The precise role for inflammation in plaque vulnerability remains enigmatic. This sub-study in the PARISK study shows an independent inverse association between an increase in leukocyte count (within fourteen days after the ischemic event) and presence of LRNC on plaque MRI in the sub-acute moment after an ischemic event in patients with symptomatic ICAS < 70%. This finding is not in line with the current hypothesis on the vulnerable plaque. However, to our understanding this is the first study that investigated the possible association between leukocyte count and vulnerable plaque characteristics.
Since the beginning of this millennium a lot of papers were published regarding plaque vulnerability and the specific vulnerable plaque characteristics on MRI in the carotid plaque. Some of these plaque characteristics were associated with a higher risk for either recurrent or first stroke (5, 9). Gupta et al. published one of the first systematic reviews on ischemic stroke risk for some of these vulnerable plaque characteristics with a meta-analysis that did show an OR for ischemic stroke or TIA combined for IPH of 4.59 (95% CI 2.91–7.24), 3.00 (95% CI 1.51-5-95) for LRNC and 5.93 (95% CI 2.65–13.20) for TRFC (9). Recently, Schindler et al. reported a hazard ratio (HR) of 11.0 for IPH specifically for ipsilateral stroke (5). These high OR/HR fuelled the interest in researching the a potential role of inflammation by investigating leukocytes and plaque characteristics.
For a long time, inflammation has been regarded a key factor in initiation and progression of atherosclerosis. Yet, recently studies focussing on treatment of this inflammatory reaction published their results (10–12). In the Canakinumab Anti-Inflammatory Thrombosis Outcomes Study (CANTOS) patients with recent myocardial infarction and elevated high sensitivity C-reactive protein were randomized between several doses of canakinumab (a monoclonal antibody against interleukin-1β) and placebo. They showed a HR of 0.85 (95% CI 0.74–0.98) with p = 0.021 for combined cardiovascular outcome measure (recurrent myocardial infarction, ischemic stroke or other cardiovascular death) (11). Alike CANTOS, in the Low-dose colchicine for secondary prevention of cardiovascular disease 2 (LoDoCo-2 trial) they tried to reduce recurrent myocardial infarction, ischemic stroke and other vascular death, but then with 0.5g colchicine once daily in patients with chronic coronary disease. They found a HR of 0.69 (95% CI 0.57–0.83) with p < 0.001 for the primary outcome (12). In contrast to canakinumab, colchicine is an old and inexpensive drug that has been used for a long time in treatment of gout. These studies show promising results for treatment of inflammation in cardiovascular disease.
All these studies were done in patients after myocardial infarction. In stroke patients the COlchiciNe for prevention of Vascular Inflammation in Non-CardioEmbolic stroke (CONVINCE) trial is currently including patients with recent non-cardio embolic TIA or ischemic stroke. In CONVINCE patients are randomised between standard stroke treatment or standard treatment plus low dose colchicine (13). It will nonetheless take some more years before these results are published.
Even though, the studies mentioned above, did show a treatment effect towards prevention of recurrent cerebrovascular events in general and most evidently in a decrease of ischemic stroke cases, no evident blood biomarkers for recurrent stroke have been identified. Leukocyte count could be considered as a potential biomarker, since we found an inverse association between leukocyte count as determined in the sub-acute moment and LRNC. However, the found association was the opposite of what was hypothesized, with a lower prevalence for LRNC in ICAS when a higher leukocyte count is found. The role of leukocytes in predicting recurrent cardiovascular events should be tested in the future.
A potential explanation for the inverse relation that was found, is the general nature of leukocyte count as a mix of several types of leukocytes. Fani et al. recently investigated the association between blood immunity markers, both from the innate and adaptive immune system, and vulnerable plaque characteristics in the Rotterdam Study (14). The Rotterdam Study is a prospective population-based cohort study that started in 1989. Fani et al. used 1602 patients with subclinical carotid atherosclerotic disease. They found that the leukocytes that are part of the innate immune system (macrophages, granulocytes and platelets) were associated with larger plaque and greater plaque thickness and that the monocyte-to-lymphocyte ratio was associated with higher incidence of LRNC. On the other hand the lymphocytes were associated with smaller carotid plaques and a lower prevalence of IPH. This could be an important reason to assess the leukocyte differential count and calculate the ratios used in their study and make a distinction between the innate and the adaptive immune system in future research. In the investigated setting of the presented study, this association shows the opposite relation as expected, this could be due to a higher proportion of lymphocytes in the researched population. This could however not be investigated since only the leukocyte count was available.
This study has some limitations. Firstly, the relatively low number of patients included in the analyses. This was mainly due to the fact that this sub study was not planned when the study was designed and so leukocyte counts were not available in all patients.
Secondly, there is a risk of confounding because of a possibly higher leukocyte count through necrotic brain tissue after the ischemic event. Correcting for this potential confounder was difficult, since NIH Stroke Scale scores (NIHSS) were not available for the included patients. Nonetheless, in this patient population the estimated effect will probably be minimal, because more than half of the patients suffered from a TIA or amaurosis fugax; and only minor stroke patients could be included (modified Rankin scale ≤ 3). In these patients the volume of necrotic brain tissue is generally low and thus the effect on leukocyte count. Finally, the last limitation of our study is the cross-sectional study design. Due to this study design we were only able to show any present associations and were unable to investigate the prediction of stroke risk.
The strengths of this study are evident as well. PARISK was a prospective, well defined, diagnostic cohort study with very complete baseline data. Moreover, it was performed in Dutch comprehensive stroke centres with large experience and specific expertise in MRI plaque imaging. The first strength of this study is the state of the art MR imaging and MDCTA that were performed in these patients (8). Both imaging modalities used in this sub study have shown high specificity and sensitivity for detecting the vulnerable plaque characteristics they are used for and so create high quality data (15–17).
Secondly, the patients included in PARISK all suffer from mild to moderate carotid stenosis, with a minimum of 30% measured with ECST criteria. This is the range in which prior research on ICAS showed the least benefit of revascularisation through carotid endarterectomy or carotid stenting (18, 19). The investigated patient population could therefore be the group in which carotid revascularisation is not performed and thus be of great interest in prevention of recurrent stroke. Potentially, patients in which the cause of stroke is stated to be cryptogenic, could be caused by vulnerable plaques in the carotid artery but with a percentage stenosis < 50%. Since these patients were included in the PARISK study, it could give a more accurate depiction of the actual situation within the overall patient population at risk of stroke out of ICAS. This is a valuable inclusion criterion of PARISK.
The final strength of this study is the use of leukocytes that were determined within two weeks around the ischemic event. With a median of one day (IQR 0.00–4.00), it decreases the risk of confounding due to necrotic brain (or eye) tissue. Furthermore, this sub-acute leukocyte count could be a valuable and easy to detect value of the general inflammatory activity within these patients, may be before the occurrence of the ischemic event as well. This hypothesis deserves future attention.
As stated above, this sub study did find an association between leukocyte count and LRNC. The direction of the relation is opposite from what was expected. This could be due to an increase in inflammatory cell count because of the cytokines that are excreted and on the plaque surface that is in contact with the streaming blood. These cytokines could increase the production of leukocytes. It could be hypothesized that plaques that excrete more cytokines are in another phase of plaque progression and thus have less LRNC compared to plaques that show LRNC more frequently and in parallel these patients could show less circulatory leukocytes. This statement is highly hypothetical and deserves further attention in future research, both in clinical as in more translational or fundamental research. At least it could be established that more research into inflammatory biomarkers seems a feasible but unexplored research field and is highly needed.
Future research should determine if the found association is useful in further risk determination. Additionally, the PARISK study could be used to determine recurrent stroke risk on basis of multiple vulnerable plaque characteristics in the form of a new risk model for ICAS. Next to these analyses, a new prospective study should be set up in which patients get blood drawn at pre-specified moments within the (sub-)acute setting (for example on the emergency department, 24 hours, three days, one week and two weeks after admission), preferably with leukocyte differential count, instead of the general leukocyte count that was used in this study. Several studies showed usefulness of using ratios between some of the components of the differential count. For example, the value of neutrophil-to-lymphocyte ratio was investigated for predicting presence of carotid atherosclerosis.(20) This use seems impractical since most patients get carotid imaging quickly, either during presentation on the emergency department or in the upcoming days, but it could be seen as indirect evidence of the importance of the inflammatory reaction inside the carotid vessel wall. This could be helpful in initiation of future studies into treatment of the inflammatory reaction in atherosclerosis. Besides this, the leukocyte differential count has been evaluated for use of predicting outcome of cardiovascular events, both stroke and myocardial infarction (21, 22).
Finally, a future prospective study should, next to leukocyte differential count and other inflammatory biomarkers, at least include carotid MRI and ideally one or several forms of imaging parameters that could detect plaque inflammation, in particular positron emission tomography (PET) CT or dynamic contrast enhanced (DCE-)MRI. In that way we could learn more on plaque inflammation in relation to time. It seems valuable to assess the course of the investigated potential biomarkers over time as well.