398P - Lacrimal canaliculus stenosis and cornea epithelium disorder induced by S-1

Date 20 December 2015
Event ESMO Asia 2015 Congress
Session Poster presentation 2
Topics Complications/Toxicities of treatment
Presenter Toshihide Onikubo
Citation Annals of Oncology (2015) 26 (suppl_9): 111-124. 10.1093/annonc/mdv531
Authors T. Onikubo, M. Nakamura, K. Nakamura, K. Uchibori, T. Miyatake, S. Suyama, N. Iwaana, K. Tauchi
  • Aizawa Comprehensive Cancer Center, Aizawa Hospital, 390-8510 - Matsumoto/JP



S-1 is widely used for patients with various solid tumors in Asian countries. Few case reports and studies have documented common ocular adverse events including lacrimal canaliculus stenosis (LCS) and cornea epithelium disorder (CED) induced by S-1 administration. We previously reported the incidence and severity of ocular adverse events induced by S-1 in ESMO 2014. We conducted further investigation about relation between existence of lacrimal canaliculus narrowing (LCN) in ophthalmologic surveillance before S-1 treatment and incidence of lacrimal canaliculus stenosis induced by S-1.


Patients who were candidates for S-1 administration were surveyed by ophthalmologists to check their lacrimal canaliculus and cornea epithelium before S-1 treatment. Ophthalmologic surveillance was kept on every 4 weeks during administration of S-1 until cessation of the treatments. The patients underwent slit lamp biomicroscopic examinations to evaluate the incidence and severity of LCS and CED.


From June 2011 to September 2014, 78 patients were treated with S-1 (gastric cancer 40 patients, colorectal 18, pancreatic 11, esophageal 4, and other cancers 5). The incidence of any ocular adverse events was 34.6% (27 cases). LCS was seen in 25.6% (20 cases. Twenty patients (15.3%) were diagnosed as grade 3 LCS (CTCAE v4.0) and needed insertion of drainage tubes in the lacrimal canaliculus. The incidence of grade3 LCS in patients with LCN detected in surveillance before S-1 treatment was statistically higher than patients without LCN (63.6% v.s. 7.5%, p < 0.01; Fisher's exact test). Fourteen cases (17.9%) were diagnosed as CED, and 2 cases with grade 3 severe vision disorder required cessation of S-1 treatment. Corneal ulcers were detected in one patient. After cessation of S-1, recovery from all adverse effects was observed in all patients.


This study showed a high incidence of ocular adverse events induced by S-1 treatment, especially in patients with LCN before S-1 treatment. In these patients, more carefully observation will be needed. Oncologists should cooperate with ophthalmologists during S-1 treatment to prevent severe ocular adverse events by detecting disorders at an early stage.

Clinical trial identification


All authors have declared no conflicts of interest.