Surgical revascularization versus endovascular therapy for chronic mesenteric ischemia: A comparative experience
Article Outline
Introduction
Endovascular therapy (percutaneous transluminal angioplasty [PTA] with stenting) has been increasingly applied in patients with chronic mesenteric ischemia (CMI) to avoid morbidities associated with open repair (OR). The purpose of this study was to compare outcomes of PTA/Stent vs OR in patients with symptomatic CMI.
Methods
During the interval of January 1991 to December 2005, 80 consecutive patients presenting with symptomatic CMI underwent elective revascularization. Patients with acute mesenteric ischemia or those with mesenteric revascularization performed as part of complex aneurysm repair were excluded. PTA/Stent (with stenting in 87%) was the initial procedure in 31 patients (42 vessels). OR was performed in 49 patients (88 vessels) and consisted of bypass grafting in 31 (63%), transaortic endarterectomy in 7 (14%), patch angioplasty in 4 (8%), or combined in 7 (15%). Mean follow-up was 15 months in the PTA/Stent group and 42 months in the OR cohort. Study end points included perioperative morbidity, mortality, late survival (Kaplan-Meier), and symptomatic and radiographic recurrence.
Results
Baseline comorbidities, with the exception of heart disease (P = .025) and serum albumin <3.5 g/dL (P = .025), were similar between PTA/Stent and OR patients. The PTA/Stent group had fewer vessels revascularized (1.5 vs 1.8 vessels, P = .001). Hospital length of stay was less for the PTA/Stent group (5.6 vs 16.7 days, P = .001). No difference was noted in in-hospital major morbidity (4/31 vs 2/49, P = .23) or mortality (1/31 vs 1/49, P = .74). Actuarial survival at 2 years was similar between the groups (88% PTA/Stent vs 74% OR, P = .28). There was no difference in the incidence of symptomatic (7/31 [23%] vs 11/49 [22%], P =.98) or radiographic recurrence (10/31 [32%] vs 18/49 [37%], P =.40) between the two groups. Radiographic primary patency (58% vs 90%, P = .001) and primary assisted patency (65% vs 96%, P < .001) at 1 year were lower in the PTA/Stent group compared with OR. Five (16%) of 31 PTA/Stent patients compared with 11 (22%) of 49 OR patients required a second intervention on at least one index vessel at any time (P = .49).
Conclusions
Symptomatic recurrence requiring reintervention is common (overall 16/80 [20%]) after open and endovascular treatment for CMI. PTA/Stent was associated with decreased primary patency, primary assisted patency, and the need for earlier reintervention. In-hospital mortality or major morbidity were similar in patients undergoing PTA/Stent and OR. These findings suggest that OR and PTA/Stent should be applied selectively in CMI patients in accordance with individual patient anatomic and comorbidity considerations.
Open surgical repair (OR) has been the standard treatment for chronic mesenteric ischemia (CMI) since the first successful repair reported by Shaw from the Massachusetts General Hospital in 1958.1 OR, however, has been associated with significant morbidity (5% to 30%) and mortality (5% to 12%) in most series.2, 3, 4, 5, 6, 7, 8 Excessive morbidity is perhaps expected in this patient population because many have significant weight loss, malnutrition, and low albumin levels, which are all predictors of increased morbidity and mortality after any major surgery.9 Symptomatic recurrence requiring reintervention is also common and occurs in 10% to 40% of patients after OR.3, 4, 5, 6, 7, 8, 10
Endovascular therapy, consisting of percutaneous transluminal angioplasty and stenting (PTA/Stent), has been increasingly applied in other vascular beds, in particular renovascular disease, to decrease the morbidity and mortality associated with OR.11 Advancements in the treatment of CMI, namely the application of PTA/Stent, have been based on small case series suggesting its efficacy.12, 13, 14, 15, 16, 17 Although consisting mostly of small single-institution series, the available literature would suggest an increasing emerging role for PTA/Stent because anticipated periprocedural morbidity is low (0% to 18%) and patency of 30% to 90% at 2 years after treatment has been reported.12, 13, 14, 15, 16 Other series have demonstrated improved morbidity and mortality compared with OR, albeit with a decreased primary patency and the need for earlier secondary intervention.10, 18, 19 No randomized, controlled trials have compared open vs endovascular treatments in CMI.
Most comparative series are limited by short-term follow-up or anatomic-related selection bias, or both; that is, more favorable lesions are treated with PTA/Stent rather than OR. The optimal initial treatment for CMI, OR vs PTA/Stent, is as yet an evolving algorithm with respect to evidence-based treatment guidelines. The purpose of this study was to review our results with OR and PTA/Stent in patients undergoing treatment for symptomatic CMI to better define the optimal first-line therapy.
Patients and methods
The Institutional Review Board of the Harvard Medical School and the Massachusetts General Hospital approved the experimental protocol. The office and hospital records of all patients undergoing treatment for CMI on the vascular surgical service at the Massachusetts General Hospital from 1991 to 2005 were retrospectively reviewed. The database included demographic information, cardiovascular risk factor profile, clinical presentation, diagnostic and radiographic information, nutritional status, details of treatment, outcomes, and long-term follow-up data (Table I). The presence of cardiac disease was defined as any patient with a history of myocardial infarction, angina, congestive heart failure, or coronary intervention to include percutaneous transluminal coronary angioplasty or coronary artery bypass grafting. Assessment also included a search of the Social Security Death Index online database and directly contacting the patients or their family when possible.
Table I. Demographic and clinical data
| Risk factor⁎ | PTA/Stent (n = 31), n (%) | OR (n = 49), n (%) | P |
|---|---|---|---|
| Age | 69 ±15 | 65±13 | .25 |
| Female | 21(68) | 32(65) | .82 |
| Smoking | 19(61) | 34(69) | .45 |
| Hypertension | 26(84) | 32(65) | .07 |
| Cardiac disease | 18(58) | 16(33) | .025 |
| COPD | 3(10) | 3(6) | .07 |
| Renal disease† | 6(19) | 10(20) | .90 |
| Hyperlipidemia | 20(65) | 22(45) | .08 |
| PVD | 12(39) | 21(43) | .71 |
| Carotid disease | 4(13) | 14(29) | .10 |
| Diabetes mellitus | 2(6) | 2(4) | .64 |
| Previous colon/SB surgery | 3(10) | 7(14) | .54 |
| Hypercoagulable disorder | 2(6) | 4(8) | .77 |
| Unintentional weight loss >5 lbs | 19(61) | 32(65) | .72 |
| Mean pre-op albumin (g/dL) | 3.73±0.56 | 3.32±0.73 | .025 |
⁎Continuous data presented as mean ± standard deviation. |
†Serum creatinine >1.5 mg/dL. |
Early study end points included technical success, in-hospital minor and major morbidity, in-hospital mortality, and transfer to a rehabilitation facility. Late study end points included symptomatic recurrence, radiographic recurrence, secondary intervention, and death after hospital discharge. Major in-hospital morbidity was defined as bowel infarction, myocardial infarction, stroke, acute renal failure, respiratory failure, or multisystem organ failure. Minor morbid events included any problem or complication after the procedure that resulted in the need for further treatment or prolonged the hospital stay.
Long-term follow-up data included radiographic evidence of hemodynamically significant stenosis, symptomatic recurrence, reintervention, and death. Radiographic follow-up was obtained yearly after discharge or earlier if recurrent symptoms developed. Radiographic recurrence was the finding of evidence of recurrent stenosis (>70%) on the vessels previously undergoing treatment by any conventional imaging modality. Symptomatic recurrence was defined as the return of the original symptoms despite intervention.
Kaplan-Meier estimates of the primary patency, primary assisted patency, and secondary patency were calculated according to reporting standards.20 Only patients with >1 month of radiograph follow-up were included in the patency analysis. Clinical patency, defined as continued relief of symptoms, was calculated on the entire study population, whether or not radiographic follow-up data were present. Failure of clinical patency was defined as evidence of radiographic restenosis or recurrent symptoms.
Univariate analysis was performed on clinical and demographic variables to assess their potential association with study end points. All variables significant at the P = .05 level were included in a multiple logistic regression analysis of the primary end points.
During the study interval, 80 consecutive patients presenting with classic symptomatic CMI underwent elective intervention. Classic CMI was defined as an antecedent history of postprandial abdominal pain associated with significant weight loss, food fear, nausea, vomiting, or diarrhea. Excluded from the study population were 33 patients who underwent urgent or emergency surgery for acute mesenteric ischemia. Patients in whom mesenteric revascularization was performed as part of an open suprarenal or thoracoabdominal aortic aneurysm repair were also excluded.
The OR cohort consisted of 49 patients in whom 88 vessels were revascularized. The PTA/Stent group consisted of 31 patients in whom 42 vessels were treated. The decision to offer endovascular repair as the first line treatment for patients presenting with CMI was based upon multiple factors, including age, comorbidities, and patient preference. The endovascular treatments occurred in the interval 2001 to 2005 as more aggressive use of PTA/Stent was gradually adopted.
A consistent history and radiographic imaging confirming a hemodynamically significant stenosis were required for the diagnosis of CMI before intervention. Many patients (25%) were transferred from outside hospitals for definitive management after a diagnosis of CMI. Therefore, the patients had a variety of preoperative imaging modalities, including duplex ultrasound, magnetic resonance (MRA) and computed tomography angiography (CTA), and angiography. Recurrent stenosis or occlusion was determined by a combination of duplex ultrasound, MRA, CTA, or repeat arteriogram.
Endovascular techniques
In the PTA/Stent cohort of 31 patients, 72 mesenteric vessels were involved. Endovascular therapy was performed on 42 vessels: balloon angioplasty alone was used in four (13%) patients and stenting was used in 27 (87%). In this group, 49 stents (45 balloon expandable, 4 self-expanding) were placed in 38 vessels. (Table II)
Table II. Endovascular procedures performed in 31 patients with chronic mesenteric ischemia
| Variables | Number |
|---|---|
| Patients, n | 31 |
| Vessels involved, n | 72 (2.3 average) |
| Vessels treated, n | 42 (1.5 average) |
| PTA/Stent procedures, n | 31 |
| Stent | 27(87%) |
| PTA alone | 4(13%) |
| Brachial approach | 28(90%) |
| Femoral approach | 3(10%) |
| Arteries involved in 31 patients, n | 72 |
| Celiac alone | 2 |
| SMA alone | 1 |
| IMA alone | 2 |
| Celiac and SMA only | 9 |
| Celiac and IMA only | 1 |
| SMA and IMA | 1 |
| Celiac, SMA, and IMA | 15 |
| Arteries occluded (n = 72) | 27(38%) |
| Arteries stenotic (n = 72) | 45(62%) |
| Arteries treated in 31 patients, n | 42 |
| Celiac alone | 9 |
| SMA alone | 7 |
| IMA alone | 4 |
| Celiac and SMA | 8 |
| Celiac and IMA | 2 |
| SMA and IMA | 1 |
| Celiac, SMA and IMA | 0 |
| Final vessel diameter (mm) | |
| Celiac | 6.4 |
| SMA | 6.0 |
| IMA | 4.4 |
| Stents per vessel | |
| Celiac | 1.3 |
| SMA | 1.1 |
| IMA | 1.0 |
All PTA/Stent procedures were performed in an angiographic suite or an endovascular operating room with a fixed imaging system. Under local anesthesia with conscious sedation, a brachial approach (28/31 [90%]) or femoral approach (3/31 [10%]) was used. The brachial approach was chosen in most patients, reflecting our preference to use the technically more facile antegrade approach.
Upon radiographic confirmation of a hemodynamically significant lesion >70% by visual inspection, a 6F sheath delivery system was advanced proximal to the mesenteric vessels. Translesional pressure gradient measurements were performed in cases where there was a question about the severity of the stenosis. Lesions were crossed with platinum-tipped (.014-inch or .018-inch) or hydrophilic (.014-inch or .018-inch) wires, usually using a multipurpose MPA catheter (Cook Inc, Bloomington, Ind). The sheath was advanced over the MPA catheter to the orifice of the mesenteric vessel. Noncompliant angioplasty balloons equal to the lesion length were inflated to at least the nominal pressure or until the lesion waist disappeared. The balloons were kept inflated for 60 seconds.
Residual lesions (>20%) were routinely dilated with progressively larger-diameter balloons to reach the diameters equal to the intended diameter of the adjacent nondiseased arterial segment. Four (13%) of 31patients underwent primary balloon angioplasty alone after a repeat arteriogram showed no residual waist or dissection. In 27 patients (87%), balloon-expandable stents were used. Most lesions involved the orifice of the mesenteric vessels, and therefore, the most proximal extent of the stent protruded into the aortic lumen. Subintimal angioplasty and recanalization techniques for occlusions were not routinely used in this patient population upon documentation of a complete vessel occlusion owing to concerns for causing a distal dissection. Distal embolic protection, atherectomy devices, or laser were not used during any of the endovascular procedures.
After the procedure, all patients were treated with 300 mg of clopidogrel (if not previously prescribed) and maintained on 75mg/day for 6 to 8 weeks. All patients also received 325 mg of aspirin on the day of the procedure and indefinitely.
Open surgical revascularization
The OR cohort included 49 patients with involvement of 124 vessels. OR was performed on 88 vessels and icluded bypass grafting in 31 (63%), transaortic endarterectomy in 7 (14%), celiac endarterectomy with superior mesenteric artery bypass used as the patch in 6 (12%), local endarterectomy in 4 (8%), or other in 1 (2%; Table III).
Table III. Open surgical procedures performed in 49 patients with chronic mesenteric ischemia
| Variables | Number |
|---|---|
| Patients | 49 |
| Vessels involved | 124 (2.5 average) |
| Vessels treated | 88 (1.8 average) |
| Retroperitoneal approach | 29(59%) |
| Transperitoneal approach | 20(41%) |
| Bypass graft | 31 |
| Antegrade bypass | 24 |
| Retrograde bypass | 7 |
| SMA + Celiac | |
| BPG | 13 |
| SMA BPG alone | 7 |
| Celiac BPG alone | 1 |
| Celiac endarterectomy with SMA BPG | 6 |
| Transaortic endarterectomy alone | 7 |
| + SMA BPG | 2 |
| + SMA patch angioplasty | 7 |
| + Celiac BPG | 1 |
| SMA BPG with IMA reimplantation | 1 |
| Endarterectomy with patch angioplasty | 4 |
| Combined procedures | |
| Renal artery bypass | 14 |
| Infrarenal AAA repair | 8 |
| Aortobifemoral bypass | 6 |
| Arteries involved (49 patients) | 124 |
| Celiac alone | 2 |
| Celiac and SMA | 15 |
| SMA and IMA | 4 |
| Celiac, SMA, and IMA | 28 |
| Arteries occluded | 61/124 (49%) |
| Arteries stenotic | 63/124 (51%) |
| Arteries treated | 88 |
| Celiac alone | 2 |
| SMA alone | 8 |
| IMA alone | 2 |
| Celiac and SMA | 37 |
| SMA and IMA | 1 |
A retroperitoneal approach was used in 29 patients (59%). Prosthetic bypass (Dacron or polytetrafluoroethylene) was used in 27 patients and autologous vein was used in four. The antegrade bypass approach from the supraceliac aorta was used in 23 (74%) of 31 patients and the retrograde bypass from the infrarenal aorta was used in eight (26%). The most frequent bypass graft size (75%) used was 6 mm (range, 5 to 8 mm).
Open mesenteric revascularization was combined with renal artery bypass or endarterectomy in 14 (29%) of 49 patients. An infrarenal abdominal aortic aneurysm repair (8/49 [16%]) or aortobifemoral bypass (6/49 [12%]), or both, were combined with open mesenteric reconstruction. In all of the cases of combined intervention, the patient’s primary presentation was symptomatic CMI and was found to have additional pathology requiring intervention.
Statistical methods
Statistical analysis was performed with SPSS software (SPSS Inc, Chicago, Ill). Univariate analysis of variables potentially associated with study end points was determined using the Fisher exact test for two-by-two contingency tables. All risk ratios (RR) were reported with 95% confidence intervals (CI). Multivariate logistical regression models were constructed by including all variables that were significant at the P = .05 level. Kaplan-Meier estimates were used to examine patency, freedom from secondary intervention, symptom recurrence, and late survival in the study populations. Significance was determined using the Cox-Mantel log-rank statistic.
Results
Patient features
The groups were similar with respect to baseline preoperative clinical and demographic features (Table I), with the exception of cardiac disease and preoperative serum albumin levels. All patients undergoing open mesenteric reconstruction underwent preoperative cardiac stress testing, and none required cardiac catheterization. In the PTA/Stent cohort, five patients (16%) had involvement of one vessel, 10 (32%) had two-vessel involvement, and 16 (52%) had involvement of all three mesenteric arteries. In comparison, the OR group had two patients (4%) with single-vessel involvement, 17 (35%) with two vessels involved, and 30 (61%) patients with all three involved (Table II, Table III).
In comparing the severity of diseased mesenteric vessels, there was no difference in the number of stenotic vessels between PTA/Stent and OR (46/73 [63%] vs 64/126 [51%], P = .28). There was also no difference between the numbers of occluded vessels between the two groups (27/73 [37%] vs 61/126 [49%], P = .08). In the PTA/Stent group, 17 patients (55%) had one vessel treated, 14 patients (45%) had two vessels treated, and no patients had all three revascularized. In comparison, nine OR patients (18%) had one vessel revascularized and 40 patients (82%) had two vessels revascularized.
Overall, the number of vessels involved was not different between the PTA/Stent group and the OR group (mean, 2.4 vs 2.6, P = .18). The number of vessels treated, however, was significantly different between the two groups (PTA/Stent 1.5 vs OR 1.8, P = .001).
Treatment outcomes
Technical success was achieved in 30 (97%) of 31 PTA/Stent procedures. After difficult sheath placement within an infrarenal abdominal aortic aneurysm, the single failure occurred as a balloon-expandable stent migrated from the superior mesenteric artery (SMA) distally into the aneurysm sac on deployment and wire access was lost and could not be re-engaged. Success was achieved in all 49 (100%) of the OR procedures based on preoperative plans for revascularization. Both endovascular and open repair were effective: 27 (87%) of 31 PTA/Stent and 44 (90%) of 49 OR patients experienced relief of symptoms initially.
Major in-hospital morbidity was not significantly different between the two groups at 13% (4/31) for PTA/Stent vs 4% (2/49) for OR (P = .14). In-hospital minor morbidity, to include any adverse event, was present in nine (29%) PTA/Stent patients vs 17 (35%) OR patients and was not significantly different (P = 0.6; Table IV). In-hospital mortality also was not different between the two groups at 3% (1/31) for PTA/Stent vs 2% (1/49) for OR (P = .74). The death in the PTA/Stent group was secondary to a myocardial infarction. The single in-hospital mortality for the OR group was secondary to recurrent pneumonia and intra-abdominal abscesses after a 2-month intensive care unit (ICU) stay.
Table IV. Major and minor complications
| Major complications | |
|---|---|
| PTA/Stent (4/31) | |
| Myocardial infarction resulting in death | 1 |
| Cerebrovascular accident | 1 |
| Acute renal failure | 2 |
| OR (2/49) | |
| Pneumonia/respiratory failure resulting in death | 1 |
| Non-Q wave myocardial infarction | 1 |
| Minor complications | |
| PTA/Stent (9/31) | |
| Brachial artery thrombosis requiring open repair | 5 |
| Cerebrovascular accident | 1 |
| Acute renal failure requiring dialysis | 2 |
| Clostridium difficile colitis | 1 |
| OR (17/49) | |
| Pneumonia | 5 |
| Non-Q wave myocardial infarction | 1 |
| Respiratory failure requiring reintubation | 2 |
| Urinary tract infection | 1 |
| Atrial fibrillation | 1 |
| Decubitus ulcers | 2 |
| Post-op bleed and fascial dehiscence | 1 |
| Enterocutaneous fistula | 1 |
| Splenectomy and pancreatic fistula | 1 |
| Acute cholecystitis requiring cholecystectomy | 1 |
| Negative abdominal exploration | 1 |
Hospital stay was significantly reduced in the PTA/Stent group at a median 4 days (range, 1 to 28 days) compared with a median of 12 days (range, 4 to 102 days) for those undergoing OR (P = .001). ICU stay was decreased as well (0.3 ± 0.7 days vs 7.25 ± 12 days, P = .002). Discharge to a rehabilitation facility was more common in the OR group (17/49 [35%]) compared with the PTA/Stent cohort (1/31[3%], P < .001).
Late outcomes
Mean follow-up in the study population was 15 months (range, 0 to 32 months) in the PTA/Stent group and 42 months (range, 0 to 149 months) in the OR cohort. Radiographic follow-up included a combination of duplex ultrasound, angiogram, CTA, or MRA. Many patients were referred from out of state and returned to their primary physicians after intervention, which limited our ability to have continued radiographic follow-up in a significant number of patients (20/31 PTA/Stent vs 29/49 OR). There was no difference in the incidence of radiographic recurrence between the two groups (12/20 vs 17/29, P = .65).
Two-year estimated Kaplan-Meier survival was not different between the PTA/Stent and OR groups (88% vs 74%, P = .28; Fig 1). Primary radiographic patency at 1 year in the PTA/Stent group was estimated at 58% ± 10% vs 90% ± 4 (P = .001) in the OR group (Fig 2). Primary assisted patency at 1 year increased to 65% ± 9% vs 96% ± 3% (P < .001; Fig 3). Secondary patency at 2 years was lower in the PTA/Stent group at 69% ± 14 vs 87% ± 5 (P = .003).
Fig 1.
Kaplan-Meier actuarial survival comparing endovascular (percutaneous transluminal angioplasty [PTA/Stent], dashed line) and open repair (OR, solid line).
Fig 2.
Kaplan-Meier primary patency (per vessel) comparing endovascular (percutaneous transluminal angioplasty [PTA/Stent], dashed line) and open repair (OR, solid line). *In patients with >30 days radiographic follow-up.
Fig 3.
Kaplan-Meier primary assisted patency comparing endovascular (percutaneous transluminal angioplasty [PTA/Stent], dashed line) and open repair (OR, solid line). *In patients with > 30 days radiographic follow-up.
There was no difference in the incidence of symptomatic recurrence (7/31 vs 11/49, P = .98) between the two groups. Kaplan Meier estimates of the freedom from symptomatic recurrence were 74% ± 9% PTA/Stent vs 91% ± 4% OR (P = .23) at 1 year (Fig 4). Although the overall incidence was not statistically different between the two groups, five (16%) of 31 PTA/Stent patients compared with 11 (22%) of 49 OR patients required a second procedure at any time (P = .49). Kaplan-Meier estimates of the freedom from reintervention were 80% ± 8% PTA/Stent vs 80% ± 6 OR (P = .54) at 2 years (Fig 5). Freedom from open surgical revision was also estimated in the two groups at 3 years and was not different (88% ± 6 vs 85% ± 6, P = .86). Clinical patency was higher in the OR group at 2 years (61% ± 10 PTA/Stent vs 83% ± 6 OR, P = .02; Fig 6).
Fig 4.
Kaplan-Meier freedom from symptomatic recurrence comparing endovascular (percutaneous transluminal angioplasty [PTA/Stent], dashed line) and open repair (OR, solid line).
Fig 5.
Kaplan-Meier freedom from reintervention comparing endovascular (percutaneous transluminal angioplasty [PTA/Stent], dashed line) and open repair (OR, solid line).
Fig 6.
Kaplan-Meier clinical patency comparing comparing endovascular (percutaneous transluminal angioplasty [PTA/Stent], dashed line) and open repair (OR, solid line).
Univariate analysis did not identify risk factors for symptom recurrence, reintervention, or radiographic restenosis. Variables associated with an increased early mortality (<365 days) included age >70 (odds ratio, 3.0; 95% CI 1.1 to 8.2) and a preoperative albumin level <3.5 (odds ratio, 4.8; 95% CI, 1.2 to 32.0). Multivariate logistic regression did not identify risk factors for any of the early or late outcomes.
Discussion
Although atherosclerotic lesions of the mesenteric vessels are common in elderly patients, the clinical syndrome of CMI is uncommon. A recent population-based study using duplex ultrasound imaging found 17.5% of persons >65 years old had a critical stenosis of at least one mesenteric vessel.21 The exact incidence of CMI is unknown, but is estimated at only a few cases per 100,000 population. CMI is a debilitating disease leading to significant weight loss and inanition. Left unrecognized or untreated, CMI may progress to intestinal infarction and death. Unfortunately, the first symptom related to atherosclerotic mesenteric occlusive disease in many patients is acute intestinal infarction. In a recently reported series of patients with acute mesenteric thrombosis, only 20% had an antecedent history of symptoms suggestive of CMI.22 Emergency treatment of acute mesenteric ischemia is associated with a 40% to 60% mortality rate.23, 24
Open surgical repair (OR) has been the gold standard treatment for CMI. Because the incidence of CMI is quite low, few surgeons and institutions have extensive experience in its management. Currently, there is no consensus on the optimal means of revascularization (open vs endovascular), bypass route (antegrade vs retrograde), conduit (prosthetic vs vein), number of vessels revascularized, or even the management of asymptomatic three-vessel mesenteric occlusive disease.
With respect to combined aortic and mesenteric revascularization, some reports suggest an increased morbidity and mortality with such operations and recommend staging the procedures.3, 24 In our series, a combined approach was performed in 19 (39%) of 49 OR patients without an increased in-hospital major morbidity or mortality (1/19 combined approach vs 0/30 mesenteric reconstruction alone, P = .74). However, concomitant treatment of associated aortic pathology is performed only in the limiting circumstances of significant aortic pathology that mandates treatment on its own merit.
OR has been reported to have excellent 5-year symptomatic cure rate of 79% to 92%, and a 5-year primary patency durability rate of 57% to 89% in many series (Table V).2, 3, 4, 5, 6, 7, 8, 24, 25 We found a 5-year freedom from symptomatic recurrence of 66% in our cohort, slightly lower than others. The primary and primary assisted patency rates reported in our series are lower than those reported elsewhere in the literature. Significant selection bias occurred when patency rates were examined among the groups in our study because we excluded from patency analysis those patients without radiographic follow-up after 1 month. Because of the low rate of radiographic follow-up in both groups, those more likely to have radiographic follow-up were also more likely to seek medical care for symptom recurrence and restenosis, thus lowering the patency rates. The clinical patency at 2 years, to include those with and without radiographic follow-up, was 83% in the OR group and 61% in the PTA/Stent group, which compares favorably with the bulk of the literature (Table VI).
Table V. Perioperative and long-term outcome after open surgical revascularization for chronic mesenteric ischemia
| Author, year | Pts (n) | Vessels revascularized | Retrograde bypass | Symptom improvement | Follow-up (months) | Complications | Peri-op mortality | Patency | Symptom recurrence |
|---|---|---|---|---|---|---|---|---|---|
| Mateo,25 1999 | 85 | 130 | 40% | 100% | 36 | 33% | 8% | 76% | 24% |
| Kihara,5 1999 | 42 | 66 | 0% | 96% | 33 | 30% | 10% | 65% | 35% |
| Cho,8 2002 | 25 | 24 | 42% | NA | 60 | 60% | 0% | 57% | 41% |
| Jimenez,6 2002 | 47 | 92 | 0% | 100% | 31 | 66% | 11% | 69% | 9% |
| Park,7 2002 | 98 | 179 | 15% | 93% | 24 | NA | 5% | NA | 20% |
| Present series, 2006 | 49 | 88 | 25% | 90% | 42 | 35% | 2% | 63%⁎ | 35% |
⁎Determined in patients with radiographic follow-up >30 days. |
Table VI. Perioperative and long-term outcome after endovascular revascularization for chronic mesenteric ischemia
| Author, year | Pts (n) | Vessels revascularized | Technical success | Symptom improvement | Follow-up (months) | Complications | Peri-op mortality | Primary patency | Primary assisted patency⁎ | Symptom recurrence |
|---|---|---|---|---|---|---|---|---|---|---|
| Kasirajan,10 2001 | 28 | 32 | 100% | 66% | 136 | 18% | 11% | 73% | NA | 34% |
| Steinmetz,14 2002 | 19 | 19 | 100% | 94% | 31 | 16% | 0% | 69% | 88% | 16% |
| Matsumoto,13 2002 | 33 | 47 | 81% | 82% | 3 | 13% | 0% | 83% | 96.6% | 17% |
| Sharafuddin,12 2003 | 25 | 26 | 96% | 88% | 15 | 12% | 0% | 92% | 92% | 8% |
| AbuRahma,18 2003 | 22 | 24 | 96% | 95% | 26 | 0% | 0% | 30% | NA | 39% |
| Brown,19 2005 | 14 | 18 | 93% | 100% | 13 | 0% | 0% | 27% | NA | 53% |
| Silva,26 2006 | 59 | 79 | 96% | 88% | 38 | NA | 12% | 71% | 83% | 17% |
| Present series, 2006 | 31 | 41 | 97% | 87% | 15 | 29% | 3% | 58%⁎ | 65% | 23% |
⁎Determined in patients with radiographic follow-up >30 days. |
Endovascular mesenteric revascularization has been associated with high technical success, low morbidity, mortality, and decreased hospital stay in many series. In the meta-analysis of seven case series reported by Kasirajan et al,10 the immediate technical success rate was 91% ± 8%. Immediate symptom relief was noted in 79% ± 9%. Follow-up in the endovascular series averaged 27 ± 8 months. Complications occurred in 18% of the cases with a periprocedural mortality rate of 4%. Subjective long-term pain relief was found in 72% ± 10% and radiographic patency was 70% ± 18%.
It is difficult to interpret several of the other recent endovascular series because clinical patency is used as a surrogate for radiographic patency.13, 26 Clearly, an absence of symptoms should not equate to a patent mesenteric vessel. Because of such confusion, we chose to report both radiographic and clinical patencies (Fig 2 and Fig 6).
The Cleveland Clinic group also reported their series of 28 patients (32 vessels) treated by PTA/Stent and compared the group with a historical series of 85 OR cases.10 They concluded that the results of PTA/Stent and OR were similar with respect to morbidity, mortality, and recurrent stenosis. They found that the PTA/Stent group had a significantly higher incidence of recurrent symptoms and recommended that OR should be preferentially offered to patients fit for open mesenteric reconstruction. Recurrent symptoms developed in 11 (39%) of 28 PTA/Stent patients during the study interval.
Recurrent symptoms developed in seven of 31 patients in our PTA/Stent cohort and all seven were found to have recurrent radiographic restenosis. Four underwent successful repeat angioplasty. One had a moderate stenosis with severe reduction of his forward cardiac output by angiogram. He died 6 weeks later from complications of congestive heart failure. Two patients required eventual conversion to OR after multiple PTA/Stent procedures. Of the 20 PTA/Stent patients with radiographic follow-up >1 month, we observed three asymptomatic patients who had evidence of recurrent radiographic restenosis. None of these patients developed worsening symptoms or acute mesenteric ischemia, and none had repeat intervention. We have not intervened in asymptomatic patients; however, these patients are being closely monitored for evidence of clinical deterioration.
As summarized in Table V, most of the literature supports the intuitively logical stance that periprocedural mortality would be diminished with PTA/Stent. The low in-hospital mortality in the present series, 3% PTA/Stent vs 2% OR, is consistent with our series of combined open aortic and renal revascularizations wherein we reported a mortality rate of 4%.27
A recent Mayo clinic series detailed 14-year results of 229 patients undergoing OR or PTA/Stent for CMI. They found that although OR was associated with higher morbidity and longer hospitalization, periprocedural mortality did not differ (2.5% OR vs 3.6% PTA). They also reported that patients undergoing PTA/Stent were five times more likely to develop restenosis, seven times more likely to have recurrent symptoms, and four times more likely to undergo reintervention. They concluded that OR was the preferential first-line treatment in good-risk patients and that PTA/Stent was reserved for those at increased risk for surgical repair.28
The radiographic recurrence rates in most series of PTA/Stent for CMI are 10% to 40%, which is significantly higher than those found in series of renal PTA/Stent, where the 1-year patency rate approached 80% and correlated with final vessel diameter.29 Although aortorenal bypass has been shown to be a more durable procedure compared with renal PTA/Stent, this has clearly not stopped the trend towards renal PTA/Stent becoming the first-line treatment for renal artery stenosis.11 Open surgical revascularization for renal artery stenosis is largely disappearing. The higher recurrence rate of mesenteric vs renal PTA/Stent may be due to the higher percentage of women (about 70%) found to have CMI and that the diameter of their mesenteric vessels may be smaller than their male counterparts. The orientations of the mesenteric vessels with respect to the aorta may also influence the anatomic failure rate, but fundamentally, the reasons for the excessive failure rate of mesenteric stents are unknown.
Limitations of the study include those factors that are inherent in any retrospective study, including interpretation bias, recall bias of patients that were contacted for follow-up, and inherent differences between the PTA/Stent and OR groups that might have influenced treatment. It is possible that major differences in overall mortality in the OR group were related to differences in the comorbid profile that were impossible to control for. The subjective nature of the symptom complex and the myriad of causes for abdominal pain make assessment of symptom recurrence difficult. A significant limitation of the study is the lack of radiographic follow-up >1 month in both cohorts, which makes impossible a comparison of our patency rates with other reported series.
Conclusion
Symptomatic recurrence requiring reintervention is common (16/80, 20%) after intervention for mesenteric ischemia, irrespective of initial endovascular or open repair. This high recurrence rate, and an apparent increased anatomic failure rate compared with other relatively large-bore, high-flow vessels, has significant implications for clinical decision-making in such patients. PTA/Stent was associated with decreased primary patency, primary assisted patency, and the need for earlier reintervention compared with OR. Mortality and major morbidity in patients undergoing PTA/Stent was similar compared with OR (Table VII). In environments where low surgical morbidity and mortality can be anticipated, OR offers superior anatomic results to PTA/Stent and could accordingly be considered first-line treatment in appropriate surgical candidates.
Table VII. Early and late outcomes after revascularization for chronic mesenteric ischemia
| End points | Interval | PTA/Stent (n= 31)⁎ | OR (n= 49)⁎ | P |
|---|---|---|---|---|
| Technical success | 30(97) | 49(100) | .8 | |
| In-hospital mortality | 1(3) | 1(2) | .74 | |
| Major morbidity | 4(13) | 2(4) | .14 | |
| Minor morbidity | 9(29) | 17(35) | .6 | |
| Freedom from | ||||
| Symptomatic recurrence | 1 year | 74%±9% | 91%±4% | .23 |
| 5 years | NA | 66%±9% | ||
| Reintervention | 1 year | 80%±8% | 91%±4% | .54 |
| 5 years | NA | 72%±8% | ||
| Patency | 1 year | 67%±9% | 91%±4% | .02 |
| Clinical | 5 years | NA | 68%±8% | |
| Primary† | 1 year | 58%±10% | 90%±4% | .001 |
| Primary assisted† | 1 year | 65%±9% | 96%±3% | .001 |
| Survival | 1 year | 88%±6% | 85%±5% | .28 |
| 5 years | NA | 65%±7% |
⁎Categoric data presented as number (%); continuous data as means ± standard deviation. |
†Includes only patients with radiographic follow-up >30 days. |
Patient preference, decreased major morbidity, and rapid recovery will continue to drive the preferential initial use of PTA/Stent in many patients, however. Indeed such considerations have driven our preference in recent years to offer PTA/Stent as the initial treatment to most patients, presuming they have suitable lesion anatomy, and we have thus far observed that initial endovascular therapy did not preclude subsequent open surgical bypass. However, the short length of the celiac axis could indeed compromise subsequent surgical options, particularly if it has been previously stented. We consider our current approach of using both PTA/Stent and OR in accordance with individual patient anatomic and comorbidity considerations to be balanced.
Author contributions
References
- . Acute and chronic thrombosis of the mesenteric arteries associated with malabsorption. N Engl J Med. 1958;258:874–878
- . Current results of surgical therapy for chronic mesenteric ischemia. Arch Surg. 1997;132:613–618discussion 8-9
- . Elective surgical treatment of symptomatic chronic mesenteric occlusive disease: early results and late outcomes. J Vasc Surg. 1999;29:821–831discussion 32
- . Mesenteric arterial bypass grafts: early and late results and suggested surgical approach for chronic and acute mesenteric ischemia. Surgery. 1995;118:1–7
- . Risk factors and outcomes following revascularization for chronic mesenteric ischemia. Ann Vasc Surg. 1999;13:37–44
- Durability of antegrade synthetic aortomesenteric bypass for chronic mesenteric ischemia. J Vasc Surg. 2002;35:1078–1084
- Current results of open revascularization for chronic mesenteric ischemia: a standard for comparison. J Vasc Surg. 2002;35:853–859
- . Long-term outcome after mesenteric artery reconstruction: a 37-year experience. J Vasc Surg. 2002;35:453–460
- . Preoperative serum albumin level as a predictor of operative mortality and morbidity: results from the National VA Surgical Risk Study. Arch Surg. 1999;134:36–42
- Chronic mesenteric ischemia: open surgery versus percutaneous angioplasty and stenting. J Vasc Surg. 2001;33:63–71
- Diffusion of new technology for the treatment of renovascular hypertension in the United States: surgical revascularization versus catheter-based therapy, 1988-2001. J Vasc Surg. 2004;40:717–723
- . Endovascular treatment of celiac and mesenteric arteries stenoses: applications and results. J Vasc Surg. 2003;38:692–698
- Percutaneous transluminal angioplasty and stenting in the treatment of chronic mesenteric ischemia: results and longterm followup. J Am Coll Surg. 2002;194(1 suppl):S22–S31
- Endovascular treatment as first choice in chronic intestinal ischemia. Ann Vasc Surg. 2002;16:693–699
- . Stent placement for treatment of mesenteric artery stenoses or occlusions. J Vasc Interv Radiol. 1999;10:861–867
- . Endovascular treatment of chronic mesenteric ischemia: report of five cases. Cardiovasc Intervent Radiol. 1998;21:305–313
- Percutaneous transluminal angioplasty in the treatment of chronic mesenteric ischemia: results and 3 years of follow-up in 23 patients. Abdom Imaging. 1998;23:358–363
- . Angioplasty/stenting of the superior mesenteric artery and celiac trunk: early and late outcomes. J Endovasc Ther. 2003;10:1046–1053
- Mesenteric stenting for chronic mesenteric ischemia. Journal of Vascular Surgery. 2005;42:268–274
- Recommended standards for reports dealing with lower extremity ischemia: revised version. J Vasc Surg. 1997;26:517–538
- Mesenteric artery disease in the elderly. J Vasc Surg. 2004;40:45–52
- . Surgical management of thrombotic acute intestinal ischemia. Ann Surg. 2001;233:801–808
- Contemporary management of acute mesenteric ischemia: factors associated with survival. J Vasc Surg. 2002;35:445–452
- Revascularization of the superior mesenteric artery alone for treatment of intestinal ischemia. J Vasc Surg. 2000;32:37–47
- . Elective surgical treatment of symptomatic chronic mesenteric occlusive disease: early results and late outcomes. J Vasc Surg. 1999;29:821–832
- Endovascular therapy for chronic mesenteric ischemia. J Am Coll Cardiol. 2006;47:944–950
- Preservation of renal function with surgical revascularization in patients with atherosclerotic renovascular disease. J Vasc Surg. 2004;39:322–329
- Oderich GS ST, Bower TC, Misra S, Noel AA, Miskulin JJ, et al. Open versus Endovascular revascularization for chronic mesenteric ischemia: risk stratified outcomes. Presented at: Society for Vascular Surgery Annual Meeting, Philadelphia, Pa, June 3, 2006.
- . Primary renal artery stenting: characteristics and outcomes after 363 procedures. Am Heart J. 2001;142:314–323
Supported in part by the Harold and June Geneen Vascular Research Fund.Competition of interest: none.CME article
PII: S0741-5214(07)00233-9
doi:10.1016/j.jvs.2007.01.067
© 2007 The Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.