| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ann Thorac Surg 1999;67:1043-1044
© 1999 The Society of Thoracic Surgeons
a Department of Surgery, Baylor College of Medicine, The Methodist Hospital, 6560 Fannin, Suite 1100, Houston, TX 77030 USA
Invited commentary
Biglioli and associates reported an encouraging trend in their 16-year experience with descending thoracic aortic aneurysm repair in 143 patients. After a 10.6% incidence of paraplegia and paraparesis in the first 94 patients, there were no further neurologic deficits among the subsequent 49. Additionally, both the mean aortic cross-clamp time and in-hospital mortality rate have decreased over time. Based on these improvements, the authors propose ligating all involved intercostal arteries before simple aortic clamping as a means of preventing spinal cord injury. The authors are to be congratulated on achieving excellent results in their recent patients, but meaningful interpretation of their comparative data is difficult. In addition to the inherent limitations of retrospective comparisons using historical control groups, the study illustrates the need for standardized reporting methods in the field of thoracic aortic operations.
When comparing a recent treatment group with a historical control group, the validity of the latter must be scrutinized carefully, particularly when groups with small numbers of patients are used. The crux of the statistically significant reduction in neurologic deficits between groups 2 and 3 lies in the surprisingly high rate of paraplegia and paraparesis in group 2 (6 of 28, 21.3%). This discrepancy between the results for group 2 and those for other contemporary series requires explanation and indicates that the subsequent comparisons must be viewed with caution.
A major obstacle in interpreting the influence of the proposed technique on outcome is the potential for susceptibility bias [1]. Inequalities between the groups in terms of risk factors for paraplegia, rather than the differences in technique, might explain the differences in outcome. The fact that emergency operations, a well-established risk factor for paraplegia and death, were nearly twice as frequent in group 2 than group 3 supports this possibility and could explain the long cross-clamp time and high paraplegia rate in the former group. The prevalence of aortic dissection, which is another established risk factor and potential confounding factor, is not reported.
As alluded to by the authors, their reported reductions in mortality and morbidity rates might be related to evolving refinements in several aspects of perioperative management during the past 15 years, including improvements in the selection and optimization of surgical candidates, surgical technique and proficiency, anesthetic care, and postoperative treatment. Taken as a whole, these improvements in care are a confounding factor that is difficult to quantify and adjust for statistically. However, when a specific aspect of treatment, such as the use of cerebrospinal fluid drainage, is not ubiquitous, the frequency of its use in each group should be reported.
To facilitate meaningful comparisons between groups of patients, both within studies and between studies, all relevant data must be reported in a standardized manner. The incidence of commonly accepted risk factors for morbidity and mortality should be provided for each group being compared. Risk models can then be used to quantify differences in overall risk [2]. Alternatively, statistical methods can adjust for these and other inequalities between groups, thereby minimizing their impact on the subsequent comparisons [1]. The use of standardized definitions for both risk factors and outcomes will further facilitate comparisons. Operative [3], in-hospital, and 30-day mortality rates should all be stated routinely. Actuarial survival graphs should include the number of patients at risk, the standard error, or both for each interval [4].
Clinical research aimed at reducing paraplegia after thoracic aortic operations depends heavily on retrospective analyses. The relative infrequency of descending thoracic and thoracoabdominal aortic operations makes it difficult to do prospective studies with sufficient numbers of patients in a timely manner. Therefore, many observational studies that suggest a beneficial technique will never be validated by subsequent randomized clinical trials. This reliance on retrospective studies makes their proper design and interpretation all the more important [1].
Regarding the practice of sacrificing all involved intercostal arteries during repair of descending thoracic aortic aneurysms, other proponents have reported success with their variations of the technique. Acher and associates [5] documented a 13-year experience with 41 patients, including 25 with acute presentation and 4 with dissection, with no paraplegia or paraparesis. Similarly, there were no neurologic deficits in 55 consecutive patients reported by Griepp and associates [6]. In contrast to these encouraging results in patients with descending thoracic aortic aneurysms, the outcome after sacrifice of the segmental arteries in patients with more extensive aneurysms is substantially worse. Despite the use of meticulous somatosensory-evoked potential monitoring, distal aortic perfusion, cerebrospinal fluid drainage, low-dose heparinization, and mild hypothermia, Griepp and associates [6] reported a 10% incidence of paraplegia and 29% in-hospital mortality rate in patients who had more than 10 pairs of segmental arteries sacrificed. This observation, coupled with the frequent need for subsequent aortic operations that put additional segmental arteries in jeopardy, suggests that preserving as many of the multiple interchangeable inputs as possible could prevent future complications.
The proposal by Biglioli and associates advocating the quick, simple clamping technique as a means of preventing spinal cord ischemia is partially based on the premise that current strategies are inadequate. However, using the following approach consistently since 1987, the senior author [7] has done graft repair of descending thoracic aortic aneurysms in 317 patients. Only 3 patients (0.9%) have had postoperative paraplegia or paraparesis, and there have been no deficits among the 136 patients who have had operations since November 1994. In most cases, graft repair was done using the combination of low-dose heparinization, permissive mild hypothermia, simple proximal aortic clamping, reattachment of critical intercostal arteries, and an open distal anastomosis. This procedure is usually accomplished with less than 30 minutes of aortic cross-clamp time. Left heart bypass is used to provide distal aortic perfusion only in patients who are likely to require longer cross-clamp times, ie, patients with aortic dissection or with aneurysms involving the entire descending thoracic aorta. In the few cases in which proximal aortic clamping is impossible or unsafe, we do the repair during hypothermic circulatory arrest.
References
Related Article
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
