PhD Defence André van Petersen

chronic mesenteric ischemia with a focus on collateral circulation

The diagnosis of chronic mesenteric ischemia (CMI) is defined as abdominal symptoms combined with occlusive disease of the mesenteric arteries in the absence of any other, more common, explanation of the symptoms. In this thesis we study patients with (suspicion) of CMI. In the first part, we describe the role and value of mesenteric duplex to determine mesenteric stenosis as well as the role of gastric tonometry (GET) in the workup of patients suspected of CMI and a specific cohort of patients suspect of the celiac artery compression syndrome (CACS) or median arcuate ligament syndrome. In the second part, we evaluate the presence and type of collaterals in mesenteric circulation and their possible compensating roles in preventing ischemia. The third part focuses on treatment of CMI (open or endovascular), and the development and clinical outcome of the endoscopic retroperitoneal approach for celiac artery release.

Part 1: Diagnostics in chronic mesenteric ischemia

In chapter 3, we describe the value of duplex ultrasound imaging in a large cohort of patients suspected of CMI. We conclude that the respiratory cycle does influence measured duplex parameters significantly. The main consequence of these observations is that cutoff values of the different duplex parameters to diagnose stenoses are respiration-dependent, and the different phases of respiration should be part of the duplex protocol. Velocity was significantly higher during expiration than during inspiration. Using ROC curve analysis, we established cutoff values with optimal accuracy, defined as minimal false-negative and false-positive results.

 In chapter 4, we describe the role of gastric tonometry in a large cohort of patients suspected of CMI. In a 7-year period, 316 patients were evaluated for possible CMI. Mesenteric stenoses were found in 50% of patients; 61% had one-vessel, 34% had two-vessel, and 5% had three-vessel disease. CMI was diagnosed in 57% of the patients with single-vessel involvement, in 83% of patients with two-vessel, and in all patients with three-vessel stenoses. This indicates that the role of GET is most crucial in the management of single-vessel disease. It can be concluded that CMI actually consists of two types of disease, single- and multiple-vessel disease. The goal of treatment in patients with single-vessel stenosis is primarily symptom relief, whereas in patients with multiple-vessel stenoses, it is primarily preventing major morbidity and death.

In chapter 5, we prospectively studied the use of gastric exercise tonometry as a key criterion for revascularization treatment in patients with otherwise unexplained abdominal complaints and significant stenosis of only the CA by compression of the arcuate ligament. We found that two-thirds of the patients with a significant compression of the CA and abdominal complaints displayed evidence of gastrointestinal ischemia. Follow-up tonometry showed improvement in all symptom-free patients, and improvement in only a minority (25%) of patients with persistent symptoms. This study strongly suggests an ischemic cause of CACS and provides support for treatment with AC release.

Part 2: Collateral circulation

Chapter 6 reports on the mesenteric collateral pathways and their influence on duplex parameters in relation to mesenteric stenosis in a large cohort. We observed that not all collaterals are “clinically” equal. Some collaterals were visible even on nonselective angiography (grade 2), while others were only visible during selective injection into the mesenteric arteries (grade 1). These grade 2 collaterals are rarely seen in normal subjects, whereas grade 1 collaterals can be seen in subjects with normal or nonsignificant stenosed mesenteric vessels. Collateral flow in patients with an isolated stenosis of the CA was facilitated mainly through the gastroduodenal pathway. In patients with an isolated SMA stenosis, this shifted more toward collateral flow through the arcade of Riolan. In patients with a stenosis of the SMA most collaterals were grade 2 compared to less than half in patients with CA stenosis. When a combined stenosis of CA and SMA was present, collateral circulation was mainly facilitated by grade 2 collaterals through the arcades of Riolan or Drummond.

In patients with a stenosis in only the CA or the SMA, the flow in the unaffected other artery was higher, supporting the idea of compensatory blood flow. An incremental effect of the stenosis was found, and compensatory flow was even greater in the presence of grade 2 collaterals. The data suggest that a 70% stenosis in the mesenteric arteries (especially the SMA) may be the cutoff for collateral development and increased compensatory blood flow. One consequence of the increased duplex parameters in the unaffected mesenteric vessel may be an overestimation of the stenosis severity in this nondiseased vessel.

In chapter 7, the influence of the presence of collaterals on clinical signs of mesenteric ischemia is studied. The pathophysiology of the collateral circulation is not fully understood, and studies on the effect of (mesenteric) collaterals on clinical symptoms are lacking. We found that patients with CACS and extensive mesenteric grade 2 collateral circulation are less likely to respond to CA release. The clinical success of AC release in patients without grade 2 collaterals was 84%. The presence of grade 2 collaterals reduced the success rate of intervention to around 50%.

Taken together, our work has provided solid evidence for an ischemic cause of CACS. We also offered recommendations for the use of a functional test, tonometry, to guide patient selection (chapter 5). It seems that fully developed grade 2 collaterals may prevent ischemic complaints in up to 50% of patients (chapter 7). This in line with the measured increased blood flow in unaffected vessels on duplex ultrasound (chapter 6). Still, 50% of patients with these grade 2 collaterals benefitted from treatment. One explanation could be that the blood flow volume in some collaterals is too low to prevent ischemia. We did not quantitate actual blood flow in the current study. Another explanation could be that metabolic demand differed between responders and nonresponders, but we are unaware of any method to measure this. Whether the presence of grade 2 collaterals is a contra-indication for treatment should be confirmed in prospective studies. Our study certainly points to better appreciation of these large collaterals. The decreased chance of relevant symptom reduction after AC release in patients suspected of CACS with grade 2 collaterals should be debated in the shared decision-making process. Whether these collaterals can also be detected by other imaging modalities (CTA, MRA) requires further investigation.

Part 3: Treatment

Chapter 8 is a single center study, which showed that patients unfit for operative revascularization benefit from endovascular treatment in more than 80% of cases, with an acceptable complication rate and without procedure or CSS-related mortality. This study suggests that the endovascular treatment of CSS is low risk, technically challenging, and has reasonable long-term outcomes. This should also be considered when operating on patients with acute-on-chronic splanchnic ischemia. In cases of ongoing peritonitis, a laparoscopy or laparotomy is advised to assess the existence of transmural necrosis.

During the study period, the role of endovascular treatment has evolved and found its place in higher risk patients. Endovascular treatment has the potential of lower morbidity compared to open surgical treatment with possible equal patency. In chapter 9, we present a summary that reviews the available evidence for endovascular (ER) or open reconstruction (OR) in chronic splanchnic syndrome.

The short-term main outcome parameters were generally similar, although the in-hospital morbidity was better after ER (11% vs 32%) and the intensive care unit and hospital lengths of stay were shorter. Long-term primary anatomic patency (86% vs 51%) and clinical outcome were in favor of OR (88% vs 74%). However, secondary patency was not significantly different between OR and ER (87% vs 83%). This was at the expense of more reinterventions in the ER group.

Analysis from uncontrolled ER series showed no significant differences between percutaneous transluminal angiography (PTA) alone and stenting. Almost all studies, including those with multiple-vessel CMI patients treated with OR, supported the statement that long-term relief of symptoms can be best achieved by the repair of more than one splanchnic artery. However, evidence is lacking. Our analysis of all uncontrolled ER series showed a slight advantage for long-term pain relief in patients treated with multiple-vessel ER at the cost of reinterventions. When looking at complications, thrombosis occurred more often in patients with more vessels treated. Duration of intervention, risk of complications, and (long-term) outcome should be weighed individually. Most complications in ER were access-site related.

On the basis of this review, endovascular repair is recommended in patients with comorbidity, which is often the case in patients with mesenteric ischemia. Open repair is the preferred option for patients who are relatively young, with age <50 years as a guideline, and otherwise fit for surgical repair. However, experience in endovascular repair is growing, and techniques and materials are becoming more advanced. This will result in a further increase of endovascular procedures and less open procedures. Based on an extension of the findings in chapter 8 and 9, we adjust the paradigm “ER is a bridge to OR” to “ER is a bridge to repeat ER”. OR is nearly exclusively reserved for patients in which ER is not feasible.

In chapter 10, we describe a new technique for the release of CA in CACS, and the results are presented. The endoscopic retroperitoneal approach for the release of the CA in CACS was performed in 46 patients. We included preoperative GET to support the CACS diagnosis and postoperative inspiration and expiration DSA to record successful CA release. The technique involved division of only the left crus, which keeps the celiac plexus and the lower esophageal sphincter complex largely intact. Since the right-side of celiac plexus is untouched, this provides a neurogenic origin for complaints in CACS patients unlikely. It also minimizes the occurrence of gastroesophageal reflux disease (GERD) caused by damage to the lower esophageal sphincter. We found an incidence of 9% for GERD after open CACS release, and none after endoscopic release.

In a previous cohort of patients treated by an open subcostal approach that included GET, clinical success was 83%. In the present study, which included endoscopic CA release, we achieved a primary clinical success of 78% and a primary-assisted anatomic and clinical success of 89%. Advantages compared with the open technique are less operative trauma resulting in a shorter mean hospital stay. The present study, which included patients with chronic abdominal symptoms, abnormal GET results, and a respiration-dependent significant compression of the CA, showed the feasibility and efficacy of endoscopic retroperitoneal CA release in combination with endovascular treatment of persisting intraluminal CA stenoses. However, even with GET included, it is still not possible to perfectly distinguish nonsymptomatic from symptomatic CA stenoses.