Secondary Logo

Journal Logo

Original Study

Conjunctival Melanoma: Outcomes based on the American Joint Committee on Cancer Clinical Classification (8th Edition) of 425 Patients at a Single Ocular Oncology Center

Shields, Carol L. MD; Yaghy, Antonio MD; Dalvin, Lauren A. MD; Vaidya, Sarangdev BA; Pacheco, Richard R. BA; Perez, Alejandro L. MD; Lally, Sara E. MD; Shields, Jerry A. MD

Author Information
Asia-Pacific Journal of Ophthalmology: March-April 2021 - Volume 10 - Issue 2 - p 146-151
doi: 10.1097/APO.0000000000000343

Abstract

Conjunctival melanoma is a rare malignancy, arising in the periocular tissues and potentially leading to metastatic disease and death.1,2 Several studies have documented that this once-rare malignancy has been increasing in frequency.3–5 Yu et al3 explored the Surveillance, Epidemiology, and End Results (SEER) database of the National Cancer Institute and reported that the incidence of conjunctival melanoma increased by 5.5% biannually between 1973 and 1999, most notably in older white males. Similar findings were documented in Finland where the age-adjusted incidence of conjunctival melanoma doubled between 1973 and 1999 from 0.40 to 0.80 per million population, relative to the United States where the increase was 0.27 to 0.54 per million population.4 In Sweden, the age-standardized incidence of conjunctival melanoma between 1960 and 2005 increased substantially from 0.10 to 0.74 per million population in men and from 0.06 to 0.45 per million population in women.5 The proposed reason for these increases has been speculated to be related to sunlight exposure.3,5

A recent comprehensive cohort analysis of >5000 consecutive conjunctival tumors in an ocular oncology practice revealed that conjunctival melanoma was among the most common malignant tumors, representing approximately 12% of all cases.6 Certain clinical features help to distinguish primary acquired melanosis (melanoma precursor) from frank melanoma as melanoma more often occurs at older patient age (P < 0.0001), male sex (P < 0.0001), location in fornix (P = 0.0016) or tarsus (P = 0.0018), greater diameter (P < 0.0001), greater thickness (P < 0.0001), and with feeder vessels (P < 0.0001), intrinsic vessels (P < 0.0001), and hemorrhage (P = 0.0001).6

Conjunctival melanoma is classified according to the American Joint Committee on Cancer Classification (AJCC) into clinical and pathologic tumor (T) categories and T subcategories. In 2012, we explored outcomes of conjunctival melanoma based on the AJCC 7th edition.7 In 2017, the AJCC was updated to the 8th edition.8 Herein, we review conjunctival melanoma outcomes based on the AJCC 8th edition clinical classification in 425 consecutive patients.

MATERIALS AND METHODS

The medical records of patients with clinical and confirmatory pathologic diagnosis of conjunctival melanoma managed on the Ocular Oncology Service at Wills Eye Hospital, Philadelphia, between April 18, 1974 and September 9, 2019 were retrospectively reviewed. Patients who underwent surgery elsewhere before referral and were lacking clinical information on tumor features were excluded due to the inability to be properly staged. This study was approved by the Institutional Review Board of Wills Eye Hospital, adhered to the tenets of the Declaration of Helsinki, and complied with the Health Insurance Portability and Accountability Act. Informed consent was obtained from all patients.

All patients were examined by ocular oncologists (C.L.S., S.E.L., J.A.S.), with slit lamp biomicroscopy, documentation on large conjunctival drawings, and clinical photographs at each visit. The recorded demographic data included age (years), race (Caucasian, African American, Hispanic, Asian Indian, Asian Oriental, others/unknown), sex (male, female), and the involved eye (right, left). Smoking history was noted. Ocular history included a history of primary acquired melanosis (PAM), conjunctival nevus, conjunctival complexion-associated melanosis, melanocytosis, conjunctival secondary acquired melanosis, and a prior history of conjunctival or eyelid surgery. Each tumor was clinically staged based on the American Joint Committee on Cancer (AJCC) Staging Manual (8th edition) (eFigure 1, http://links.lww.com/APJO/A63).8

The clinical features at presentation included best-corrected visual acuity, largest tumor basal diameter (millimeters), tumor thickness (millimeters), number of quadrants involved (1–4), number of clock hours involved (0–12), tissues involved (bulbar conjunctiva, limbus, cornea, plica, caruncle, fornix, tarsal conjunctiva, eyelid, orbit), presence of feeder and intrinsic vessels, and tumor pigmentation (absent, partial, complete).

The management of conjunctival melanoma included complete surgical resection under dry conditions with superficial keratectomy for the corneal component, partial lamellar scleroconjunctivectomy (wide “no touch technique” surgical resection) for the conjunctival component, and supplemental double freeze-thaw cryotherapy to surrounding conjunctival margins.9 Other treatment options included medical treatment (cryotherapy, mitomycin C, interferon α-2b), radiotherapy (external beam radiation therapy, plaque radiotherapy), or surgical treatment (enucleation, exenteration). The types of primary, secondary, and tertiary treatments, and the total number of treatments was recorded. Primary treatment was defined as the initial treatment.

The outcomes included visual acuity loss (≥3 Snellen lines), local recurrence/new tumor after referral to our service, need for enucleation or exenteration, development of metastasis and site of metastasis (locoregional lymph nodes, systemic), and death. Local tumor control was defined as no clinical and/or histopathologic evidence of tumor at any site in the periocular region. Tumor recurrence/new tumor was defined as the presence of new melanoma at the same site of previous melanoma or at any site in the conjunctiva, adnexa, or orbit. Assessment of melanoma-related locoregional lymph nodes and systemic metastasis was provided by the family physician and general oncologist through history, physical examination, and imaging.

Statistical analysis was performed using SPSS Statistics Software (version 23; SPSS Inc., Chicago, IL). Continuous variables were expressed as mean (median, range). The 1-sample Kolmogorov-Smirnov test was used to assess for normality. Comparison between the 3 AJCC T categories (T1 vs T2 vs T3) was performed using the one-way analysis of variance test for continuous variables with normal distribution and Kruskal-Wallis H test for continuous variables without normal distribution. Comparisons for categorical variables were performed using the χ2 test and Fisher exact test when indicated. Kaplan-Meier analysis was performed to determine cumulative probability of outcomes, including vision loss, tumor recurrence/new tumor, enucleation, exenteration, locoregional lymph node and systemic metastasis, and death. The log-rank test was performed to assess differences in survival distribution between AJCC T categories. Hazard ratios and 95% confidence intervals were calculated for the AJCC T categories at each time point of interest using the Cox proportional hazards model. A P value <0.05 was considered statistically significant.

RESULTS

There were 629 consecutive patients with conjunctival melanoma managed at the Ocular Oncology Service at Wills Eye Hospital over this 45-year period, of which 425 were included in this study. The 204 patients excluded from this study were those in whom tumor staging could not be determined because of surgical management elsewhere before referral to our service. On the basis of AJCC clinical classification, there were 266 (63%) patients classified as T1, 75 (18%) as T2, 84 (20%) as T3, and 0 (0%) as T4.

The demographic features are listed in eTable 1, http://links.lww.com/APJO/A64. The mean age at presentation was 62.7 years (median 66.0, range 4.6–95.3) with 216 (51%) males and 209 (49%) females. There were 393 (93%) whites, 13 (3%) African Americans, 9 (2%) Hispanics, 1 (<1%) Asian Indian, 6 (1%) Asian Orientals, and 3 (1%) patients classified as other/unknown. There were 354 (83%) patients with history of conjunctival PAM, 69 (16%) with history of conjunctival nevus, 15 (4%) with complexion associated melanosis, and 19 (5%) with a history of melanocytosis. A comparison (T1 vs T2 vs T3) revealed that history of PAM was more common in T2 (81% vs 96% vs 81%; P = 0.01) and previous conjunctival nevus was more common in T1 (20% vs 9% vs 11%; P = 0.03). With increasing T category, history of prior conjunctival surgery was more common (60% vs 72% vs 74%; P = 0.02), mean number of conjunctival surgeries was higher (1.6 vs 2.1 vs 2.4; P = 0.03), and prior eyelid surgery was higher (2% vs 9% vs 18%; P < 0.001).

The clinical features are listed in eTable 2, http://links.lww.com/APJO/A65. Visual acuity at presentation was assessed in 422 patients, categorized as 20/20-20/40 in 361 (86%), 20/50-20/150 in 44 (10%), and 20/200 or worse in 17 (4%). The mean logMAR visual acuity at presentation was 0.2 (median 0.1, range 0.0–4.0) [Snellen equivalent 20/30 (20/25, 20/20—light perception)]. Tumor features included mean basal diameter 12.5 mm, mean thickness 2.7 mm, mean number of quadrants involved 1.9, and mean clock hours involved 4.6. The tumor location involved the bulbar conjunctiva in 408 (97%) patients, limbus in 290 (69%), cornea in 226 (54%), plica in 53 (13%), caruncle in 46 (11%), fornix in 120 (29%), tarsal conjunctiva in 112 (27%), eyelid in 60 (14%), and orbit in 15 (4%). Comparison between AJCC T categories (T1 vs T2 vs T3) revealed that increasing T category demonstrated worse mean visual acuity at initial presentation (20/30 vs 20/30 vs 20/40; P = 0.003), and increase in tumor basal diameter (10.2 vs 15.2 vs 17.2 mm; P < 0.001), thickness (2.0 vs 3.4 vs 4.3 mm; P < 0.001), number of quadrants involved (1.6 vs 1.9 vs 2.6; P < 0.001), and number of clock hours involved (3.9 vs 4.7 vs 6.6; P < 0.001). The presence of feeder vessels decreased with increasing T category (71% vs 43% vs 23%; P < 0.001).

The treatment modalities are listed in eTable 3, http://links.lww.com/APJO/A66. There were 425 patients who received primary treatment, 411 patients who received secondary treatment, and 303 patients who received tertiary treatment. Patients received a mean of 4 treatments (median 3, range 1–31). Overall, 72 (17%) patients received medical therapy, 425 (100%) received surgical therapy, and 48 (11%) received radiation therapy. A comparison (T1 vs T2 vs T3) revealed an increase in number of treatments per increase in T category (2.9 vs 4.1 vs 5.0; P < 0.001).

The treatment outcomes are listed in Table 1. There were 381 patients who had follow-up after their initial referral to our service with a mean of 57.6 months (median 33.4, range 0.03–318.5 months). A comparison (T1 vs T2 vs T3) revealed T1 with longer follow-up duration (66.9 vs 47.2 vs 37.7 months; P = 0.002). Snellen visual acuity was taken in 347 patients during their last visit. Comparison (T1 vs T2 vs T3) revealed worsening logMAR visual acuity with increasing T category [(0.4 vs 0.5 vs 0.8) (Snellen equivalent 20/50 vs 20/60 vs 20/125); P = 0.04]. There were 88 (25%) patients who lost at least 3 lines of vision. At date last seen, 344 (90%) patients had local tumor control. There was melanoma recurrence/new tumor in 136 (36%) patients with a mean of 2 recurrences/new tumors per patient. Comparison (T1 vs T2 vs T3) revealed an increase in the number of patients with recurrence/new tumor per increasing T category (30% vs 43% vs 49%; P = 0.004). Enucleation was required in 3 (1%) patients and exenteration in 33 (9%). Comparison (T1 vs T2 vs T3) revealed an increase in the number of patients requiring exenteration per increasing T category (3% vs 9% vs 28%; P < 0.001). In 72 (19%) patients, there was evidence of metastatic disease. Comparison (T1 vs T2 vs T3) revealed increasing T category with an increase in percentage of patients with melanoma-related locoregional lymph node metastasis (2% vs 7% vs 12%; P = 0.001) and percentage of patients with melanoma-related systemic metastasis (9% vs 25% vs 23%; P < 0.001). Of 19 cases with locoregional lymph node metastasis, the involved lymph node included preauricular node in 10 patients, submandibular node in 5, and cervical node in 5. Of 56 cases with systemic metastasis, the location included liver in 18, lung in 16, brain in 11, parotid gland in 15, bone in 4, and other locations of the nasopharynx, lacrimal apparatus, mediastinal lymph node, breast, skin, stomach, or small intestine in 10. Comparison (T1 vs T2 vs T3) revealed an increase in the percentage of patients who died from conjunctival melanoma per increasing T category (4% vs 12% vs 18%; P < 0.001).

TABLE 1 - Conjunctival Melanoma: Outcomes Based on the AJCC Classification (8th edition) of 425 Tumors
Outcomes T1(n = 266)n (%) T2(n = 75)n (%) T3(n = 84)n (%) T4(n = 0)n (%) P All Patients(N = 425)N (%)
Follow-up n=237 n = 68 n = 76 n = 0 N = 381
 Duration, mo, mean (median, range) 66.9 (46.9, 0.03–318.5) 47.2 (31.2, 0.1–251.1) 37.7 (19.3, 0.1–304.1) 0 (0) 0.002 57.6 (33.4, 0.03–318.5)
 Patients without follow-up visits 30 (11) 8 (11) 8 (10) 0 (0) NA 44 (10)
Snellen visual acuity at date last seen n = 227 n = 63 n = 57 n = 0 N = 347
 20/20–20/40 170 (75) 44 (70) 31 (54) 0 (0) 0.02 245 (71)
 20/50–20/150 34 (15) 10 (16) 11 (19) 0 (0) 55 (16)
 20/200 or worse 23 (10) 9 (14) 15 (26) 0 (0) 47 (14)
 Visual acuity (Snellen), mean (median, range) 20/50 (20/25, 20/20-LP) 20/60 (20/30, 20/20-HM) 20/125 (20/40, 20/20-NLP) NA 0.04 20/60 (20/30, 20/20-NLP)
 Visual acuity (LogMAR), mean, (median, range) 0.4 (0.1, 0.0–4.0) 0.5 (0.2, 0.0–3.0) 0.8 (0.3, 0.0–5.0) NA 0.5 (0.2, 0.0–5.0)
Visual loss n = 228 n = 63 n = 61 n = 0 N = 352
 Visual acuity loss of >3 lines 49 (22) 17 (27) 22 (36) 0 (0) 0.06 88 (25)
Local tumor control
 Tumor controlled at date last seen 218 (92) 62 (91) 64 (84) 0 (0) 0.13 344 (90)
Recurrence/new tumor after primary treatment
 Patients with recurrences/new tumor 70 (30) 29 (43) 37 (49) 0 (0) 0.004 136 (36)
 No. recurrences/new tumor, mean (median, range) 1.9 (1.0, 1.0–8.0) 1.5 (1.0, 1.0–4.0) 2.8 (1.0, 1.0–27.0) NA 0.32 2.1 (1.0, 1.0–27.0)
Surgical outcome
 Enucleation 3 (1) 0 (0) 0 (0) 0 (0) 0.24 3 (1)
 Exenteration 6 (3) 6 (9) 21 (28) 0 (0) <0.001 33 (9)
Metastasis n = 238 n = 69 n = 77 n = 0 N = 384
 Total no. Patients 25 (11) 21 (30) 26 (34) 0 (0) <0.001 72 (19)
Locoregional lymph nodes 4 (2) 5 (7) 10 (12) 0 (0) 0.001 19 (5)
 Preauricular 4 (100) 2 (40) 4 (40) 0 (0) NA 10 (53)
 Submandibular 0 (0) 1 (20) 5 (50) 0 (0) 5 (26)
 Cervical 0 (0) 2 (20) 3 (30) 0 (0) 5 (26)
systemic locations 21 (9) 17 (25) 18 (23) 0 (0) <0.001 56 (15)
 Liver 5 (24) 6 (35) 6 (33) 0 (0) NA 18 (32)
 Lung 6 (29) 5 (29) 4 (22) 0 (0) 16 (29)
 Brain 3 (14) 2 (12) 6 (33) 0 (0) 11 (20)
 Parotid 8 (38) 3 (18) 4 (22) 0 (0) 15 (27)
 Bone 1 (5) 1 (6) 1 (6) 0 (0) 4 (7)
 Others§ 4 (19) 3 (18) 3 (17) 0 (0) 10 (18)
Death n = 237 n = 69 n = 76 0 (0) N = 382
 Total number of patients 9 (4) 8 (12) 14 (18) 0 (0) <0.001 31 (8)
Bold values indicate statistical significance.
Follow-up data were not available in 3 patients who had metastasis as they were followed-up elsewhere.
Multiple patients had metastasis to more than 1 lymph node.
Multiple patients had metastasis to more than 1 systemic location.
§Others include metastasis to one of the following locations: nasopharynx, lacrimal apparatus, mediastinal lymph node, breast, skin, stomach, or small intestine
Follow-up data was not available in 1 patient who died after being followed-up elsewhere.AJCC indicates joint committee on cancer classification; HM, hand motion; LP, light perception; NA, not applicable; NLP, no light perception.

Kaplan-Meier results are listed in Table 2. Comparison (T1 vs T2 vs T3) at 10 years showed that there was a higher probability of visual acuity loss ≥3 lines per increasing T category (32% vs 42% vs 63%; P < 0.001) (Fig. 1A), melanoma recurrence/new tumor (47% vs 70% vs 74%; P < 0.001) (Fig. 1B), enucleation (<1% vs 0% vs 0%; P = 0.74) (Fig. 1C), exenteration (4% vs 24% vs 46%; P < 0.001) (Fig. 1D), melanoma-related locoregional lymph node metastasis (3% vs 13% vs 25%; P < 0.001) (Fig. 2A), melanoma-related systemic metastasis (13% vs 45% vs 40%; P < 0.001) (Fig. 2B), and melanoma-related death (8% vs 22% vs 37%; P < 0.001) (Fig. 2C).

TABLE 2 - Conjunctival Melanoma: Outcomes BASED on the AJCC Classification (8th edition) of 425 Tumors: Outcomes by Kaplan-Meier Analysis
Vision Loss >3 Lines Tumor Recurrence/New Tumor Enucleation Exenteration Locoregional Lymph Node Metastasis Melanoma-Related Systemic Metastasis Melanoma-Related Death
Outcomes AJCC Classification CPE (%) HR 95% CI P CPE (%) HR 95% CI P CPE (%) HR 95% CI P CPE (%) HR 95% CI P CPE (%) HR 95% CI P CPE (%) HR 95% CI P CPE (%) HR 95% CI P
6 mo T1 5 Ref 6 Ref <1 Ref 1 Ref 0 Ref 1 Ref 0 Ref
T2 7 1.44 0.45–5.60 0.54 13 2.26 0.92–5.52 0.08 0 0.02 0–>1000 0.76 3 3.48 0.49–24.70 0.21 2 >1000 0.00–>1000 0.96 2 1.15 0.12–11.10 0.90 0 NA NA NA
T3 19 4.15 1.73–9.98 <0.001 15 2.97 1.28–6.88 0.01 0 0.02 0–>1000 0.75 12 15.20 3.28–70.39 0.01 3 >1000 0.00–>1000 0.96 3 2.22 0.37–13.30 0.38 0 NA NA NA
12 mo T1 8 Ref 10 Ref <1 Ref 1 Ref <1 Ref 2 Ref <1 Ref
T2 18 2.17 0.99–4.74 0.05 25 2.64 1.35–5.15 0.01 0 0.02 0–>1000 0.76 5 3.48 0.70–17.25 0.13 2 3.42 0.21–54.74 0.38 5 2.61 0.59–11.68 0.21 2 3.36 0.21–53.71 0.39
T3 24 3.18 1.52–6.66 0.002 31 3.61 1.92–6.76 <0.001 0 0.02 0–>1000 0.75 16 12.90 3.60–46.27 <0.001 5 10.31 1.07–99.09 0.04 5 2.55 0.57–11.41 0.22 2 3.60 0.23–57.51 0.37
24 mo T1 13 Ref 17 Ref <1 Ref 2 Ref 2 Ref 4 Ref 2 Ref
T2 23 1.89 0.94–3.77 0.07 27 1.82 1.00–3.32 0.05 0 0.02 0–>1000 0.76 5 2.64 0.59–11.79 0.20 6 3.41 0.69–16.90 0.13 17 4.03 1.55–10.44 0.004 8 4.47 1.00–19.97 0.05
T3 30 2.74 1.42–5.31 0.003 49 3.58 2.14–5.98 <0.001 0 0.02 0–>1000 0.75 20 11.84 3.86–36.37 <0.001 12 7.21 1.80–28.85 0.01 17 4.08 1.58–10.59 0.004 7 3.65 0.74–18.07 0.11
36 mo T1 15 Ref 23 Ref <1 Ref 2 Ref 2 Ref 6 Ref 2 Ref
T2 26 1.84 0.95–3.57 0.07 37 1.88 1.10–3.21 0.02 0 0.02 0–>1000 0.76 5 2.67 0.60–11.95 0.20 9 4.75 1.06–21.23 0.04 20 3.38 1.44–7.97 0.01 11 5.86 1.40–24.52 0.02
T3 33 2.67 1.42–5.03 0.002 59 3.54 2.21–5.67 <0.001 0 0.02 0–>1000 0.75 28 15.15 5.06–45.40 <0.001 15 8.87 2.29–34.37 0.002 32 5.24 2.40–11.41 <0.001 20 10.22 2.71–38.54 0.001
48 mo T1 20 Ref 26 Ref <1 Ref 4 Ref 2 Ref 7 Ref 2 Ref
T2 30 1.72 0.92–3.21 0.09 50 2.10 1.28–3.44 0.003 0 0.02 0–>1000 0.76 5 1.87 0.47–7.48 0.38 9 4.75 1.06–21.23 0.04 26 3.86 1.73–8.61 0.001 11 6.06 1.45–25.38 0.01
T3 45 2.81 1.58–5.00 <0.001 65 3.55 2.26–5.58 <0.001 0 0.02 0–>1000 0.75 32 11.39 4.47–28.98 <0.001 15 8.87 2.29–34.37 0.002 35 5.30 2.51–11.23 <0.001 26 13.20 3.63–48.02 <0.001
60 mo T1 21 Ref 30 Ref <1 Ref 4 Ref 2 Ref 10 Ref 3 Ref
T2 34 1.80 0.98–3.29 0.06 57 2.15 1.33–3.46 0.002 0 0.02 0–>1000 0.76 16 3.39 1.03–11.13 0.04 13 6.55 1.56–27.53 0.01 30 3.52 1.67–7.41 0.001 11 4.69 1.26–17.49 0.02
T3 55 3.07 1.77–5.33 <0.001 65 3.38 2.16–5.27 <0.001 0 0.02 0–>1000 0.75 46 15.14 6.03–37.99 <0.001 25 12.59 3.39–46.72 <0.001 35 4.41 2.18–8.95 <0.001 26 10.20 3.19–32.61 <0.001
84 mo T1 27 Ref 37 Ref <1 Ref 4 Ref 3 Ref 13 Ref 5 Ref
T2 42 1.79 0.99–3.20 0.05 70 2.20 1.40–3.45 0.001 0 0.02 0–>1000 0.76 24 4.16 1.34–12.94 0.01 13 5.07 1.35–18.97 0.02 36 3.38 1.67–6.82 0.001 22 4.87 1.63–14.58 0.01
T3 63 3.05 1.79–5.20 <0.001 69 3.19 2.08–4.90 <0.001 0 0.02 0–>1000 0.75 46 15.53 6.18–38.99 <0.001 25 9.70 2.97–31.69 <0.001 40 4.12 2.12–8.02 <0.001 31 8.22 3.03–22.33 <0.001
120 mo T1 32 Ref 47 Ref <1 Ref 4 Ref 3 Ref 13 Ref 8 Ref
T2 42 1.70 0.95–3.03 0.07 70 2.04 1.31–3.18 0.002 0 0.02 0–>1000 0.76 24 4.16 1.34–12.94 0.01 13 5.07 1.35–18.97 0.02 45 3.67 1.84–7.32 <0.001 22 3.84 1.38–10.64 0.01
T3 63 2.92 1.72–4.95 <0.001 74 3.01 1.99–4.57 <0.001 0 0.02 0–>1000 0.75 46 15.53 6.18–38.99 <0.001 25 9.70 2.97–31.69 <0.001 40 4.18 2.14–8.13 <0.001 37 7.14 2.90–17.56 <0.001
AJCC indicates American Joint Committee on Cancer Classification; CI, confidence interval; CPE, cumulative probability of event; HR, hazard ratio; NA, not applicable; Ref, reference.Bold P values indicate statistical significance.

FIGURE 1
FIGURE 1:
Kaplan-Meier analysis showing (A) time to vision loss of ≥3 Snellen visual acuity lines, (B) time to recurrence/new tumor, (C) time to enucleation, and (D) time to exenteration.
FIGURE 2
FIGURE 2:
Kaplan-Meier analysis showing (A) time to locoregional lymph node metastasis, (B) time to melanoma-related systemic metastasis, and (C) time to melanoma-related death.

DISCUSSION

In 2012, we reported outcomes of conjunctival melanoma in 343 cases based on the AJCC 7th edition, focusing on 5-year outcomes of local tumor recurrence/new tumor, lymph node metastasis, distant metastasis, and death.7 A comparison of the distribution (AJCC 7th vs current 8th edition) disclosed T1 (57% vs 63%), T2 (32% vs 18%), T3 (11% vs 20%), and T4 (0% vs 0%). The 5-year Kaplan-Meier outcomes for the AJCC 7th edition (T1, T2, T3) included local tumor recurrence/new tumor (44%, 78%, 76%), lymph node metastasis (17%, 52%, 49%), melanoma-related metastasis (11%, 35%, 42%), and melanoma-related death (5%, 20%, 23%). Those classified as T2 or T3 (compared to T1) carried significantly greater risk for lymph node metastasis, melanoma-related metastasis, and melanoma-related death.

In this analysis we explored the AJCC 8th edition, for longer-term outcomes at 10 years and we included visual outcomes and rate of enucleation and previous outcomes. We found that the 10-year Kaplan-Meier estimates (T1 vs T2 vs T3) were significantly different for several outcomes, including visual acuity loss ≥3 lines (32% vs 42% vs 63%; P < 0.001), melanoma recurrence/new tumor (47% vs 70% vs 74%; P < 0.001), exenteration (4% vs 24% vs 46%; P < 0.001), melanoma-related locoregional lymph node metastasis (3% vs 13% vs 25%; P < 0.001), melanoma-related systemic metastasis (13% vs 45% vs 40%; P < 0.001), and melanoma-related death (8% vs 22% vs 37%; P < 0.001). Based on our results, the AJCC 8th edition for conjunctival melanoma is reliably predictive of these outcomes.

We have experienced no cases of T4 conjunctival melanoma in this series regarding the AJCC 8th edition and in our previous series7 on the AJCC 7th edition. Perhaps given the low frequency of T4 (tumor of any size with central nervous system involvement), the classification should be reconsidered and T4 could be merged with T3 as the most advanced subcategory (T3e) rather than having an entirely new category for a nearly absent subset of tumors.

In 2019, Jain et al provided a multicenter assessment of the AJCC 8th edition for conjunctival melanoma in 288 cases pooled from 10 ocular oncology centers and found, similar to our data, that there were no T4 cases.10 They found metastatic disease in 8.5% at relatively short median follow-up time of 4.3 years. They additionally noted significantly higher cumulative mortality rate of clinical T2 and T3 compared to T1 category, similar to our results. Esmaeli et al reviewed 88 cases per the 8th edition AJCC classification and indicated that greater clinical T category was associated with greater risk for metastasis and death, quite similar to our results.11

There are limitations to our analysis due to the retrospective nature of our data collection and to our referral practice. Some patients preferred follow-up close to home and returned only when recurrence/new tumor was found, which could negatively have biased our results to these adverse events. Additionally, the mean follow-up period was relatively short at 58 months, even though our total cohort dated back to 1974 with potentially 4.5 decades of care. This was likely also attributable in part to travel distance and patient preference for follow-up close to home. Some physicians kept us informed of outcomes with letters or phone correspondence, but others did not. It should be noted that patient care and data recording at our center is kept in meticulous order and this single-center analysis can provide accurate results due to the reliable data collection by trained ocular oncologists familiar with this malignancy. Furthermore, with the introduction of new immune-related medications like the checkpoint inhibitors, we might see future improvement in outcomes that can be compared to these data in which no patient received first-line checkpoint inhibitor.

In conclusion, we have provided relatively long-term outcomes of conjunctival melanoma per the AJCC 8th edition. Overall, findings revealed visual acuity loss of >3 lines (25%), tumor recurrence/new tumor (36%), exenteration (9%), lymph node metastasis (5%), melanoma-related systemic metastasis (15%), and melanoma-related death (8%). At 10 years, these outcomes were all significantly worse in those classified as T3 (vs T1) with hazard ratio of 2.92 for visual acuity loss of >3 lines (P < 0.001), 3.01 for tumor recurrence/new tumor (P < 0.001), 15.53 for exenteration (P < 0.001), 9.70 for melanoma-related locoregional lymph node metastasis (P < 0.001), 4.18 for melanoma-related systemic metastasis (P < 0.001), and 7.14 for melanoma-related death (P < 0.001).

Acknowledgements

The authors acknowledge contributions to this manuscript for data collection from Tyler M. Kaplan, BA, Ryan S. Kim, BA, Jordan L Hamburger, BS, and Shaleen R. Oza, BS.

REFERENCES

1. Shields CL, Chien JL, Surakiatchanukul T, et al. Conjunctival tumors: review of clinical features, risks, biomarkers, and outcomes. The 2017 J. Donald M. Gass Lecture. Asia Pac J Ophthalmol (Phila) 2017; 6:109–120.
2. Shields CL, Markowitz JS, Belinsky I, et al. Conjunctival melanoma: outcomes based on tumor origin in 382 consecutive cases. Ophthalmology 2011; 118:389–395.
3. Yu GP, Hu DN, McCormick S, Finger PT. Conjunctival melanoma: is it increasing in the United States? Am J Ophthalmol 2003; 135:800–806.
4. Tuomaala S, Kivelä T. Conjunctival melanoma: is it increasing in the United States? Am J Ophthalmol 2003; 136:1189–1190.
5. Triay E, Bergman L, Nilsson B, et al. Time trends in the incidence of conjunctival melanoma in Sweden. Br J Ophthalmol 2009; 93:1524–1528.
6. Shields CL, Alset AE, Boal NS, et al. Conjunctival tumors in 5002 cases. Comparative analysis of benign versus malignant counterparts. The 2016 James D. Allen Lecture. Am J Ophthalmol 2017; 173:106–133.
7. Shields CL, Kaliki S, Al-Dahmash SA, et al. American Joint Committee on Cancer (AJCC) clinical classification predicts conjunctival melanoma outcomes. Ophthalmic Plast Reconstr Surg 2012; 28:313–323.
8. Coupland SE, Barnhill R, Conway RM. Amin MB, Edge S, Greene F, et al. Conjunctival melanoma. American Joint Committee on Cancer. AJCC Cancer Staging Manual 8th ed.New York, NY: Springer; 2017. 803–811.
9. Shields JA, Shields CL, DePotter P. Surgical management of circumscribed conjunctival melanomas. Ophthalmic Plast Reconstr Surg 1998; 14:208–215.
10. Jain P, Finger PT, Damato B, et al. Multicenter, international assessment of the eighth edition of the American Joint Committee on Cancer Cancer staging manual for conjunctival melanoma. JAMA Ophthalmol 2019; 137:905–911.
11. Esmaeli B, Rubin ML, Xu S, et al. greater tumor thickness, ulceration, and positive sentinel lymph node are associated with worse prognosis in patients with conjunctival melanoma: Implications for future AJCC classifications. Am J Surg Pathol 2019; 43:1701–1710.
Keywords:

AJCC 8th edition; American Joint Committee on Cancer Classification 8th edition; Classification

Supplemental Digital Content

Copyright © 2020 Asia-Pacific Academy of Ophthalmology. Published by Wolters Kluwer Health, Inc. on behalf of the Asia-Pacific Academy of Ophthalmology.