Fig. 39.3 “Fishmouthing” the femoral vein to equalise diameter with the aorta

Fig. 39.4 Femoral vein reconstruction from the infrarenal aorta (a) to the right iliac (b) and left common femoral artery (c). The right ureter (d) overlies the right limb of the graft

Question 8

If the patient had presented with an exposed prosthetic graft in the groin how would this have altered management?

A.Prolonged antibiotic therapy.

B.Use of vacuum dressings.

C.Simple coverage with a muscle flap without graft replacement.

D.Addition of a muscle flap to graft replacement with autologous vein.

b

Exposedgraftspresentadifficultproblemforachievingskinclosureandtheadditionofalocal muscleflapinthegrointograftreplacementwithautologous(femoropopliteal)veinisthemost certain method of cure. The author´s preference is a rectus femoris flap for this (Fig. 39.5).

Question 9

What complications might occur following this operation?

A.  Anastomotic haemorrhage.

B.  Graft rupture.

C.  Femoral deep vein thrombosis (DVT).

D.  Limb swelling.

E.  Seroma in the thigh.

F.  Intestinal obstruction.

G.  Wound infection.

Thepatienthadthemostcommoncomplicationofthisoperation,whichwasalargeseroma in the left thigh and a smaller one in the right thigh despite prolonged suction drainage. These were aspirated repeatedly and resolved after 3 weeks, although in other cases reoperation and reinsertion of a drain may be required. Intestinal obstruction is no more common following this operation than other abdominal procedures.

The patient otherwise made a good recovery and was discharged from hospital at 14 days. The graft and pus swabs were sterile so most of the antibiotics were stopped at 1 week but the co-amoxiclav was continued for 6 weeks on empirical grounds.

Question 10

What routine follow-up investigations should be performed?

A.  Abdominal CT scan. B.  Abdominal ultrasound. C.  Graft duplex scans.

D.  Abdominal X-ray.

E.  Erythrocyte sedimentation rate.

Routine 3-monthly duplex scans were performed over 1 year for vein graft surveillance. The patient remains well without further intervention at 5 years.

39.1  Commentary

Aortic graft infection is thankfully rare, occurring in 1–5% of reconstructions,1 but is one of the most feared complications in vascular surgery because of its high mortality and morbidity.2 In a UK multicentre audit of 55 graft infections 31% died, 33% underwent amputation and only 45% left hospital alive without amputation.3 If left untreated there is a high risk of graft occlusion and anastomotic haemorrhage, which may lead to aortoenteric fistula. Prompt curative treatment is therefore wise in patients sufficiently fit to withstand major surgery.

Graft infection may present at any time from a few days to many years after surgery. It canfollowawoundinfection,particularlyinthegroinwherewoundbreakdownmayresult in exposure of the graft, or it may present later with a perigraft fluid collection or sinus at the femoral anastomosis. Infection of wholly intra-abdominal grafts may present with backache and fever but more often remain undetected until anastomotic haemorrhage or graft thrombosis occurs. Duplex ultrasound scanning or CT angiography is wise in all cases of graft thrombosis not only to confirm the occlusion but also to demonstrate any perigraft fluid which would indicate graft infection. [Q1: B, D]

The most common causative organism is Staphylococcus aureus in most series and such infections tend to present in the early postoperative period. Methicillin-resistant strains (MRSA) are said to be particularly virulent and have been associated with a high mortality in some series.3–5 Staphylococcus epidermidis infections tend to be less virulent

404 C.P. Gibbons

and often present many years later. They produce a slime or biofilm around the graft or occasionally thin pus. Isolation of Staph. epidermidis is more difficult and may require agitation of the extirpated graft with ultrasound to release it for culture. Other infections are caused by coliforms, Salmonella, Serratia, Pseudomonas, enterococci, streptococci or Bacteroides.6 Gram-negative organisms may be more likely to present with anastomotic haemorrhage.7 In many cases no causative organism can be isolated despite obvious infection. Possible causes of this are previous antibiotic administration or failure to isolate

Staphylococcus epidermidis.

A preoperative diagnosis of graft infection is usually secured by ultrasound followed by CT or MRI. Aspiration of the perigraft fluid may secure a bacteriological diagnosis prior to surgery, although in many cases the responsible organism cannot be isolated. Fluid is often present in the aneurysmal sac after aortic aneurysm replacement and can be seen in smallerquantitiesaroundanaorto-bifemoralprosthesisperformedforocclusivediseaseon ultrasound or CT for a few weeks after surgery. However, persistence of fluid around an aortic prosthesis for more than 3–6 months after surgery is highly suggestive of infection. Similarly, perigraft gas may be present for up to 10 days after surgery but indicates infection beyond this time.9, 10

If a groin abscess develops in relation to an aorto-bifemoral graft, aspiration under asepticconditionsintheclinicwillconfirmthepresenceofgraftinfectionandmayprovide preoperative bacteriology. Perigraft fluid or gas may be absent in low-grade chronic infection orif asinus in one or other groin allows thepus to escape. Exploration of a sinus under anaesthesia will demonstrate a connection with the infected graft and gently passing a bougie alongside the graft will determine whether or not the infection is confined to the anastomosis. If there is no sinus or perigraft fluid, a 99technetium-labelled leucocyte scan may demonstrate increased activity over an infected graft.11 However, this investigation haspoorsensitivityandspecificityandisonlyusefulforchronicgraftinfectionasincreased leucocyte adherence is demonstrated by most prostheses for up to 6 months after insertion. The ultimate diagnosis of graft infection is made at operation by the lack of tissue incorporation into knitted Dacron or polytetrafluoroethylene (PTFE) prostheses and the presence of perigraft pus from which organisms may be cultured. Preoperative angiography is helpful for operative planning by delineating the vascular anatomy but adds no useful information about the presence of graft infection. [Q2: A, B]

There are multiple treatment options: Antibiotic therapy may buy time, but is rarely curative because the graft acts as a foreign body rendering the responsible organisms inaccessible to antibiotics. There have been occasional reports of successful treatment by drainage of the abscess around the graft followed by irrigation with antibiotic or iodine solutions or implantation of gentamicin-impregnated beads or foam but these are anecdotal.15, 16 Simple excision of an aortic graft is unwise unless it has already occluded without critical ischaemia as subsequent limb loss or severe lower body ischaemia is likely. Excision of the infected graft with debridement and replacement with a rifampicin-bonded or silver-impregnated graft has been advocated but most would reserve this for chronic low-grade infections because of the risk of reinfection of the new graft. Despite their in vitro effectiveness21 encouraging individual series, there is no convincing clinical evidence that either rifampicin-bonding or silver-impregnated Dacron grafts are less susceptible to reinfection after replacement of infected grafts.

Moreover, randomised clinical studies have failed to show that either rifampicin-bond- ing or silver impregnation prevents primary infection in vascular grafts, although these studies were somewhat underpowered.22, 23

An alternative approach, which avoids direct reimplantation of prosthetic material, is in situ replacement with fresh or cryopreserved aortic allografts. The reported results have been variable but in all series there have been instances of early or late graft disruption or aneurysm formation particularly with fresh allografts and when used for aorto-enteric fistula.24–29

Until the last 5 years, the mainstay of treatment has been excision of the graft with extra-anatomical reconstruction. For infected aortoiliac grafts reconstruction can be performed with an axillo-bifemoral or bilateral axillofemoral grafts. However, in those patients with infected aorto-bifemoral grafts, the lower anastomosis must be performed at the level of the superficial femoral or popliteal artery to avoid placing the new graft in an infected field. Good results can be obtained with this approach but there remains a 10–15% risk of graft reinfection.2, 30–32 If this option is used, the extra-anatomic bypass should be performed before graft excision to reduce the risk of irreversible limb ischaemia and amputation.2

Morerecently,Claggett33, 34 andNevelsteen35, 36 independentlyadvocatedaorticreplacement with femoropopliteal veins for infected grafts. Femoropoliteal veins are much wider and thicker-walled than long saphenous or arm veins and have adequate length, making them ideal for aortoiliac reconstruction. Iliac veins are too short and their excision would result in severe limb swelling. Results were excellent with reduced mortality, limb loss and reinfection rates. Subsequently other authors have confirmed the effectiveness of this approachineliminatingreinfection,withmortalityandamputationratessimilartoorlower than reports using other techniques.37, 38 This is now recognised as the procedure of choice in most situations. The procedure is demanding and may take several hours to perform but can be made easier by the use of two or more operative teams working together. Femoral veins may be harvested even after the removal of the long saphenous vein but it is generally advised that the profunda femoris vein should be left intact and that the popliteal vein should not be removed below the knee joint.39, 40 Fears of venous morbidity from femoral vein harvest have not been borne out in practice although Valentine reported an 18% incidence of prophylactic or therapeutic fasciotomy for compartment syndrome.41 However, neither Nevelsteen42 nor the present author has found this necessary. Femoral vein harvest should be the initial step in the operation to avoid prolonged abdominal exposure or aortic clamping. Partial graft replacement is best avoided, as the remaining graft usually requires laterreplacement.43 [Q3:E,F,G,H][Q4:C][Q7:B]Iffemoropoplitealaorticreconstruction is planned, it is wise to perform a preoperative venous duplex scan of the legs to confirm that the femoral veins are patent and of adequate calibre (1 cm). [Q5: D]

Whichevertechniqueisused,theimportanceofadequatedebridement,antisepticwashouts and drainage cannot be overstressed. Combination antibiotic cover (beginning immediately preoperatively) to cover any cultured organism and the common pathogens is essential to eliminate infection and prevent catastrophic haemorrhage from anastomotic breakdown. Routine preoperative investigations such as full blood count, urea and electrolyte estimation, chest X-ray and ECG are indicated. Compression stockings are used by some surgeons after femoropopliteal vein harvest to limit ankle swelling but cannot be