Replies to post #773391 on NorthWest Biotherapeutics Inc (NWBO)

[dennisdave]

They absolutely did not use patient level data

You are right my mistake. That was a copy-and-paste error from another subject I made from posting all day. I apologize, good you called that out.

I rephrase what I was trying to say:
Northwest Biotherapeutics used an External Control Arm (ECA) made from pooled, aggregate-level data of contemporaneous, peer-reviewed randomized controlled trials (RCTs) in glioblastoma. These were not patient-level datasets and not from real-world healthcare settings like EHRs or registries

Despite the above, MHRA’s current draft guidance (May–July 2025) includes this specific line:

“Many of the general principles would be relevant for external controls drawn from other sources, such as previously completed clinical trials.”

This means:
Even if the control arm isn’t derived from RWD (e.g. it's from past RCTs), MHRA explicitly chooses to apply similar standards to such ECAs.
So while NWBO’s ECA is not RWD, it still falls within the regulatory scope outlined in the draft guidance.

Critics (such as ATL) arguing “DCVax-L doesn’t qualify because it didn’t use RWD” are technically right on data source, but wrong in implication. MHRA is saying: “It doesn’t matter whether the control data is from real-world sources or prior RCTs the same bias mitigation and statistical principles apply.”

[Lykiri]

Ex,

which type of external control group do you think the MHRA will consider more reliable: one assembled before unblinding without patient-level data, or one assembled after unblinding with patient-level data?

[RRP60]

Ex, i always read your replies. Why do you only see dcvaxl as an ineffective drug. Over the years this drug showed positive signals in studies and now its cementing that notion in the data coming out in newer studies. But you dont want to acknowledge it. I see you as a person with good understanding on nwbo and dcvax. But the animosity you show is beyond comprehension. Looks like one of us is biased. Hopefully we will know the answer in the very near future. to your defense, everything gets delayed with nwbo. My conviction is none of these delays are from things you observe. 

[Doc logic]

exwannabe,

I was waiting for you to “clean up” this claim about patient level data. The claim that MAIC on its own may not be the best method has some merit but when used properly along with other checks and balances as NWBO’s enlisted folks putting together the ECAs did, the composite does look to stand up very well and this is what is being considered; ). Best wishes.