OUP user menu

The diabetes conundrum: despite increasing incidences of coronary disease in diabetic type II patients, their first cathlab presentation is later than expected

Nico Bruining
DOI: http://dx.doi.org/10.1093/eurheartj/ehs441 715-718 First published online: 21 December 2012

This editorial refers to ‘Severity of coronary atherosclerosis in patients with a first acute coronary event: a diabetes paradox’, by M. Niccoli et al., on page 729

Niccoli et al. assessed the severity of coronary artery disease (CAD) in the culprit vessels of diabetic type II as well as non-diabetic patients (n = 167) at the time of their first coronary event employing standard coronary angiography and intracoronary optical coherence tomography (OCT) imaging (n = 72).1 To assess the severity of CAD, they applied the Bogarty, Gensini, and Sullivan scores, and to evaluate the development of collaterals they used the Rentrop score, all based on coronary angiography. Intracoronary imaging by OCT was performed to study coronary plaque morphology, more particularly to quantify the amount of lipids and calcification present in the plaques. This was assessed at the site of the minimal lumen area (MLA) and at the cross-section showing the largest arc of calcium in the culprit vessel. Furthermore, the OCT images were inspected for the presence of calcified nodules.2

The findings of the authors are: more collateral development towards the culprit vessel, less lipids at the MLA site, more severe calcified segments, and more superficial calcified nodules in diabetic type II patients as compared with non-diabetics at the time of their first event. As the author's state: despite more severe coronary calcification, including calcified nodules, potent pro-inflammatory, pro-oxidant, and pro-thrombotic stimuli in diabetic type II patients, these patients seem to experience their first event at a later stage in the process of CAD as compared with non-diabetic patients. This finding is seemingly at odds with general convention, i.e. that the first event for all patients happens at the same stage of CAD. The authors hypothesized that there might be, besides an increased presence of collaterals, also other as yet unidentified protective factors operating in diabetic type II patients, which is an intriguing finding. The authors collected and investigated a patient cohort showing an interesting, and possibly very important, outcome for which they should be complimented.

The authors hypothesize that the lower levels of lipids and higher levels of calcium in diabetic type II patients could help identify plaque characteristics which might be less thrombogenic. This could be one possible explanation, however, and, in contrast to the findings of Niccoli et al., some other studies also employing OCT could not show any difference in plaque characteristics between diabetic and non-diabetic patients.3,4 The authors attribute this to heterogeneous populations and also due to the use of different imaging and analysis methodologies. This observation justifies a further consideration of what has been employed in this study with regard to imaging methods used in other studies and which are currently clinically available.

This current study employed two imaging methods to assess the severity of CAD: angiography and intracoronary imaging by OCT. Both methods have proven their usefulness to image CAD, within their limitations. The analysis of the OCT data has been performed extensively, applying a method that can be easily adapted by others, which warrants praise, although knowledge and experience are required to interpret the image data correctly. This study raises the important question of whether the two imaging methods employed reveal all aspects of CAD and therefore confirm the hypotheses of the authors. A future reproducible study employing additional imaging methods could corroborate their current data.

Visualizing the amount and extent of CAD is not an easy task, and all currently available coronary imaging methods have their strengths and weaknesses both from a methodological point of view and from a more pragmatic, i.e. clinical, point of view. Ideally, one would like to apply all current imaging methods available that could address the hypotheses as raised by the authors, and these are: mutislice computed tomography (MSCT), angiography including fractional flow reserve (FFR), intravascular ultrasound (IVUS), OCT, and near infrared spectroscopy (NIRS) (Figure 1). Angioscopy is, with the exception of Japan, not often employed and hence falls outside the scope of this Editorial. These above-mentioned modalities are often presented as competitors; however, they are actually complementary to each other (Figure 1). In brief, what are some of the strengths and weaknesses of coronary imaging with respect to CAD today?

Figure 1

Imaging of coronary artery disease with in (A) coronary angiography showing collaterals. In (BF), images of an explanted coronary imaged at approximately the same location with histology in (B), magnetic resonance imaging in (C), optical coherence tomography in (D), intravascular ultrasound in (E), and multislice computed tomography in (F). Calcium is visualized differently in all modalities. Histology and magnetic resonance imaging show calcium accurately, but are not available clinically. Optical coherence tomography visualizes calcium also well but cannot visualize large plaques due to high attenuation of the light signal. For large plaques, intravascular ultrasound is a good modality but it cannot penetrate and look ‘through’ and thus behind, calcium. Multislice computed tomography visualizes the complete coronary artery tree; however, due to the ‘blooming’ effect of calcium, assessment of the lumen is difficult in heavy calcified areas. In (G and H), images of a combined intravascular ultrasound and a near infrared spectroscopy catheter showing lipids into the coronary artery wall identified by the yellow colour on the top of the intravascular ultrasound images (H).

The technique of MSCT has become a popular coronary imaging modality as it can show with a relatively easy calcium scan the amount of calcium present in the complete coronary artery tree and this can be quantified by an accepted standard method, which is the Agatston score.5 Applying MSCT, large population studies have been performed and they showed that in diabetic patients there is more calcium in the coronary arteries, identifying higher levels of CAD.6,7 However, MSCT visualizes calcium so well (Figure 1) that it overestimates the ‘real’ amount of calcium. This affects visualization and quantification of the coronary lumen in contrast-enhanced MSCT scans in segments which are heavily calcified. There is also the concern of extra ionizing radiation for the patients, but, today, with continuously decreasing radiation levels of newer generations of MSCT scanners, that might become less of a problem.8

Angiography is always needed to perform the intervention, and perhaps a functional test such as FFR, although not an imaging modality, might possibly be a better indicator to quantify the physiological and functional assessment of the severity of the culprit stenosis.9

Until a few years ago, in the pre-OCT era, IVUS was the reference method for visualizing and quantifying the extent and morphology of CAD and to evaluate new therapies. It is capable of visualizing most of the coronary plaque; however, it cannot ‘look’ behind heavily calcified areas10 (Figure 1). It shows the calcium interface as seen from the coronary lumen only, and it is mostly only the length of the calcified segment and the arcs of calcium in individual cross-sections that are measured. Unfortunately, there is no widely accepted standard calcium quantification method developed yet for IVUS data analogous to the Agatston score for MSCT today.

With the arrival of OCT, this imaging modality has become available in the cathlab, allowing visualization of the coronary artery with cross-sectional images at a resolution comparable with that of histopathology using near infrared light11 (Figure 1). Its high resolution and the current rapid non-occlusive acquisition technique explain its popularity in the cathlab armamentarium. Light can easily travel through calcified areas, showing them clearly on the cross-sections; however, it can only penetrate the first 1–3 mm of other tissue types and is very limited in visualizing thrombi and lipidic areas due to a high attenuation of the light in these tissue types (Figure 1).

The NIRS catheter-based method is capable of making a chemogram of the coronary artery, operating more or less in a similar fashion to IVUS and/or OCT in the cathlab.12 During the pullback of the catheter, the chemical composition of the coronary plaque is collected and the presence of lipids is quantitatively visualized. It does not, however, visualize the plaque, as do IVUS or OCT, which could be overcome by a combination catheter which uses IVUS. The method has recently received Food and Drug Administration (FDA) approval. A combination of this method with a visualization method such as IVUS (Figure 1) is a great tool to assess the morphology of CAD.

Reviewing the above-mentioned imaging modalities with respect to this and other studies and as also stated by the Niccoli et al., a future additional study is warranted and should include some if not all of these imaging modalities, naturally accompanied by biochemical analyses. This future study could therefore overcome most of the mentioned limitations. However, from both a clinical and a financial perspective, it could be difficult to justify and execute such a multimodality imaging study today.

Over the past decade, there has been a growing interest in trying to identify the so-called vulnerable plaques (VPs) in the coronary artery.13 However, the results were not optimal and interest shifted towards identifying the vulnerable patient. None of the applied imaging modalities in the VP studies, even when combined with biomarker evaluation, was capable of unambiguously identifying VPs.14 One could argue that the current study exceeds that objective by instead of trying to identify plaques having a high causal characteristic, tries to identify why severe plaques in diabetic patients did not cause a coronary event earlier. This is a difficult objective to address. The complexity lies not only in having access to the right tools, but also in the timing, e.g. at what point imaging should be performed in order to avoid missing the crucial stages during the natural progression of CAD. In addition to these questions, and what has been suggested in the past, might diabetic patients have limited levels of warning signs?15 Also, when they ultimately are in the cathlab for the first time, the problems are larger as compared with other patients in terms of the extra development of collaterals, and thus should ischaemic areas also be quantified for a thorough assessment of the stage of CAD and the damage it has caused?

Regarding invasive imaging and as in other cardiovascular treatment fields, new technological developments are currently ongoing; knowledge is growing. In spite of this, remarkably some old methodologies are rediscovered, updated, and improved, and applied again in the clinical field. Seen in that light, and the previous somewhat disappointing VP imaging studies, FFR, non-occlusive rapid-scan OCT, the NIRS method, and minimized radiation with MSCT generally enjoy broad acceptance amongst practitioners.

It is timely, based on this current study, to re-evaluate if the technological developments together with the reported findings warrant defining a future ‘ideal’ study to observe the natural history of CAD, in all different patient groups, and to try to identify if there might indeed be protective mechanisms in place, or that it is something else?


The author wishes to express his gratitude to Mr P. Cummins for his editorial work on this manuscript.

Conflict of interest: none declared.


  • The opinions expressed in this article are not necessarily those of the Editors of the European Heart Journal or of the European Society of Cardiology.