Intralesional 5-fluorouracil (5-FU) for the treatment of eyelid squamous cell carcinoma in 5 horses



Eyelid squamous cell carcinoma in equine patients often presents a therapeutic challenge to practitioners due to the generally large area affected upon presentation. Surgical excision can be curative if wide enough margins are achieved, but this is not often attainable without enucleation. Other alternatives have been examined including cryotherapy, radiotherapy, brachytherapy, intralesional chemotherapy and photodynamic therapy. Intralesional chemotherapy using cisplatin, mitomycin-C and bleomycin have been shown to be successful in treating eyelid squamous cell carcinomas but may be prohibitive to some owners due to the cost of therapy. To the authors' knowledge, this is the first case series to illustrate the effectiveness of intralesional 5-fluoruracil in treating large equine eyelid squamous cell carcinomas. Macroscopic reduction in the size of the affected areas was noted in each case together with regression of clinical signs associated with the mass. This, coupled with the relative inexpensive nature of the procedure, makes this technique an attractive therapy for either primary treatment of eyelid squamous cell carcinoma or as a cytoreductive technique prior to surgical excision.


Squamous cell carcinoma (SCC) is the most common tumour of the equine eye and adnexa (Lavach and Severin 1977; Dugan et al. 1991a). It is usually unilateral but can present bilaterally. Common ocular SCC locations include the nictitating membrane (28%), cornea/limbus (28%) and eyelids (23%) (Lavach and Severin 1977; Schwink 1987; Dugan et al. 1991b). Numerous factors have been labelled as predisposing to the development of SCC but the most commonly implicated is the role of ultraviolet radiation (UV) exposure. Other predisposing factors which have been proposed include any chronic irritation, lack of periocular pigmentation, age, breed, sex and geographic location (Rebhun 1998; Hendrix 2005; Giuliano 2010a,b).

Treatment options and prognosis vary depending on the location of the tumour. Squamous cell carcinoma of the eyelids and nictitating membrane have the highest recurrence rate and are associated with the worst prognosis in cases of recurrence (Dugan et al. 1991b). Complete excision of eyelid SCC can be curative but excision is limited by the size of the mass and the ability to adequately preserve a functional eyelid aperture. A functional eyelid is crucial for proper maintenance of ocular health and protection. Eyelid SCC that is too large to excise in situ requires alternative therapy for cytoreduction. Options currently available for large eyelid SCC include CO2 laser ablation, cryotherapy, hyperthermia, brachytherapy, radiotherapy, photodynamic therapy and various chemotherapeutic drugs (Hendrix 2005; Giuliano 2010a,b). The overall cost of therapy together with the need for specialised equipment has led to the increased use of intralesional chemotherapy for eyelid SCC.

Cisplatin is the chemotherapeutic drug most commonly used to treat eyelid SCC (Giuliano 2010a). Other chemotherapeutics have been used including bleomycin, mitomycin-C and 5-FU (Theon et al. 1997; Hendrix 2005; Rayner and Van Zyl 2006). The use of these drugs intralesionally offers the advantages of direct chemotherapy administration, concentrated levels of the drug at the site and minimal systemic adverse effects. Cisplatin and bleomycin are relatively expensive to use which may prohibit their use in certain patients. The use of 5-FU in equine eyelid SCC is an inexpensive option with relatively few adverse effects and can provide beneficial macroscopic cytoreduction along with resolution of clinical signs as evidenced in this case series. Although review articles contain anecdotal comments regarding the use of intralesional 5-FU for eyelid SCC in horses, to our knowledge this is the first report of treatment and outcome in a series of cases.

Case details


Five horses were referred to the Centre for Veterinary Health Sciences at Oklahoma State University with a complaint of suspected eyelid SCC and treated with intralesional 5-FU between 2005 and 2008. The horses were middle aged (8–14 years) and included various breeds (Table 1). All horses lacked periocular pigmentation around the affected eye(s), were geldings and primarily kept outside. Three of the 5 horses had the tumour diagnosed histologically after biopsy. Diagnosis for the other 2 horses was based on characteristic clinical appearance including lack of eyelid pigment, actinic changes of the eyelid margins, eyelid/conjunctival location and red, ulcerated, friable, cobblestone surface. In all cases the size of the eyelid mass was deemed too large for complete excision and primary closure. Medical and surgical therapies were discussed with the owners and intralesional 5-FU was ultimately chosen due to the extent of eyelid involvement, cost consideration and preference to maintain a visual eye.

Table 1. Summary of case details, location of tumour, treatment and outcome in 5 horses with eyelid squamous cell carcinoma treated with intralesional 5-FU
Case No.BreedAge (years)SexLocation of eyelid SCCBiopsyNo. 5-FU treatmentsDuration of treatment (months)Response to treatmentAdditional therapy
  1. G = gelding; OD = right eye; OS = left eye; LTFU = lost to follow-up; UV = ultraviolet.

1Missouri Fox Trotting Horse13GOS superior and inferiorNo21Slight reduction in size after first treatment then LTFUTopical 1% 5-FU, UV mask
2American Paint Horse8GOD inferior and lateral canthusYes48Moderate reduction in sizeTopical 1% 5-FU
3American Quarter Horse14GOS inferiorYes735Marked reduction in sizeCO2 laser excision, topical 1% 5-FU, UV mask, enucleation
4American Quarter Horse14GOS superior and inferiorNo44Marked reduction in sizeTopical 1% 5-FU
5American Paint Horse12GOD inferiorYes913Marked reduction in sizeFly mask

Clinical findings

A thorough ophthalmic examination including fluorescein stain, biomicroscopy and indirect ophthalmoscopy was performed on each horse. All adnexal SCC tumours in these cases were confined to the eyelids. The remaining ophthalmic examination findings were normal with the exception of Case 2 that also had focal, limbal, lymphoplasmacytic keratitis (histological diagnosis via biopsy). Tumour location and size were diagrammed or measured. Photographs were taken at most visits to record efficacy of therapy. Tumours were unilateral and involved only the inferior eyelid in 3 cases and both inferior and superior eyelids in 2 cases (Table 1). Case 2 had extension from the lower lateral eyelid into the lateral canthus. All affected eyes had considerable mucoid to serous ocular discharge, blepharospasm and sensitivity to touch. Tumours had a grossly ulcerated surface. Palpation of all eyelid margins revealed only eyelid involvement without extension to the nictitans or orbital rim. A complete physical examination, including palpation of regional lymph nodes, was performed on all horses with no abnormalities noted.

Histopathological findings

Biopsy was recommended in all cases to confirm the clinical diagnosis but was declined initially by the owner in 3 of the cases. Histopathology was performed on biopsy samples taken at the initial visit before treatment in Cases 2 and 5. Case 3 had a total of 3 biopsies performed. The first biopsy was obtained during excisional surgery following 4 intralesional 5-FU treatments. Histopathology of all biopsy samples exhibited marked anisocytosis and anisokaryosis. The mitotic index per 10 random high power fields (HPF) was 24 in Case 2 and 4 in Case 5. The initial biopsy in Case 3 following the four 5-FU treatments showed only rare mitotic figures. A second biopsy was taken 10 months after the first and exhibited a mitotic rate of 4 per 10 HPF and a third biopsy taken 31 months after the first exhibited a mitotic rate of 11 per 10 HPF.


Prior to each treatment the tumour sites were clipped and prepped using 1% povidone-iodine scrub and solution. All safety protocols were followed with regards to administration and disposal of chemotherapeutic drugs and administration sets. Each of the 5 horses underwent intralesional 5-FU1 (50 mg/ml) therapy with standing sedation and topical anaesthetic (lidocaine 2%)2 administered over the ulcerated surface of the tumour. Each tumour location was injected until saturation was noted, either by inability to inject more drug and/or drug leaking out of previous injection sites within the friable tissue. Volume of drug used per tumour depended on tumour size and varied between 4 and 12 ml. All owners were instructed to avoid contact with the treated tissue and ocular discharge for 48 h after 5-FU administration. The frequency of injections varied from every 2–5 weeks depending on clinician preference, owner's ability to return and response to therapy. Intralesional therapy was discontinued when clinical signs were controlled including reduction of tumour size, no ulcerated tissue, and no blepharospasm (Cases 3 and 5 and the lower eyelid tumour in Case 4). In Cases 1, 2 and 4 the therapy was discontinued prematurely due to owners not returning for recommended recheck examinations. In Cases 2, 3 and 5 injections were repeated when there was a return of clinical signs such as increased swelling or ulceration. Duration between treatments for relapses was 5 months (Case 2); 9, 17 and 7 months (Case 3) and 9 months (Case 5).

Three of the 5 horses also were prescribed once daily topical 1% 5-FU ointment3 (Table 1). In Case 1 the ointment was discontinued at the second visit due to a skin reaction (yellow discolouration and mild erosion). In Case 2 the owner used the ointment sporadically and discontinued on his own at an undetermined time. In Case 4 the ointment was used for at least 2 months.

Use of a mask that blocks UV light (Guardian mask)4 was recommended in 4 of the 5 cases. Of those 4, the ultraviolet protective mask was only utilised in Cases 1 and 3. Case 5 wore a fly mask regularly.


Side effects of the 5-FU injections were minimal. Swelling at the injection sites from the volume of drug administered, minor bleeding from injection sites and focal hyperaemia of the treated areas were immediate side effects that all resolved within the first week.

In each case, the 5-FU was successful in reducing the size of the tumour as well as decreasing clinical signs (Table 1). Complete regression of the tumours was not achieved; however, the 5-FU injections were very effective at resolving the ulcerative component of the tumour and alleviating discharge and blepharospasm. Assessing response to therapy in these 5 cases is difficult due to unreliable owner compliance with respect to returning for recommended treatments. Case 1 showed improvement in the superior eyelid SCC after the first 5-FU treatment but never returned after the second treatment. Case 2 showed marked reduction in the area affected after 2 injections (Figs 1a,b), but did not return following the third injection for 5 months at which time the area affected was similar to the initial presentation. Case 3 was followed for 3.5 years. The horse received 3 initial injections and did well for 9 months following the last injection (Figs 2a,b). The horse re-presented when the owner noticed swelling without ulceration in the lower eyelid. The mass was injected with 5-FU and examination 4 weeks later revealed an increase in size of the mass. The mass was excised with a CO2 laser and the surrounding tissue ablated with the CO2 laser followed by cryotherapy. Sixteen months later the medial, inferior eyelid was thickened and had a small ulceration on the surface. Repeat 5-FU injections resulted in shrinking of the thickened tissue and resolution of the ulceration. Seven months following the last injection mild swelling in the same area was noted and 5-FU was injected. The owner did not return for 7 months at which time the mass in the medial canthus had enlarged and on palpation was adhered to the orbital rim. Enucleation was performed. Case 4 showed a steady reduction in the size of the tumours but did not return following the fourth treatment (Figs 3a–c). Case 5 received 4 initial injections over 2 months resulting in a marked reduction in size of the tumour. Nine months following the last injection swelling of the eyelid was noted again. Five more 5-FU injections were given over the next 2 months resulting in marked reduction in size.

Figure 1.

a) Case 2 inferior eyelid and lateral canthus SCC at initial presentation. b) Case 2 three weeks following the first 5-FU treatment. The affected area is smaller and flatter with less ulceration and a smoother eyelid margin.

Figure 2.

a) Case 3 presumed inferior eyelid SCC at Week 3, 3 weeks following the first 5-FU treatment. b) Case 3 at Week 10, 4 weeks following the third 5-FU treatment. The inferior eyelid margin is smooth, nonulcerated and functional.

Figure 3.

a) Case 4 at initial presentation. Presumed SCC involving the medial one third of the superior eyelid and the lateral two thirds of the inferior eyelid. b) Case 4 at Week 4, 4 weeks following the first 5-FU treatment. The thickness of the tumours and ulcerated areas are markedly reduced. c) Case 4 at Week 9, 5 weeks following the second 5-FU treatment. The eyelid thickening at both sites has almost completely resolved (solid arrows); the remaining lesion at the eyelid margin is white and nonulcerated (open arrow).

Four of the 5 owners were available for follow-up and a standard set of questions was asked. The length of time from the last visit to the follow-up contact for each case was 4 years (Case 1), 3.5 years (Case 2), 6 months (Case 3), 4 years (Case 4) and 2 years (Case 5). None of the horses were treated with any other medication for the SCC other than the 5-FU therapy instituted at Oklahoma State University. All owners reported a visible reduction in the size of the mass and clinical signs during treatment. All owners believed the cost of therapy was reasonable and the duration of time between treatments was acceptable. Owners agreed they would use 5-FU again if another SCC developed on the eyelids in the future. The only complaint voiced by the owners was that they would have preferred to receive treatments closer to their home.

Long-term follow up information was not available for Case 1 (died of an unrelated cause), Case 2 (sold shortly after last visit and new owner had the eye enucleated so as to not need additional therapy) or Case 4 (owner died so follow-up limited to last visit). Case 3 follow-up ended with enucleation and the owner of Case 5 reported no obvious recurrence of the tumour 2 years since the last injection but the area affected was slightly reddened and hairless.


A number of predisposing factors have been implicated in the formation of eyelid SCC including UV exposure, lack of periocular pigmentation, age, sex and breed (Hargis 1981; Dugan et al. 1991a; Giuliano 2010b). Breeds more commonly affected with eyelid SCC include those with little or no periocular pigmentation. The age of an animal can also affect their risk of developing SCC as older animals have more prolonged UV exposure compared to younger animals. Approximate age at initial diagnosis ranges from 3–26 years, with a mean of about 10 years in other reports (Lavach and Severin 1977; Schwink 1987; Dugan et al. 1991a; Giuliano 2010a). All of the horses reported here had no periocular pigmentation of the affected eyelids. Age range for the horses was 8–14 years (mean 12.2 years) and all were geldings. Geldings are reportedly twice as likely than intact females and 5 times more likely than intact males to develop ocular SCC. The reason for this is unknown at the present time but it is postulated that androgens, oestrogens or both may play a protective role when present (Dugan et al. 1991a).

The use of intralesional 5-FU has been described for dermatological use in human patients with varying degrees of success (Kraus et al. 1998; Longley et al. 2003; Morse et al. 2003; Good et al. 2010; Kirby and Miller 2010). It is an alternative to other intralesional chemotherapy drugs that does not require specialised preparation or equipment to deliver. 5-FU is a fluoropyrimidine anti-metabolite drug that exerts its effect on both DNA and RNA during the S-phase of the cell cycle. By inhibiting thymidylate synthase 5-FU is able to halt the production of thymidylate, thereby allowing for misincorporation of nucleotides into DNA strands and subsequent strand breakage. The effects of 5-FU on RNA are a result of the incorporation of the 5-FU metabolite, fluorouridine triphosphate (FUTP), into RNA which serves to disrupt RNA processing at multiple levels (Miller 1971; Schmoll et al. 1999; Longley et al. 2003; Scartozzi et al. 2011). Some authors have reported the co-administration of epinephrine with 5-FU can improve local concentrations (Kraus et al. 1998), but this was not used in these cases. In the cases reported here, treatment with intralesional 5-FU resulted in a decrease in tumour size which was variable between cases. The most consistent result of intralesional 5-FU was resolution of the ulcerated surface of the tumour, resolution of the ocular discharge, softening of the tumour tissue and markedly improved comfort. Following treatment, the blepharospasm and increased sensitivity to touch around the eye resolved. Intralesional injections were simple to give and resulted in no clinically significant side effects.

Compared to other chemotherapeutic agents 5-FU is also relatively inexpensive therefore multiple injections are feasible. Although the cost of therapy will vary based on the amount of drug used and the frequency of administration, the average wholesale price of 5-FU is 5 United States dollars (USD) per 500 mg (10 ml) vial compared to 25 USD per 50 mg (50 ml) vial of cisplatin and 80 USD for 15 iu of bleomycin, the smallest quantities available. In addition, both cisplatin and bleomycin require mixing with sterile water and sesame oil before administration (Theon et al. 1997). One concern of the owners in this study was making the return trips to the referral facility. To ameliorate this concern additional 5-FU injections could readily be performed by the referring or local veterinarian. Practitioners performing 5-FU injections would need to follow routine chemotherapy protocol including protective attire (gloves, gown and eye protection) and appropriate disposal of 5-FU drug and supplies. Client instructions should include avoiding contact with the treated area for 48 h.

It is important to note that 5-FU did not cure the SCC tumours in this study. It should be presented to owners as a palliative measure to control growth of the tumour and provide relief from clinical signs. It also has a role in decreasing the size of the tumour making excision and retention of a functional eyelid margin possible. Case 3 is an example of 5-FU controlling the tumour growth and enabling surgical resection of the lateral tumour curing all but the tumour in the medial canthus. This horse maintained a functional and comfortable eye for 3.5 years following diagnosis.

The negative aspects of intralesional 5-FU include empirical dosing, off-label use of the drugs, variability in volume per treatment, multiple treatments required and variability in tumour response. The limitations in this case series include the small number of cases, unconfirmed diagnosis in 2 horses, lack of a standard treatment protocol resulting in variable times between 5-FU treatments and variations in ancillary therapies, limited owner compliance with the treatment schedule recommendations and lack of a control group. Given these limitations it is not possible to compare responses between the cases which therefore does not allow for providing a reliable, long-term prognosis for owners.

Only 3 cases had histopathology performed. The limited number of biopsy samples along with variable lengths of treatment limit any conclusions regarding efficacy of 5-FU with respect to aggressiveness of the tumour. Subjectively, Cases 3 and 5 with low mitotic indexes responded better than Case 2 with the highest mitotic index. Case 3 had 3 biopsies performed throughout the course of treatment. Interestingly, each biopsy showed a progression in the mitotic rate of the tumour. This raises the question whether repeated 5-FU injections can induce a more aggressive neoplasm or if the change in mitotic rate was a natural progression of the tumour over time independent of 5-FU effects.

The results noted in these cases show that intralesional 5-FU could potentially provide another reasonable alternative for the management of eyelid SCC in horses. It may be particularly useful in cases where cost of treatment is a concern and/or when the presenting tumour is too large for excision, either as the sole therapy or to decrease tumour size prior to surgical excision. Future studies using a randomised controlled trial to compare 5-FU to other intralesional chemotherapy drugs would provide a more objective assessment of the efficacy of 5-FU.

Manufacturers' addresses

1 Parenta Pharmaceuticals, West Columbia, South Carolina, USA.

2 Sparhawk Laboratories, Lenexa, Kansas, USA.

3 The Prescription Center, Fayetteville, North Carolina, USA.

4 Guardian Mask, Burnet, Texas, USA.