I found a piece of research on breast cancer diagnosis...

I found a piece of research on breast cancer diagnosis seasonality; looks like a reputable journal https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2741038/  

On a graph the seasonality looks like ...



Note that I have added a lag of a month from diagnosis to treatment so that the trough in breast cancer diagnoses in December impacts treatment in January (THIS IS MY ASSUMPTION). This analysis means that the flat unit sales for the March Qtr for the US is a combination of downwards seasonality of 15% and a unit sales increase of approximately the same amount.

In Australia I suspect the seasonality is significantly greater; on my numbers the seasonality means a 25% decrease in sales and therefore a unit sales growth rate of a similar amount. It is worth looking at the unit sales numbers in Australia for the 2016-17 years; 84, 107, 83, 121 for Sept 16, Dec 16, March 17 and June 17 Qtrs respectively. They generally support this hypothesis.

I present this for the sake of the sanity of those AXP retail shareholders unfortunate enough to have discovered the company in mid- 2016 before US FDA approval (like me)!