You say that a mistake on the website does not invalidate the study. Just to be clear, do you believe the WPI has mistakenly posted that no samples from the original 101 patient cohort had lymphoma, when in reality "WPI-1125" did have lymphoma (therefore, the info on their website is wrong)?
As for what I believe, any time I see a conflict, I experience doubt. All I can do after that is weigh things. In my opinion, the bigger weight is in favor of this being a web site error. That doesn't mean I'm right. But that is where the weight is coming down for me.
The only other point that I think is worth mentioning is the inclusion of a patient with lymphoma. I disagree for a couple reasons; I think the sample size is too small to be throwing "curve balls" in there