Perforations of the upper gastrointestinal tract

Management

Initial management should focus on resuscitation in a systematic process. In some instances, it may be necessary to initiate resuscitative measures at the same time as making the initial clinical assessment. Consideration should be given to involve critical care teams early. Critical care support is indicated in the presence of organ dysfunction, where there is evidence of abnormal physiology (e.g. elevated serum lactate), and where the patient is predicted high risk on relevant scoring systems (e.g. NELA or P-POSSUM risk predictors). Oxygen therapy should be administered, intravenous access achieved and adequate doses of parenteral analgesia should be provided. Intravenous fluids should be prescribed and broad-spectrum antibiotics should be commenced as soon as possible in accordance with the principles of the ‘Sepsis 6’ care bundle. Surgery is the mainstay treatment for most patients with perforated PUD. Patients should proceed to theatre as soon as is practical and following appropriate resuscitation. Delay to surgery is a clear prognostic factor and mortality increases by approximately 2.5% per hour after admission.

Omental patch repair and thorough peritoneal lavage are the principles of surgical management and are sufficient for most patients. The patch is usually a well vascularised omental pedicle but in instances where there is paucity of omentum or it is absent, the falciform ligament may be used. It is the author’s opinion that no attempt should be made to primarily close the ulcer but simply to cover with a viable omental patch ( Fig. 12.2 ). A thorough and extensive peritoneal washout is vital. The value of placing an abdominal drain remains controversial, although there is some evidence that drainage provides no benefit and may indeed be harmful. A drain should therefore be used selectively.

There are several studies including a meta-analysis of randomised controlled trials as well as a large propensity score-matched series favouring minimally invasive surgery over conventional open surgery for selected patients. ^, These reviews confirm a significant reduction in postoperative morbidity (particularly surgical site infection rates), reduced length of hospital stay and a possible mortality benefit. Indeed, there is now sufficient evidence that laparoscopic repair is the standard of care in centres with appropriate equipment and expertise. However, appropriate patient selection is vital to ensure good outcomes. Limitations of minimally invasive surgery are the requirement for a pneumoperitoneum, difficulty removing large food debris and possibly longer operating times. Therefore, patients who are clinically unstable, the elderly patient and in situations where there is gross peritoneal contamination may be better managed with open surgery. Whatever approach is used, the priority should always be the provision of safe and effective surgery.

Gastric ulcers should always be biopsied as they carry a small risk of malignancy. In some circumstances ulcer excision may be a better option where there is sufficient redundant stomach to allow limited resection of the ulcer to ensure healthy tissue for closure and to maximise the amount of tissue for histological assessment. This remains the preferred approach even in circumstances where there is a high index of suspicion that the gastric perforation is malignant. Retrospective analyses demonstrate increased morbidity and mortality when a gastrectomy is performed rather than a simple repair and urgent rather than elective gastrectomy results in an increased positive surgical margin rate, increased inpatient mortality and decreased overall survival. ^, Major oncological resections are best done electively after normal physiology is restored, appropriate work-up including staging is completed, and the full range of treatment options has been discussed at a multidisciplinary team meeting (such as considering a role for preoperative chemotherapy). Similar oncological outcomes including overall survival can be achieved for selected patients by adopting this approach when comparing perforated and non-perforated gastric cancers treated with a curative resection, suggesting there is no reason to rush. ^,

However, there are circumstances where a simple omental patch or limited ulcer excision is not possible or is considered likely to fail. This may occur with large duodenal perforations where the defect may be several centimetres and the surrounding tissue may be extremely friable. There are several procedures described to manage this scenario. The simplest procedure is to create a controlled fistula by placing a tube drain or T-tube into the duodenal defect and approximating the surrounding tissue ( Fig. 12.3 ). This creates a definite leak and therefore should probably be reserved for instances where the patient is very unwell and abnormal physiology dictates a simpler and faster approach. It may be a better option to attempt to repair the duodenum, possibly with a double layer repair or an overlying omental patch, and placing adjacent drains. If the repair leaks, then the drains will create a controlled fistula. In some circumstances, it might be appropriate to consider a resection of the ulcer tissue. This would normally involve an antrectomy with Roux-en-Y reconstruction. This procedure carries the advantage of removing the ulcerated tissue and allowing earlier introduction of enteral nutrition. However, managing the duodenal stump remains challenging and this procedure should only be considered in patients with normal physiology. There are many other techniques described to manage this challenging scenario including a Finney pyloroplasty and creating a duodenojejunostomy. Many of these procedures are now considered historical. Selecting the best approach for an individual patient is the key to maximising chances of a favourable outcome.

Non-operative management

Non-operative management of peptic ulcer perforation may be appropriate for selected patients. Patients who are clinically well, with normal physiology and vital signs and without evidence of peritonitis, may be considered suitable for a trial of non-operative management. Cross-sectional imaging with oral contrast is also useful to support a conservative treatment approach. Minimal peritoneal contamination and no leak of oral contrast would favour a successful outcome. Patients should be maintained nil by mouth, have nasogastric drainage, intravenous fluids, intravenous proton pump inhibitors (PPI) and antibiotics. Frequent review by an experienced clinician is mandatory and any deterioration in the condition of the patient should be interpreted as a failure of non-operative management. Up to 70% of patients may avoid surgery by adopting this strategy and with appropriate patient selection, although it is important to note that non-operative management in patients over 70 years old is associated with a poorer outcome. ^, There is no role for endoscopic management, such as application of clips, in the management of these patients.

Although there are relatively few trials of PPI in relation to UGI perforations including PUD, there is a considerable body of evidence showing a large reduction in bleeding and re-bleeding risk in PUD, and it seems logical that there would be both early and late benefits in relation to perforation. There are data that suggest intermittent dosing is as effective as an infusion and that oral is as effective as intravenous dosing. ^,

In the absence of randomised trials, experienced clinical judgement and local practice should guide antimicrobial choice. Antibiotics are generally continued for 7 days and should be administered by the intravenous route until oral intake is established. In most centres, antifungal agents are reserved for high-risk patients (such as immunosuppressed patients) or those with a positive culture and clinically severe infection.