Punctal and Canalicular Stenosis Following Topical 1% 5-Fluorouracil Eye Drop Therapy for Ocular Surface Squamous Neoplasia

Abstract

Purpose:

To present the first known cases of punctal and canalicular stenosis following topical 5-fluorouracil (5-FU) eye drops for ocular surface squamous neoplasia (OSSN).

Methods:

A retrospective chart review of patients with OSSN receiving topical 1% 5-FU eye drops as a primary or adjuvant treatment between 2013 and 2021 was performed. Individuals diagnosed with punctal or canalicular stenosis during or after topical 5-FU use were included in this study. Four individuals met these criteria; we report clinical findings and treatment course for each patient.

Results:

We reviewed 303 patients (316 eyes) with OSSN. All patients were treated with topical 1% 5-FU eye drops in cycles, given 4 times daily for 1 week with 3 weeks off. Four patients (5 eyes) developed punctal stenosis, for a frequency of 1.3%. The mean age at 5-FU initiation in the 4 individuals was 61 years (range 53–69). Most individuals were male (75%), White (75%), and non-Hispanic (100%). The most common presenting symptom of stenosis was epiphora, noted 4.6 ± 3 months after initiating topical 5-FU (after 4.9 ± 2.0 cycles). This occurred in 4 eyes while on therapy (3, 2, 4, and 4 months since 5-FU initiation) and in 1 eye after stopping therapy (10.3 months since 5-FU initiation). In 1 eye, punctal stenosis and epiphora resolved spontaneously after stopping 5-FU. Punctal dilation led to symptom resolution in 2 eyes. Surgical intervention (punctoplasty and external dacryocystorhinostomy) was needed in 2 eyes for punctal and canalicular stenosis, respectively.

Conclusions:

Punctal or canalicular stenosis is a rare and previously unreported adverse effect of topical 5-FU that may require surgical treatment.

Ocular surface squamous neoplasia (OSSN) is the most common nonpigmented malignancy of the ocular surface and represents a spectrum of disease ranging from dysplasia, to carcinoma in situ, to squamous cell carcinoma.¹ Traditional treatment consists of surgical excision with wide margins and a no-touch technique; however, surgical excision has been associated with sequelae such as conjunctival scarring, symblepharon, conjunctival hyperemia, and limbal stem cell deficiency.^2–5 In addition, surgical excision with resultant positive margins has been associated with up to 56% recurrence frequency.⁶ 5-fluorouracil (5-FU) is a topical chemotherapeutic agent that has recently emerged as a valuable primary treatment for OSSN and as an effective alternative to the traditional treatment of surgical excision.^7–9 Several studies on the use of topical 5-FU as primary treatment for OSSN have demonstrated a high frequency of resolution (82%–100%) and low frequency of recurrence (10%–14%) with a tolerable side effect profile.^8–10 It has also been used as a postoperative adjunct post excision. Punctal stenosis has been associated with the systemic use of 5-FU for diseases such as breast, colon, and other cancers,^11–15 and a recent study found canalicular stenosis in rabbits receiving topical 5-FU.¹⁶ However, there is a knowledge gap on whether this side effect can occur with topical 5-FU eye drop therapy in humans. To bridge this gap, we examined individuals treated with topical 5-FU eye drops as a primary or adjuvant therapy for OSSN and identified 4 cases of punctal or canalicular stenosis that occurred during or after treatment, a previously unreported adverse effect.

MATERIALS AND METHODS

This retrospective study was approved by the institutional review board of the University of Miami, and the methods adhered to the tenets of the Declaration of Helsinki and complied with the Health Insurance Portability and Accountability Act. A chart review of patients with OSSN treated with topical 5-FU as a primary or adjuvant treatment between 2013 and 2021 was performed. At our institution, 5-FU is prescribed in a cyclical approach with 1 cycle consisting of 1 week of topical 1% 5-FU eye drops 4 times per day followed by 3 weeks of no treatment. No punctal plugs were used. Individuals diagnosed with punctal or canalicular stenosis during or after the use of topical 5-FU were identified and included in this study. Patients were excluded if they had a history of epiphora or stenosis, or if they were diagnosed with punctal or canalicular stenosis after using other topical chemotherapies, such as mitomycin C (MMC) or interferon (IFN) α-2b after the 5-FU use. In this case series, we first compared demographic and treatment information of the 4 individuals with that of the entire cohort and then reported the clinical findings and course for each patient who was diagnosed with punctal or canalicular stenosis.

RESULTS

A total of 303 patients (316 eyes) were reviewed who received topical 5-FU as a primary or adjuvant treatment for OSSN. Four patients (5 eyes) were identified who developed punctal or canalicular stenosis during or after topical 5-FU eye drop therapy, for a frequency of 1.3%. The mean age of the 4 patients at the beginning of 5-FU treatment was 61 years (range 53–69). Most individuals were White (75%), male (75%), and non-Hispanic (100%). These demographics were mostly similar to those of the entire cohort [mean age 67.4 years, range 31–98, with a White (84.2%) and male (74.9%) predominance]. However, the entire cohort had more individuals of Hispanic ethnicity (54.1%) compared with the 4 individuals who developed punctal or canalicular stenosis (0%, P = 0.03). At our institution, the 5-FU 1% eye drops are given in cycles of 4 times a day for a week with 3 weeks off treatment. The 5 eyes of individuals who developed stenosis were treated with topical 5-FU for a mean of 4.9 cycles (SD 2.0, range 2–8), which is a similar treatment length to the entire population treated primarily for OSSN (4.9 cycles, SD 2.1, range 1.05–15.5).

The most common presenting symptom of punctal or canalicular stenosis was epiphora, which presented at a mean of 4.6 months (SD 3.0, range 1.6–10.3), or 4.9 ± 2.0 cycles, after initiating topical 5-FU. None of the patients included in this study had a history of epiphora or punctal or canalicular stenosis. This occurred in 4 eyes while on therapy (3, 2, 4, and 4 months since 5-FU initiation) and in 1 eye after stopping therapy (10.3 months since 5-FU initiation). The stenosis necessitated surgical intervention in 2 eyes [puncto-plasty and external dacryocystorhinostomy (DCR)] and dilation of the puncta in 2 eyes, and spontaneously resolved in 1 patient after cessation of the drops. Fortunately, symptoms improved in 4 of 5 eyes. No factors were identified that could be predictive of spontaneous resolution.

Case Summaries

Patient 1

A 67-year-old female patient presented to the clinic with a history of OSSN in the left eye. She had previously undergone 1 month of topical IFN α-2b followed by surgical excision 6 months before presentation. On examination, the clinical evaluation revealed an opacity on the inferior cornea, suggesting residual disease. The original pathology was reviewed and found to be conjunctival intraepithelial neoplasia with moderate and severe dysplasia and foci of carcinoma in situ, but clear margins could not be confirmed. Given the indeterminate margins and visible residual tumor, the patient elected for empiric treatment with topical 1% 5-FU eye drops. After completing 2 cycles, she returned to the clinic with left eye epiphora. Clinical examination revealed mild punctal stenosis of the left lower punctum. The left punctum was dilated in the office at the slit lamp, and the patient was advised to see an oculoplastic specialist if the problem persisted. At this point, there was no residual tumor evidenced by clinical examination or anterior segment optical coherence tomography (AS-OCT). Having completed the planned 2 cycles of 5-FU for uncertain margins, the treatment with 5-FU was suspended and the patient continued with close follow-ups. Fortunately, after this visit, the epiphora resolved and the punctal stenosis did not recur in the subsequent follow-up of 48 months.

Patient 2

A 69-year-old male patient presented to the clinic with a pigmented conjunctival lesion of the left eye. Clinical evaluation and AS-OCT suggested OSSN, and the patient underwent excisional biopsy with wide margins and a no-touch technique. Pathological analysis confirmed conjunctival intraepithelial neoplasia with negative conjunctival margins and melanosis without atypia. Empirically, 2 months after excisional biopsy, the patient began postoperative adjuvant topical 1% 5-FU eye drops. After 2 cycles, the patient returned to the clinic for routine follow-up of the left eye with mild hyperreflectivity on AS-OCT; therefore, 5-FU was continued for 2 more cycles in the left eye. In addition, at this visit, a new lesion was observed in the right eye that was consistent with OSSN. As such, the patient began topical 5-FU therapy in the right eye as well. Two months later, the patient returned to the clinic complaining of left eye epiphora after having completed a total of four 1-week cycles in that eye; he had also completed 2 5-FU cycles in the right eye but did not have right eye epiphora at this time. The clinical examination and AS-OCT were completely normal in the left eye at that visit, and thus, the 5-FU was discontinued in the left eye. Still under treatment for his tumor in the right eye, the patient continued on topical 5-FU. On 2 months of follow-up later, the left eye epiphora resolved spontaneously, but the patient now complained of epiphora in the right eye and an elevated tear lake was noted (Fig. 1). At this visit, the patient was evaluated by an oculoplastic specialist, who noted a patent punctum but stenosis in the proximal canalicular system on the right side. Punctal dilation was performed on the right lower punctum. Following dilation, irrigation demonstrated complete passage with no reflux. At this time, the clinical examination and AS-OCT had normalized in the right eye and 5-FU was discontinued after having completed 4 cycles of treatment. Unfortunately, the patient had persistent epiphora in the right eye and required dilation again 2 months later. After the second dilation, the patient reported a resolution of symptoms in the right eye within 2 weeks and had no tearing of either eye for 32 months of follow-up.

FIGURE 1.

Patient 2. A, Pigmented OSSN (arrow) in the patient’s left eye. B, Resolved lesion after surgical excision and then 7 months of 5-FU drops. C, Left eye with blue filter shows mildly elevated tear lake after 5-FU. D, Conjunctival lesion (arrow) in the patient’s right eye. E, Resolved lesion after 4 months of 5-FU drops. F, Right eye with blue filter shows significantly elevated tear lake after 5-FU.

Patient 3

A 55-year-old male patient was referred to the clinic for evaluation of a left papillomatous conjunctival lesion. Clinical examination and AS OCT were both consistent with OSSN and the patient elected topical treatment with 5-FU eye drops. After the fourth cycle of 5-FU, the patient returned to the clinic for the routine follow-up and reported epiphora of the left eye that started during the third 5-FU cycle; the right eye was unaffected. At the visit, the punctum was noted to be open; however, a high tear lake was observed. As the tumor was still present, he was offered surgical intervention or continued medical treatment. Preferring medical therapy, the patient was offered punctal plugs but preferred to manually occlude the punctum with future 5-FU applications. The patient returned to the clinic after the sixth cycle of 5-FU with resolution of the tumor. The epiphora persisted, and he was referred for an oculoplastic evaluation. On probing, the patient was found to have canalicular stenosis of both upper and lower left eyelids. The patient chose to proceed with an external DCR with placement of a canalicular stent; however, the stent was replaced after 1 month due to stent extrusion. His symptoms resolved over 28 months of follow-up after the second surgery.

Patient 4

A 53-year-old male patient was referred for evaluation of a left corneal lesion. Clinical examination and AS-OCT were both consistent with OSSN and biopsy confirmed conjunctival intraepithelial neoplasia. The patient began topical 5-FU eye drops in cycles consisting of 1 drop of 1% 5-FU 4 times a day for 1 week followed by 3 weeks of no drops. The patient completed 2 cycles of 5-FU with excellent tumor response of clinical resolution. While we usually treated for 4 cycles, the patient deferred any further 5-FU. Approximately 10.3 months after initiating treatment with 5-FU (and 9 months after completing treatment), the patient presented to the clinic with left eye epiphora; the right eye was unaffected. He was referred to an oculoplastic specialist who diagnosed stenosis of both upper and lower puncta of the left eye. The patient elected surgical reconstruction and subsequently underwent left upper and lower lid punctoplasty with an unsuccessful attempt at donut stent placement due to extensive canalicular scarring. After surgery, the patient was given the choice to continue with conjunctivodacryocystorhinostomy with a Jones tube, but the patient chose observation instead. At his last follow-up, the patient’s epiphora remained unresolved. He is since lost to follow-up.

DISCUSSION

We present the first reported cases of punctal or canalicular stenosis following topical 5-FU therapy. Four individuals (5 eyes) with OSSN developed punctal or canalicular stenosis on average 4.6 ± 3 months (or 4.9 ± 2.0 cycles) after commencing topical 1% 5-FU eye drop therapy, representing a frequency of 1.3%. None of the patients had a history of epiphora or punctal/canalicular stenosis. Prior reports have described punctal or canalicular stenosis with systemic 5-FU therapy. In addition, a previous study found 6 of 10 rabbits receiving topical 1% 5-FU therapy to have canalicular stenosis.¹⁷ However, after a computerized search through the PubMed database, we were unable to find any reports of punctal or canalicular stenosis in humans receiving topical 5-FU therapy. Therefore, to our knowledge, we are the first to report this finding.

In a study, Eiseman et al¹¹ identified 3 cases of 52 who developed punctal-canalicular stenosis after receiving systemic 5-FU therapy, for a frequency of 5.8%. Epiphora began on average 3 weeks (SD, 60.9) after commencing systemic 5-FU therapy. Outcomes were not well described in this case series, but similar to 2 of our cases, 1 individual reported symptom resolution after probing and irrigation.

In a second study, Hassan et al¹³ identified 3 cases of 20 who developed punctal-canalicular stenosis after receiving systemic 5-FU therapy for gastrointestinal carcinoma, for a frequency of 15%. Tearing began at a mean of 2.7 months after chemotherapy initiation. The epiphora was conservatively managed with massaging and a short course of topical corticosteroids. All 3 patients declined surgical intervention and remained symptomatic.

In addition, a few cases have noted spontaneous resolution of epiphora and stenosis after cessation of systemic 5-FU. In a review of 18 individuals with systemic 5-FU–associated punctal-canalicular stenosis, 2 patients experienced resolution of their symptoms with topical corticosteroids, probing, and cessation of 5-FU,¹⁵ similar to 2 of our cases. Four individuals underwent surgical intervention (DCR) with varying levels of success: two experienced symptom relief, one failed DCR with silicone tubing, and details for one were not reported. Similarly, one of our cases required DCR and resulted in resolution of symptoms. Overall, punctal-canalicular stenosis following systemic or topical 5-FU appears to have similarly varied treatment outcomes. We were unable to identify any features that could be predictive of whether an individual would have spontaneous resolution or would require surgery. In addition, our findings from topical 5-FU show a lower prevalence (1.3%) and generally faster onset (1.6–10.3 months) when compared with systemic 5-FU (5.8%–15% prevalence; 1 week–26 months onset).^11,13,15,18

5-FU may cause stenosis of the punctum or the canalicular system, and possibly both. It is important to note that the punctum may be patented and that the inflammatory response leading to stenosis may be in the canalicular system. For example, patient 2 was evaluated by an oculoplastic specialist and was found to have a patent right lower punctum at the time he presented with epiphora of the right eye with proximal canalicular stenosis. The resolution of his symptoms occurred following 2 dilation procedures of the proximal canalicular system.

As mentioned previously, the only existing evidence in the literature of punctal or canalicular stenosis following topical 5-FU use is in rabbits.¹⁷ This study found 6 of 10 rabbits receiving topical 1% 5-FU eye drops to have clinical evidence of canalicular stenosis. However, it is important to note that these rabbits received topical 1% 5-FU eye drops 4 times a day for 4 consecutive weeks, whereas the typical regimen for human use is 4 times a day for a week followed by 3 weeks of no medication. This difference in administration may help explain the large difference in frequency of punctal and canalicular stenosis after topical 5-FU in rabbits (60%) and our study in humans (1.3%).

The exact mechanism by which the puncta and canaliculi stenosis following systemic or topical 5-FU remains unknown. Studies have confirmed the presence of 5-FU in the tear film of patients receiving systemic 5-FU therapy, which indicates that the drug may directly affect the puncta and canaliculi.¹⁹ One study in rabbits found histopathological evidence of epithelial hyperplasia, subepithelial scarring, atrophic changes, and hyalinized collagen bundles in the stroma of canalicular tissue of the sacrificed animals receiving 4 weeks of topical 1% 5-FU eye drops.¹⁷

Beyond rare punctal stenosis, topical 5-FU is generally well-tolerated, with mostly transient side effects that include redness, pain, edema, and keratopathy.⁸ Other topical chemotherapies for OSSN include IFN α-2b and MMC.⁷ IFN is usually very well-tolerated with minimal side effects.⁸ MMC is more toxic to the ocular surface and is more commonly associated with punctal stenosis and other adverse effects, as has already been established in numerous reports. A punctal plug is generally placed before treatment. When no plug is used, the frequency of punctal stenosis with topical MMC has been reported to be as high as 43% (range 4%–43%),^20–23 which is much higher than the 1.3% frequency noted in this study. In addition, the onset of punctal or canalicular stenosis is faster with MMC compared with 5-FU. In a study of 91 individuals who received topical MMC for OSSN (0.04% eye drops 4 times a day for 1 week followed a 1 week holiday; without punctal plugs), 14 presented with punctal stenosis a mean of 1.9 months after initiating therapy (range 0.75–4).²¹ In fact, symptoms of epiphora have been noted as early as 1 month with topical MMC,^22,23 as reported by 9 of 14 patients who reported epiphora when interviewed 1 month after starting treatment without the use of punctal plugs.²² Six of the 9 patients were confirmed to have punctal or canalicular stenosis by probing and lacrimal irrigation. Similar to our findings, symptoms improved in 2 individuals after probing, 3 individuals proceeded with surgical intervention (silicone intubation left in place for 2–4 months), and 1 individual declined surgical intervention. Given the well-documented, high frequency of punctal or canalicular stenosis with MMC, many physicians insert punctal plugs to protect the puncta during treatment with topical MMC.

Although the frequency of punctal or canalicular stenosis with 5-FU is low, it is nonetheless important for physicians and patients to understand the risks associated with its use. When a patient presents with epiphora secondary to punctal or canalicular stenosis, our first step would be to hold the 5-FU, as sometimes this can resolve the issue. In some cases, a simple punctal/proximal canalicular dilation may solve the problem. If persistent, we have the patient seen by an oculoplastic specialist, as punctal or canalicular stenosis may require further surgical intervention. Options to minimize the risk of stenosis include manually occluding the punctum immediately after instilling 5-FU and/or punctal plug placement. Given the ease and affordability of manual punctal occlusion, it may be reasonable to err on the side of caution and recommend this to patients taking topical 5-FU; other patients may opt for punctal plugs depending on insurance coverage and personal preferences. We do not routinely place punctal plugs with topical 5-FU eye drop therapy, but since this novel observation of punctal and canalicular stenosis described in this study, we recommend manual punctal closure for 30 seconds with each 5-FU drop instillation. No additional cases of punctal or canalicular stenosis have been identified.

In conclusion, 5-FU is a topical therapeutic agent with a low (1.3%) frequency of punctal and/or canalicular stenosis that has previously not been reported. Despite this potential adverse effect, 5-FU remains a highly effective and safe treatment option for OSSN with tolerable side effects.