Emphysematous pyelonephritis (EPN) is a complication seen in patients with long-standing diabetes and is characterised by formation of gas in the renal parenchyma.1 Such gas may extend into the peri-nephric or para-renal spaces and additionally in advanced case, there may be bilateral and extensive involvement.1 The primary cause of EPN is urinary tract infection caused by either Gram-negative organisms or anaerobes in some cases. The diagnosis of EPN is confirmed by non-contrast computed tomography (CT) and treatment is usually based on the CT classification. In the case of class IIIa EPN, the current recommended treatment of choice is nephrectomy or percutaneous drainage. Herein, however, we managed our patient with class IIIa EPN conservatively with only medical management, without percutaneous drainage or even haemodialysis.

In August 2010, a 54-year-old male with diabetes for the past 20 years presented to the Dr Rajendra Prasad Government Medical College in Himachal Pradesh, India. His chief complaints were pain in left flank for 5 days and fever for 1 day. The symptom of pain was gradually progressive, continuous type localised in left flank. The patient developed fever on the fourth day which was recorded as 102°F (38.9°C). He was given ornidazole and ofloxacin tablets orally at a primary health centre and was then referred to our medical department because of uncontrolled blood glucose. He reported a history suggestive of osmotic symptoms and burning micturition for 5 days, and there were no other significant present or past complaints. The vital signs showed blood pressure of 90/60 mm Hg at presentation, pulse rate of 106 beats/min, respiratory rate of 16 breaths/min, and temperature of 103°F (39.4°C) was recorded. On examination, the patient was conscious, oriented, and there was tenderness in the left renal angle, but all the other systems were normal. The biochemistry results showed a haemoglobin level of 101 g/L (reference range, 120-150 g/L), total leukocyte count of 17 700/mm³, neutrophils of 88%, and erythrocyte sedimentation rate of 68 mm/h (reference range, 0-22 mm/h), with normal platelet counts and liver functions. Metabolic control was very poor with glycated haemoglobin level of 12.8%, urea of 19.27 mmol/L (reference range, 4-8.3 mmol/L), and serum creatinine of 132.6 µmol/L (reference range, 50-110 µmol/L). Urine examination showed no albumin, traces of sugar, while urine microscopy showed 15 to 20 pus cells per high-power field, and therefore urine culture and sensitivity was ordered. However, the urine and blood culture reports later turned out to be sterile.

The patient was managed for uncontrolled blood glucose and a possible diagnosis of acute pyelonephritis of the left kidney was made clinically. He was started on human mixed insulin along with ceftriaxone (1 g administered intravenously every 12 hours). Chest X-ray showed raised left dome of diaphragm. Ultrasound examination of abdomen revealed EPN of left kidney with destruction of upper portion and extending into the perinephric space near the upper pole. The treatment given was continued and the abdominal CT scan revealed evidence of a small renal calculus on the left side with extensive EPN of the left kidney and the presence of small amount of fluid in the left perinephric space (Fig). In this patient, the CT classification was class IIIa EPN.

With regard to further management, we gave an option of nephrectomy to the patient, but he refused to give consent for any surgical intervention. The patient refused even for percutaneous nephrostomy, so we continued with medical management alone. To our surprise, throughout the hospital stay, the patient remained afebrile and fully conscious. He showed a steady recovery in clinical and laboratory parameters without any new complaints or complications. After 1 week of insulin and antibiotic (ceftriaxone) treatment, his renal functions remained normal with a serum creatinine of 106.1 µmol/L and blood glucose was adequately controlled with human mixed insulin injections over the next few days. The patient was started on intravenous ceftriaxone empirically on admission and this was continued for a total of 2 weeks. During this period, the patient became fully active and mobile. Due to absence of any complications or features of sepsis after 2 weeks of treatment, the patient was discharged on oral antibiotics for a further 2 weeks. After 4 months, ultrasound examination showed normal kidneys and even the small calculus had disappeared.

Emphysematous pyelonephritis is a well-known condition which mainly affects the diabetic population (90%) and seen in patients with chronic diabetes.1 Some interesting facts about EPN are that it has a strong female preponderance with female-to-male ratio of 5:1, mean age of occurrence is around the fifth decade, and most often it involves left kidney in almost 60% of cases.1 The gas contains carbon dioxide mainly because of bacterial fermentation of glucose along with some hydrogen and nitrogen. Presentation is often vague with non-specific abdominal pain, fever, and tenderness in the renal angle.1

Since most cases are sporadic and spread over time, it is not easy to formulate definitive guidelines for all cases of EPN. The best evidence available for EPN is dependent on a few large prominent studies by Huang and Tseng2 in 2000 with 48 patients and a review by Pontin and Barnes1 in 2009 with 52 patients. Such studies serve as a good guide for the rest of the world, but may suffer from bias from the treating physicians and surgeons. So it is prudent to report all cases of EPN with details of presentation, extent of involvement, line of management, and outcomes.

The medical management of EPN was proposed a long time ago when it was first described by Schultz and Klorfein in 1962.3 In fact, in their report it was stated that “Vigorous treatment of uncontrolled diabetes and the use of appropriate antibiotics, avoiding anaesthesia and surgery seems a more rational approach.”3

The current management recommendations are based on the CT-based classification used by Huang and Tseng2 in which they divide EPN into classes I, II, IIIa, IIIb, and IV (Table).

Class I and II EPN can be managed by medical therapy alone or combined with percutaneous drainage. Huang and Tseng2 found that class IIIa EPN showed a 71% failure rate following percutaneous drainage with a 29% mortality rate, and those belonging to class IIIb had a 30% failure rate with a 19% mortality rate. Thus surgery became the norm for extensive disease except where there was a single kidney or bilateral involvement. Unilateral nephrectomy can be done with a low mortality risk in most centres.1

The most recent updates, however, recommend a more conservative approach. Recently, reports of successful management using percutaneous drainage and medical management alone have started coming from across the world even with bilateral and extensive disease.4 5 6 A recent case series describes eight cases of EPN with medical management alone.6 Of these eight cases, four were class IV EPN, five cases required haemodialysis, whereas four needed percutaneous drainage. The authors successfully used injections of imipenem for 10 days in five patients, cefoperazone + sulbactam for 14 days in two patients, and piperacillin + tazobactam for 14 days in one patient.6 Antibiotics that are effective in treating EPN include quinolones, beta-lactamase inhibitors, cephalosporins, and aminoglycosides, alone or in combinations mainly to cover Gram-negative bacteria like Escherichia coli, Klebsiella pneumoniae, Proteus spp, Streptococcus spp, Pseudomonas spp, and anaerobes like Bacteroides fragilis, and Clostridium septicum.7 8

The reason for successful shift in management is probably because nowadays EPN gets picked up early on CT showing very small air pockets. Such patients are therefore more likely to respond to conservative management. Even for larger EPN, percutaneous drainage rather than nephrectomy can be the standard of care in a majority of patients.1

Since India is one of the epicentres of diabetes epidemic, we wish to emphasise that systematic reporting of all EPN cases from this region can aid in formulating guidelines, especially for poor resource settings like ours. Most of our cases go unreported due to lack of proper record maintenance. Our patient had received ciprofloxacin tablets for 5 days before presenting to our hospital. Early initiation of antibiotics may have limited the progress of EPN in this patient and may have also been responsible for the sterile cultures that were reported. The patient had long-standing diabetes and also had a renal calculus, both of which can predispose to infection and subsequently lead to EPN. The patient was categorised as class IIIa EPN, who responded exceptionally well to antibiotics alone and had complete resolution of EPN on follow-up.

In conclusion, this is a rare case report because our patient recovered without drainage, surgery, or even haemodialysis. There have been reports that in class IIIa EPN that medical management alone with percutaneous drainage has a 92% failure rate.2 We wish to emphasise sensitisation of treating physicians regarding early investigations using contrast CT scan in diabetes patients with fever, renal angle tenderness, and vague abdominal pain. We do not recommend medical treatment for all such cases, but follow-up should be based upon patient’s response to the initial treatment.

1. Pontin AR, Barnes RD. Current management of emphysematous pyelonephritis. Nat Rev Urol 2009;6:272-9. Crossref

2. Huang JJ, Tseng CC. Emphysematous pyelonephritis: clinicoradiological classification, management, prognosis, and pathogenesis. Arch Intern Med 2000;160:797-805. Crossref

3. Schultz EH Jr, Klorfein EH. Emphysematous pyelonephritis. J Urol 1962;87:762-6.

4. Flores G, Nellen H, Magaña F, Calleja J. Acute bilateral emphysematous pyelonephritis successfully managed by medical therapy alone: a case report and review of the literature. BMC Nephrol 2002;3:4. Crossref

5. Tahir H, Thomas G, Sheerin N, Bettington H, Pattison JM, Goldsmith DJ. Successful medical treatment of acute bilateral emphysematous pyelonephritis. Am J Kidney Dis 2000;36:1267-70. Crossref

6. Kolla PK, Madhav D, Reddy S, Pentyala S, Kumar P, Pathapati RM. Clinical profile and outcome of conservatively managed emphysematous pyelonephritis. ISRN Urol 2012;2012:931982. Crossref

7. Chen MT, Huang CN, Chou YH, Huang CH, Chiang CP, Liu GC. Percutaneous drainage in the treatment of emphysematous pyelonephritis: 10-year experience. J Urol 1997;157:1569-73. Crossref

8. Shokeir AA, El-Azab M, Mohsen T, El-Diasty T. Emphysematous pyelonephritis: a 15-year experience with 20 cases. Urology 1997;49:343-6. Crossref

Oliguria or anuria early after cadaveric renal transplant (CRT) is not uncommon and can be related to numerous causes. We report here a rare case of complication of ureteric implantation in the peritoneum in CRT causing anuria. Diagnosis was suspected early on day 1 after operation by clinical examination and abdominal X-ray (AXR), and confirmed with non-contrast computed tomography (CT) of the abdomen and pelvis. The complication was managed with immediate reimplantation of the ureter into the bladder and the patient had an uneventful recovery thereafter with good graft function. To our knowledge, this is the first reported case in Hong Kong and the fifth in the world. We believe the actual incidence is underreported and would like to share our suggestions on how to avoid this rare complication.

A 29-year-old man with Alport’s syndrome developed end-stage renal failure in 2009 and was started on intermittent peritoneal dialysis in 2011. His serum creatinine level and estimated glomerular filtration rate were 1458 µmol/L and 5.1 mL/min, respectively. In September 2013, he received CRT from a 60-year-old man with brain stem death due to haemorrhagic stroke. The donor’s serum creatinine was 88 µmol/L and there were no hypotensive episodes or inotrope infusion before organ harvesting.

The cadaveric right kidney was transplanted into the right iliac fossa of the recipient, with good perfusion and turgor after release of vascular clamps. No urine was noted at cut-end of the graft ureter at the time of implantation. A needle test to aspirate pre-filled gentamicin solution from the bladder was done and extravesical ureteroneocystostomy with Lich-Gregoir technique was performed with a 7-French, 15-cm long, double J ureteric stent in situ. The total operating time was 2 hours and 20 minutes; with cold ischaemic time of 7 hours 41 minutes and second warm ischaemic time of 33 minutes. The patient was haemodynamically stable but remained anuric 12 hours after operation with a serum creatinine level of 1268 µmol/L. An urgent Doppler ultrasound of the graft kidney was done and showed good perfusion of graft with patent renal artery and vein. Repeated physical examination revealed a slightly distended abdomen and AXR showed the distal coil of double J stent above the pelvis level (Fig a), leading to the suspicion of implantation of the ureter into the peritoneum. Subsequent non-contrast CT of the abdomen and pelvis confirmed placement of the ureteric stent outside the bladder (Fig b and c).

Exploration and reimplantation of ureter were performed immediately. It was noted that the ureter was implanted into the thickened peritoneum at a level just above the right upper lateral bladder wall, with urine draining into the intra-abdominal cavity. The detrusor layer was thin and not well developed. The anastomosis was taken down and ureteroneocystostomy was refashioned with the same Lich-Gregoir technique. There was immediate return of good urine output and his serum creatinine levels improved to 186 µmol/L and 126 µmol/L on postoperative day 3 and week 4, respectively.

Post–renal transplant urological complications are not uncommon. Commonly reported complications include thromboembolic events of vascular anastomosis, acute tubular necrosis, lymph leak, and urinary reflux. Recent retrospective series have reported incidence rates of 2.8% to 15.5% of urological complications after CRT.1 2 3 These are significantly decreased rates compared with those in an earlier series after introduction of various modified techniques of implantation and use of ureteric stents.1 Peritoneal implantation of ureter constitutes an incidence of 0.1% to 0.2% only.3 4 Gibbons et al4 reported a case of ureteric implantation into an ovarian cyst in addition to two cases into the peritoneum in a series of 1000 CRT recipients. We believe the actual incidence is underreported, given the general impression that this complication is solely technically related. The Table3 4 5 6 summarises all reported cases of peritoneal implantation of graft ureter in the current literature.

All reported patients presented with postoperative anuria, with one patient developing ascites, abdominal pain, anuria, and sudden shock.5 Timing of diagnosis had been reported from immediate postoperation to few weeks later. A high level of suspicion remains the key for reaching the diagnosis. Common contributing factors identified from the literature and our case include long-term peritoneal dialysis with thickened peritoneum, and the presence of residual peritoneal fluid mimicking urine in the bladder. Therefore, this complication should be suspected in such a patient with unexplained delayed graft function. Tan et al6 suggested that an unexplained rise in ultrafiltration volume in transplanted peritoneal dialysis patients accompanied by a fall in baseline serum creatinine is highly suggestive of the diagnosis. If ureteric stenting was employed, imaging such as AXR and CT scan can help identify the position of the ureteric stent and confirm the diagnosis. Definitive diagnosis can only be established upon exploration.

A note of caution on ways to prevent this rare surgical complication would be more beneficial than treating it. We suggest several measures to avoid it, which have not been discussed in previous reports. Regarding the technique of ureteric implantation, the classic transvesical Leadbetter-Politano technique in which two cystostomies are required, is now replaced by the extravesical Lich-Gregoir technique which requires only one cystostomy and, hence, less bladder dissection, shorter ureteral length, and no interference with native ureteral function.7 This technique, however, may pose challenges in recipients previously on peritoneal dialysis, as the peritoneum is thickened due to exposure to dialysis fluid and episodes of peritonitis, if any. A thickened peritoneum, with the presence of residual peritoneal fluid upon incision, can easily be mistaken as the bladder during transplantation. In our practice, the bladder is filled with 80 to 100 mL of gentamicin solution before the procedure and needle aspiration test is done before ureteric implantation to aid identification of the bladder. Our case illustrated that even with these standard precautions, one may not be able to completely prevent this complication. Tagging up the extravesical tissue with parallel stay sutures on both sides of the 2 to 3 cm submucosal tunnel before creating the cystostomy will help identify the bladder and peritoneum vigilantly, avoiding shifting of the incision site to the peritoneum instead of the bladder wall after the needle test. If the bladder is scarred and non-compliant due to neuropathic bladder or prolonged anuria, identification of the junction between the peritoneum and bladder wall may become difficult. Preoperative emptying of all peritoneal dialysis fluid is advocated so that if there is nil drainage of bladder content after making the incision, entry into the abdominal cavity instead of the bladder can be suspected. Direct visualisation of the urethral catheter should be the best way to ensure the correct cavity is entered. A larger cystostomy, however, is often required and is not preferred.

Another innovative trick to pick up the complication, should the implantation be done already, is to notice the colour of effluent from the bladder after ureteroneocystostomy. Any urine or antibiotic solution drained in the early postoperative period is at least lightly blood-stained because of the disturbance to the mucosal edges during cystostomy. The absence of blood-stained effluent after ureteric implantation should raise the suspicion that the peritoneum, and not bladder, was opened.

Our report suggests that peritoneal implantation of ureter in CRT is not only technically related but also involves multiple contributing factors. A high index of suspicion is required to pick up this complication and meticulous measures should be adopted to avoid its occurrence.

1. Zavos G, Pappas P, Karatzas T, et al. Urological complications: analysis and management of 1525 consecutive renal transplantations. Transplant Proc 2008;40:1386-90. Crossref

2. Praz V, Leisinger HJ, Pascual M, Jichlinski P. Urological complications in renal transplantation from cadaveric donor grafts: a retrospective analysis of 20 years. Urol Int 2005;75:144-9. Crossref

3. Davari HR, Yarmohammadi H, Malekhosseini SA, Salahi H, Bahador A, Salehipour M. Urological complications in 980 consecutive patients with renal transplantation. Int J Urol 2006;13:1271-5. Crossref

4. Gibbons WS, Barry JM, Hefty TR. Complications following unstented parallel incision extravesical ureteroneocystostomy in 1,000 kidney transplants. J Urol 1992;148:38-40.

5. Blanco Parra M, Calviño J, Romero Burgos R, et al. Ureteral implantation in peritoneum. Exceptional complication in renal transplantation [in Spanish]. Actas Urol Esp 2002;26:579-80. Crossref

6. Tan SY, Lim CS, Teo SM, Lee SH, Razack A, Loh CS. Peritoneal implantation of ureter in a cadaveric kidney transplant recipient. Med J Malaysia 2003;58:769-70.

7. Zhao JJ, Gao ZL, Wang K. The transplantation operation and its surgical complications. In: Ortiz J, Andre J, editors. Understanding the complexities of kidney transplantation. China: InTech; 2011: 466-7. Crossref

Hepatic artery aneurysm (HAA) is an uncommon condition that can present with life-threatening haemorrhage. In case of catastrophic intraperitoneal rupture, laparotomy with ligation of the aneurysm is often the treatment of choice due to its rapid access and ability to secure control of the extravasation site. Controversy exists whether hepatic artery should be reconstructed after ligation of the aneurysm, especially when distal ligation is performed on the hepatic artery proper. We herein describe the technical considerations during the operative management of a patient with ruptured HAA.

A 69-year-old male presented to Queen Mary Hospital, Hong Kong, with a 3-week history of epigastric pain in December 2010. A contrast-enhanced computed tomography (CT) of the abdomen showed a common HAA with involvement of the root of gastroduodenal artery (GDA; Fig 1a) but there was no evidence of contrast leakage or haematoma formation at the time of diagnosis. Complete blood profile showed evidence of inflammation that included mild leukocytosis with neutrophil predominance, thrombocytosis, and mildly elevated erythrocyte sedimentation rate. Serum biochemistry showed elevated γ-glutamyl transpeptidase and C-reactive protein. Screening for vasculitis and infection was unremarkable.

Three days later, however, the patient developed generalised peritonitis complicated by hypovolaemic shock during the workup for this condition in the hospital. There was an acute drop in haemoglobin level from 150 g/L to 125 g/L. A diagnosis of rupture of the common HAA was made clinically and an emergency laparotomy was performed.

On entering the peritoneal cavity, 5 L of hemoperitoneum was recovered, and after four-quadrant packing of the peritoneal cavity, supracoeliac aortic clamping was performed to secure temporary haemostasis. The lesser sac was entered and a ruptured 4 x 3 cm fusiform aneurysm of the common hepatic artery surrounded by inflamed tissues cranial to the pancreatic head region was identified. The proximal and distal neck of the aneurysm was subsequently isolated and encircled by tape without further dissection at the hepatoduodenal ligament. After release of the aortic clamp, the distal neck of the aneurysm at the hepatic artery proper was test clamped and Doppler ultrasonography was performed, which confirmed good biphasic flow in both the right and left hepatic arteries. The aneurysm was then ligated and suture-transfixed at its proximal and distal neck region (Fig 2).

In the postoperative period, the alanine and aspartate transaminase levels were elevated to 1607 U/L and >3000 U/L, respectively. Both the liver parenchymal enzymes returned to normal levels 3 weeks after the procedure. A short course of renal replacement therapy was required for the acute renal injury sustained during the shock. Otherwise, the patient made an uneventful recovery. Pathological examination of the aneurysm wall showed infiltration of lymphocytes and neutrophils but there was no evidence of infection.

Follow-up CT performed at 2 and 8 months postoperatively showed arterial collateral formation along the porta hepatis, and there was no evidence of previous liver infarction (Figs 1b and 1c).

Hepatic artery aneurysm is a rare condition with an estimated prevalence of 0.1% according to autopsy series,1 and incidence of 0.002%.2 The majority (75%) of cases are incidentally diagnosed, and pain in the right upper quadrant or epigastrium is the most common presenting symptom.1 Rupture is the second most common mode of presentation and accounts for 14% of the newly diagnosed HAAs.2

The reported mortality rate for ruptured HAA is 21% to 43%,2 3 and the key to successful resuscitation is timely control of the ruptured site. Although endovascular exclusion of HAAs has a reported success rate of up to 100%,4 open surgery is often the treatment of choice if the exsanguinating haemorrhage does not permit a failed trial of endoluminal haemostasis.

The objectives of open surgery are to exclude the diseased artery segment and preserve the hepatic arterial inflow. It is generally accepted that aneurysms of the common hepatic artery which do not involve the GDA can be simply ligated without reconstruction of inflow to the hepatic artery proper because the latter is supplied by the reverse flow from the GDA.5 For aneurysms involving GDA or distal to its origin, the conventional advocate is to reconstruct hepatic arterial inflow after ligation or resection of the aneurysm to avoid the risk of ischaemic injury to the liver.6 Depending on the morphology and size of the aneurysm, inflow restoration may take the form of aneurysmectomy with patch graft repair or exclusion of aneurysm with coeliac or aorto-hepatic bypass grafting. Autogenous saphenous vein is the preferred graft, although a dacron prosthetic graft can also be used.6 In our patient, because of the clinical suspicion of an infective aetiology for the ruptured aneurysm and due to the satisfactory hepatic arterial flow during intra-operative ultrasound, arterial reconstruction was not considered.

Graham and Cannel7 had reviewed that accidental hepatic artery ligation would result in greater than 50% mortality when the site of ligation is distal to the GDA. In later studies, however, such a high mortality rate was not observed with hepatic artery ligation during iatrogenic injuries or therapeutic procedures to ‘de-arterialise’ tumour-laden livers.5 8 9 It is believed that the liver is protected from ischaemic insults after hepatic artery ligation by a buffer response of the portal venous system. The portal vein maintains the supply of 70% of hepatic inflow and 50% of hepatic oxygen consumption during acute hepatic artery interruption, before compensatory arterial supply develops.6 With time, a multitude of arterial collateral channels can develop to compensate for the interruption of original arterial inflow. In fact, postoperative CT scan at 8 months in our patient clearly showed collateral formation via the pericholedochal route (Fig 1c). Besides, Michels10 has described 26 potential sites of collateral arterial supplies to the liver originating from the celiac trunk, superior mesenteric artery and intrathoracic arteries, routing via the hepatoduodenal ligament, hepatogastric ligament, omentum, retroperitoneum, falciform ligament, and diaphragmatic connections. Collateral formation is observed after ligation of hepatic arteries in angiographic studies as early as 10 hours after ligation.5 Moreover, replaced or accessory arterial supplies to the left or right hepatic lobes are present in about 30% of the general population,5 11 rendering feasible supply to the contralateral lobe via the interlobar anastomosis at the liver hilum. The clinical relevance of such vascular recovery potential was demonstrated by Chirica et al,12 wherein the technique of simple aneurysm ligation with minimal dissection in the hepatoduodenal ligament was adopted so as to preserve the potential arterial collaterals and hence avoiding the need for reconstruction of the extirpated hepatic artery in three out of four patients in their series of HAAs with GDA involvement.

Establishment of arterial collaterals, however, is a latent process and the anatomical information on the presence of aberrant or replaced hepatic arterial supply may not be readily available at the time of laparotomy for a ruptured aneurysm. In this situation, intra-operative Doppler ultrasonography is a convenient tool for evaluation of intrahepatic arterial flow. In transplanted livers, a normal Doppler waveform of the hepatic artery includes a rapid upstroke with an acceleration time of <80 ms, a continuous diastolic flow giving a resistive index between 0.5 and 0.7, and a peak systolic velocity of <2 m/s; any deviation from normality from any of these parameters is suggestive of hepatic artery stenosis and predictive of subsequent ischaemic bile duct injury.13 Perhaps, the same criteria can be adopted to assess the arterial sufficiency in a native liver. Back bleed from distal cut end of hepatic artery and cyanosis of hepatic lobe should be counted with caution as these have been reported to be inaccurate in predicting adequate perfusion or subsequent ischaemic injury, respectively.14

Simple ligation of the hepatic artery is considered safe and effective in the treatment of ruptured HAA. The establishment of arterial collaterals should be expected given that the hepatoduodenal ligament is not severely disrupted. The use of intra-operative ultrasound facilitates the decision-making in devising the operative strategy for this condition.

No conflict of interest was declared by the authors.

1. Arneson MA, Smith RS. Ruptured hepatic artery aneurysm: case report and review of literature. Ann Vasc Surg 2005;19:540-5. Crossref

2. Abbas MA, Fowl RJ, Stone WM, et al. Hepatic artery aneurysm: factors that predict complications. J Vasc Surg 2003;38:41-5. Crossref

3. Shanley CJ, Shah NL, Messina LM. Common splanchnic artery aneurysms: splenic, hepatic, and celiac. Ann Vasc Surg 1996;10:315-22. Crossref

4. Tulsyan N, Kashyap VS, Greenberg RK, et al. The endovascular management of visceral artery aneurysms and pseudoaneurysms. J Vasc Surg 2007;45:276-83. Crossref

5. Mays ET, Wheeler CS. Demonstration of collateral arterial flow after interruption of hepatic arteries in man. N Engl J Med 1974;290:993-6. Crossref

6. Tygstrup N, Winkler K, Mellemgaard K, Andreassen M. Determination of the hepatic arterial blood flow and oxygen supply in man by clamping the hepatic artery during surgery. J Clin Invest 1962;41:447-54. Crossref

7. Graham RR, Cannel D. Accidental ligation of the hepatic artery. Br J Surg 1933;20:566-79. Crossref

8. Ariyan S, Cahow CE, Greene FL, Stansel HC Jr. Successful treatment of hepatic artery aneurysm with erosion into the common duct. Ann Surg 1975;182:169-72. Crossref

9. Brittain RS, Marchioro TL, Hermann G, Waddell WR, Starzl TE. Accidental hepatic artery ligation in humans. Am J Surg 1964;107:822-32. Crossref

10. Michels NA. Collateral arterial pathways to the liver after ligation of the hepatic artery and removal of the celiac axis. Cancer 1953;6:708-24. Crossref

11. Covey AM, Brody LA, Maluccio MA, Getrajdman GI, Brown KT. Variant hepatic arterial anatomy revisited: digital subtraction angiography performed in 600 patients. Radiology 2002;224:542-7. Crossref

12. Chirica M, Alkofer B, Sauvanet A, Vullierme MP, Levy Y, Belghiti J. Hepatic artery ligation: a simple and safe technique to treat extrahepatic aneurysms of the hepatic artery. Am J Surg 2008;196:333-8. Crossref

13. Crossin JD, Muradali D, Wilson SR. US of liver transplants: normal and abnormal. Radiographics 2003;23:1093-114. Crossref

14. Mathisen DJ, Athanasoulis CA, Malt RA. Preservation of arterial flow to the liver: goal in treatment of extrahepatic and post-traumatic intrahepatic aneurysms of the hepatic artery. Ann Surg 1982;196:400-11. Crossref

Venous thromboembolism (VTE) is a condition that includes both deep vein thrombosis (DVT) and pulmonary embolism (PE). The formation of thrombi is associated with the Virchow’s triad, ie circulatory stasis, vascular wall injury, and a hypercoagulable state. Most data of VTE in orthopaedic surgeries are based on studies of patients who undergo hip or knee arthroplasty. The risk of VTE following lower-limb surgery is noted to be considerably higher than those following upper-limb surgery. To date, there has been no evidence to support the prescription of VTE prophylaxis in upper-limb operations.

In June 2013, a 56-year-old, non-obese, Chinese carpenter, who enjoyed good past health and was a non-smoker, sustained injuries to both his left ring finger and little finger while using an electric saw. His left ring finger was amputated at the level of the middle phalanx and his little finger was nearly amputated at the distal interphalangeal joint level. No other site of injury was noted (Fig a). He presented to our hospital 1 hour after the injury. Upon arrival he was haemodynamically stable.

Due to the severe soft tissue contamination of the finger stumps (Fig b), options of either replantation or revision amputation for the ring finger was discussed with the patient. Subsequent emergency operation was arranged for attempted replantation of his ring finger, and wound debridement and fixation of his left little finger. Revascularisation of the ring finger was attempted thrice, including use of a vein graft. However, due to the inability to achieve sustained blood flow to the amputated stump, revision amputation was performed. Meanwhile, the open wound of the little finger was debrided, and the fracture of the finger was temporarily fixed with axial K wire. The operation lasted 6 hours 25 minutes, with a total anaesthetic time of 7 hours 12 minutes. The fluid balance was +2450 mL. He regained full consciousness in the recovery room, with good breathing and oxygen saturation. Antibiotics were continued and he was transferred back to a general ward.

The operation was completed at 4 am the next day and his vital signs remained stable throughout. He offered no complaints in the morning round and was able to mobilise and walk to the washroom where he sustained an unwitnessed fall. He had neither preceding chest pain, headache, nor swelling and pain over his lower limbs, and did not lose consciousness. He developed tachycardia, tachypnoea, and desaturation after being helped to his bed. Echocardiogram revealed sinus tachycardia. He started developing chest discomfort with subsequent witnessed arrest 40 minutes later. Cardiac monitoring revealed pulseless electrical activity. Cardiopulmonary resuscitation was started immediately with injection of adrenaline. Bedside echocardiogram revealed no pericardial effusion but there was no response to resuscitation and the patient succumbed 35 minutes later. Postmortem examination revealed bilateral DVT in the calf muscles. Both lungs were markedly congested, with occluding thromboemboli noted in the hilar main pulmonary arteries and their main branches. No other thromboemboli were noted.

Lower limb surgery is a risk factor for VTE events. The prevalence of DVT in patients who undergo hip fracture surgery and hip or knee arthroplasty ranges from 40% to 60%.1 2 Clinical or fatal PE in Hong Kong Chinese patients is even rarer.2

With regard to VTE following upper-limb surgery, nine VTE events have been reported. These comprised seven PEs following total elbow arthroplasty (of which three were fatal), one non-fatal PE after distal radius fixation following acute fracture, and one non-fatal PE following revision osteosynthesis of the proximal diaphysis of ulna.3 There has been no reported incidence in the literature of VTE following microsurgery.

It is difficult to determine whether microsurgery is completely risk-free for VTE. Prolonged surgery may involve prolonged immobilisation and blood loss that in turn increases the risk of VTE, based on Virchow’s triad. It is thus possible that prophylaxis for VTE is necessary in microsurgery when the operating time is expected to be long, for instance, in finger replantations or free-flap surgeries.

Currently there is no guideline on VTE prophylaxis for microsurgeries. The British Society for Surgery of the Hand4 considers upper-limb procedures under general anaesthesia for more than 90 minutes and/or with one risk factor for VTE (Box3) as moderate-risk procedures. It recommends use of mechanical compression devices in the operating room and/or until the patient becomes ambulatory. With the risk of bleeding in mind, low-molecular-weight heparin (LMWH) may be started no less than 6 hours postoperatively in selected patients and continued until they are fully ambulatory.3 None of these recommendations apply specifically to hand or microsurgeries.

In addition to LMWH, the National Institute for Health and Care Excellence clinical guideline and the American College of Chest Physicians (ACCP) also mention the newer non–vitamin K antagonist oral anticoagulants (NOACs).5 This group of drugs is associated with a rapid onset of action and predictable pharmacokinetics and pharmacodynamics. There are also fewer interactions with food and other drugs. The ACCP recommends the use of VTE prophylaxis for a minimum of 10 to 14 days following major orthopaedic surgery. In patients who undergo total joint replacement, the ACCP suggests the use of LMWH in preference to other agents (eg fondaparinux, apixaban, dabigatran, rivaroxaban, low-dose unfractionated heparin, adjusted-dose vitamin K antagonist or aspirin), irrespective of the concomitant use of an intermittent pneumatic compression device. For those patients who refuse injections, apixaban or dabigatran is recommended. However, the available studies (eg RE-MODEL trial, RECORD trial and ADVANCE-2 study) and guidelines for VTE prophylaxis focus on total joint replacement surgeries.6 There are no recommendations on the use of NOACs in microsurgeries.

When prescribing VTE prophylaxis, it is important to balance the associated benefits and risks. In the case of mechanical compression, in view of its low-risk profile, it should be offered to all patients who undergo prolonged microvascular procedures, for example replantation, until they are fully ambulatory. The application of mechanical compression on both lower limbs may nonetheless be limited if donor sites for blood vessels and other tissues are anticipated from a lower limb.

Based on available evidence, we suggest that all patients who undergo microsurgery should have mechanical compression prophylaxis. Additional pharmacological prophylaxis should be considered for those who are at relatively high risk of developing VTE, for example, patients who have more than one risk factor, those in whom upper limb or tumour surgery will exceed 90-minute duration, and/or those in whom there will be prolonged postoperative immobilisation.

1. Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of venous thromboembolism: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th Edition). Chest 2008;133(6 Suppl):381S-453S.

2. Mok CK, Hoaglund FT, Rogoff SM, Chow SP, Ma A, Yau AC. The incidence of deep vein thrombosis in Hong Kong Chinese after hip surgery for fracture of the proximal femur. Br J Surg 1979;66:640-2. Crossref

3. Roberts DC, Warwick DJ. Venous thromboembolism following elbow, wrist and hand surgery: a review of the literature and prophylaxis guidelines. J Hand Surg Eur Vol 2014;39:306-12. Crossref

4. The British Society for Surgery of the Hand. VTE Guidelines. Available from: http://www.bssh.ac.uk/education/guidelines/vteguidelines. Accessed 2 Jul 2012.

5. Beyer-Westendorf J, Ageno W. Benefit-risk profile of non–vitamin K antagonist oral anticoagulants in the management of venous thromboembolism. Thromb Haemost 2015;113:231-46. Crossref

6. Saraf K, Morris P, Garg P, Sheridan P, Storey R. Non–vitamin K antagonist oral anticoagulants (NOACs): clinical evidence and therapeutic considerations. Postgrad Med J 2014;90:520-8. Crossref