Background In SSc a dysfunction of peripheral nervous system (PNS) has been showed. SP instillation in human eye induces a cholinergic-independent pupil miosis via specific receptors on neuroirideal junction.

Objectives In order to study PNS function in SSc, we assessed by pupillometry pupil basal diameters and pupil ability to respond to SP instillation in SSc patients and in controls.

Methods 32 SSc patients [17 with limited (lSSc) and 15 with diffuse SSc (dSSc)] and 30 controls underwent pupillometry for evaluating basal iris diameters in both eyes. Then, SP was instilled in one eye [treated eye (TE)] and placebo in the other eye [not treated eye (NTE)] and the pupil response to SP instillation was assessed both in TE and in NTE. In order to verify the existence of a dose-dependence activity of SP, two different SP concentrations (10–3 and 10–5 M) were tested separately. For both eyes, t0 (pupil basal diameter) and AUC (iris pharmacological response to SP instillation) were considered. 2-tailed t-test was used to compare TE with NTE within groups and SSc with HC. ANOVA, Scheffè and Dunnett post-hoc tests were utilised to compare lSSc, dSSc and HC.

Results At t0, basal pupillary diameters of SSc resulted significantly minor in respect to HC (p < .001) and significant differences between lSSc and dSSc and HC (p < .05), but not between dSSc and HC, were found. Significant differences between SSc and controls were found in the response to SP 10–3 and 10–5, both for TE and for NTE (p < .05 in all cases). The TE of lSSc showed a significant major miosis to SP 10–3 and 10–5 instillation vs dSSc and controls (p < .05), whereas in the TE of dSSc pupillary miosis was not different from controls. In NTE, a significantly different miosis between lSSc and controls was induced by both SP concentrations (p < .001), while a significant difference between lSSc and dSSc was found only when SP 10–3 M was used (p = .002). In dSSc, the instillation of SP 10–3, but not of SP 10–5 M, induced in NTE a more intense miosis in respect to controls NTE (p = .004). The comparison between TE and NTE was significantly different in controls (p < .001), but not in lSSc and in dSSc.

Conclusion Our results show that lSSc is characterised by the involvement of neuroirideal junction. In fact, the irideal basal diameter is reduced, and SP induces a more intense miosis in respect both to dSSc and to HC. The lack of difference between TE and NTE could be due to a dysfunction of peripheral nervous system, more likely involving central control mechanisms.