Aneurysmal Disease in Patients With Takayasu Arteritis

DISCUSSION

In this monocentric retrospective study of patients with TA with extended follow-up, aneurysmal disease was demonstrated in almost a third of patients. Prior hypertension, seizure disorders, low-grade fever, and aortic valve regurgitation were more common in patients with aneurysms. Aorta vasculitis involvement had been diagnosed previously in all patients with subsequent arterial aneurysmal lesion. Thoracic aortic aneurysm was most common, followed by abdominal aortic aneurysm, though subclavian, brachiocephalic, and carotid aneurysms were demonstrated. Most patients with aneurysms were treated with GCs, and a substantial proportion also required further immunosuppressives, including IFX. Despite this, patients with TA with aneurysms suffered more vasculitis relapses. To our knowledge, the latter has not been previously described.

Aneurysm prevalence in TA varies considerably. Approximately 31% of our cohort showed arterial aneurysms. Multiple factors may come into play. Race and ethnicity have been shown to affect clinical presentation, as well as angiographic involvement. Our patient population consisted primarily of White patients (43%), with less than a quarter (24%) being of Asian descent (other than India). Two Chinese monocentric retrospective studies reported prevalences of 4.2% to 16.6%.^6,11 Our cohort included only 5 patients of Chinese or Korean descent, of which only 2 had arterial aneurysms. Interestingly, all 4 Indian patients had aneurysmal disease (P < 0.05). Older reports suggested a predilection for thoracic aorta aneurysm in Indian patients with TA,²⁴ although more contemporary studies showed varying frequencies (9-22%).^12,25 Similar to our findings, mixed race cohorts reported higher incidences (France [12-24%],^21,26 South Africa [47%],⁸ and the US [27%]²⁰). Further, in our cohort, aneurysms were found in 10 pediatric patients, with a mean age at disease onset of 13.4 (SD 3.7) years. Aneurysmal lesions have previously been reported in 13% to 65% of pediatric patients with TA.^11,27,28 In 1 study, the interval between pediatric TA onset and aneurysm detection was only 2 years.¹¹

Imaging modalities and duration of follow-up may also contribute to variable frequencies. Radiological^7,29 and surgical^16,30 TA cohorts, in which there is systematic screening at diagnosis and periodic imaging thereafter such as in our center, report aneurysm frequency of 32% to 45% and 48% to 64%, respectively. This is further supported by the high frequency of arterial aneurysm (up to half) recorded in autopsy series.^2,31 The duration of follow-up in our study was considerable (median 11.5 [IQR 7.0-19.0] years), as opposed to other cohorts with a median of 36.0 to 72.0 months.^6,11 Only 1 other small surgical retrospective study reported outcomes of patients with TA with operated thoracic aortic aneurysm, with a follow-up of 7.4 years.³²

Aneurysm development in TA is likely multifactorial, and possibly not solely due to vessel wall inflammation. In patients who are nonarteritic, aneurysms are most often degenerative and arise in association with traditional CV risk factors. High blood pressure was diagnosed in more than two-thirds of our patients with aneurysms, significantly more often than in patients without. Similar rates of hypertensive disease in patients with TA with aneurysms have previously been described.^{6,7,11,12,25,29,30} Interestingly, and as previously reported by others, renovascular hypertension was not more frequent in our patients who experienced aneurysms.^5,7,11,12 Though 1 recent Chinese retrospective study demonstrated increased risk of aneurysm development in patients with TA with elevated low-density lipoprotein,¹³ our cohort, as well as others, did not demonstrate such an association.^6,7,11 Yang et al hypothesized whether tobacco use, at least in men, was a possible contributor to aneurysmal disease, as 10 of 14 male patients with TA with aorta aneurysms were active smokers.⁶ In our cohort, as well as others,^11,13 no difference was found in tobacco use between groups.

Large-vessel inflammation likely contributes to aneurysmal disease process. Similar to our findings, Pan et al reported that a third of their patients with TA with aneurysms presented fever, significantly more than in patients without (OR 6.79).¹³ Previous reports also suggested higher inflammatory markers in patients with TA with arterial aneurysm,^6,11,25,32 with 1 other study demonstrating increased risk of aneurysm (OR 2.85) with ESR elevation.¹³

Previous reports have noted variable aneurysm location, with some demonstrating increased rates of thoracic aorta aneurysm (1-37%),^{6,8,13,26,29,33-36} whereas others showed the abdominal aorta (1-47%) to be more frequently involved.^{7,11,12,25,37} At the last follow-up, close to a quarter of our cohort was diagnosed with large-vessel aneurysmal disease. Thoracic aorta aneurysms (18%) were most prevalent, notably the ascending aorta, followed by abdominal aorta aneurysms (11%). Cheng et al described a higher proportion of Numano Classification type IIa (ie, ascending aorta, aortic arch, and branches) vasculitic involvement in patients with TA with arterial aneurysms, which is similar to our findings.¹¹ Aortic valve regurgitation, a known poor prognostic risk factor,^15,37-39 was also noted more frequently in our cohort with aneurysms, in accordance with previous reports (5-33%).^11,21,25,40 No differences in total burden of arterial stenosis and/or occlusion, nor in peripheral vessel involvement, were found, barring a nonsignificant decrease of iliac artery involvement. However, clinical limb claudication was less frequent in our patients with aneurysmal disease. Such a finding has previously been reported.^11,13 Subclavian and iliac artery involvement were present, respectively, in up to 10% and 3% of our cohort, and 1% to 15% and 0% to 12% in previous studies.^6,35,37 Seizures were more frequently recorded in our cohort of patients with aneurysmal disease. Rates of carotidynia, visual complaints, headaches, and cerebrovascular accidents did not differ significantly between groups. Carotid and vertebral artery aneurysms were found in up to 9% and 2% of our patients, respectively. Rates of extracranial artery aneurysm in TA are quite variable, with 1 report describing them as extremely rare, whereas carotid arteries were described as the second most common site of aneurysm in another report.^11,37,41,42

Management of aneurysms in TA seldom used immunosuppressive therapy, other than systemic GCs. In previously presented studies,^6,11 GCs were used in 62% to 86% of patients with TA with aneurysms. However, immunosuppressives were given in only 3% to 15%. Only Pan et al employed immunosuppressants in two-thirds of their cohort, mostly cyclophosphamide (43%), TCZ (29%), and methotrexate (24%).¹³ Of note, use of TNFi in patients with TA with aneurysms was not extensively reported. The majority of our patients with TA with aneurysms (91%) were treated, at some point, with at least GCs. Conventional immunosuppressants were given in close to three-quarters of these patients, whereas biologics were prescribed in more than a third, with IFX being most frequent. In our cohort, tocilizumab was given only to patients with aneurysms. Revascularization was required in less than half of patients, of which aorta repair was necessary in less than a quarter. No aneurysmal rupture, nor deaths, were recorded. It is not well known whether immunosuppression can play a role in preventing aneurysm growth. In Pan et al, after a median follow-up of 18.3 months on immunosuppressive therapy, 15% of aneurysms showed no change and up to 39% actually demonstrated reduction in size.¹³ Sueyoshi et al studied aortic aneurysm outcome in TA by imaging and reported aneurysm expansion in 53% of their cohort.⁷ After a median of 36.0 months, 3 out of 9 patients with TA with an increase of aneurysm size demonstrated rapid expansion and eventual rupture.⁷ In patients with TA with aneurysmal disease, frequency of rupture is estimated at 3% to 11%.^8,29 The majority of patients with TA with aneurysm (80%) eventually require revascularization,¹¹ with aneurysm operations being performed in more than half (51%).⁶ Although a progressive increase in the number of aneurysms was noted in our study, considering the lack of aneurysmal rupture with low rate of surgical aorta repair, one may wonder whether immunosuppressive therapy may indeed reduce, at least partially, aneurysm progression.

Despite comparable clinical course, vasculitis relapses were more frequent in our patients with aneurysms. Such a finding, to our knowledge, has not previously been published. Patients with TA with aneurysmal disease may indeed suffer from more severe vasculitis. In the latter group, we demonstrated more frequent fever, seizures, and higher ESR at diagnosis. Other cohorts have also reported such findings,^{6,11,13,25,32} as well as a higher prevalence of elevated Indian Takayasu Clinical Activity Score (ITAS2010).¹¹ Prescription of TNFi and interleukin 6 receptor antagonist could also be considered as a surrogate marker of severity, likely explaining increased use of biologics in patients with TA with aneurysms. Similar to previous reports,^20,21,35 more than half of our patients with TA experienced relapse during their disease course, even though more than a third were able to taper off therapy for at least 6 months.⁴³ Successful treatment discontinuation did not seem to alter the rate of aneurysm burden.⁴³ Chronic recurrent granulomatous arterial inflammation, with progressive thickening of aneurysm wall in TA is a common finding.²⁹ One study indeed noted wall thickening in most patients (78%) with TA with aneurysm expansion, including all patients with rapid aneurysm progression and subsequent rupture.⁷ Yang et al also noted that onset of vasculitis preceded aortic aneurysm in most patients, with a median delay of 66.0 months,⁶ whereas Pan et al reported the risk of aneurysm to be 4.2-fold higher in patients with a disease duration > 5 years.¹³ This is further supported by persistent histologically active disease found in surgical arterial specimens from patients with clinically and biologically inactive disease.^20,44

Our study has several limitations, including its retrospective design and small number of patients. Imaging techniques and follow-up were not standardized, especially in the early years of the clinic. Survival analysis could not be undertaken, as individual relapse dates were not recorded. The latter would have indeed been informative, as length of follow-up varied. Use of antiplatelet drugs, anticoagulants, statins, and/or angiotensin-converting enzyme inhibitors was also not recorded.

In conclusion, a significant proportion of our patients with TA were found to have aneurysmal lesions, most often involving the thoracic aorta. Patients with TA with aneurysms experienced more vasculitis relapses. Considering the risk of rupture, aneurysmal disease should be reported systematically in all studies of TA.