Optical coherence tomography tissue tissue typing - Clinical validation

Cardiovascular disease is responsible for 30% of all deaths, in The Netherlands
and worldwide, while ischemic heart disease is the single largest cause of
death in The Netherlands. In spite of a growing range of treatment options,
nearly half of all cardiac deaths are due to acute coronary syndromes. Most of
those are triggered by rupture of a so-called vulnerable plaque. The
histopathology of a vulnerable plaque is largely known: a mildly stenotic,
positively remodeled, eccentric lesion, containing a lipid-rich necrotic core
under a thin fibrous cap that is weakened by inflammation. Rupture of the cap
releases the thrombogenic contents into the bloodstream.
Nevertheless, our understanding is far from complete, resulting in an inability
to accurately identify high-risk plaques in vivo. This lack of diagnostic
capacity is aptly illustrated by the outcomes of the recent PROSPECT trial,
presented at TCT2009. Only about half of the cardiac events in this trial were
caused by lesions that fit the classical vulnerable plaque characteristics, as
assessed by IVUS. In addition, many apparent vulnerable plaques were found that
remained silent during the 3 years of the study. The significant gaps remaining
in our understanding of disease processes in the coronary tree originate in the
limited sensitivity and specificity of in vivo vulnerable plaque detection
techniques, and in the post-mortem nature of the studies that led to the
present picture of the vulnerable plaque.
Accurate data on vascular wall condition can provide valuable data for guidance
of interventions. A recently published case suggests a relationship between
plaque type and long-term clinical outcome of drug-eluting stent placement. In
addition, the occurrence of post-procedural microemboli and post-MI cardiac
enzyme levels have been linked to culprit lesion composition.
Coronary atherosclerosis will be imaged in vivo, using a novel multimodality
imaging validation scheme. We will use independent imaging data as a validation
standard. IVUS clearly shows the presence of calcified tissue, but has a
limited specificity for different soft tissue types. Dense calcium shows up in
IVUS images as a bright reflection, with a dark shadow behind it. The Lipiscan
NIRS technology has a high specificity for detection of lipid-core plaques, and
was shown to produce similar spectra ex vivo and in vivo. It processes a
measured reflection spectrum into a probability for the presence of lipid-core
containing plaque (LCP), which is displayed as a color-coded fold-out map of
the imaged artery segment, called a chemogram. Lipiscan/IVUS collects both data
sets in one pullback, and indicates the probability of LCP circumferentially
around the IVUS data. In addition, the probability of LCP per IVUS cross
section is displayed as a color code, the block chemogram. Joint interpretation
of the data sets allows the operator to identify the location of the lipid
concentration in the vessel wall.
The NIRS data collected by the Lipiscan system will serve as the standard for
the presence of lipids in the artery wall, while grayscale IVUS provides the
standard for dense calcium. A combination catheter Lipiscan/IVUS system was
recently introduced commercially. This dual modality system will be used for
the study, reducing the number of catheters and complication risk, while
ensuring registration between the validation data sets. OCT will be performed
using the Lightlab Imaging C7XR imaging system and Dragonfly catheters (CE mark
& FDA approval since 2010). Lipiscan/IVUS catheters are expected to receive
CE-mark in early 2011.
Pullbacks will be acquired by OCT and by Lipiscan/IVUS. The data sets will be
longitudinally and circumferentially matched landmarks such as side branches.
In the OCT pullbacks, the segments of interest will be identified (2*3 per
study vessel). The imaging catheter will be positioned at such a segment, with
auxiliary information from x ray angiography. A stationary recording will be
made during 4 to 5 heart cycles, with simultaneous ECG recording. Using the
synchronous ECG data, we will select frames in the OCT recording during
end-diastole, when catheter motion is minimal. The end-diastolic frames will be
averaged and processed to tissue optical attenuation images. This procedure was
shown to result in very reproducible attenuation data. The IVUS and NIRS cross
sections, corresponding to the analyzed OCT images, will be scored for
calcifications and NC respectively. Note that we do not anticipate to
distinguish calcium from fibrous tissue based on the optical attenuation alone,
but further development of the quantitative analysis may produce other
quantifiers of the OCT signal that are specific for calcium. No independent
validation exists for macrophage proliferation in vivo. We will apply variance
analysis of the OCT signal, which has been proposed for this purpose.
Further development of the analysis will include the design of a procedure to
eliminate the averaging step. This will allow the analysis of full pullbacks,
which can be matched at regular intervals to create a much larger data set.

We aim to validate the quantitative OC3T tissue characterization method in a
clinical setting, using independent imaging as a validation standard. The
development of this technique was developed based on ex vivo data, and in vivo
validation is needed to demonstrate its clinical value.

This is a single-center, investigator-initiated, prospective, observational,
cross-sectional study. A total of 80 patients with documented stable or
unstable coronary artery disease, including non-ST segment elevation myocardial
infarction (NSTEMI), unstable angina (rest pain without troponin elevation), or
stable (effort) angina pectoris who are scheduled to undergo PCI and who meet
all inclusion/exclusion criteria, will be included.
All patients will be enrolled in the Thoraxcenter, Erasmus MC, Rotterdam.

Introduction of a catheter in the coronary circulation can cause vasospasm,
dissection, or stent damage. These complications can be managed well in the
catheterization room. Although every catheterization inherently carries a risk,
an enhanced risk of the techniques applied in this study have not been
demonstrated. The complication rate of PCI is approximately 1%. The recent
PROSPECT study, in which three coronary arteries per patient were imaged, had a
complication rate of 1.6% (11 in a population of 697 pts with acute coronary
syndrome). In this study, only one vessel will be investigated, hence the
overall risk is probably smaller than this number.
The burden to the participants will be an extension of the PCI procedure by 15
minutes.