Cancer survival

We underwriters tend not to look back too much at how the outlook for the various medical conditions we deal with has changed over the years. This is often because papers reviewing how progress in diagnosis and treatment over time have affected prognosis are few and far between. The exceptions, perhaps, are coronary heart disease (CHD) and cancer, the two big killers in developed societies.

Statistics that are meaningful for risk evaluation are sometimes difficult to find. Often one can determine that population mortality overall from a particular cause of has reduced but the impact on case mortality or survival remains unclear. But here are some comparative statistics for cancer showing relative ten-year survival rates of cases diagnosed in 1971-2, 1980-1, 1990-1 and 2007 – the ten-year figures for the last group have been projected based on experience to date, if you were wondering. They were compiled by Cancer Research UK based on data for England and Wales, but should be reasonably indicative of progress – or the lack of it – in prosperous countries generally.

To clarify the definition of ‘relative survival’, this is the survival of patients after taking into account the background mortality that the group would have experienced had they not had cancer. Thus the figures show pretty much the impact of cancer alone.

To these figures applies the health warning that is standard for all statistics. Nevertheless, they provide a reasonable indication of how the prognoses have altered over a period of about 35 years – and we are now seven years on from 2007, the last reference point for diagnosis in the analysis.

As some of the comments in the table suggest, not all these improvements are down to better treatment. Greater health-awareness has surely played a part. Screening programs (participation in which would have been boosted by growing health-awareness) have undoubtedly helped here and there, enabling earlier diagnosis and a better chance of cure, but one can be pretty certain that any improvement in the prognosis of aggressive or non-localized prostatic malignancy is less than implied by the enormous leap in survival shown for diagnoses in 2007 compared with 1990-1. Other than where commented, improvements in survival have been steady, implying that more successful treatment has played a significant role.

But despite the optimism implied by the figures for most of the tumour sites, there is a group that appears to remain stubbornly resistant to improvement. Survival rates for myeloma, and cancers of the stomach, oesophagus, brain, lung and pancreas remain grouped at the bottom of the survival league table and have shown little improvement over three-and-a-half decades owing to difficulty in treatment or late diagnosis (by which time the disease is non-localized), or both.

Of course, figures such as these do not enable the derivation of mortality rates – or even survival rates – for groups of individuals who were diagnosed with a particular stage of a particular cancer from a point at which they remained disease-free after treatment. (That is the Holy Grail for those researching evidence-based ratings for cancer.) Equally they do not reveal the effect of secondary factors, and of course they are subject to confounding influences.

But they do suggest a very encouraging picture, to which must be added the various significant strides forward made during the 21st century, such as those relating to genetics, stem cells and ‘silver bullets’. They also show that, as medicine never stands still, underwriting philosophy must move with the times. Important to ensure that cancer risk really is worth x per mil for y years.