Who’s the faster runner: the cheetah or the rabbit?

Abstract

A malignant tumour is usually defined as a tumour which gets progressively worse and ultimately results in death. In this issue of Acta Ophthalmologica, Kujala et al. (2009) argue that, based on the difference in size (and rate of local tumour recurrence), conjunctival melanoma may be more deadly than uveal melanoma. Because we know that, ultimately, nearly half of patients with uveal melanoma (Bergman et al. 2003) but ‘only’ one-third or less of patients with conjunctival melanoma (Seregard & Kock 1992) die from widespread disease, this claim is thought-provoking and controversial. Conjunctival melanoma is on average much smaller than uveal melanoma at the time of presentation. After statistically adjusting for the difference in size (so that we can assume the two types of tumour are of equivalent size), the results of this study suggest that tumour-related mortality is higher for patients with conjunctival melanoma than for those with uveal melanoma. However, a number of other tumour-related parameters have not been taken into the equation. These include mitotic rate, location, genetic abnormalities and many others. Notably, monosomy 3 and gene expression profiling appear to predict outcome better than tumour diameter in uveal melanoma (Prescher et al. 1996). In addition, the microenvironments of the conjunctiva and uvea are far from identical. Although it is largely unclear how this affects the clinical course of disease in patients with these tumours, we know that the microenvironment is important in other cancers (Albini & Sporn 2007). We do not know what the results of this study would be were all the factors known to influence survival adjusted for. However, such information would still leave us with the need to identify a number of other factors of potential importance. Both conjunctival and uveal melanoma arise from neural crest-derived melanocytes. Whereas conjunctival melanoma originates from melanocytes located in the basal layers of the epithelium, uveal melanoma arises from the stromal melanocytes of the choroid, ciliary body or iris. Furthermore, the clinical characteristics of melanoma arising from the conjunctival sac are distinctly different from those of melanoma of the uveal tract. Notably, conjunctival melanoma typically spreads via local lymphatics and then causes seeding to the regional lymph nodes; uveal melanoma spreads haematogenously and preferentially to the liver. Indeed, conjunctival melanoma seems to have much more in common with (other) mucous membrane melanoma and (probably) with cutaneous melanoma than with uveal melanoma. Arguably, comparing two distinctly separate clinical entities such as conjunctival and uveal melanoma and then statistically adjusting for only a few of their features is somewhat like comparing the cheetah − known to be the fastest runner in the world − with the rabbit. If we were to adjust for the difference in size, the rabbit would actually appear to outrun the cheetah. Mind you, we are speaking statistically. In real life not many cheetahs, not even cubs, come in the size of a rabbit. At the end of the day, all this can be boiled down to semantics. Which type of cancer is the deadliest? For the individual patient this may actually be of little interest as the variation in clinical behaviour in both these types of cancer is significant. Surprisingly, a few patients even with quite large conjunctival melanoma do very well, probably because the tumour can be completely excised before tumour cells have seeded. Similarly, some patients with uveal melanoma filling the entire globe mysteriously fail to contract metastatic disease. We clearly need to find better ways of identifying high-risk patients and better ways of managing them.