years and long-standing osteoarthritis of the right knee presented to the orthopedic clinic with painful swelling of the right knee of four days' duration and exacerbation of osteoarthritis was diagnosed. Therapeutic aspiration was performed and he was sent home on nonsteroidal anti-inflammatory drugs and cloxacillin. Next morning, he returned to the emergency room complaining of severe pain in the right knee, fever and shortness of breath. On examination, he was ill-looking and confused. Temperature was 39.8_C, blood pressure 100/60 mm/Hg, pulse rate 117/min and respiratory rate 26/min. The right knee was swollen, painful and warm. There was diffuse crepitus in the region of the right knee and the adjacent areas.

FIGURE 1. X-ray of the knee joint showing gas in the joint and soft tissues.

Aspiration of the knee produced foul-smelling pus. The Gram's stain showed many polymorphs and gram-negative rods. Laboratory findings were as follows: leukocyte count 9x109/L with normal differential, hemoglobin 91 g/L, platelets 66x109/L, prothrombin time 14.4 sec (control):12 sec, partial thromboplastin time 41 sec (control 26-40 sec) fibrinogen degradation products 400 mg/mL, serum urea 28.3 mmol/L, serum creatinine 133 mmol/L, pH 7.34, PO2 60 mm/Hg, PCO2 20 mm/Hg, HCO3 14 and serum lactic acid 3.8 mmol/L (N=5 to 20 mmol/L). Urine culture was negative. X-ray of the right knee confirmed the presence of gas in the soft tissues (Figure 1).

The patient was treated with parenteral ceftriaxone, cloxacillin and metronidazole and the joint was surgically drained. The blood culture and the aspirated pus later revealed pure growth of E. coli in both aerobic and anaerobic media. The antibiotics were changed to ceftriaxone and penicillin. The patient improved gradually and eventually made a full recovery.

Case 2

A 70-year-old Saudi male diabetic presented with a 25-day-history of low back pain and fever. The pain was progressive and it was aggravated by movement. The patient denied lower limb weakness, fecal or urinary incontinence. Physical examination on admission showed temperature of 39.3_C and blood pressure of 130/70 mm/Hg. There was local tenderness over the lower lumbar region. The ankle jerks were depressed bilaterally, muscle power was normal in the lower limbs and the plantar reflexes were downgoing bilaterally. Other system examinations were within normal limits.

Initial investigations revealed a leukocyte count of 25.0x109/L, platelet count of 90x109/L, hemoglobin of 95 g/L and a normal serum urea, creatinine and electrolytes; urine culture was negative. The patient was started on ceftriaxone and cloxacillin intravenously. Blood culture taken on admission grew E. coli sensitive to ceftriaxone in both aerobic and anaerobic bottles. On the second day of admission, the patient's general condition deteriorated and he developed paraplegia. Myelography and CT scan of the dorsal and lumbar spine showed extrinsic epidural collection with gas in and surrrounding the vertebral bodies extending from D12 to L5 (Figure 2). Aspiration under CT scan guidance revealed pus. The Gram's stain showed many polymorphs but no organisms. Metronidazole was added to cover for the possibility of anaerobes.

The patient was taken to surgery where L3-L4 laminectomy and evacuation of the epidural space and L4-L5 disc space abscesses were performed. Aerobic and anaerobic culture of the evacuated pus revealed pure growth of E. coli.

The patient's condition deteriorated further; he developed septic shock, disseminated intravascular coagulation and multisystem failure. He expired in the intensive care unit 16 days after initial presentation.

Discussion

Since the initial report of gas-forming infections in an infected gangrenous leg of a diabetic patient by Chiari in

FIGURE 2. CT of lumbar vertebrae showing destruction and gas in the vertebral body, paravertebral tissue and in the spinal canal.

1883,4 many reports appeared in the literature describing cases with this condition.1,3,5,6 These reports clearly illustrated that infections accompanied by gas in the soft tissue may be caused by a wide variety of microorganisms other than Clostridium perfringens and other clostridial species.

Spring and Khan5 reported three diabetic patients with non-clostridial gas-associated infections. E. coli was the causative agent in two and no organism was isolated in the third. In 1974, Van Beck et al.1 reported seven cases of non-clostridial gas-forming infection and reviewed the 72 cases that appeared in the English literature until then. Later, Bessman and Wagner3 reported another 48 cases. Bacteria sited in these two reports as possible agents of gas-forming infections included E. coli, Klebsiella, Pseudomonas, Streptococci, Staphylococci and Bacteriodes, either singularly or in combination. More recently, gas-forming infection was described in a patient with salmonella group infection.10

In the majority of cases, mixed organisms were implicated; however, approximately 20% of the cases were due to single organisms, particularly E. coli and Klebsiella.1-3,7-9

Most non-clostridial gas-forming infections were described in diabetic patients.3,4-6 The increased susceptibility of this group of patients to infection is well recognized. It is postulated that the presence of high tissue glucose level combined with impaired circulation favor the anaerobic metabolism by strict or facultative anaerobic bacteria resulting in gas production in the infected tissue.1-9

Although gas-forming infections in diabetic patients are most frequently associated with diabetic foot ulcers,3-6 other sites can be affected including abdominal wall, renal pelvis, gluteal muscles, thigh muscles, scrotum, labia majora and intra-abdominal, perirectal and anal regions.1-3,9 The involvement of the knee joint, the epidural space and the vertebral body were not described before. The source of the organism in our two patients is not clear; both had a negative urine culture. The possibility of iatrogenic infection in the first patient cannot be ruled out since the synovial fluid from the initial aspiration was not analyzed. Management of patients with such infection should be prompt and aggressive; adequate surgical debridement is a cornerstone of therapy.2 Initial antibiotic coverage should be guided by the Gram's stain result and the knowledge about the possible etiologic agents; but even with the institution of such measures, some patients will not respond and the mortality of this illness remains high. It ranges from 4% to 100% with higher mortality in patients with extensive disease.1,3,5,6 The first case described in this paper responded to treatment. The second had extensive disease at presentation and the patient died in spite of early surgical drainage and proper antibiotic coverage. The possibility of co-existing anaerobic organisms in both cases was entertained. Strict anaerobic organisms are difficult to grow unless special techniques are followed during transporting and culturing the specimens. For this reason, both patients continued to receive antibiotics covering anaerobes, even after the culture results were available.

In conclusion, we stress that gas-forming infections may be caused by a wide variety of organisms other than clostridial species and that prompt management including surgical drainage and appropriate antibiotic coverage is of paramount importance in order to improve the outcome of this potentially fatal condition.

Acknowledgment

The authors would like to thank Ching Seno for secretarial assistance.

References

1. Van Beck A, Zook E, Yaw P, Gardner R, Smith R, Glover JL. Non -clostridial gas-forming infections. Arch Surg 1974;108:552-7.

2. Weisenfeld LS, Luzzi A, Picciotti J. Non-clostridial gas gangrene. J Foot Surg 1990;29:141-6.

3. Bessman AN, Wagner W. Non-clostridial gas gangrene. JAMA 1975;233:958-63.

4. Chiari H. Zur Bacteriiologie des septischen emphysema. Prog Med Wuchenschr 1893;18:1-4.

5. Spring M, Khan S. Non-clostridial gas infection in the diabetic. Arch Intern Med 1951;88:373-7.

6. Wills MR, Reece MW. Non-clostridial gas infection in diabetes mellitus. Br Med J 1960;2:566-8.

7. Digioia RA, Kane JG, Parker RH. Crepitant cellulitis and myonecrosis caused by Klebsiella. JAMA 1977;237:2097-8.

8. Napgezek MR, Hall WH. Non-clostridial crepitant cellulitis due to Klebsiella. Minn Med 1975;58:377-8.

9. Bruno-Martha LA, Sedghiraziri MA, Arbeit RD. Crepitant myonecrosis caused by Klebsiella pneumoniae in an immuno-compromised diabetic patient. J Infect Dis 1990;160:1416-7.

10. Quale JM, Ianano F. Salmonella myonecrosia in a patient with diabetes mellitus. A J Med Sci 1991;301:335-6.



1