Actinic keratosis (solar keratosis): sun damage, not cancer

An actinic keratosis is a patch of keratinocytes that have accumulated ultraviolet damage. Two coding systems agree about what it is not. In ICD-10 it is L57.0, inside the chapter on diseases of the skin and subcutaneous tissue, in the block for skin changes due to chronic exposure to nonionizing radiation; that is not a neoplasm code, and the neoplasm chapter, where skin cancer lives as C44 and carcinoma in situ of skin as D04, is a different chapter of the same book. In ICD-O, the morphology system registries use, it is 8070/0, where the final digit means benign, and the classification pointedly did not give it 8070/2, the code that exists and means squamous cell carcinoma in situ. Both the fourth and the fifth editions of the WHO skin classification file it under carcinoma precursors rather than under carcinomas. No UK cancer registry registers actinic keratoses as cancer. That is why they are a glossary entry here and not a cancer page, and it is worth being explicit about, because being told you have a pre-cancer is frightening out of proportion to what it means.

How often one becomes a cancer is a question the literature answers carefully rather than confidently. In the largest prospective study, 7,784 actinic keratoses on the faces and ears of 169 participants in a chemoprevention trial were followed, and the risk of one progressing to a primary squamous cell carcinoma, invasive or in situ, was 0.60 percent at 1 year and 2.57 percent at 4 years; 55 percent of the lesions had disappeared by the 1-year follow-up and 70 percent by 5 years (Criscione 2009). The classic Australian study of 21,905 solar keratoses put the risk of one transforming within a year at less than 1 in 1,000 (Marks 1988). A systematic review found published rates from 0 to 0.075 percent per lesion-year, up to 0.53 percent per lesion in people with a previous skin cancer, regression rates for single lesions of 15 to 63 percent after a year, and concluded that 'currently, no reliable estimates concerning the frequency of AK developing into invasive carcinoma can be given' (Werner 2013). The British guideline quotes both the under 1 in 1,000 per annum figure and a modelled probability of about 10 percent of developing a squamous cell carcinoma within ten years for a person with an average of 7.7 lesions (de Berker 2017).

The reason those two framings differ is the reason the lesions matter at all. Per lesion the risk is tiny; per person with many lesions it is not, and the same study that found a low per-lesion risk also found that about 65 percent of all the squamous cell carcinomas and 36 percent of the basal cell carcinomas diagnosed in the cohort arose where an actinic keratosis had previously been seen (Criscione 2009). So an individual patch is very unlikely to be the one, and the skin that grows them is where the cancer will come from.

What they do mean is that the skin around them is damaged too. Actinic keratoses usually come in numbers rather than singly, and the useful way to think about them is as visible marks in a wider field of sun-damaged skin, the phenomenon called field cancerisation. That is why treatment is often aimed at an area rather than at a spot, and why having many of them raises the chance of a squamous cell carcinoma somewhere in that area rather than at any particular lesion. The European consensus that adopted the term keratinocyte cancer includes actinic keratosis in the group as a keratinocyte-derived precursor, which is an accurate description of where it sits: on the same biological road as squamous cell carcinoma, a long way from the end of it.

Where actinic keratosis sits relative to the two pages beside it is worth stating plainly. Actinic keratosis is dysplasia in sun-damaged skin and is not registered as cancer. Bowen's disease is squamous cell carcinoma in situ, a carcinoma confined to the epidermis, and has its own page here. Invasive cutaneous squamous cell carcinoma has broken through into the dermis. The three are a sequence in principle and not a timetable in practice: most actinic keratoses never become anything, and most Bowen's disease never becomes invasive.