Hi Aly,
Firstly embedded organoids are salvageable from this so definitely not the end of the world!
We found this seems to be because of batch to batch variation in protein content in matrigel (and sometimes how well other gel factors polymerise). There are a couple of solutions:
(1) spiking the gel with low level fibrin helps improve stability (but low ammounts as too much makes for an overly stiff gel that doesn't induce sprouting).
(2) switching to the granular method in the comBO paper has proven to solve this (so far... touch wood).
Hope this helps! Thanks for tooting,
Abs