The 100 NCI samples

Investigators at NCI received 100 samples from individuals without knowing their health status; furthermore, the samples were sent to NCI directly without passing through the WPI laboratory space. … NCI used plasma from all 100 samples they received in infection experiments with LNCaP cells.

This allows for different scenarios for the interpretation of the Ruscetti/NCI results:

1. Face Value: The above sentence is true and the results from NCI show the actual immune status of patients and controls.
2. It was the Phlebotomist: The above sentence is true, but the phlebotomists manipulated the samples.
3. It came from WPI space: The above sentence is not true and the samples came from WPI space.
4. Broken Code: The above sentence is not true and someone at NCI knew (or guessed) the code.
5. Changed Code: The above sentence is true, but the code was changed at WPI after the result came in from NCI.
6. Contaminated Phlebotomist Vials: The above sentence is true, but the way samples were collected allowed for contamination specifically of some of the patient samples.

(I think I cane safely rule out the "random contamination at NCI" scenario)

I think #2 (Phlebotomist) is unlikely. An "open" conspiracy like that would simply involve too many people.

I don't think #3 (WPI space) happened. Someone at NCI should have noticed that. But then again, it doesn't look like Ruscetti reads what is published in his name.

#4 (broken code) is interesting. If the samples arriving at NCI had numbers like 1103, 1104, 1105, etc. and then 2760, 2832, 2958, etc. a person at NCI who might have been under pressure to come up with positive results might have been able to successfully guess which sample is a patient sample and which's a control sample. Plus, if all control samples come from one place (like Maryland, maybe even in one shipment), while the patient samples came from different places, this would have been an easy task.

#5 (changed code) is possible as well, if one looks at how numbers coming from WPI seem to be in a state of flux.

#6 (contamination at the phlebotomist) is interesting as well. The patients were in Nevada, New York state and so on, while the controls were from one state (AFAIK Maryland). So if the phlebotomist for the control samples used clean vials, while one or two of the patient phlebotomists used contaminated vials… If this is the case, then this is one point were patients and controls were not treated the same, BTW.

So what about #1, that the reported results represent reality, or something close to it? After that many inconsistencies (and while I can not completely rule out that the NCI results represent reality) I can nevertheless safely say:

Falsus in uno, falsus in omnibus.

(And I think if one looks more closely, one could find more inconsistencies between Lombardi et al. 2009 and its addendum.)