Pubmed: A Love Letter

Dear Pubmed,
I love you so much.   The "Results by year" bar graph on the right is really really cool.   When I make any queries, you always check how many papers you got.   If you have enough data, you will make a bar graph to show me how many papers on published in the most recent decades.   This way, I get a quick gauge on the research interests of thousands of breast cancer researchers, and how they wax and wane.

http://www.ncbi.nlm.nih.gov/pubmed?term=Insulin%20breast%20cancer
http://www.ncbi.nlm.nih.gov/pubmed?term=triple%20negative%20breast%20cancer

For example, someone tells me to take DIM for breast cancer.   So I asked you:
http://www.ncbi.nlm.nih.gov/pubmed?term=DIM%20breast%20cancer

You said: "83 papers total" over the years, 3-4 from 2012, many are talking about other things than DIM.
I said: OK.  I will pass.

Then I asked you:  should I consider metformin for breast cancer?
http://www.ncbi.nlm.nih.gov/pubmed?term=metformin%20breast%20cancer
You said: "160 papers total", with 20+ papers from 2012 alone.
So I figure that metformin must be interesting.

Then I asked you:  some patients swore by "oligometastatic breast cancer" and clobbered me for suggesting maybe it's rather uncommon:
http://www.ncbi.nlm.nih.gov/pubmed?term=oligometastatic%20breast%20cancer
You said 32 papers over the years, none from 2012
So I say, hmmm.   This doesn't seem a very active field, maybe it reflect the difficulty of clinical trials in this area, maybe it reflects the rarity of this subtype.    

Then I asked:  is HGF an interesting area for understanding?
http://www.ncbi.nlm.nih.gov/pubmed?term=HGF%20breast%20cancer

Guess what pubmed says?   Anyway, I suspect that one day you will save my life.    I just wish clinicaltrial.gov is as cool as you.