key: cord-352901-ia34l2ml
authors: Natalello, Gerlando; De Luca, Giacomo; Gigante, Laura; Campochiaro, Corrado; De Lorenzis, Enrico; Verardi, Lucrezia; Paglionico, Annamaria; Petricca, Luca; Martone, Anna Maria; Calvisi, Stefania; Ripa, Marco; Cavalli, Giulio; Della-Torre, Emanuel; Tresoldi, Moreno; Landi, Francesco; Bosello, Silvia Laura; Gremese, Elisa; Dagna, Lorenzo
title: Nailfold capillaroscopy findings in patients with coronavirus disease 19: Broadening the spectrum of covid-19 microvascular involvement()
date: 2020-09-17
journal: Microvasc Res
DOI: 10.1016/j.mvr.2020.104071
sha: 
doc_id: 352901
cord_uid: ia34l2ml

OBJECTIVE: Increasing evidence points to endothelial dysfunction as a key pathophysiological factor in coronavirus disease-2019 (COVID-19). No specific methods have been identified to predict, detect and quantify the microvascular alterations during COVID-19. Our aim was to assess microvasculature through nailfold videocapillaroscopy (NVC) in COVID-19 patients. METHODS: We performed NVC in patients with a confirmed diagnosis of COVID-19 pneumonia. Elementary alterations were reported for each finger according to a semi-quantitative score. Capillary density, number of enlarged and giant capillaries, number of micro-hemorrhages and micro-thrombosis (NEMO score) were registered. RESULTS: We enrolled 82 patients (mean age 58.8 ± 13.2 years, male 68.3%) of whom 28 during the hospitalization and 54 after recovery and hospital discharge. At NVC examination we found abnormalities classifiable as non-specific pattern in 53 patients (64.6%). Common abnormalities were pericapillary edema (80.5%), enlarged capillaries (61.0%), sludge flow (53.7%), meandering capillaries and reduced capillary density (50.0%). No pictures suggestive of scleroderma pattern have been observed. Acute COVID-19 patients, compared to recovered patients, showed a higher prevalence of hemosiderin deposits as a result of micro-hemorrhages (P = .027) and micro-thrombosis (P < .016), sludge flow (P = .001), and pericapillary edema (P < .001), while recovered patients showed a higher prevalence of enlarged capillaries (P < .001), loss of capillaries (P = .002), meandering capillaries (P < .001), and empty dermal papillae (P = .006). CONCLUSION: COVID-19 patients present microvascular abnormalities at NVC. Currently ill and recovered subjects are characterized by a different distribution of elementary capillaroscopic alterations, resembling acute and post-acute microvascular damage. Further studies are needed to assess the clinical relevance of NVC in COVID-19.

involvement and the damage related to uncontrolled systemic phlogosis [3] [4] [5] [6] .

Dysfunctional endothelial activation is a phenomenon now recognized during COVID-19 [7] . The pathogenetic mechanisms of this process are still not clear and could depend both on a stimulus secondary to the hyperinflammatory state and a direct pathogenetic role of the viral agent, as suggested by the finding of indirect signs of endothelial activation already in the first days of the disease [8] . Even the first pathological evidence seems to support the centrality of these phenomena, and the histological finding of acral vasculitis with the presence of fibrin thrombus has been reported [9, 10] .

Clinically, patients easily develop thromboembolism. The greatest evidence poses for a prevalent commitment at the level of the lung compartment, the target organ of the virus, while the association with thromboembolic phenomena at different districts -which have also been described -is more nuanced [11] .

Disseminated intravascular coagulation is mainly to refer to the severe cases with an advanced deterioration of the clinical picture.

As yet, no specific methods have been identified to predict, detect and quantify the endothelial alteration during COVID-19. In clinical practice, the dosage of humoral markers of fibrin degradation and systemic inflammation could be useful, also for the prognostic role [12] . Some case series have shown a significant prevalence in both acutely ill and convalescent patients of antibodies related to interference with the phospholipid-dependent coagulation pathways, and in particular of lupus anticoagulant [13, 14] . The anticoagulant treatment with heparin and derivatives, even in the absence of overt thromboembolism, is increasingly widespread in the therapeutic schemes for hospitalized and non-hospitalized patients, with promising results [15] .

Nailfold capillaroscopy (NVC) is an investigation capable of identifying the morphological alterations of microcirculation at the level of the capillaries of the nail bed. This method is widely used in the rheumatology field for the diagnostic classification of vascular acro-syndromes, thanks to its ability to The present study aimed to assess microvasculature and to characterize NVC abnormalities in acute and recovered COVID-19 patients

Eighty-two adult patients affected by COVID-19 pneumonia, diagnosed by laboratory test (positive nasopharyngeal-swab and/or SARS-CoV2 serology) and suggestive chest imaging, were enrolled at IRCCS San Raffaele Hospital (Milan, Italy) and Fondazione Policlinico Agostino Gemelli IRCCS (Rome, Italy).

In a first group of consecutive patients, the NVC was performed during hospitalization, while in a second group of consecutive patients was performed after discharge (i.e. patients' recovery based on clinical improvement and after repeated negative swabs confirmation) in an outpatient setting.

A comprehensive medical and pharmacological history was obtained for all patients at each center.

Specifically, the following data were collected: COVID-19 onset symptoms, duration and outcome, clinical and radiologic characteristics of COVID-19 pneumonia, lung parameters and need for supplemental oxygen therapy, inflammatory and coagulation blood markers, thromboembolic events or other COVID-19 related complications, comorbidities, body mass index (BMI), concomitant therapies and COVID-19 specific treatments. The presence of pre-existing or recent acral symptoms with respect to COVID-19 was also investigated.

Patients with a positive history of traumatization or repeated micro-traumatization of the hands in the last two weeks (e.g. onychophagy, use of vibrating instruments and use of musical instrument), and patients with diagnosed or suspected CTDs were excluded from the study.

All patients provided written informed consent. The study was approved by the Ethics Committees at each Institution (Milan: approval no. 34/int/2020; Rome: approval no. 0024185/20).

A specialized VideoCap 3.0 tissue-contact type biomicroscope providing color images at 200x magnification (DS Medica, Milan, Italy) and dedicated software for image analysis (version 10.0) with standard settings for NVC was used at both sites. The procedure was conducted at each site by two physicians at the same level of experience, already practicing capillaroscopy (image acquisition and analysis) for many years in their routine clinical practice, and applying standard recommendations for patient preparation and execution of the exam [17] . All fingers except the thumb were examined J o u r n a l P r e -p r o o f bilaterally. Four different images were consecutively recorded for each finger, in order to accurately characterize the central area of the nail bed (32 images per patient in total). The images were assessed dynamically and, at the end of the examination, statically by the executing physician. Considering the novelty of the pathology under study and the absence of specific standardization and previous data, to better distinguish all the possible microvascular alterations, we considered a variety of elemental alterations of capillaries at the distal row, as previously described or defined in details below: 11. avascular area: distance greater than 500 μm between two adjacent capillary loops from the distal rows [17] ;

12. empty dermal papilla: one or more missing capillary at the expected place inside dermal papilla which does not reach the extent to define an avascular area.

A widely used semi-quantitative scoring for features 1-7 was applied [19] . Specifically, for each capillary abnormality the score for each digit was calculated as follow: absence of alterations (0), less than 33% of examined capillaries altered (1), 33-66% of examined capillaries altered (2), more than 66% of examined capillaries altered (3) . For the definition of lower capillary density, the percentage refers to reduction with respect to a normal mean value of 9 capillaries per linear millimeter. The average score values from the eight digits were added, and the final value divided for eight fingers; the resulting value represents the score for each parameter analyzed. Since the score was previously developed for scleroderma patients, in whom a high frequency of major alterations is expected, we decided to design and apply in parallel a further semi-quantitative scoring approach, to avoid losing sensitivity. Indeed, for each feature from 1 to 12, a synthetic score was then assigned according to the following criteria: absent (0), present at least one time in only one finger (1), present at least one time in 2 to 4 fingers (2), present at least one time in 5 to 8 fingers (3). Subpapillary plexus visualization was qualitatively scored as previously reported [24] , while neoangiogenesis was noted in the presence of a suggestive combination of elementary alterations, based on the physician's evaluation (e.g. several deranged or bizarre capillaries which develop inside the same dermal papilla accompanied by a reduction in capillary density). Regarding crude quantitative data, mean and minimum capillary density were registered for each patient, as per maximum diameter and number of enlarged and giant capillaries, number of micro-haemorrhages and micro-thrombosis (NEMO score) [25] , and length of representative capillaries. Finally, for each patient, an overall qualitative assessment was made by the clinician distinguishing between normal pictures and the presence of significant abnormalities, based on previous observations and definitions [16, 26] . Mann-Withney test were used for comparison between groups, as appropriate. Statistical significance was defined as a p-value (P) <0.05.

We enrolled 82 patients: 28 during hospitalization, and 54 post-recovery. Considering the entire cohort, the mean age was 58.8 ± 13.2 years; 56 (68.3%) were male, 25 (30.5%) had hypertension, 50 (61.0%) were overweight or obese (BMI>25 kg/m 2 ; mean BMI 25.9 ± 3.5 kg/m 2 ), 11 (13.4%) were current smoker and 9 patients (11.0%) had diagnosed diabetes. Four patients (4.9%) had a diagnosed rheumatic disease (2 rheumatoid arthritis, 1 crystal-induced arthritis and 1 granulomatosis with polyangiitis). None of the patients had active cancer at the time of evaluation. Eight patients (9.8%) reported acral symptoms temporally related to COVID-19 (1 Raynaud's phenomenon, 1 acrocyanosis, 1 "puffy" hands, 5 distal paresthesia/dysesthesia) and 3 reported long term Raynaud's phenomenon, not associated with other symptoms suggestive for a CTD. Population's characteristics are reported in Table 1 . Overall, at the NVC examination we found a high prevalence of abnormalities classifiable as non-specific pattern (53 patients, 64.6%), without pictures suggestive of scleroderma pattern. Common abnormalities, found in more than two fingers in at least 50.0% of cases, were pericapillary edema (80.5%), enlarged capillaries (61.0%), sludge flow (53.7%), meandering capillaries and capillary density below 9 capillaries per linear millimeter (50.0%) ( Table 2 ). Neoangiogenesis, n (%) 6 (7.3) 0 (0) 6 (11.1) .09

In the patients evaluated during hospitalization (n=28) the mean time from the onset of symptoms to the evaluation was 7. At the NVC examination, 25 patients (46.3%) showed abnormalities classifiable as non-specific pattern.

The most frequent abnormalities, found in more than two fingers, were enlarged capillaries (85.2% of patients), meandering capillaries (81.4% of patients), pericapillary edema (70.4% of patients) and capillary density below 9 capillaries per linear millimeter (63.0% of patients). Furthermore, according to the semi-quantitative evaluation, the involvement of a percentage greater than 33% of all the capillaries analyzed was found for the following alterations: meandering capillaries in 10 (18.5%), enlarged capillaries in 3 (5.6%), capillary density below 9 capillaries per linear millimeter in 3 (5.6%) and microvascular derangement in 2 (3.7%) of the patients. Giant capillaries and avascular areas were rare (both found in 3 cases) and always encountered as an isolated alteration (only one digit for each patient), therefore they did not delineate a scleroderma pattern. Hemosiderin deposits as a result of microhaemorrhage or micro-thrombosis were found at least in one digit in 17 patients (31.5%, of whom 1 with micro-thrombotic aspect) and more than in one finger concurrently in 6 (11.1%) cases, with a number of individual lesions ranging from 1 to 11 capillaries per linear millimeter (P=.002), and empty dermal papillae (P=.006) ( Table 2 ). Figure 1 shows representative pictures of the alterations commonly found in acutely ill and recovered COVID-19 patients.

In both groups, we have not found any meaningful association between available markers of inflammation and endothelial impairment and NVC findings. In particular, there was no correlation between NEMO score and CRP and DDU levels.

Patients evaluated during hospitalization and after discharge were not significantly different for mean age 

To the best of our knowledge, this is the first study investigating the presence of microvascular alterations evaluated by NVC in COVID-19 patients. We performed NVC and described in detail the morphological appearance in patients who needed hospital admission for COVID-19 pneumonia, in two different stages of the disease: patients still hospitalized with acute pneumonia, and patients already discharged and evaluated at outpatient clinics. The most striking result was that, while a clear scleroderma pattern was absent in both cohorts of patients, a significant number of microvascular abnormalities was depicted in the vast majority of both groups. Interestingly, the prevalent abnormalities that we observed were different J o u r n a l P r e -p r o o f according to the stage of the disease. Specifically, the prevalence of hemosiderin deposits was significantly higher in acute COVID-19 pneumonia patients compared to patients already discharged from hospital. Conversely, recovered patients out of the acute phase had a higher prevalence of altered capillary bed with enlarged capillaries, meandering capillaries and frequent capillary density below 9 per millimeter with empty dermal papillae.

Among all abnormalities, the presence of hemosiderin deposits is particularly interesting in the context of acute patients as it suggests a higher prevalence of micro-hemorrhages and micro-thrombosis in this phase of the disease. Also, consistently with a pro-thrombotic alteration of the microcirculation in acute COVID-19, we observed more frequently a sludge flow in these patients. This is in keeping with the proposed cardinal role of diffuse lung micro-thrombosis in the physiopathology of SARS-CoV-2 infection [27] ; recent pathological studies [9, 10] have indeed shown how a diffuse microvascular thrombosis is a common finding in the lungs of COVID-19 patients at post-mortem examination. Moreover, coagulation abnormalities and increased risk of thrombosis have been widely recognized as peculiar clinical features of COVID-19, highlighting the use of anticoagulation as fundamental for the treatment strategy [15] . It is then not surprising that even the nailfold capillary bed is affected by microvascular abnormalities in the acute phase of the disease. In addition, the frequent finding of enlarged capillaries and capillary density at the lower limit in post-acute patients strengthens this concept as these findings could reasonably be the result of previous thrombotic capillary damage in the acute phase.

Our results suggest therefore a continuum spectrum of microvascular alterations in the nailfold capillary bed that goes from micro-thrombosis to altered capillary structure, according to COVID-19 phase. It is interesting to note that only in post-acute patients there was an involvement greater than the cut-off of Given its exploratory nature, the present study nevertheless presents several shortcomings that deserve to be discussed. The main limitation of our study is the small number of patients evaluated, especially in the acute phase. Fortunately, this is due to the timing of our study (mid-May 2020) when the peak of new COVID-19 cases had already passed in Italy. Another major limitation is the cross-sectional design.

Indeed, the evaluation of the same patients in the acute and post-acute phase, which would provide more solid and clear data on the possible evolution of microvascular damage, is missing. For these reasons, we cannot rule out the absence of correlation between the severity of the clinical status and the NVC results.

Similarly, the impact of some inhomogeneities in clinical features such as the presence of diabetes and active smoking cannot be clearly defined.

Of note, in our cohort there were no patients with CTD classically linked to specific alterations of the microcirculation (i.e. scleroderma pattern), however, it should also be considered that non-specific capillaroscopic changes have been described in several non-rheumatological diseases such as arterial hypertension, diabetes, Alzheimer's disease, glaucoma, and interstitial lung disease, among others [29, 30] . Some of these comorbidities (e.g. arterial hypertension and diabetes) have been reported as risk factors for COVID-19 pneumonia [12, 31] and therefore, having applied only a generic exclusion criterion (hand trauma), they are fairly represented in our cohort, representing a further variable to consider.

Regarding the influence of pharmacological treatments, the difference in exposure to enoxaparin in the two groups needs to be addressed: since the recovered patients had been hospitalized on average 5-8

weeks before the patients of acute group, should be emphasized that the observed difference probably reflects the change in clinical attitude in relation to the use of enoxaparin in patients with COVID-19 pneumonia, with use gradually increased during the pandemic. Exposure to other specific drugs such as antivirals, hydroxychloroquine and anti-IL6R and oxygen therapy during hospitalization was homogeneous as a result of defined treatment protocols within the clinics. In general, it should be emphasized that at present there is no clear evidence regarding the impact of these or other common treatments (e.g. lipid-lowering and antihypertensive drugs) on capillary morphology detected with NVC.

Longitudinal studies with a substantial number of patients, including both patients with asymptomatic or mild SARS-CoV2 infection (e.g. patients without pneumonia) and patients actively treated in ICU, are thus needed to make the application of more stringent exclusion criteria sensible and correct for possible confounding factors, such as known determinants of microvascular damage, and get further clues on the physio-pathogenetic implications of microvascular alterations seen at nail bed.

For the same purpose, the comparison between the capillaroscopic data of patients before and after COVID-19 would also be of great interest. Clearly, this kind of evaluation is difficult to implement since generally the capillaroscopy has restricted clinical indications. However, careful monitoring of patients followed up at rheumatology clinics, who regularly perform capillaroscopy for acral symptoms (e.g.

primary RP without an underlying CTD) for re-evaluation of those with COVID-19, would be desirable.

Another possibility would be to carry out capillaroscopic assessment in an at-risk population included in follow-up clinical studies.

Finally, some technical limitations inherent to the NVC must be considered. Considerable efforts for the standardization of the methodology and the definition of capillaroscopic changes have been made recently [32] , providing a fundamental framework for the diagnosis and follow-up of CTDs; however, a part of the alterations considered in this work, even though previously described, could have uncertain reliability [18] . Specifically, the frequent finding of pericapillary edema, particularly in acute patients, may have marginally influenced the finding of milder alterations such as meandering capillaries, contributing in part to the difference observed between the two groups. Furthermore, bearing in mind that NVC is not the J o u r n a l P r e -p r o o f examination of choice for the evaluation of the flow, the reported assessments on rheological aspects of the microcirculation (i.e. sludge flow), must be taken cautiously, and future studies should include dedicated methods such as sidestream dark field imaging [33] for better peripheral perfusion measurements along with the NVC morphological assessment.

COVID-19 patients present microvascular abnormalities at direct assessment through NVC. Acutely ill and recovered subjects are characterized by a different distribution of elementary capillaroscopic alterations, resembling an acute and post-acute type of microvascular damage. These preliminary results could pave the way to further longitudinal studies in larger cohorts.

COVID-19 illness in native and immunosuppressed states: A clinical-therapeutic staging proposal

Sarilumab use in severe SARS-CoV-2 pneumonia

Interleukin-1 blockade with high-dose anakinra in patients with COVID-19, acute respiratory distress syndrome, and hyperinflammation: a retrospective cohort study

Efficacy and safety of tocilizumab in severe COVID-19 patients: a single-centre retrospective cohort study

GM-CSF blockade with mavrilimumab in severe COVID-19 pneumonia and systemic hyperinflammation: a single-centre, prospective cohort study

Thrombosis, Kidney Failure, and Diabetes: Is COVID-19

Endothelial Disease? A Comprehensive Evaluation of Clinical and Basic Evidence

The Enigma of Endothelium in COVID-19

Pulmonary and cardiac pathology in African American patients with COVID-19: an autopsy series from New Orleans

Chilblain-like lesions in children following suspected COVID-19 infection

Pulmonary Vascular Endothelialitis, Thrombosis, and Angiogenesis in Covid-19

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: a retrospective cohort study

Lupus anticoagulant is frequent in patients with Covid-19

on behalf of Gemelli Against COVID-19 Post-Acute Care Study Group

COVID-19 and Thrombotic or Thromboembolic Disease: Implications for Prevention, Antithrombotic Therapy, and Follow-up

Nailfold videocapillaroscopy assessment of microvascular damage in systemic sclerosis

Nailfold Capillaroscopy in Rheumatic Diseases: Which Parameters Should Be Evaluated?

Standardisation of nailfold capillaroscopy for the assessment of patients with Raynaud's phenomenon and systemic sclerosis

Scoring the nailfold microvascular changes during the capillaroscopic analysis in systemic sclerosis patients

Nailfold videocapillaroscopy micro-haemorrhage and giant capillary counting as an accurate approach for a steady state definition of disease activity in systemic sclerosis

The Impact of Nailfold Capillaroscopy in the Approach of Microcirculation

Quantitative nailfold capillary microscopy findings in patients with acrocyanosis compared with patients having systemic sclerosis and control subjects

Feasibility of different capillaroscopic measures for identifying nailfold microvascular alterations

Quantitative nailfold capillaroscopy findings in a population with connective tissue disease and in normal healthy controls

The cumulative number of micro-haemorrhages and micro-thromboses in nailfold videocapillaroscopy is a good indicator of disease activity in systemic sclerosis: a validation study of the NEMO score

Scleroderma pattern of nailfold capillary changes as predictive value for the development of a connective tissue disease: a follow-up study of 3,029 patients with primary Raynaud's phenomenon

Microvascular COVID-19 lung vessels obstructive thromboinflammatory syndrome (MicroCLOTS): an atypical acute respiratory distress syndrome working hypothesis

Nailfold capillaroscopy in common non-rheumatic conditions: A systematic review and applications for clinical practice

Nailfold videocapillaroscopy in internal medicine

COVID-19 in people with diabetes: understanding the reasons for worse outcomes

Fast track algorithm: How to differentiate a "scleroderma pattern" from a "non-scleroderma pattern

Non-invasive imaging of microcirculation: a technology review

The authors want to thank the members of the Gemelli Against COVID-19 Post-Acute Care Study Group for their dedication in taking care of patients sadly affected by the COVID-19. All members and their specific roles are listed below.