The mid-term results of the Investigation of Stent grafts in patients with type B Aortic Dissection (INSTEAD) study, presented recently at the Charing Cross International Symposium (London, UK) and 5th Vienna Interdisciplinary Symposium on Aortic Repair (VISAR) meeting (Vienna, Austria) have left uncertainly as to the role of endovascular treatment of chronic type B aortic dissections following the disclosure that the outcomes indicated no benefit for intervention as a firstline treatment. The results were presented at VISAR by Professor Christoph A Nienaber, Head of the Division of Cardiology and Internal Medicine Department, Rostock University, Germany and principal investigator of the study. Dr Mike Dake, Chair of the Department of Radiology at the University of Virginia Health System, Charlottesville, presented the results of Neinaber’s behalf at the Charing Cross meeting.
The multi-centre INSTEAD study is the first prospective, randomised thoracic clinical study, comparing the outcomes of patients treated for aortic dissection with the Talent thoracic stent graft system (Medtronic) against outcomes for patients managed only with conventional anti-hypertensive treatment. The study recruited 136 patients with uncomplicated type B aortic dissections; 70 patients received a stent graft plus best medial therapy and 66 best medial therapy. The patients were enrolled 11 clinical sites in four European countries (Austria, Germany, France and Italy) from February 2002 to May 2005.
Patients older than 18 years with type B aortic dissection as diagnosed by computed tomography or magnetic resonance angiography were randomised to either a thoracic aortic endoprosthesis and antihypertensive treatment, or a tailored antihypertensive treatment. Only patients in a clinically stable condition and without spontaneous thrombosis of the false lumen after 14 days of the index dissection are considered eligible for study inclusion. The primary outcome measure is all-cause mortality. Secondary outcome variables include conversion to stent and/or surgery, induced thrombosis of the false lumen, cardiovascular morbidity, aortic expansion (N5 mm/y of maximum diameter including true and false lumina), quality of life, and length of intensive care unit and hospital stay.
The primary outcome analysis was a comparison of survival curves for all-cause mortality between the two study groups on an intent-to-treat basis. Survival curves were plotted and the statistical difference between the curves will be assessed by nonparametric log rank test statistics.
The study was designed to find an 18% difference in mortality over one year between the two types of treatment. The null hypothesis would mean no difference in mortality between the two treatment groups. The null hypothesis would be rejected in favour of one alternative if a difference in mortality between the two treatment arms reveals a P value <.05 and/or if the 95% CI of the difference will not include zero. Allowing for a type I error of 5% (a = .05), it was calculated that a sample of 136 patients (68 per arm) would give an 80% power to reject the null hypothesis, calculated under the assumption that the group undergoing stent-graft treatment plus antihypertensive medications has an expected one-year mortality of 5%.
The results
Nienaber revealed that there were a total of nine cross-over patients; seven from medical treatment to stent graft (six for rapid expansion beyond 5.5cms and one for late malperfusion) and two from stent-graft to medical treatment. Interestingly, he noted that when patients crossed over from the best medical to endovascular treatment these patients did ‘remarkably well’. At one-year there was a slightly higher, but not statistically significant, rate of all-cause mortality observed in the stent graft arm (7/70%) versus tailored medical treatment (2/66). These outcomes can be explained by the significant improvement in medical therapy in recent years.
“For uncomplicated type B dissection a primary strategy of tailored anti-hypertensive medical treatment and serial imaging is justified,” concluded Neinaber, “with deferred stent-graft implantation as an option for patients failing to respond to medical management.”
However, Dake in his analysis of the study commented that saying medical therapy of type B dissection is inherently superior to intervention (surgical) is not so simple as only patients (15 to 20%) who went to surgery were those with life-threatening complications. He explained that in reality, the intent to treat was medical in all patients, with a failure rate for this strategy of 15 to 20%. As a result, the magnitude of potential early beneficiaries from an effective less invasive therapy is the 15 to 20% surgical group and the 12 to 14% medical mortality group. Therefore, he suggested that the late beneficiaries (those requiring emergent or elective surgery post-discharge) require a controlled investigation to establish the best form of treatment.
Some commented that continued follow-up of the INSTEAD trial was desirable and funding should be sought. However, Professor Roger Greenhalgh, London, UK, commented that the problems seem to be to define the circumstances for protocol for intervention in uncomplicated type B dissections, which do benefit from intervention. He commented that the situation was similar to the UK Small Aneurysm Trial in which the protocol determined, tested if it was safe to follow patients either if the aneurysm reached 5.5cm or if it became tender or it grew more than one centimetre per year. He suggested progress could be made by attempting to define a protocol and testing it against early intervention. “If early intervention does not win hands-down, then following-up until those criteria are satisfied, would be justified. That would be an entirely different approach to the subject,” he added.
Later in the week at the VISAR meeting, Johannes Lammer also pressed Nienaber along these lines. It is clear that the INSTEAD results have left some uncertainty as to the future of intervention for uncomplicated type B aortic dissection.
