Surgical treatment of iris and ciliary body melanoma: follow-up of a 25-year series of patients


Professor Jan U. Prause
Eye Pathology Section
Institute for Neuroscience and Pharmacology
University of Copenhagen
Frederik V’s Vej 11
DK 2100 Copenhagen
Tel: + 45 3532 6070
Fax: + 45 3532 6080


Purpose:  To evaluate outcome of surgical resection of iris and irido-ciliary melanomas.

Method:  Retrospective analysis of all cases treated in Denmark 1975–1999 with clinical follow-up in 2002 and death certificate analysis in 2008. A quality of life questionnaire was completed at follow-up.

Results:  A total of 53 patients were identified. Of these, 47 were examined at follow-up. Median observation time was 7.15 years (range 0.3–27.4 years). Five patients had died of nonmelanoma causes, and one could not be reached because of immigration. None of the patient had melanoma metastases, and none had died of melanoma-related causes. Only one patient had a local recurrence, which was successfully treated by cryotherapy. The quality of life-related questions demonstrated that most patients (40) suffered from photophobia, and eight patients had changed their driving habits, not driving at night time. However, none had changed job as a consequence of the surgical treatment. Only two patients were emotionally affected by the diagnosis of iris melanoma.

Conclusion:  Resection of small iris and irido-ciliary melanomas is a safe and efficient procedure, provided that strict diagnostic and surgical procedures are followed and the preoperative intraocular pressure is normal.


Melanoma of the iris constitutes approximately 10% of all uveal melanomas, and despite that some ocular cancer centres also apply radiotherapy in the treatment (Damato et al. 2005), the standard treatment for iris melanoma is iridectomy, and irido-cyclectomy for tumours also involving anterior parts of the ciliary body and angle (Conway et al. 2001). Outcomes of local excision surgery vary among reports (McGalliard & Johnston 1989; Naumann & Rummelt 1996; Conway et al. 2001; Shields et al. 2001) and depend on histological, therapeutic and surgical criteria. We have applied the same strict criteria for surgical treatment of anterior uveal melanomas since 1975. The aim of this study is to present our long-term results.

Materials and Methods

The melanoma database at the Eye Pathology Section, University of Copenhagen, and the clinical database of the Eye Department, Rigshospitalet, were screened for patients, surgically treated for melanomas of the iris and ciliary body during the period 1975–1999. All retrieved Danish patients were included in the study. Clinical, biochemical and imaging data were obtained from the hospital files retrospectively. A clinical follow-up examination was carried out in 2002. Re-examination consisted of best visual acuity (VA; Snellen chart 6 m), Javal keratometry, intraocular pressure (IOP), slit lamp examination including Posner impression gonioscopy for recording of possible recurrence of tumour, development of satellites and inflammation. Slit lamp photos of the surgical edges were obtained through the Posner lens or the Goldman gonioscope. If indicated, the patients were examined with ultrasound biomicroscopy (UBM).

The patient-completed questionnaire is related to quality of life. The questions were:

  • Do you experience increased photophobia after the surgery?

  • Has the treatment affected your job situation?

  • Has the treatment affected your driving habits?

  • Has the malignant diagnosis of your disease affected your quality of life?

Clinical diagnostic criteria for the selection of treated melanomas

Iris melanoma

Elevated, localized solid tumour mass, most often pigmented and sometimes with superficially gelatinous areas; pupillary ectropion; localized cataract adjacent to the tumour; invasion of adjacent tissue and/or satellites in the anterior chamber angle; documented growth.

Ciliary body melanoma

Circumscribed elevated, tumour mass with shadow on transillumination; ± invasion of anterior chamber angle; localized cataract adjacent to tumour; characteristic echo pattern on UBM; documented tumour growth.

Criteria for selected type of surgery

Tumours were selected for surgery if it was technically possible to incise normal tissue ≥1 mm from the border of the melanoma. The same surgeon (SVK) performed all operations. Iris melanomas located in the central and middle part of iris, extending up to one quadrant and with normal IOP and satellites up to two quadrants, were treated with iridectomy (Fig. 1).

Figure 1.

 Iridectomy and iris sutures for localized iris melanoma. (A) Sclero-corneal flap and deep keratectomy with the tumour accessible. (B) Tumour excised. (C) Iris sutures and closed incisions.

Peripheral iris melanomas involving the iris root, but not the ciliary band or trabecular meshwork and extending up to one quadrant with normal IOP and satellites up to two quadrants, were treated by partial anterior irido-trabeculo-cyclectomy.

Irido-ciliary melanomas extending up to one quadrant of the iris/ciliary body and Schlemms canal, with normal IOP and not reaching the ora serata (Fig. 2A), were treated by complete anterior irido-trabeculo-cyclectomy. Tumours not fulfilling the criteria mentioned above lead to enucleation of the eye and were excluded from the present study.

Figure 2.

 Lamellar block excision of ciliary body melanoma with irido-trabeculo-cyclectomy. (A) Preoperative gonioscopic view of chamber angle invasion. (B) Sclero-corneal and scleral flaps at the tumour area. (C) The lamellar block excised tissue (chamber angle visualized) placed upon the partially closed eye. (D) Closed incisions.

Surgical techniques


After a conjunctival peritomy, a thick sclero-corneal flap (3/4 of the sclera/cornea thickness) was created with the scleral, limbal parallel incision 2–3 mm posterior to the limbus and the radial incision about 2 mm outside the border of the tumour. A keratectomy was made over the tumour through the corneal floor (Fig. 1A). Then, incisions were made in the iris ≥1 mm outside the tumour (Fig. 1B), in case of melanoma near the pupil as a sector iridectomy, otherwise as a peripheral iridectomy. Closure was achieved with 10-0 nylon in the cornea, virgin silk 9-0 in the sclera. Iris sutures (Fig. 1C) were used after 1995. Satellites were treated postoperatively with focal diode laser photocoagulation (0.5 second, 200 μ spot diameter, 500 mW).

Partial anterior irido-trabeculo-cyclectomy

This includes a peripheral iridectomy with the adjacent chamber angle and anterior part of pars plicata (ciliary band) of the ciliary body. After a conjunctival peritomy, a sclero-corneal flap was created as described earlier, covering the tumour area. The anterior chamber was opened with a limbal parallel incision in the corneal floor near the anterior tumour margin. From this incision, two radial corneal incisions were made beyond the chamber angle through the scleral floor, just including the anterior part of the ciliary body. Then iris incision on both sides of the tumour was made, beginning ≥1 mm anterior to the tumour and extending posterior to the ciliaris sulcus. This was followed by an incision parallel to the limbus made through the sulcus and the scleral floor. The incision is closed as mentioned earlier. Satellites were treated as indicated earlier.

Complete anterior irido-trabeculo-cyclectomy

This includes a peripheral iridectomy with the adjacent chamber angle, a cyclectomy with a part of pars plana and the adjacent corneal and scleral lamellas (Fig. 2C). The technique was a modification of the deep lamellar block excisions technique of Barraquer (1972).

After a wide conjunctival peritomy, a sclero-corneal flap was created over the tumour area as mentioned earlier and advanced into the cornea covering the anterior part of the tumour. From the limbal parallel scleral incision, a posterior radial, deep scleral incision (3/4 of the scleral thickness) was made over the estimated centre of the tumour (T-incision) in the ciliary body followed by two posterior scleral flaps created to both sides of the radial incision (Fig. 2B). The anterior chamber was opened as mentioned earlier, and incisions into the iris ≥1 mm outside the tumour were made. These were continued with radial cuts through the chamber angle on both sides of the tumour as far as the ciliary sulcus. Then, 10-0 nylon sutures were placed in the corneal part of the superficial sclero-corneal flab. Further, radial cuts were made through the scleral floor and the underlying pars plicata on both sides of the tumour, and these were extended into limbus. Parallel incisions were made through the scleral floor and pars plana posterior to the tumour. The radial scleral part of the sclero-corneal flap and the radial part of posterior scleral flaps were sutured after each step (Virgin silk 09), leaving the limbal parallel, scleral incision between the anterior and posterior flaps without sutures. Through this slit, the tumour was slowly extracted (Fig. 2C). The surgery was completed with suturing of the conjunctiva (Fig. 2D).


Paraffin-embedded specimens were oriented according to a surgical sketch and divided along a plane through the centre of the corneal lamella, via the chamber angle and through the centre part of the iris and adjacent ciliary body. The two parts of the specimens were embedded in the same paraffin block and sliced with the resection margins in the first slice. Subsequent slices included the more central parts of the tumour and adjacent tissue. Each slice comprised several 4-μm sections. Sections were stained with Haematoxylin and Eosin and examined by light microscopy. Immune-histochemistry was not applied as part of this study.

Criteria for the diagnosis melanoma were: loss of naevoid growth pattern with loss of discrete margins; plaque-like growth of tumour cells on the mesodermal surface of the iris; invasion of the trabecular meshwork and corneal endothelium; mixture of epithelioid and spindle cells.


A total of 53 patients (53 tumour-affected eyes) were identified. There were 28 women and 25 men. Their median age at surgery was 52 years (range 17–77 years). At follow-up, 47 (89%) of the patients participated. Five patients had died before follow-up in 2002 of causes not related to the ocular tumour. One had emigrated and could not be reached.

Preoperative operative characteristics

The left eye was affected in 29 of the 53 tumour eyes. A primary iris tumour was present in 45 eyes. Of these, 11 were located at the pupillary margin and 34 in the peripheral part of the iris. A primary ciliary body tumour was noted in eight cases. The median radial tumour length of the iris tumours was 3.6 mm (range 1–14 mm), and the median transversal tumour diameter was 3.0 mm (range 1–10 mm). Invasion of the iris root, trabecular meshwork or ciliary body was noted in 23 of the 45 iris tumours. Nine of the melanomas were amelanotic, and gelatinous areas within the tumour were seen in 20 cases. In 19 eyes, the tumours were adherent to the cornea. Pigment deposition in the trabecular meshwork was recorded in 14 eyes, but real tumour satellites were found in five eyes only. A local cataract was found preoperatively in 18 eyes.

Surgical results

All 11 pupillary-near and three peripheral iris tumours could be treated by iridectomy alone. A partial anterior irido-trabeculo-cyclectomy was performed in 24 patients, and a complete procedure was performed on 15 patients. Thirty-seven patients had no intra- or postoperative complications. Hyphaema was the most frequent problem, occurring in 10 patients (Table 1). Only one patient had retinal detachment, and no patients lost their eye as a result of the surgery. The five patients with tumour satellites in the inferior trabecular meshwork were treated with focal diode laser photo coagulation. In five specimens, histology demonstrated tumour extension to the surgical margins; however, clinically, there was no evidence of residual tumour. These patients did not receive supplementary therapy.

Table 1.   Complications.
Type of complicationsNumber
  1. IOP = intraocular pressure.

 Severe preoperatively haemorrhage2
 Shallow anterior chamber2
 Subluxation of the lens2
 Retinal detachment1
 Hyphaema, IOP >24 mmHg5
 Hyphaema, IOP <24 mmHg5
 Cataract, requiring extraction4
 Wound leakage3
 Vitreous haemorrhages1
 Vitreous loss1
At follow-up
 Cataract progression postoperatively10
 Cataract development postoperatively4
 Local recurrence1
 Retinal detachment1

Histological results

All 53 specimens were suitable for histological analysis and fulfilled the criteria for the diagnosis of melanoma. Tumour-free margins were found in 47 specimens; however, five of the irido-trabeculo-cyclectomy specimens had at least one margin containing tumour cells.

Follow-up results

The median observation time of the 47 patients examined was 7.15 years (range 0.3–27.4 years). No patient had melanoma metastases, and none died of melanoma-related causes. Five patients died of other causes and without metastases. The VA of the patients was in general not affected by the treatment (Fig. 3). The number of patients loosing vision was in balance with the number of patients gaining vision. Cataract was the main complication, occurring in 34% of patients. We found 18 eyes with preoperative, localized cataracts; of these, 10 progressed. At follow-up examination, we found further four cataracts that were not present preoperatively (Table 1).

Figure 3.

 Correlation between preoperative and postoperative follow-up visual acuity.

One of these patients had a complicated cataract extraction that was followed by a retinal detachment resulting in phthisis and enucleation. None of the five patients with tumour satellites demonstrated growth of the satellites. Most of these satellites presented as flat pigmented scars. None of the five patients with histological evidence of incomplete resection developed clinical signs of recurrence. One patient developed a focal, histologically verified recurrence adjacent to a scleral channel for an aqueous vein. The lesion was successfully treated by excision and cryotherapy. Three years later, there was no recurrence or metastases. One patient had reactivating of an anterior uveitis not related to the tumour. One patient, in whom the iris tumour touched the corneal endothelium and trabecular meshwork, developed increased IOP, which fluctuated between 20 and 30 mmHg (Table 1). The same patient developed corneal endothelial decompensation. During the 27-year follow-up, there has not been any deterioration in this patient’s condition, and the eye is still retained. The patient with retinal detachment was a high myope. She had a small tumour-related detachment before surgery. After surgery, she developed total retinal detachment, which was successfully treated by buckling and cryotherapy.

The quality of life-related questionnaire demonstrated that most patients (40) suffered from photophobia, and eight patients had changed their driving habits, not driving at night time (Table 2). However, none had changed job as a consequence of the surgical treatment. Only two patients reported being emotionally affected by the diagnosis of iris melanoma.

Table 2.   Quality of life questionnaire.
QuestionsHighly affected
Number of patients (%)
Moderately affected
Number (%)
Number (%)
Have your quality of life been negatively affected by your disease?2 (4)14 (30)31 (66)
Have you developed photophobia?40 (85)7 (15)0
Have you changed job?0 (0)47 (100)0
Do you have driving problems?8 (17)39 (83)0


The study consisted of a follow-up examination of a consecutive series of patients with iris- and ciliary body melanoma. However, a prospective, double blind comparison of surgical and plaque therapy of these rare tumours is not possible in a population of five million people. Our material is complete for the entire nation within the study period. A high proportion of the patients (47 of 53, 89%) participated in the follow-up examination. Five patients had died from causes that were unrelated to the ocular melanoma, according to death certificates. One patient had emigrated and could not be reached. The diagnostic criteria that led to surgery were firm, and the subsequent histopathological analysis confirmed the clinical diagnosis of an iris- or ciliary body melanoma in all cases. Our study involves only histopathologically verified malignant iris and/or ciliary body melanomas suitable for local resection. Other studies have used either clinical or histological diagnoses (Shields et al. 1983; Damato et al. 2005) or have included a high number of cases with associated secondary glaucoma (Shields et al. 2001). Naumann & Rummelt treated larger tumours and applied a range of surgical methods. At follow-up, two of their 47 melanoma patients had local recurrence, four eyes had been enucleated, and four patients had died of metastasis (Naumann & Rummelt 1996). All published studies have found low metastatic- and mortality rates compared to choroidal melanomas. Iris tumours are easy to observe by patients, and at diagnosis, they have a small volume compared to choroidal tumours (55 mm3 and 300 mm3, respectively); this may be one of the causes for the low metastatic prognosis at time of diagnosis of only 3% (Davidorf 1981). A retrospective study of 80 patients treated between 1961 and 1985 found a 10% death rate from iris melanoma metastasis (Jensen 1993). However, half of these patients had ring melanoma. None of the patients with a pure iris melanoma died from metastasis. We only had one patient with local tumour recurrence. A reason for this low rate of recurrence may be that the median tumour diameter in our cohort was only 3.6 mm when compared to a median of 5.0 mm in other studies (Shields et al. 2001). Our median patient age at time of treatment was 52 years, which is in accordance with that of previous studies (Jensen 1993; Conway et al. 2001; Shields et al. 2001; Damato et al. 2005).

The quality of life of the patient at the follow-up was influenced by a disturbing photophobia caused by the large iris coloboma in 85% of patients. However, only 17% of patients changed their driving habits because of the photophobia, which suggested that the photophobia was comparatively mild. Conway et al. (2001) recorded troublesome photophobia among 25% of their patients. This difference could be explained by the difference in latitude between Denmark and Sydney, Australia.

Preoperative cataract progressed in 10 patients at follow-up. This may have been caused by the melanoma surgery or the long time to follow-up. We consider the progression in cataract a minor complication and a low cost in the treatment of a malignant disease.

The results of the quality of life questionnaire demonstrated that the patients received sufficient information on their long life expectancy and low recurrence rate. Only two patients (4%) were highly affected by their malignant diagnosis, while 66% were unaffected.

This study demonstrates that resection of small iris and ciliary body melanomas is a safe procedure if strict diagnostic and surgical procedures are followed and only eyes with normal preoperative IOP are included. VA remains good after surgery. Only one eye had to be enucleated, and this was not related to the iridocyclectomy. No patients developed metastasis.


Mauritzen La Fontaine Foundation, Copenhagen, generously supported the study.