What do cancer survival rates mean if youre told 5 years, for example?

Clinical trials for. Many cancers use the 5 yr survival for a benchmark. As with most statistics from trials, they report on a group rather than individuals. 5 yr is no t a good endpoint for prostate: it needs 10 & 15 year data, but 5 yrs is often used with lung, breat and colorectal. These cancers comprise 80% of cancer.
Cancer-Dependent. It is very common to report survival statistics based on 5 and 10 yr anniversaries after cancer diagnosis. These are very meaningful but somewhat arbitrary end-points since some cancers have a tendency to recur very early after diagnosis while others have the potential to recur even 20-30 years later. The line between remission and cure can sometimes be a bit fuzzy.
No crystal ball. Two very different questions are: 1) What are my chances of being cured? 2) What is my life expectancy? For #1, we often give the percentage of patients who were alive at some fixed time point (e.g. 60% after 5 years). For #2, we may give the median number of years lived by previous patients. A median is the 50th percentile. It means that half of patients lived longer and other half shorter.