Crohn and a few other GIs in the 1920s, 1930s tried vaccinating people with anti-dysentery serum (dysenteric poly-valent vaccine) and found excellent results, even in patients they knew had never had dysentery. They found that patients who had the best responses were those who suffered from a low level systemic septic shock from the vaccine - they further tested whether it was specifically the dysentery vaccine working on the disease itself or a general mechanism invoking a systemic response, by treating people with other things that also induced a septic shock (horse serum, mercury even) and found similar results, but only when such a shock occurred.
Effectively, the Qu biologic trial is a similar in some ways, but they are using a killed microbe (not serum from patients), specific to the gut ( e coli), but I suspect the mechanism is the same. The only difference is dose (I worked out that one doctor who put a patient in remission (who had UC), was injecting almost 300 mL of this stuff every 2nd day!!!), Qu does subcut injections of micromolar concentrations of e coli antigen.
I posited in both cases it is causing a reset of the upper cytokine/chemokine network that governs pro/anti inflammation (that happens after septic shock - dense systemic immunosuppression is how septicaemia kills by systemic organ failure in some cases).
Personally I think we can broadly push Crohn's into a disease with two over-arching mechanisms (general) - one in which mutations in a gene cause the anti-inflammatory cytokines/chemokines (Il-10 etc), to go down, which releases inhibition of pro-inflammation (so when a trigger comes on, be it a bug, a drug, a food antigen, etc), we get the slow chronic inflammation + flare ups. And on the other hand, we have mutations/dysregulations in genes that generally favour and drive inflammation (be it Il-6 etc). It's quite interesting to note that under this idea, where crohn's is mechanistically two entities, one expects that direct inhibition of a master regulator of inflammation would only generally work in 50 % of patients (if the hypothesis were true) ; which is pretty much the standard response rate (at least in Australia) for the anti-TNF Humira.
One of the biggies of septic shock (or serum shock) is TGF-B. I suspect the Qu trial, the Crohn's dysentery experiments etc, and possibly this study, induces that low level sepsis that, over time, "resets" this network of cytokines/chemokines into a more wild-type state, but bringing TGF-B and it's associate immuno-suppresors up, and aids in driving down the pro-inflammatory sneaky buggers.
I could write and write and write, but I already did 6000 words and pumped it off to a very unknown journal. When the reviews and edits come back I'll post the link, the literature points very sternly towards this idea, once you step back a bit and try to look at the whole picture.