Skin cancer prevention 1

Given that skin cancer is the commonest human cancer in many populations, that we know what the cause is, and that most skin cancers are readily visible, you might think that the possibilities for prevention are large. As we will see, whilst this a reasonable expectation, in practice things are are not so straightforward.

Prevention can usefully be demarcated into primary and secondary prevention. Primary, in this context would mean stopping skin cancers from developing, and secondary prevention would refer to early detection or the minimisation of any harm from early dysplastic lesions. This chapter deals with primary prevention: secondary prevention is discussed in the next chapter.

I have explained that the two principal causes and risk factors for most types of skin cancer are

• a relative lack of melanin pigmentation, and
• exposure to ultraviolet radiation.

Human pigmentation is almost entirely genetically determined, and we can do little about it. True, melanin pigmentation can be increased by tanning in response to UVR, but as was explained in an earlier chapter, such changes are of modest effect in terms of protecting against further UVR damage and to acquire a tan means acquiring DNA damage. If we are going to reduce skin cancer our focus has to be on minimising exposure to UVR.

Conceptually, we can think about human exposure to UVR under two headings. First, there is the amount of ambient UVR — this is, if you like, the ‘fact of life’ physics bit: there is more UVR in Australia than the UK; there is more UVR in summer than winter etc. The second, relates to human behaviour. Why do some people sunbathe, whereas others do not. Why have many humans changed their view of sunshine over the last century. Whilst the ‘physics’ and human behaviour seem very different, in practice, the interactions between these two domains are such that it makes sense to talk about them together.

In chapter 4 I discussed the following key figure (below). If you cannot remember what an action spectrum is, or the difference between an action spectrum and  spectral power distribution, check out the video ‘Physics Matters’ below this figure (this is the same video that was in Chapter 4). 

Human beings cannot sense UVR exposure in real time, whereas we can sense heat and light. As we will see, this often means we have some cognitive biases that lead to erroneous beliefs about UVR exposure.

Furthermore, we cannot distinguish (at any time) between exposure to UVR of different wavelengths. As shown in the graph above and discussed earlier,  UVB is mainly responsible for UVR induced burning and we know that the rates of cancer mimic the erythema action spectrum. This central role of UVB does not negate any role for UVA (especially in melanoma). It is just that the carcinogenic role of UVA has been harder to pin down, and is still open to different interpretations of what data there is.

As the following two graphs make clear UVA and erythemal UVR (mainly UVB) vary by time of day and throughout the year. The first graph (below) shows the amounts of UVR during the day. UVA is shown on the ‘Y’ axis on the left; and ‘erythemal’ UV (which is mainly UVB) on the right. Peak exposure is given as solar noon. Beware that ‘solar noon’ in the UK in summer is 1.00 pm, and in much of mainland Europe, 2.00pm. (based on data provided by Prof Brian Diffey).

The graph below, again showing UVA and erythemal UVR (‘UVB’) across the year for Northern England.  The large variation from day to day in the second graph, reflects variation in cloud cover, and will be greater in areas far from the equator. As a rough estimate, clouds reduce UVR by about 30% over the whole year. (based on data provided by Prof Brian Diffey). 

Clouds reduce UVR due to scattering, but also reduce — as anybody who lives in Edinburgh knows — infrared radiation (‘heat’). However, clouds attenuate infrared more than UVR, therefore temperature is not a perfect proxy for UVR. In practice this means people may underestimate their chances of burning if the weather is cloudy.

The effect of heat is important in another way. Increases in temperature tend to make most of us wear fewer clothes, and spend more time al fresco. Just look at the Meadows on a sunny day in summer compared with a sunny day in winter. Or compare the seating outside coffee bars in winter versus summer.

The graph of UVR throughout the day shows that irradiance is greater at midday. Close to the equator, UVR is compressed into a shorter day. Avoidance of the sun in the middle of the day proportionately has a greater effect in areas close to the equator than at higher latitudes.

Note again, that sunlight and UVR are not perfectly correlated. The winter sunlight on a clear day even in Edinburgh in the early morning may be intense, but there is little UVB or UVA present.

As the sun lowers in the sky, radiation from the sun is distributed over a larger area. In addition, the longer length of atmosphere the UVR has to pass through reduces UVR of all wavelengths but especially those in the UVB range. In the UK, the maximum solar altitude is around 60° degrees in summer: in winter, the maximum is about 16° degrees (!).

For the months with the highest UVR, ambient UVR at midday, assuming a clear sky, is close to three times higher at the equator than at higher latitudes (such as Edinburgh) . However, the cumulative UVR through out each day only varies for the same regions by a factor of two, because the day is ‘shorter’ at the equator. Differences in cloud cover will magnify these differences. Winter to summer variation in ambient UVR varies by a factor of about 20 for Edinburgh. For subtropical ares the figure is only 3.

One way to think about this is to realise that on a summer’s day the amount of UVR may vary by 3 between high and low latitudes, but that some low latitudes’ winters are effectively summer like.

Longer wavelengths (UVA) pass through glass but UVB does not. In normal individuals, if you go red in a parked car on a summer’s day, it is likely due to heat induced vasodilation, not sunburn. UVR does pass through water so those who enjoy snorkelling may burn on their back, unless they take precautions. Water also reflects UVR, but the percentage reflected is small (5%) although this will increase if the water is choppy 20% (i.e. waves).

By contrast snow can reflect up to 90% of UVR. So, whereas sunburn in December in skiers in the Alps on a cloudy day is unlikely, the same cannot be said in Spring. Most other surfaces reflect UVR poorly, although sand can reflect 15-30%. Remember reflection means that even in the shade you are still exposed to UVR.

The above makes clear that variation in UVR reflects a number of factors, and is quite complicated. It is easy to fall into stereotypes such as all Australians have high sun exposure whereas people in Scotland do not. Reality is messier, and behaviour plays a large part in this variation.

Within a population daily human exposure may vary 1000 fold throughout the year. This relates to seasonal variation in ambient UVR, differences in hobbies, shade seeking and holiday behaviour. The figure below shows average figures for the contribution of different activities throughout the year for an average UK resident. Person to person variation would need to be added to this. As a caricature, the Australian teenager who spends all his spare time gaming on a PC may get less exposure that the rugby mad kid from Llanelli who spends most of his time emulating Barry John, by perfecting his sidestep and drop goal.

Exposure to UVR as a percentage of annual total exposure. I have not normalised for the length of each activity (the summer holiday accounts for 30% annual exposure, but might only occupy 4% of the year). (Redrawn from data provided by Professor Brian Diffey).

Short of living underground there are a number of strategies that can be employed to reduce exposure to UVR even when spending time in sunny climates.

• In latitudes like the UK, sun protection may make sense between April to September. For the tropics protection will be needed all year round.
• Avoiding exposure during the middle three hours of the day dramatically reduces (>50%) exposure — more so in the tropics.
• Shade from trees or human structures such as umbrellas can reduce exposure ten-fold depending on the design
• Most summer clothing can be used to afford UVR protection of greater than 50-fold. Hats with large brims can protect 5-fold.

The role of sunscreens and the meaning of the sun protection factor (SPF) is discussed in the following section

It is important you understand the principles behind the concept of sun protection factor (SPF). This means resisting advertising, and thinking logically.

The SPF is a measure of the ability of a substance or behaviour to reduce skin erythema from sunlight.

UVR induced erythema is chosen as an endpoint, because of its proven relevance to skin cancer. If you have been paying attention, you will know that it is UVB that accounts for most erythema (80%) from natural sunlight. Therefore, the SPF is almost (but not entirely) a measure of the ability to block UVB. What about UVA? UVA makes a smaller contribution to UVR induced erythema (20%),and this effect is taken into account in the SPF measure. There are other measures of ‘pure’ UVA protection (an ordinal star system for UVA protection is employed within the EU) but I will say little more on this topic. Let’s return to the SPF.

Imagine we have a source of sunlight (a UV lamp with the same wavelength distribution as sunlight), and a machine that measures erythema. We irradiate two patches of skin on a volunteer’s back. On one of these patches we have applied a sunscreen. If we needed to irradiate with twice the dose of UVR to get the same degree of erythema on both patches, we would say the sunscreen has a SPF of 2. A SPF of 2 means that 50% of the photons have been ‘blocked’. So, the SPF number is the reciprocal, expressed as a percentage, of the number of photons transmitted. If the SPF is 4, then only 1/4 (25%) photons get through. If only 25% of photons get through, then 75% are blocked. If the SPF is 8, then only 12.5% of photons get through to the skin; and 87.5% are blocked.

The two figures below show the relation between SPF and the percentage of photons that pass through, or alternatively the percentage of photons that are blocked. Why do you think I have plotted them for you?

The graph above shows how the amount of UVR transmitted to the skin varies with the SPF.

The graph below shows the same data plotted differently: here, the amount of UVR blocked is plotted agains the SPF.

The reason for showing both these graphs is to remind you what functions based on reciprocals look like.

The practical consequence is that you can see that the higher the SPF the smaller the absolute effect is on the number of photos transmitted or blocked (screened). The difference between a SPF of 2 and 4, is greater in absolute terms than between 4 and 8. So, why the obsession with larger numbers?

In reality, the standardised testing conditions used to calculate SPFs, do not mirror human usage. The amount of sunscreen per unit area is critical (the more there is, the more UVR that is blocked). When real people use sunscreens, on average, they apply them at around a 1/3 of the quantity per unit area that sunscreens are tested under. As a rough rule of thumb, this means a sunscreen with a SPF of 30, is applied such as to provide a SPF of 10. Sunscreens do not last forever! They need to be reapplied as they get wiped away, especially after swimming.

If you use two sun protection strategies, say a broad brimmed hat with a SPF of 4, and a sunscreen cream with an effective SPF of 10, the total SPF is the product of the two numbers = 4 times 10 = 40 (for this example I assume the sunscreen is applied as it was in testing)

Second, fake tans are not sun-screens, rather they are body paint. Just because a fake tan contains sunscreen with a SPF of X, does not mean is provides a SPF of X (the sunscreen has been diluted with the fake tan).

The only proven medical benefit of UVR is the synthesis of vitamin D. Vitamin D can be obtained from the diet, although the diet of many is deficient in vitamin D, in which case supplementation might be considered. There is some observational data relating vitamin D levels or proxy measures of sun exposure to systemic malignancies or cardiovascular health (more UVR exposure seemed beneficial). Many believe these conclusions are mistaken,  reflecting confounding, and note the absence of any convincing causal mechanism. By contrast the role of UVR in causing skin cancer is clear cut and supported by a variety of mechanistic studies as well as human observational studies.

I discuss this topic a little more in the short video (8 min) below with my doppelgänger, Ieuan Llywelyn.

Skincancer909 by Jonathan Rees is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. Where different rights apply for any figures, this is indicated  in the text.