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

Systemic sclerosis (SSc) is a connective tissue disease clinically characterized by Raynaud's phenomenon, and progressive fibrotic changes in the skin and internal organs, such as heart, lung, and kidney. Subsets of SSc have been defined, that is, limited cutaneous (lc) SSc, diffuse cutaneous (dc) SSc and SSc without skin involvement [1].

As in other systemic autoimmune diseases, SSc has a clinical course characterized usually by early phases of activity. A small vessel vasculopathy, characterised by peripheral vascular and interstitial mononuclear cell infiltration, represents the hallmark of the disease in the early phase. Conversely, in the late disease stages, the entire process may evolve to irreversible fibrotic changes of the involved tissues and organs [2]-[4]. This should induce the clinician to define an early diagnosis and recognize any active phase of the disease, in order to contrast the irreversible final damage by introducing an appropriate treatment [5]-[7].

The assessment of disease activity (DA) in SSc is a rather difficult challenge, and only a composite index that includes several clinical, instrumental, and serological items has, so far, been proposed and validated [8],[9].

Nailfold videocapillaroscopy (NVC) is a simple method that allows the clinician to observe and follow up the disease-related micro-vascular changes in an easily accessible capillary bed [10]. NVC plays an important role in the diagnosis of SSc and is included as an item even in the 2013 classification criteria for SSc [11]. In addition, different NVC patterns have been defined and their prognostic value underlined [12]-[14]. Finally, NVC has been used to define the different phases of the disease. This allows distinguishing patients with early, active, and late NVC pattern, where the active pattern is defined by the presence of numerous ectasic and giant capillaries (GC), ramified capillaries, micro-haemorrhages (MHE) and micro-thrombosis (MT), and initial loss of capillaries with avascular areas [15].

In the present study, we reconsider the specific NVC abnormalities that have been suggested to be increased (MHE, MT, and GC), or progressively reduced (number of capillaries, Cs) during the active phases of disease, to verify whether some quantitative combination of these features could be useful to identify patients with a relevant level of DA.

We decided to assess the presence of MT together with that of MHE separately from that of GC.

This statement was based on the present interpretation of micro-vascular NVC abnormalities in SSc. It has been suggested that MT and MHE are strictly related to each other, and represent the final evolution of enlarged capillary loops in a rapidly progressive pathological process [16]. Conversely, GC formation is certainly the first response to micro-vascular aggression and damage [17], but they may remain substantially unchanged when repeatedly observed in NVC in a slowly evolving or stable disease [18]. Consequently, the progressive reduction of the numbers of Cs can be regarded as the final result of capillary bed pathological aggression, and also of the substantial failure of capillary reconstruction attempts [17],[18].

The NEMO score appeared to be a good predictor for DA since NEMO values were strictly correlated with both ESSG index scores (Spearman's R =0.65, P <0.0001), and with mRSS (R =0.59, P <0.001), but not with patients' age and disease duration. When the other components of ESSG activity index were considered, NEMO score appeared to be significantly higher in patients presenting with scleredema (Mann-Whitney U Z value =4.59, P <0.0001), worsening of skin (Z =5.79, P <0.0001), cardio-pulmonary (Z = 3.59, P <0.0005), and vascular features (Z =3.18, P <0.002), current digital ulcers (Z = 2.91, P <0.005), and erythrocyte sedimentation rate (ESR) over 30 mm/h (Z = 3.79, P <0.0001).

Similarly, also the GC score was significantly correlated with ESSG index score and with mRSS (R = 0.47, P <0.0001, and R = 0.34, P <0.001, respectively). Among the different components of ESSG scoring system, GC score was significantly associated with the presence of scleredema (Z = 4.8, P <0.0001), digital ulcers (Z = 2.80, P <0.01) and worsening of cutaneous (Z = 3.72, P <0.0005), vascular (Z = 2.64, P <0.01), and cardio-pulmonary (Z = 2.12, P <0.05) features.

As expected, negative correlations were found between Cs score and both ESSG index and mRSS (R = -0.27, P = 0.004, and R = -0.26, P = 0.007). Furthermore, Cs score was significantly lower in patients with scleredema (Z = -3.02, P = 0.003), digital ulcers (Z = -3.08, P = 0.002), and diffusing capacity (DLCO) <80% of predicted value (Z = -3.4, P = 0.001).

The ROC curves constructed by plotting the sensitivity and specificity values of different NEMO, GC and Cs scores in correctly classifying active patients defined by ESSG index cut-off values of both 3.5 and 3.0 were represented in Figure 2. Generally speaking, NEMO score works better with respect to GC and Cs scores in identifying patients with active disease, and the area under the curve (AUC) of the NEMO score ROC plots were higher than the corresponding AUC of either GC or Cs ROC curves. Both differences reached the level of significance (P <0.03 and P <0.0006, respectively, considering the two-tailed Hanley-McNeil test) when the ESSG index cut-off value of 3.5 was taken into account. Similarly the AUC of the NEMO score was significantly higher than that of the GC score and Cs score when the ESSG index cut-off value of 3.0 was considered (P <0.05 and P <0.0001, respectively; see Figure 2 for details).

When the distribution of NEMO score according to the NVC patterns was analysed, it appears evident that the highest NEMO score values and majority of patients with NEMO scores ≥6 were found in the population of patients classified as having the `active’ NVC pattern [15] (27/43, 62,8%). However, a relevant number of patients with a NEMO score ≥6 could also be found in the `early’ and `late’ SSc population [15]. Eight out of eleven patients with the early and all of the five with the late scleroderma pattern having a NEMO score ≥6 had an ESSG index ≥3.5.

With this figure in mind, we tested sensitivity, specificity and accuracy of a modified (m) NEMO score where a classification of the patients as active was indicated by the presence of ≥6 MHE/MT, or - alternatively - by the contemporary presence of <6 MHE/MT plus a variable number of GC. A mNEMO defined by the presence of <6 MHE/MT, or alternatively by the presence of three to five MHE/MT plus at least three GC showed the best accuracy (88.7%) in correctly classifying patients with either moderate or high active disease (with an ESSG score ≥3), with a sensitivity of 95.1% and a specificity of 84.8%., and an odds ratio of 109.2 (95% CI 22.7 to 526.1, P <0.0001).

In this study, we demonstrated that the presence of a given number of MHE and MT is highly indicative of an active phase of disease. Even the presence of a sufficient number of GC, and a reduced number of Cs may suggest the presence of DA, although the strength of correlation is higher for the former NVC abnormalities. A score where the presence of a defined number of MHE plus MT, and of a given amount GC were combined - derived by a logistic regression model - showed the best predictive value in selecting patients with a current level of moderate and high DA.

NVC is a simple technique largely adopted in clinical assessment of patients with SSc and other connective tissue diseases [24]. Different applications have been described for NVC in SSc. First of all, NVC has an important diagnostic value in SSc, and a specific pattern has been described as characteristic of the disorder, generically defined as belonging to the scleroderma spectrum [17]. Moreover, the recent ACR-EULAR classification criteria for SSc include NVC among the proposed diagnostic items [11]. A relevant prognostic value has also been ascribed to NVC, and different NVC patterns have been defined to be predictive for a worse outcome [12]-[14], and for some specific complications of the disorders [25],[26]. Finally, NVC has also been used to subclassify patients with SSc according with the phases of disease course, and different NVC patterns described for early, active, and late disease [15]. Conversely, so far, no studies have been carried out specifically aimed at evaluating the potential utility of NVC in stratifying patients according to the level of DA.

The fact that, according to our results, MHE and MT are the NVC abnormalities more strictly indicative of a current DA status is not completely surprising. SSc is usually defined as a disease of the micro-vascular bed of the skin compartment, but potentially compromising the small vessels of internal organs. The endothelial involvement is considered the `primum movens’ of the pathological process in SSc, while migration of mononuclear inflammatory cells in the interstitium, activation of myo-fibroblast and fibroblast lineage, with the consequent increased production of extracellular matrix and then fibrosis, could be considered as the following stages [2]-[4]. A large body of evidence indicates that the typical response to the initial micro-vascular damage and capillary loss in SSc is represented by the compensatory dilatation of remaining capillary loops with formation of GC [17]. In slowly evolving or stable disease, it is not unusual to observe GC unchanged over time. On the contrary, in progressive disorder, the destiny of enlarged capillaries is thrombotic obliteration followed by extravasation. When these phenomena are synchronous in many capillaries, serial extravasations with hemosiderin deposits can be observed in the NVC, aligned distally in the cuticle [16].

The finding that the presence of a given amount of MHE and MT (more strongly than that of GC) is associated with DA phases in SSc seems to confirm that these specific abnormalities could represent the NVC counterpart of a rapidly evolution of micro-vascular involvement in active disease phases. The fact that the correlation is less strong for GC can be ascribed to the well-know finding that some GC may appear stable over time in sequential NVC examinations, and then their presence cannot reflect an active phase of the disease. Conversely, the reduction in the number of Cs may be regarded as the final outcome of pathological process in the capillary bed, once the capillary reconstruction attempts have failed. Of course, in a longitudinal study, where a rapidly progressive reduction of capillary number may be assessed in two or more consecutive NVC examinations, one can certainly consider this finding as expression of an active evolving phase of the disease. This is not the case of the present cross-sectional study in which we demonstrated that the current state of DA is more closely represented by the presence of MH and MT. This statement is certainly reinforced by the observation that MHE and MT, but not GC, have been described in other disorders where the involvement of micro-vascular compartment is an important pathological aspect, although induced by partially different mechanisms [27]-[30].

Since in a specific logistic model the presence of a certain number GC appear to be also predictive for DA, although to a lesser extent, we checked whether a given combination of all these NVC abnormalities could better identify patients with a relevant DA. The modified score (mNEMO score), in which both MHE/MT and GC were included, demonstrated to be actually more accurate and valid for this purpose.

We are, of course, aware that any instrument devoted to assess DA should be not only valid and reliable but also sensitive to change. Not a cross-sectional study like the present one, but a longitudinal study, where the instrument is used to assess changes of DA over time, is certainly needed to verify this latter characteristic.

The here proposed NVC scoring system named mNEMO appears a feasible instrument to assess DA in SSc, since it is possible to predict an active phase of disease during an outpatient visit and in a rather short time. On the contrary, to evaluate the same entity with the previously proposed methods implies a complicate and time-consuming patient work-up. Of course, patients in whom this specific NVC approach may suggest an active phase of disease should be addressed to a more careful clinical, instrumental and serological evaluation to confirm the suspicion and more precisely define the specific clinical profile. Once a patient has been definitely classified as being in an active phase of disease, a more aggressive therapeutic strategy should be taken into consideration, in accordance with the present recommendations that suggest treating active patients as soon as possible, in order to tentatively modify the natural course of SSc.

Some specific features observed during NVC examination, such as MHE/MT and GC are strongly predictive of moderate or high level of DA in SSc. A scoring system derived by these findings has been developed and proposed.

DS made substantial contributions to the conception and design of the work; acquisition, analysis, and interpretation of data; drafting the paper; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. GS made substantial contribution to the design of the work; acquisition, analysis, and interpretation of data; revising the work critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. EZ contributed to acquisition of data; revising the work critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. WM contributed to the acquisition of data; revising the work critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. RP contributed to the interpretation of data; to revising the work critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. AMVA contributed to the interpretation of data; to revising the work critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. CV made substantial contribution to the design of the work; analysis and interpretation of data; drafting the work and revising it critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. NDP made substantial contributions to the design of the work; acquisition and interpretation of data; revising the work critically for important intellectual content; and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. All authors read and approved the final manuscript.