Trends in Uveal Melanoma Presentation and Survival During Five Decades: A Nationwide Survey of 3898 Swedish Patients

Abstract

Background: In contrast to most other cancers, uveal melanoma (UM) is characterized by an absence of major improvements in patient survival during the last several decades. In this study, we examine changes in incidence rates, patient age and tumor size at diagnosis, treatment practices and survival for patients diagnosed in Sweden during the period 1960-2010.

Methods: All patients diagnosed with posterior UM between January 1st, 1960, and December 31st, 2009, in Sweden, were included (n = 3898). Trends in incidence, primary treatment modality, patient age and tumor size were analyzed. Disease-specific survival was plotted in Kaplan-Meier curves and the cumulative incidence of UM-related mortality was evaluated in competing risk analysis.

Results: Crude (6.5-11.6 cases/million/year) and age-standardized incidence rates (5.6-9.6 cases/million/year) varied between individual years during the study period, but both had a stable linear trend overall (p ≥ 0.12). Gradually, plaque brachytherapy with ruthenium-106 replaced enucleation as the most common primary treatment. The mean patient age at diagnosis increased from 59.8 years in 1960 to 66.0 in 2009. Conversely, the mean tumor size became gradually smaller during the period. In linear regression, the basal diameter and tumor apical thickness decreased with a slope coefficient of -0.03 mm (p = 0.012) and -0.05 mm (p = 1.2 × 10^-5) per year after 1960, respectively. Patients diagnosed after 1990 had significantly better disease-specific survival than patients diagnosed before 1990 (p = 2.0 × 10^-17). Similarly, the cumulative incidence of UM-related mortality was highest for patients diagnosed 1960-1969 and 1970-1979, with slightly lower incidences for patients diagnosed 1980-1989 and even lower for those diagnosed after 1990 (p = 7.1 × 10^-13). The incidence of mortality from other causes than UM did not differ between periods (p = 0.16).

Conclusion: In the period from 1960-2010, crude and age-standardized incidence rates of UM have remained stable in Sweden. Several other aspects have changed: Plaque brachytherapy with ruthenium-106 has replaced enucleation as the most common primary treatment modality; patients have become older and their tumors smaller at the time of diagnosis; and their survival has improved. This might indicate a beneficial survival effect of earlier diagnosis and treatment, but the potential influence from lead-time bias should be taken into consideration.

Keywords: Sweden; cancer; melanoma; ophthalmology (MeSH); survival; time-trend; treatment; uveal melanoma.

FIGURE 1

Primary treatment modality for uveal melanoma in the period 1960–2010. (A) Number of treatments with the respective modality. Plaque brachytherapy with ruthenium-106 was introduced in 1979 and plaque brachytherapy with iodine-125 was introduced in 1999. (B) Proportion of all tumors treated with the respective modality.

FIGURE 2

Mean patient age and tumor dimensions at diagnosis in the period 1960–2010. (A) Uveal melanoma patients were gradually older at diagnosis during the period (red). The linear trend for their mean age (dashed red) increased with 0.1 years per calendar year, which was similar to the linear trend for the mean age of the Swedish population (green). (B) Reversely, the mean largest basal tumor diameter (blue) and mean tumor apical thickness (yellow) were gradually smaller at diagnosis, with linear trends (dashed) indicating a decline of 0.03 and 0.05 mm per calendar year, respectively.

FIGURE 3

Incidence rates in the period 1960–2010. The crude (blue) and age-standardized (green) incidence rates varied between 6.5 to 11.6 cases/million/year, and 5.6–9.6 cases/million/year, respectively. In linear regression (dashed), neither crude nor age-standardized incidence rates changed significantly over time.

FIGURE 4

Kaplan–Meier disease-specific survival in the period 1960–2010. (A) Patients diagnosed in different decennia had dissimilar disease-specific survival. (B) Patients diagnosed after 1990 had significantly better disease-specific survival than patients diagnosed before 1990. Colored areas represent 95% confidence intervals.

FIGURE 5

Cumulative incidence of uveal melanoma-related mortality and mortality from other causes in the period 1960–2010. The uveal melanoma-related mortality was highest for patients diagnosed 1960–1969 and 1970–1979, with slightly lower incidences 1980–1989 and even lower after 1990. The incidence of mortality from other causes than uveal melanoma (dashed) did not differ between the periods.