Download this article's Factoid (PDF & PPT for Gold Subscribers)

In patients with very late stent thrombosis, optical coherence tomography (OCT) often reveals advanced atherosclerosis with thrombosis at the site of neointimal rupture, according to a small observational study published in the June 2013 issue of JACC: Cardiovascular Imaging.

Investigators led by Seung-Jung Park, MD, PhD, of Asan Medical Center (Seoul, South Korea), used OCT to analyze lesions in 33 patients who presented to their institution with Academic Research Consortium-defined definite very late stent thrombosis.

Just over half of patients (52%) presented with STEMI, while 33% presented with NSTEMI and 15% with unstable angina. The median time from stent implantation was 74.1 months: 61.5 months (IQR 38.7-80.3 months) in 27 DES-treated lesions and 109.1 months (IQR 93.7-135.6 months) in 6 BMS-treated lesions.

Overall, 94% of lesions showed intraluminal thrombi, with similarly high proportions for DES (93%) and BMS (100%). Including both stent types, red thrombi were seen in 79% and white thrombi in 94% of lesions.

Neointimal Rupture Predominates

Neointimal rupture was present in 70% of all lesions, and thin-cap fibroatheroma (TCFA)-containing neointima in 64%, with more ‘vulnerable’ plaque seen with BMS than DES, especially proximal to the minimal lumen area (MLA). More than half of DES had at least 1 strut that was uncovered or malapposed, while no BMS showed those defects (table 1).

Table 1. OCT Findings

 
While in the overall cohort, very late stent thrombosis was associated with neointimal rupture in more than two-thirds of lesions, stent malapposition was seen in fewer than half (42%), and only about two-thirds of those had thrombi within the malapposed segments. In addition, 18% of lesions with malapposition also showed neointimal rupture within the stented segment.

Fracture was detected in 3 Cypher stents, and all of the lesions in those stents had abundant TCFA-containing neointima and intimal rupture near the fracture site. Although expansive remodeling was seen in all of these lesions, only 1 was linked to minor malapposition.

Compared with lesions without neointimal rupture, lesions with rupture were more likely to be associated with STEMI (65% vs. 20%; P = 0.040) and a higher peak CK-MB (163.1 ng/mL vs. 15.7 ng/mL; P = 0.017). In addition, TIMI flow grade 3 was less common in lesions with neointimal rupture (22% vs. 63%; P = 0.044). Malapposition was far more likely to be associated with NSTEMI and unstable angina than with STEMI (44% vs. 6%).

Possible Reasons for Discrepancy from Earlier Studies

The authors observe that the current study differs from previous studies in the prevalence of neointimal rupture, especially in proximity to thrombus. There are 2 plausible explanations for the discrepancy, they say.

First, since the time from implantation to stent thrombosis was longer here, it may have provided more time for the development of neoatherosclerosis, which is the precursor of neointimal rupture and stent thrombosis. Second, in the previous studies, OCT was performed after thrombus aspiration. Recent evidence suggests that aspirate from very late stent thrombosis often contains fragments of atherosclerotic intima and thin fibrous cap, they note, and as a result these may be absent on post-aspiration OCT. Moreover, in pre-aspiration OCT, the presence of red thrombus may obscure neointimal rupture or malapposition, leading to underestimation of “vulnerable intima.”

“Thus, our study suggests that advanced neoatherosclerosis with neointimal rupture is the primary mechanism of [very late stent thrombosis],” Dr. Park and colleagues conclude.

Because patients showing neointimal rupture were more likely to present with STEMI, the underlying mechanism of very late stent thrombosis appears to be as important as the event itself, the investigators say.

The authors caution that this observational study was aimed at identifying the sources of very late stent thrombosis and does not address the predictive power of OCT in asymptomatic patients. Furthermore, the study was underpowered to perform subgroup assessment or multivariable analysis, they add.

OCT ‘the Way to Go’ to ID Thrombosis Pathology

In a telephone interview with TCTMD, Renu Virmani, MD, of CVPath Institute (Gaithersburg, MD), called OCT “the way to go” to examine the morphology behind stent thrombosis. “Now when a patient presents, we have a tool that we can use to look at the plaque and say, ‘What went wrong?’ That is very exciting.”

Dr. Virmani noted that although the median time from implantation to thrombosis was about 5 years for DES, that is about 3 years faster than for BMS and is “dramatically accelerated when you think of the decades it takes to get atherosclerosis in native arteries.”

Most patients in the study received first-generation DES, Dr. Virmani acknowledged, so it will be necessary to determine whether second- and third-generation devices are susceptible to similarly accelerated development of neoatherosclerosis. But investigators need not wait 5 to 10 years for clinical events to occur, she insisted. Instead, they should monitor patients with OCT to see if aggressive anti-inflammatory and statin therapy can prevent or delay in-stent neoatherosclerosis.

Compared with plaque erosion or rupture, stent malapposition as a cause of in-stent thrombosis is “overblown,” Dr. Virmani asserted, adding that  today one “can easily look at OCT and tell if a stent was deployed properly.” In fact, she added, when malapposition does occur in DES, it is likely the result of the [physiological effects of the] drug.

Dr. Virmani concluded that for DES patients who are 5 years out from implantation, OCT monitoring may help prevent a catastrophic event.

Study Details

The DES group included 22 lesions with Cypher stents (Cypher Select, Cordis/Johnson & Johnson, Miami Lakes, FL), 3 with Taxus stents (Boston Scientific, Natick, MA), 1 with the Xience stent (Abbott Vascular, Santa Clara, CA), and 1 with the Pico Elite stent (amg International GmbH, Raesfeld-Erle, Germany).

For most of the study period, OCT images were acquired using the proximal occlusive technique, the 0.019-inch ImageWire, and a commercially available system (LightLab Imaging/St. Jude Medical, Westford, MA). For the last 6 months, OCT images were acquired using a nonocclusive technique with the C7XR system (also LightLab Imaging/St. Jude Medical).

Note: Coauthor Gary S. Mintz, MD, is editor-in-chief of TCTMD. He is also a faculty member of the Cardiovascular Research Foundation, which owns and operates TCTMD.

Related Stories:

• Stent Thrombosis Linked to Uncovered, Malapposed Stent Struts
• Registry Provides Comprehensive Look at Stent Thrombosis Predictors
• Portrait of DES Thrombosis Varies Over Time