Serotonin is certainly related to mood, but I think the point the doctor is making is that is not as simple as previously thought. That is, the layman's way of explaining MDMA-induced depression/anxiety is just 'your serotonin is low, which modulates your mood, so you feel sad.' However, this theory really isn't that supported (hence why ADs are controversial). Some of the more contemporary theories suggest that actually it is a downstream effect resulting from numerous chemical imbalances (one being serotonin), causing the problems. One of my preferred theories is the Neurotrophic Growth Factor (NGF) theory (google it if you're interested). These theories are more suitable for explaining why ADs appear to only work after weeks of administration, despite causing peak serotonin levels after only hours of consumption. Surely if it was just serotonin, then increasing it would instantly relieve symptoms, which of course is not the case. These theories are also more pertinent for explaining why MDMA-induced depression doesn't occur until days after the MDMA.
Loads of research also fails to really consider the cognitive side of depression. If you are certain you are going to have a panic attack after MDMA, then there is a good chance you will influence it to happen while in a serotonin-deprived state. I'm kind of a believer that emotions/behaviours are a result of both your brains chemical balance/functional connectivity (the brain-behaviour theory) and how you cognitively deal with the chemical balance of the brain (the behaviour-brain theory). They both influence each other, instead of just being a one-direction sort of thing. I mean, if we were to assume that everything we do is entirely dominated by chemical balances, then CBT wouldn't work. The two are able to interact. CBT could help you to reduce your stress response to situations you previously found anxiety-provoking. You could probably trace this all the way the back to CBT altering the biological basis of that situational stress response (Corticotrophin theory etc).
Just another consideration, if 5 people out of 1000 did experience mental issues from MDMA then that would probably be statistically insignificant to a researcher. Considering researchers can only really generate conclusions based upon their data, they would conclude that MDMA does not cause problems. Whether you subjectively view that as significant is a different issue. But should you really attribute a panic attack to MDMA if only 5 in 1000 people actually report it? While this is purely speculation, I'd imagine that about 5 in 1000 people (if not more) probably report regular panic attacks without even taking MDMA. A small amount of people within a sample is not exactly compelling evidence, especially considering how bad the Ecstasy market actually is. If you wanted to be even more pro-MDMA, you could probably argue that the small selection of the sample that do report problems are a result of drug impurities and not even MDMA itself. But, i'm not convinced by that. High doses of MDMA regularly could cause depression, anxiety and panic attacks (in my opinion). However, moderate doses or one-off usage - i'm not so convinced.