**1. Introduction**

Millions of people worldwide are affected by dry eye disease (DED), a heterogeneous, complex disorder of the ocular surface [1]. Within the current concept of a staged treatment, lubricating, hydrating teardrops are the standard long-term therapy for DED [2]. Hyaluronan (HA) eye drops, aiming to increase tear viscosity and enhancing lubrication, are one of the options favored, particularly in Europe and Asia [2]. The combination of concentration and chain length of the HA molecules contained in these eye drops determines their viscoelastic and mucoadhesive properties, resulting in more or less entanglement and rheological synergism with the mucins dissolved in the muco-aqueous layer of the tear film. These physical properties of HA eye drops contribute to minimizing the friction between the moving eyelid and the surface of the eyeball during blinking, thus reducing known stimuli of ocular surface inflammation [3]. High molecular weight hyaluronan (HMWHA) has an anti-inflammatory e ffect, whereas low molecular weight hyaluronan (LMWHA) promotes inflammation [4,5]. A recent study confirmed in an environmental dry eye stress model in mice that HMWHA eye drops protect the ocular surface from mechanical damage and inflammation better than LMWHA [6]. Future clinical investigations of HMWHA eye drops in humans su ffering from chronic ocular surface inflammation should, therefore, include inflammation markers.

The current treatment for severe DED is mainly based on the model of the self-maintaining circle of chronic inflammation [7–10]. The underlying pathomechanism of severe dry eye disease focuses on inflammation in various situations such as in autoimmune diseases, as well as damage of corneal nerves, for example, in diabetes mellitus or aging [11]. The cornea is by far the most densely innervated tissue of the human body [12]. Nerves provide important trophic support to the corneal epithelium and contribute to ocular surface homeostasis [12–16]. Activated corneal nerves release neuropeptides that contribute to neurogenic inflammation [17–19]. Denervation eliminates the neurotrophic support causing neuroparalytic keratitis and breakdown of the corneal epithelium [17,20]. On the other hand, trophic interactions are essential for neuronal survival [21–23]. Moreover, there is cross-talk between glia, the extracellular matrix, and neurons [24]. Attrition within the ocular epithelia has been recognized as a lubrication deficit induced factor, enhancing inflammation [25]. Due to these complex interactions, severe DED is regularly associated with compromised corneal nerves [26–30]. This, in turn, results in dysregulation of tear production and blink reflex [31]. Corneal innervation disorders as a primary pathogenic mechanism are due to the absence of ocular pain only diagnosed in a late-stage, although they are often accompanied by keratopathy and delayed epithelial wound healing, sometimes leading to corneal ulcerations and vision loss [32–34]. There is a lack of treatments targeting nerve regeneration [34,35].

Patients su ffering from neuropathic ocular pain tend to respond poorly to the treatment with lubricant eye drops [36,37]. Experimental evidence suggests that HMWHA, but not LMWHA, can suppress pain in nociceptive a fferent nerves [38–40], but it is not ye<sup>t</sup> proven whether or not topically applied HMWHA can reduce ocular pain. Moreover, the possible role of hyaluronan in the proliferation of nerve cells has raised attention [24,41,42]. Therefore, we decided to study the potential influence of HMWHA on the corneal nerves within the HYLAN M study. The main intention of the HYLAN M study was to investigate if symptoms and/or signs of patients su ffering from severe DED could be improved by substituting the best treatment lubricant eye drops with HMWHA eye drops. In vivo confocal microscopy (IVCM), in particular, when performed as confocal laser scanning microscopy (CSLM), is the gold standard in assessing the subbasal corneal nerve plexus [43–46]. Within the HYLAN M study, CSLM images were taken at baseline and after eight weeks of treatment and were sent to a masked reading center for evaluation.

#### **2. Experimental Section**

#### *2.1. Study Design*

The HYLAN M study, a multicenter prospective randomized, open-label study, was performed in 11 centers in eight countries. Details have been published elsewhere [47]. The study adhered to the Declaration of Helsinki, was approved by ethics committees of all countries involved, and registered

on the database of the European Database for Medical Devices (EUDAMED) under the registration number CIV-16-06-015964.

Patients suffering from severe DED were randomized into two parallel arms. The control group continued with their currently-used therapy by the time of inclusion. In the verum group (Comfort Shield group), the individual lubricant eye drops used by each patient by the time of inclusion were replaced by eye drops containing 0.15% HMWHA (Comfort Shield® eye drops, i.com medical GmbH, Munich, Germany). Concomitant treatment for dry eye, like cyclosporine eye drops, remained unchanged in both arms.

Demographic data and medical history were recorded during the baseline visits. Symptoms and signs associated with DED were assessed at the baseline visit, at week 4, and week 8 follow-up visits, respectively (see Table 1).


**Table 1.** Diagnostic testing schedule with optional tests in round brackets.

Abbreviations: Ocular Surface Disease Index (OSDI), best corrected visual acuity (BCVA), corneal fluorescein staining (CFS), tear film break-up time (TBUT), intraocular pressure (IOP), and lid wiper epitheliopathy (LWE).

The study centers were suggested to optionally take CSLM images at the baseline and week eight visits and provide them to a masked reading center for assessment. Four out of 11 study centers participated in this optional test. These four study centers provided CSLM images of all their per-protocol patients; thus, the electronic randomization used throughout the HYLAN M study also applied to the optional confocal microscopy study. The results of the assessment of the CSLM images of these four study centers are the subject of this report. The results of the other diagnostic tests performed, such as the Ocular Surface Disease Index (OSDI), dropping frequency, best corrected visual acuity (BCVA), corneal fluorescein staining (CFS), tear film break-up time (TBUT), Schirmer 1, tear osmolarity, intraocular pressure (IOP), lid wiper epitheliopathy (LWE), and Yamaguchi score of all 84 per-protocol patients included in the HYLAN M study have been previously reported [47].
