309 2. Abrutyn E (1991) Tetanus. In: Wilson JD, Braunwald E, Isselbacher KJ et al (eds) Harrison's principles of internal medicine, 12th edn. McGraw-Hill, New York, pp 577-579 3. Fleshner PR, Hunter JG, Rudnick J (1988) Tetanus after gastrointestinal surgery. Am J Gastroenterol 83:298-300 4. Myers KJ, Heppell J, Bode WE, Culp CE, Thurber DL, Van Scoy RE (1984) Tetanus after anorectal abscess. Mayo Clin Proc 59:429-430 5. Editorial (1984) Postoperative tetanus. Lancet II:964 Dr. S. Ruiz-Santana, Hospital del Pino, Unidad de Cuidados Intensivos, E-35005 Las Palmas de Gran Canaria, Spain

Gas gangrene and purulent pericarditis during Clostridium septicemia revealing a cecal carcinoma Dear Sir, Gas gangrene occurs infrequently nowadays and is generally due to infection by Clostridium perfringens. Infection due to Clostridium septicum is much more rare [1]. In this clinical setting, purulent pericarditis due to Clostridiurn septicum is an exceptional finding. We report the case of a patient with Clostridium septicurn pericarditis with gas gangrene leading to the diagnosis of a cecal malignancy. A 62-year-old non-insulin-dependent male diabetic complained of increasing fatigue, a 10-kg weight loss, and frequent episodes of nonbloody diarrhea of 3 months duration. He was hospitalized because of the onset of chest and abdominal pain and symptoms suggestive of a partial intestinal obstruction. Upon admission, the physical examination revealed the abdomen was diffusely tender without guarding, and the absence of bowel sounds. A pericardial friction rub was heard. The blood pressure was 120/80 mmHg and the patient was afebrile. Laboratory test revealed a microcytic anemia (hemoglobin 7.7 g/dl, MCV 62 la3), a normal white cell count, elevated platelets (771000/mm3), and an elevated CPK serum level (3 098 IU/1; normal < 230). The chest X-ray demonstrated enlargement of the heart shadow. The fiat plate of the abdomen revealed no fluid levels or other abnormalities. The electrocardiogram demonstrated ST-segment elevation in all leads without diminution in voltage. Shortly following admission, the patient had a sudden cardiac arrest which responded to resuscitation. The state of shock persisted. Echocardiography revealed a small pericardial effusion. Extensive gas gangrene of the left flank rapidly appeared, prompting an emergency laparotomy on account of the initial digestive tract symptomatology. Inside the abdominal cavity, there were no signs of peritonitis, no abscess, tumor. The cecum appeared normal. A wide excision of the necrotic tissue was performed; bacteriological cultures were obtained and treatment with penicillin G and metronidazole was initiated. The patient was transferred to the intensive care unit. A severe state of shock resistant to treatment persisted, however, and the gas gangrene extended beyond the excised tissue. The patient soon died, 48 h after admission from multiorgan failure. At autopsy, a 300-ml purulent pericardial effusion was found without adherences and the peritoneal cavity was normal without evidence of intestinal perforation. There were no signs of necrosis of the left internal abdominal wall; an exophytic, hemorrhagic endoluminal mass was found in the cecum. Culture of the pericardial fluid revealed Clostridium septicum recovered also in the blood cultures and the intraoperative skin cultures. Histological examination of the cecal mass revealed an adenocarcinoma with numerous vascular emboli. Clostridium septicum, a Gram-positive anaerobic organism, has long been considered an intestinal tract saprophyte originating from the soil, with the potential of causing opportunistic infection in compromised hosts [2]. It is motile by means of peritrichous flagella which allows it to cross the intestinal membrane into the blood stream through areas of tissue abnormality [3]. This capability differentiates it from Clostridium perfringens and partly explains why gas gangrene due to Clostridium septicum occurs in locations distant from the malignant lesion.

The majority of cases of Clostridium septicum septicemia are secondary to malignant lesions (70- 85070) [3, 4] in particular, cancer of the right colon where the pH and osmolarity favor bacterial proliferation [5]. Thus, rapid hematologic spread occurs more frequently than local extension. Clostridium septicurn multiplies rapidly and metastatic localizations occur frequently [3]. Excluding trauma, in which Clostridiurn perfringens infection predominates, the most frequent clostridial species to cause gas gangrene is Clostridiurn septicum. Microangiopathic lesions caused by tissue hypoxia and localized acidosis favor Clostriclium septicum proliferation especially in cutaneous tissue [6]. As soon as the diagnosis of gas gangrene is established and after all obvious causes have been eliminated, a search for a digestive tract portal of entry must be rapidly performed (laparotomy). If this yields no results, other diagnostic studies must be undertaken. It is often at autopsy [2, 4], however, that the diagnosis is established. Treatment of gangrene is an emergency and must include wide debridement and antibiotics of which penicillin G remains the antibiotic of choice for Clostridium septicum infections [5]. The efficacy of hyperbaric oxygen is not clear [2]. Several cases of Clostridium perfringens pericarditis have been described, however, only one case due to Clostridium septicurn could be found in the literature [7]. The prognosis of systemic infection due to Clostridiurn septicum remains very poor despite early treatment (45-70% mortality according to the series). This case report focuses once again on the importance of making a careful search for a digestive tract cancer when confronted with a patient with gas gangrene without a history of trauma or with a blood culture that is positive for Clostridium septicum. Although rare, systemic infection by Clostridium septicum can cause pericarditis. Yours faithfully, B. Francois, L. Delalre, P. Vignon, L. Niquet, R. F. Gobeaux and R. Gay

References I. Veidenheimer MC, Compton CC (1990) A 63-year-old woman with diarrhea and bullous lesions of the buttock. N Engl J Med 323:1406- 1412 2. Collier PE, Diamond DL, Young JC (1983) Nontraumatic Clostridiurn septicum gangrenous myonecrosis. Dis Colon Rectum 26: 703 - 704 3. Sjolin SU, Hansen AK (199I) Clostridium septicum gas gangrene and an intestinal malignant lesion. J Bone Joint Surg [Am] 73: 772- 773 4. Alpern R J, Dowell VR (1969) Clostridium septicurn infections and malignancy. JAMA 209:385-388 5. Katlic MR, Derhac WM, Coleman WS (i981) ClostrMium septicum infection and malignancy. Ann Surg i93:361-364 6. Hallock GG (1984) Delayed antemortem diagnosis of adenocarcinoma of the cecum presenting as lower extremity gas gangrene. Dis Colon Rectum 27:I31-133 7. Brahan RB, Kahler RC (1990) Clostridium septicum as a cause of pericarditis and mycotic aneurysm. J Clin Microbiol 28:2377-2378 Dr. B. Francois, Service de Rdanimation Polyvatente, CHU Dupuytren, 2, Avenue Martin Luther King, F-87042 Limoges Cedex, France

Real or apparent entrapment of a Swan-Ganz pulmonary artery catheter after cardiac surgery? Dear Sirs, There are many instances where the placement of a pulmonary artery flotation catheter has been cited as the cause of patient morbidity and mortality. These problems have included pulmonary artery haemor-

310 rhage [1, 2], aneurysm formation [3], arrhythmia [4], knotting [51 and entrapment [6, 7]. We would like to describe the course o f events in a patient who h a d apparent entrapment of a Swan-Ganz type catheter and how in other circumstances a sternotomy m a y have been avoided. Case report. A 62-year-old male underwent routine mitral valve replacement surgery for mitral incompetence. A Swan Ganz p u l m o n a r y artery flotation catheter was placed via the right internal jugular route at the beginning of surgery. After post-operative intensive care was complete the patient, w h o had been extubated and was otherwise welt, was found to have catheter entrapment. Several attempts were made to remove the catheter which could be withdrawn to about the 30 cm marking but no further. Strong pulling on the catheter caused chest discomfort. Arrangements were made to further evaluate the situation with fluoroscopy. This demonstrated that the catheter appeared to be caught at the level of the right atrium and deformation of the mediastinal contents could be seen when traction was applied to the catheter. The patient was scheduled for a resternotomy and exploration. At surgery the catheter was found to be free in the right atrium and superior vena cava and not trapped as we had thought would be the case. The whole arrangement, introducer sheath and catheter were then removed together with ease. The patient made an uneventful recovery. On closer inspection the catheter appeared to have been caught by a suture. This had happened at the original operation. The surface of the catheter had clearly been deformed by something cutting across it. The catheter can be seen in Fig. 1. Point A on the figure shows how the surface of the catheter was cut through by the suturing (probably of the right atrium) and point B shows the deformation caused on the introducer tip. Unfortunately the surface had be roughened so m u c h that withdrawal through the introducer cannula had been impossible. The catheter snagged on the introducer if an attempt was made to withdraw it. Our previous attempts to withdraw the catheter had in fact deformed the end of the introducer to such an extent that this was visible on the X-ray screening and we interpreted these findings as evidence o f catheter entrapment. Had the introducer been removed first this would have very likely freed the catheter for subsequent removal. Discussion. After discovery of apparent entrapment the best course of action would be to determine whether the catheter is in fact really trapped or just appears to be so. This could easily be done by sliding off the introducer sheath over the catheter and proceeding to withdraw both from the patient. If, after the withdrawal of the introducer sheath the catheter still appears trapped, then it is more likely that it really is caught somewhere in the heart or great vessels. This case demonstrates that cases of apparent entrapment are best dealt with by first removing the introducer and then removing the catheter. This is especially the case after cardiac surgery. Yours faithfully, J. Benson and V. Patla

References 1. Feng WC, Singh AK, Drew T et al (1990) Swan-Ganz catheter-induced massive hemoptysis and p u l m o n a r y artery false aneurysm. A n n Thorac Surg 50:644-646 2. Jondeau G, Lacombe P, Rocha P e t al (1990) Swan-Ganz catheter-induced rupture of the pulmonary artery: successful early management by transcatheter embolization. Cathet Cardiovasc Diagn 19:202 - 204 3. Pezzulli F, Goldberg N, Posner D (1992) P u l m o n a r y artery pseudoaneurysm secondary to Swan-Ganz catheterization - a case report. Austral Radiol 36:56-61 4. Weissman MS, Altus P (1989) Heart block after Swan-Ganz insertion. Hosp Pract 24:36 5. Aguilar JL, Mazo V, Mata JM et al (1988) Double-knot formation of a Swan-Ganz catheter in the thoracic great venous vessels. Can J Anaesth 35:539- 540 6. Lazzam C, Sanborn TA, Christian F Jr (1989) Ventrieular entrapment of a Swan-Ganz catheter; a technique for nonsurgical removal. J A m Coll Cardiol 13:1422-1424 7. Schregel W, Linge C, Melis W (1991) Entrapment of a Swan-Ganz catheter in tricuspidal valve. Acta Anaesth Belg 42:117-119 Dr. J. Benson, University Department o f Anaesthesia and Intensive Care, Queen Elizabeth Hospital, Vincent Drive, Edgbaston, Birmingh a m B 15 2TH, UK