Patients

Surgery was performed in 70 eyes of 70 cases (43 men and 27 women), and patient age at diagnosis ranged from 32 to 91 years (median age: 71.6 ± 12.6 years). The postoperative follow-up period ranged between 2 months and 13.8 years (median period: 38.6 ± 38.6 months).

Tumor size and location

At the initial examination, the tumor was found in the limbus in 59 patients, in the palpebral conjunctiva in 8 patients, and from the bulbar to the palpebral conjunctiva in 3 patients. Among the cases in which the tumor was located on the limbus, it was found to be present within 1 quadrant in 31 patients, within 2 quadrants in 15 patients, within 3 quadrants in 8 patients, and in all quadrants in 5 patients (Table 1).

Table 1 Demographic characteristics, diagnoses, and treatments of the study participantsHistological classification

Histopathological examination of the tumor revealed that among the 70 patients, there was dysplasia in 8, carcinoma in situ in 26, and invasive SCC in 36 (Table 1).

Surgical treatment

Surgical excision of the tumor was successfully carried out in all 70 patients, and MMC was administered intraoperatively in 58 of the 70 patients. Epithelial transplantation was combined when the tumor was larger than 3 quadrants of the corneal limbus (Fig. 1). When the tumor was less than 2 quadrants of the corneal limbus, a patient-specific appropriate treatment method was selected. If the limbal epithelial defect is less than half the circumference, the epithelial supply is not needed and it is better to combine corneal transplantation to prevent pseudopterygium. In all cases, the surgeon made the final decision, taking into consideration the patient’s age, surgical time, and the availability of a donor eye. In 4 of 8 patients in which the tumor was found in the palpebral conjunctiva, and in all 3 patients in which the tumor was found from the bulbar to the palpebral conjunctiva, AMT was combined (Fig. 2; Table 2). The epithelial transplantation procedures performed in this study were limbal transplantation/keratoepithelioplasty (N = 29 cases), cultivated oral mucosal epithelial transplantation (N = 3 cases), and auto-conjunctival epithelium transplantation (N = 2 cases).

Fig. 1

Images of ocular surface squamous neoplasia (OSSN) tumors of the corneal limbus obtained before and after epithelial transplantation. a Preoperative appearance of a tumor involving 1 quadrant of the corneal limbus. b Appearance of the eye shown in a after tumor excision and epithelial transplantation. c Preoperative appearance of a tumor involving 2 quadrants of the corneal limbus. d Appearance of the eye shown in c after tumor excision, epithelial transplantation and amniotic membrane transplantation. e Preoperative appearance of a tumor involving 3 quadrants of the corneal limbus. f Appearance of the eye shown in e after tumor excision and epithelial transplantation

Fig. 2

Images of OSSN tumors of the palpebral conjunctiva obtained before and after amniotic membrane transplantation (AMT). a Appearance of the tumor in the palpebral conjunctiva. b Appearance of the eye shown in a after tumor excision and AMT

Table 2 Combination of ocular surface reconstructionPostoperative topical chemotherapy

Topical 5-FU was administered postoperatively in 30 cases (i.e., in 18 cases with invasive SCC and in 12 cases with carcinoma in situ). Topical IFN was administered postoperatively in 5 cases (i.e., in 2 cases with invasive SCC and 3 cases with carcinoma in situ).

Visual prognosis

When the tumor size exceded 3 quadrants, it covered the pupillary area and produced visual dysfunction. In such cases, the preoperative BCVA was less than 20/20, yet in 25 of 31 cases it improved postoperatively through the combination of epithelial transplantation (in 3 cases that underwent epithelial transplantation, the preoperative BCVA data was missing) (Fig. 3).

Fig. 3

Pre- and postoperative best-corrected visual acuity of the 31 patients who underwent epithelial transplantation

Recurrence

Fluorescein staining and scleral scattering were performed at every follow-up visit in order to detect any new abnormal appearances of the epithelium, and “recurrence” was defined as the appearance of a new tumor lesion post surgery. Post surgical excision, recurrence of the tumor was clinically observed in 19 (27.1%) of the 70 patients at from 2 months to 50 months (median: 12.5 months) postoperative (Tables 1 and 2). For histological classification, the rate of recurrence was: dysplasia, 1 (12.5%) of 8 patients; SCC in situ, 7 (26.9%) of 26 patients; and invasive SCC, 11 (30.6%) of 36 patients.

The average tumor size in the patients in whom recurrence occurred was 10.5 ± 3.4 mm, while that of the patients without recurrence was 8.0 ± 3.3 mm. Recurrence occurred in 5 (16.1%) of 31 cases in which the tumor was localized to less than 1 quadrant of the limbus, in 4 (26.7%) of 15 cases in which it involved 2 quadrants, in 3 (37.5%) of 8 cases in which it involved 3 quadrants, in 4 (80.0%) of 5 cases in which it involved all quadrants, in 1 (12.5%) of 8 cases in which it involved the palpebral conjunctiva, and in 2 (66.7%) of 3 cases in which it involved both the bulbar and the palpebral conjunctiva.

Although MMC was administered intraoperatively in 58 patients, recurrence occurred in 15 (25.9%) of those patients. However, recurrence also occurred in 4 (33.3%) of the 12 patients in whom MMC was not administered intraoperatively.

Among the 30 cases in which topical 5-FU was administered postoperatively (i.e., 18 cases with invasive SCC and 12 cases with carcinoma in situ), recurrence was observed in only 2 (6.7%) of those cases. Of note, patients in whom topical 5-FU was not administered post surgery had a significantly higher incidence of recurrence compared to the patients in whom it was administered (Fig. 4).

Fig. 4

Kaplan-Meier survival curve analysis estimate of no recurrence for all patients. Results are shown for the group in which topical 5-fluorouracil (5-FU) was administered postoperatively (solid line) and the group in which topical 5-FU was not administered postoperatively (dotted line)

Uni- and multivariate analyses of risk factors for OSSN recurrence

Among the prognostic factors, the rate of postoperative recurrence was found to be related to patient age (hazard ratio [HR] = 1.05, 95% confidence interval [CI] 1.00 to 1.09, P = 0.028), lesion location (i.e., in bulbar and palpebral conjunctiva: HR = 7.17, 95% CI 1.10 to 46.73, P = 0.039), and tumor size (HR = 1.20, 95% CI 1.01 to 1.41, P = 0.037) (Table 3). All eyes in which the tumor location was in the bulbar and palpebral conjunctiva presented an extended large size of OSSN.

Table 3 Uni- and multivariate analyses of risk factors for OSSN recurrence (prognostic factors)

The rate of postoperative recurrence was found to not be statistically related to the detection of HPV-16 in the patients’ tear samples. Among the treatment factors, including surgical intervention, postoperative 5-FU administration was the only factor found to reduce the risk of recurrence (HR = 0.12, 95% CI 0.02 to 0.60, P = 0.010) (Table 4). No significant difference was observed between the patients in whom MMC was, or was not, administered intraoperatively (HR = 1.16, 95% CI 0.29 to 4.59, P = 0.834).

Table 4 Uni- and multivariate analyses of risk factors for OSSN recurrence (treatment)Adverse events

In the eyes that underwent ocular surface reconstruction, no complications such as epithelial defects, graft rejection, or infections occurred. Moreover, in all cases in which MMC was administered intraoperatively, no complications occurred. Among the 30 eyes in which 1% topical 5-FU was administered, a large corneal epithelial defect with associated eye pain suddenly occurred in 1 case at 16 days of treatment, yet the defect in that eye healed within 1 week after the treatment was discontinued. In eyes that underwent topical IFN, no complications occurred.

At the final follow-up examination, 67 of the 70 cases had no serious visual disturbance, and there were no cases with scleral invasion and no cases in which metastasis developed. In all cases, the appearence of the ocular surface was favorable; i.e., no conjunctival injection and no synechia.