The Penn classification, a risk assessment system for acute type A aortic dissection (AAAD), is based on preoperative ischemic conditions. We investigated whether Penn classes predict outcomes after surgery for AAAD. Three hundred fifty-one patients with DeBakey type I AAAD treated surgically, January 1997 to January 2011, were divided into 4 groups per Penn class: Aa (no ischemia, n = 187), Ab (localized ischemia with branch malperfusion, n = 67), Ac (generalized ischemia with circulatory collapse, n = 46), and Abc (localized and generalized ischemia, n = 51). Early and late outcomes were compared between groups. In-hospital mortality was 3% (6 of 187) for Penn Aa, 6% (4 of 67) for Penn Ab, 17% (8 of 46) for Penn Ac, and 22% (11 of 51) for Penn Abc. Multivariate logistic regression analysis showed Penn classes Ac and Abc, operation time >6 hours, and entry in the descending thoracic aorta to be risk factors for in-hospital mortality. Incidences of neurologic, respiratory, and hepatic complications differed between groups. Five-year cumulative survival was 85% in the Penn Aa group, 74% in the Penn Ab group (p = 0.027 vs Penn Aa), 78% in the Penn Ac group, and 67% in the Penn Abc group (p <0.001 vs Penn Aa). In conclusion, morbidity and mortality are high in patients with generalized ischemia. The Penn classification appears to be a useful risk assessment system for AAAD, predictive of outcomes.
Acute type A aortic dissection (AAAD) remains one of the most catastrophic cardiovascular events, and the risk of death is high. The surgical outcomes for AAAD are highly influenced by patients’ preoperative conditions, with hemodynamic shock and organ ischemia, increasing mortality and morbidity. Therefore, appropriate evaluation of a patient’s preoperative risk is crucial for determining the optimum surgical strategy, managing postoperative care, and predicting the surgical outcome. The Penn classification for AAAD, advocated by the Thoracic Aortic Research Group at the University of Pennsylvania, classifies AAAD patients into 1 of 4 groups based on preoperative ischemic presentation: absence of ischemia, localized ischemia with branch malperfusion, generalized ischemia with circulatory collapse (hemodynamic shock), or both types of ischemia. This classification is reported to be a good system for predicting in-hospital mortality associated with AAAD, and a recent study confirmed generalized ischemia and localized plus generalized ischemia, as described by the Penn classification, to be independent predictors of intraoperative mortality. However, only a few studies have investigated clinical outcomes on the basis of the Penn classification, and postoperative complications and late outcomes have not been fully assessed in relation to this classification system. We investigated organ ischemia and hemodynamic shock in relation to early-to-late mortality and morbidity after surgery for AAAD using the Penn classification.
Methods
Between January 1997 and January 2011, 393 consecutive patients with AAAD underwent surgery at Saitama Medical Center, Jichi Medical University, Saitama, Japan. Emergency surgery was performed within 14 days of acute onset of symptoms in all patients, and 95% (372 of 393) of the surgeries were performed within 48 hours of symptom onset. Aortic dissection was diagnosed by enhanced computed tomography (CT) or transthoracic echocardiography. To maintain homogeneity of the study group, we excluded 42 patients with DeBakey type II AAAD (in-hospital mortality: 5%) because these patients are the least likely to have ischemia involving a limb, viscera, kidney, or spinal cord. The remaining 351 patients (193 men, 158 women; mean age, 62.5 ± 11.7 years) with DeBakey type I AAAD made up the study group. Approval of the ethics committee of Jichi Medical University Saitama Medical Center was obtained (reg. no. 2012-72). The need for individual informed consent was waived.
The study patients were divided into 4 groups representing the 4 Penn classes ( Table 1 ). The first is class Aa, characterized by absence of ischemia. The patient is hemodynamically stable, and there is no branch vessel malperfusion. The second is class Ab, characterized by branch vessel malperfusion resulting in localized visceral ischemia causing stroke, paraplegia, renal dysfunction (serum creatinine elevation, oliguria, or both), mesenteric ischemia (abdominal tenderness, intestinal bleeding, marked liver transaminase elevation), and/or ischemic extremity (loss of pulse with or without sensory or motor deficits). Radiographic or surgical evidence of dissection involving the appropriate aortic branch was needed for confirmation of malperfusion. The third is class Ac, characterized by hemodynamic shock that produces generalized ischemia because of cardiac tamponade, coronary malperfusion, acute aortic insufficiency, and/or aortic rupture. Hemodynamic shock was defined as systolic blood pressure <80 mm Hg before surgery. Coronary ischemia was included in Penn class Ac because it is usually associated with pump failure and generalized ischemia. Coronary ischemia was diagnosed on the basis of new ST-segment elevation >0.1 mV, abnormal left ventricular wall motion on the echocardiogram, and significant preoperative serum creatine kinase elevation. In cases in which coronary ischemia developed at the time of weaning from cardiopulmonary bypass (CPB), transesophageal echocardiography was used to confirm diagnosis. The fourth is class Abc, characterized by both localized organ ischemia and generalized ischemia.
| Clinical Presentation | |
|---|---|
| Class Aa | Absence of branch vessel malperfusion or circulatory collapse |
| Class Ab | Branch vessel malperfusion with ischemia |
| Class Ac | Circulatory collapse with or without cardiac involvement |
| Class Abc | Branch vessel malperfusion and circulatory collapse |
The surgical techniques we followed have been described previously. Briefly, we applied the tear-oriented strategy, that is, ascending aorta replacement (or hemiarch replacement) and resection of the primary entry were performed by means of open distal anastomosis during deep hypothermic circulatory arrest at a rectal temperature of 20°C. Total or partial (reconstruction of 1 or 2 branches) aortic arch replacement was performed in patients whose tear was located or extended into the aortic arch. A selective anterograde cerebral perfusion technique was used. When the entry site could not be identified or was identified in the descending thoracic aorta by transesophageal echocardiography, we simply replaced the ascending aorta. The site of arterial cannulation changed with time, preferential using of the subclavian artery over the femoral artery. Recently, apical cannulation has been used in selected patients. The placement technique included the interposition of woven collagen-impregnated or albumin-sealed grafts with Teflon felt reinforcement of the aortic stumps. Aortic root replacement with a composite prosthesis and reimplantation of the coronary arteries by the modified Bentall technique was performed in patients with dilation of the aortic root or an aortic root damaged by the entry tear.
A neurologic deficit was defined as a permanent neurologic abnormality that was confirmed by CT. Temporary neurologic dysfunction and complete resolution by the time of discharge without a suggestive CT abnormality were not diagnosed as a neurologic deficit. Respiratory failure was defined as a need for intubation >48 hours. Renal failure was defined as a newly developed need for renal replacement therapy. Liver dysfunction was defined as hepatocellular jaundice with a bilirubin level of at least 3 times the upper limit of normal (total bilirubin > 3.0 mg/dl), transaminase elevation to at least 5 times the upper limit of normal (alanine aminotransaminase [ALT] > 225 U/L and aspartate aminotransaminase [AST] > 200 U/L), or both. In addition, we assessed the time course of perioperative serum bilirubin and transaminase levels per Penn class, after excluding patients who died. Two patients whose postoperative laboratory data were missing were excluded from liver dysfunction study.
In-hospital data were obtained by a review of hospital records. Serum bilirubin, ALT, and AST levels had been recorded preoperatively and daily or every other day until 7 days after surgery. Preoperative hematologic data (hematocrit level, platelet count) had also been recorded. Follow-up data (median, 50.3 months) included general health condition, aortic rupture or reoperation, survival time, and cause of death, which were ascertained through our outpatient clinic, through written or telephone contact with the patient or a relative or through the local cardiologist. Aortic events were defined as aortic rupture, redissection, reoperation, endovascular stenting, or sudden death. Follow-up data were obtained for 99% (347 of 351) of the study patients.
Data are presented as mean ± SD, median (interquartile range) values, or percentages, as appropriate; p <0.05 was considered statistically significant. Differences in intraoperative and postoperative variables between the Penn classes were analyzed by Kruskal-Wallis test, chi-square test, or Fisher’s exact test, as appropriate. Differences in serum transaminase and bilirubin levels between the Penn Aa class and other Penn classes were analyzed by Mann-Whitney rank sum test, and Bonferroni correction was used for multiple comparisons. To identify risk factors for in-hospital mortality, multivariate stepwise forward logistic regression analysis was performed. Variables with a p value ≤0.10 in univariate analysis were included in a logistic regression model. The following factors were examined age, sex, obesity (body mass index > 30 kg/m 2 ), Marfan syndrome, bicuspid aortic valve, hypertension, diabetes, smoking, chronic obstructive pulmonary disease, hemodialysis, previous coronary disease, previous cerebral vascular disease, peripheral artery disease, chronic kidney disease, cirrhosis, liver dysfunction, previous cardiac surgery, severe aortic insufficiency, Penn classes (Penn Aa, Ab, Ac, and Abc), coma, hemopericardium, location of the entry tear, arterial cannulation site, hemiarch replacement, aortic arch replacement, aortic root replacement, aortic valve replacement, coronary artery bypass grafting, prolonged CPB time (>180 minutes), prolonged myocardial ischemia time (>120 minutes), prolonged operation time (>360 minutes), resection of the primary entry tear, and the amount of blood transfused. Cumulative 5-year survival per Penn class was plotted by the Kaplan-Meier method and analyzed by log-rank test. All statistical analyses were performed with SPSS 17.0 for Windows software (SPSS, Inc., Chicago, IL).
Results
Patients were divided into Penn classes as follows: class Aa, 53% (187 of 353); class Ab, 19% (67 of 353); class Ac, 13% (46 of 351), and class Abc, 14% (51 of 351). Of the 97 patients with generalized ischemia (Penn classes Ac and Abc), 8 underwent preoperative pericardiocentesis before transfer to our hospital, 20 underwent cardiopulmonary resuscitation, and 5 required implantation of an arteriovenous femorofemoral extracorporeal membrane oxygenation device because of sustained hemodynamic shock. Causes of generalized ischemia included cardiac tamponade (n = 69, 71%), followed by coronary ischemia (n = 17, 18%), severe aortic insufficiency (n = 4, 4%), and combined etiologies (n = 7, 7%). Preoperative characteristics are listed in Table 2 . Patient backgrounds, co-morbidities, and medical histories were similar between groups. Operative characteristics are listed in Table 3 . There was no significant between-group difference in the proximal reconstruction technique or the distal extent of replacement. Coronary bypass was performed in 28% of Penn Ac and 24% of Penn Abc patients. Femoral arterial cannulation was performed most frequently in the Penn Ac and Abc groups for prompt institution of CPB. CPB time and aortic clamp time were prolonged, and the amount of intraoperative blood transfusion was increased in the ischemia groups.
| Variable | Total (n = 351) | Class Aa (n = 187) | Class Ab (n = 67) | Class Ac (n = 46) | Class Abc (n = 51) |
|---|---|---|---|---|---|
| Age (yr), mean (SD) | 63 ± 12 | 62 ± 12 | 60 ± 13 | 65 ± 10 | 64 ± 12 |
| Men | 193 (55%) | 96 (51%) | 47 (70%) | 22 (48%) | 28 (55%) |
| Marfan syndrome | 10 (3%) | 7 (4%) | 2 (3%) | 0 | 1 (2%) |
| Bicuspid aortic valve | 3 (1%) | 2 (1%) | 0 | 0 | 1 (2%) |
| Hypertension | 251 (72%) | 132 (71%) | 46 (69%) | 35 (76%) | 38 (75%) |
| Diabetes mellitus | 26 (7%) | 16 (9%) | 2 (3%) | 4 (9%) | 4 (8%) |
| Current smoker | 118 (34%) | 68 (37%) | 27 (40%) | 10 (22%) | 13 (26%) |
| Previous coronary disease | 17 (5%) | 11 (6%) | 1 (2%) | 2 (4%) | 3 (6%) |
| Previous cerebrovascular disease | 27 (8%) | 19 (10%) | 4 (6%) | 1 (2%) | 3 (6%) |
| Hemodialysis | 7 (2%) | 2 (1%) | 2 (3%) | 2 (4%) | 1 (2%) |
| Previous surgery for valvular disease | 4 (1%) | 4 (2%) | 0 | 0 | 0 |
| Previous surgery for thoracic aortic disease | 1 (0.3%) | 1 (0.5%) | 0 | 0 | 0 |
| Severe aortic insufficiency | 34 (10%) | 18 (10%) | 5 (8%) | 7 (15%) | 4 (8%) |
| Organ ischemia | |||||
| Brain | 50 (14%) | 0 | 22 (33%) | 0 | 28 (55%) |
| Spinal cord | 1 (0.3%) | 0 | 0 | 0 | 1 (2%) |
| Coronary | 22 (6%) | 0 | 0 | 11 (23%) | 11 (22%) |
| Visceral | 17 (5%) | 0 | 13 (19%) | 0 | 4 (8%) |
| Kidney | 20 (6%) | 0 | 15 (22%) | 0 | 5 (10%) |
| Extremity | 54 (15%) | 0 | 40 (59%) | 0 | 14 (28%) |
| Preoperative laboratory values | |||||
| Hematocrit (%), mean (SD) | 36 ± 6 | 37 ± 6 | 36 ± 5 | 36 ± 7 | 36 ± 6 |
| Platelet (10 3 /μl), mean (SD) | 19 ± 7 | 20 ± 8 | 18 ± 7 | 19 ± 7 | 18 ± 7 |
| Bilirubin (mg/dl), mean (SD) | 0.9 ± 0.5 | 0.9 ± 0.5 | 0.9 ± 0.6 | 0.9 ± 0.5 | 0.7 ± 0.4 |
| Alanine aminotransaminase (U/L), median (IQR) | 22 (15–39) | 21 (14–31) | 19 (14–32) | 27 (16–67) | 42 (15–120) |
| AST (U/L), median (IQR) | 27 (20–52) | 24 (19–35) | 25 (20–42) | 45 (26–99) | 65 (24–214) |
| Variable | Total (n = 351) | Class Aa (n = 187) | Class Ab (n = 67) | Class Ac (n = 46) | Class Abc (n = 51) | p Value ∗ |
|---|---|---|---|---|---|---|
| Aortic arch replacement | 57 (16%) | 27(14%) | 13 (19%) | 5 (11%) | 12 (24%) | 0.27 |
| Hemiarch replacement | 84 (24%) | 39 (21%) | 21 (31%) | 9 (20%) | 15 (29%) | 0.23 |
| Aortic root replacement | 14 (4%) | 10 (5%) | 1 (2%) | 2 (4%) | 1 (2%) | 0.47 |
| Aortic valve replacement | 4 (1%) | 2 (1%) | 1 (1%) | 0 | 1 (2%) | 0.82 |
| Coronary bypass | 27 (8%) | 2 (1%) | 0 | 13 (28%) | 12 (24%) | <0.01 |
| Entry resection | 259 (74%) | 132 (71%) | 53 (79%) | 32 (70%) | 42 (82%) | 0.23 |
| Arterial cannulation site | ||||||
| Subclavian | 148 (42%) | 90 (48%) | 30 (45%) | 14 (30%) | 14 (28%) | 0.018 |
| Femoral | 110 (31%) | 48 (26%) | 10 (15%) | 23 (50%) | 29 (57%) | <0.01 |
| Subclavian + femoral | 82 (23%) | 46 (25%) | 23 (34%) | 7 (15%) | 6 (12%) | 0.017 |
| Apex | 11 (3%) | 3 (2%) | 4 (6%) | 2 (4%) | 2 (4%) | 0.32 |
| Operation time (min), median (IQR) † | 355 (290–450) | 350 (285–425) | 350 (296–450) | 368 (290–486) | 420 (310–470) | 0.10 |
| CPB time (min), median (IQR) † | 133 (115–170) | 126 (112–155) | 135 (116–160) | 142 (128–202) | 147 (123–204) | <0.01 |
| Aorta clamp time (min), median (IQR) † | 97 (84–119) | 93 (82–114) | 96 (84–111) | 106 (91–126) | 107 (82–139) | 0.02 |
| Blood transfusion (ml), median (IQR) † | 1,800 (1,000–2,920) | 1,560 (880–2,360) | 1960 (680–2,880) | 2,100 (1,510–3,510) | 2,750 (1,560–3,640) | <0.01 |
∗ For differences between Penn classes as analyzed by chi-square test unless otherwise indicated.
Overall in-hospital mortality was 8% (29 of 351). In-hospital mortality differed significantly between Penn classes ( Table 4 ). In fact, 79% (23 of 29) of patients who died presented localized and/or generalized ischemia preoperatively. In-hospital mortality per specific organ ischemia was 13% (7 of 52) in patients with central nervous system ischemia, 27% (6 of 22) in patients with coronary ischemia, 6% (1 of 17) in patients with mesenteric ischemia, 0% (0 of 20) in patients with kidney ischemia, and 13% (7 of 54) in patients with extremity ischemia. Results of multivariate logistic regression analysis for in-hospital morality are listed in Table 5 . Penn classes Ac and Abc were identified as independent risk factors for in-hospital mortality. Prolonged operation time (>6 hours) and primary entry site in the descending thoracic aorta were also identified as risk factors for in-hospital mortality. Early outcomes are listed in relation to Penn classes in Table 4 . There were significant between-group differences in the length of ICU and hospital stays for hospital survivors. Additionally, the incidence of postoperative complications, including neurologic deficits, extracorporeal membrane oxygenation support, prolonged intubation, and liver dysfunction, differed significantly between groups, whereas the incidence of postoperative renal replacement therapy did not. The leading cause of in-hospital mortality was cardiac failure, followed by cerebrovascular disease, bleeding, aortic rupture, and multiple organ failure. There was no significant between-group difference in the cause of in-hospital mortality. The time course of perioperative changes in bilirubin and transaminases is shown in Figure 1 . The Penn Aa group showed neither preoperative nor postoperative serum transaminase or bilirubin elevation. The other Penn groups, the Abc group especially, showed marked increases in preoperative ALT and AST levels, and these elevations remained until postoperative days 4 to 5.
| Variable | Total (n = 351) | Class Aa (n = 187) | Class Ab (n = 67) | Class Ac (n = 46) | Class Abc (n = 51) | p Value ∗ |
|---|---|---|---|---|---|---|
| In-hospital mortality | 8% (29/351) | 3% (6/187) | 6% (4/67) | 17% (8/46) | 22% (11/51) | <0.01 |
| ICU stay (d), median (IQR) † | 6 (4–8) | 5 (4–7) | 6 (4–9) | 6 (4–8) | 6.5 (4–10) | <0.01 |
| Hospital stay (d), median (IQR) † | 23 (18–31) | 22 (17–29) | 22 (17–31) | 23 (15–31) | 24 (15–36) | <0.01 |
| Morbidity | ||||||
| Re-exploration for bleeding | 19 (5%) | 6 (3%) | 5 (8%) | 4 (9%) | 4 (8%) | 0.28 |
| Neurologic deficit | 28 (8%) | 5 (3%) | 7 (10%) | 5 (13%) | 11 (22%) | <0.01 |
| Postoperative ECMO | 9 (3%) | 0 | 2 (3%) | 4 (9%) | 3 (6%) | <0.01 |
| Prolonged intubation >48 h | 135 (39%) | 57 (31%) | 29 (43%) | 21 (46%) | 28 (55%) | <0.01 |
| Liver dysfunction † | 50 (14%) | 14 (8%) | 12 (18%) | 9 (20%) | 15 (29%) | <0.01 |
| Renal replacement therapy ‡ | 4% (15/344) | 2% (4/185) | 9% (6/65) | 4% (2/44) | 6% (3/50) | 0.13 |
| Cause of in-hospital mortality | ||||||
| Cardiac failure | 10 (36%) | 1 (20%) | 1 (25%) | 4 (50%) | 4 (44%) | 0.69 |
| Cerebrovascular disease | 7 (25%) | 1 (20%) | 1 (25%) | 1 (13%) | 4 (44%) | 0.68 |
| Bleeding | 4 (14%) | 0 | 0 | 2 (25%) | 2 (22%) | 0.49 |
| Aortic rupture | 4 (14%) | 1 (20%) | 2 (50%) | 1 (13%) | 0 | 0.10 |
| Multiple organ failure | 3 (11%) | 2 (40%) | 0 | 0 | 1 (11%) | 0.11 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
