CAA - WPI only lab reporting false positive result. (Is this correct?)

Messages
1,471
Likes
5
Location
UK
http://www.cfids.org/xmrv/default.asp

The CAA reported the following on their XMRV page:
The U.S. Department of Health and Human Services announced in October 2009 that it would support studies to assess the potential risks to the blood supply and that will also help standardize tests for XMRV. At the July 26, 2010 meeting of the Food and Drug Administration's (FDA) Blood Products Advisory Committee, seven groups made informational presentations about XMRV, including researchers at the FDA, NIH and CDC. The Department of Health and Human Services Blood XMRV Scientific Research Working Group, of which Association scientific director Suzanne D. Vernon, PhD is a member, reported results from Phase I of its four-phase study. Analytical panels of blinded samples of XMRV and negative controls were tested by six laboratories (including the Whittemore Peterson Institute) to assess results using different methods. All six laboratories were able to detect XMRV in whole blood using nucleic acid testing and four of five plasma RNA assays performed well. CDC's whole blood assay was the most sensitive, while WPI was the only lab reporting an unexplained false positive result on a negative sample. (Presentations made at the July 26 meeting will be posted on the FDA's website soon.) Further work on the analytical panels will continue as the other three phases of the study are completed. Results of Phase II testing of blinded clinical panel samples are expected by fall 2010.
A number of question have been raised by various people about this statement. Is it correct to say that the WPI was the only lab. Firstly, the chart next to this statement shows the FDA to also have found XMRV in the sample that was not spiked. Secondly, the test was to look for spiked samples, not real live XMRV virus. Therefore the sample that was not spiked may actually have XMRV, but would not be detected by tests which use primers and probes set for an artificial clone.
 

Esther12

Senior Member
Messages
13,774
Likes
28,348
Hmmm...

Is there a copy of that chart which is easier to read than the tiny one?

Will be interesting to see what happens here, but currently I'm more interested in seeing the NIH paper.
 
Messages
1,471
Likes
5
Location
UK
Hmmm...

Is there a copy of that chart which is easier to read than the tiny one?

Will be interesting to see what happens here, but currently I'm more interested in seeing the NIH paper.
I'm not sure if there is, but if you look at the top row, where the sample is, only two groups have this painted green. The FDA (Lo) and the WPI
 

Esther12

Senior Member
Messages
13,774
Likes
28,348
Here it is for people.



Thanks for posting this V99. It is a bit worrying that the WPI got a false positive - but you're right that it seemed the FDA did too. Maybe this is why (some guy - forgotten his name) was talking about the possibility that identifying XMRV from patient samples is different to spiked samples? It would be extraordinary if they sent out a XMRV positive sample in a sample that was meant to be clean, but who knows at this point. If the FDA and WPI picked out the same sample, but that is what happened.

Were the FDA involved with the positive study we're waiting for? I'm getting my three letter acronyms mixed up - FDA, NIH, NCI...

If both positive studies were properly blinded, it would be strange that the contamination would be consistently more likely with the CFS samples.

ps: The CDC kicked arse at the spiked samples.
 

omerbasket

Senior Member
Messages
510
Likes
13
Okay, I have e-mailed Dr. Mikovits and her reply explains it all!
This is what she said:

The statement made by on the CAA website is a misinterpretation of the data and the results of the analytical panel. The analytically panel consisted of spiking blood (confirmed XMRV negative blood with synthetic XMRV (not a natural isolate. Our job was to use QPCR to quantify XMRV in these artificial panels in both whole blood and plasma. We did not do any PCR assays from whole blood or plasma in the Science paper. Moreover, we isolated infectious virus from all positive PCR samples in the Science paper and since its publication in the more than 600 XMRV positive cases we have identified. We also sequence every PCR product and do at least one other method of confirming a positive infection. It is unfortunate that the CAA posted a statement implying the results from the analytical panels presented at BPAC indicate that the WPI and Lo labs have significant problems with false positivity. The phase I study was not designed or powered to evaluate specificity, which as discussed by the group, will require blinded testing of more replicates from a larger number of pedigreed negative control donors, which is planned for phases III and IV
So there is your answer: These tests have not been done by the WPI in the science paper or later, until now. It says absolutley nothing, zero, about the validity of the findings in Science, which were confirmed by many tests - and the best evidence for this in my view is that such a prestigous magazaine, "Science", after a rigorous peer-review, published it.

It is sad, very sad, that the organization that is supposed to help us, the CAA, chose not to mention these stuff and to cause people to believe, falsely, that perhaps the WPI found XMRV so many times when it was not actually there (but in the helathy cohorts the false positives didn't come near as much... but that is something the critics don't want to address...).

People, can you update the title of this thread so that people would know that there is an explanation for all of that in the 4th page? Or perhaps it's even possible to put that message inside the first message or something?
 
Messages
1,471
Likes
5
Location
UK
If there is no official test, it is definitely possible that this sample does have XMRV. Especially when the prevalence rate could be 3% to 7%. These samples will have been confirmed negative with synthetic XMRV, not the real thing. Therefore it cannot be claimed it was a false negative.

Yes, the FDA lab in question is the one involved in the positive study, under Dr Lo. There is another FDA lab also taking part in this testing process.

ps: The CDC kicked arse at the spiked samples.
That's very true, but again not the same as real samples.
 

anciendaze

Senior Member
Messages
1,841
Likes
4,784
With a reported 3% to 7% background rate, the chance of having a positive in an assumed negative control cannot be dismissed out of hand. With rapid mutation taking place in infected individuals, due to action of APOBEC3, spiked samples may well present an easier target.

We're back to the question of starting with a panel of infected individuals, which some involved deny exist. Couldn't they begin with asymptomatic infected, for which part of the controversy disappears?
 

Esther12

Senior Member
Messages
13,774
Likes
28,348
Thanks for that Omerbasket. I didn't think the CAA were being as alarmist as Mikovits implied - the CAA didn't really do any interpretation for there to be a misinterpretation (other than not mentioning the apparent FDA positive). I understand it being a touchy issue though.

Still though - if it was a false positive, it does raise concerns about contamination while things are so uncertain.

If there are two independent blinded studies, I'm not too worried - I just want to see this new one!
 
Messages
1,471
Likes
5
Location
UK
Still though - if it was a false positive, it does raise concerns about contamination while things are so uncertain.
Officially no on would know if it was a negative sample.
 

bakercape

Senior Member
Messages
210
Likes
1
Location
Cape Cod. Mass
If there is no standardied test

How can anyone know if a blood sample was neg. To begin with. Maybe the WPI test was more sensitve and picked up the virus. Isn't that a possibility?
 
Messages
1,471
Likes
5
Location
UK
This is what the CAA said:
while WPI was the only lab reporting an unexplained false positive result on a negative sample.
But the FDA (Lo) also detected xmrv in this sample, and the result cannot be said to be a false negative, if they, as yet, have no way of knowing if it contained xmrv.
 

Esther12

Senior Member
Messages
13,774
Likes
28,348
That's all true.

I wonder how these samples were created?

They could have all been from the same individual's blood?
 

VillageLife

Senior Member
Messages
674
Likes
36
Location
United Kingdom
Perhaps they should have said, although WPI got a false positive, it doesnt matter because actually The phase I study was not designed or powered to evaluate specificity. (that would of been better!)

Instead its a short comment which implies many things and leaves us all thinking WPI have indeed done something seriously flawed! which of course they haven't.
hope it can be edited!!:cool:
 

jspotila

Senior Member
Messages
1,099
Likes
778
I am seeking clarification of what was reported at the meeting, the diagram on the Association's website, and the details of this phase 1 study of the analytic panels. The Association always seeks to provide accurate information, so if there is an error or miscommunication then that will be cleared up and corrected as necessary. I appreciate everyone's patience while I do this.

Edited to add: The statement on the Association site says that there WPI reported only one false positive in this phase 1 study. The Association did not say anything about WPI's testing of samples in general. There was some confusion on another thread because the statement says 1 false positive but the diagram looks like FDA had a false positive as well. As I said, I am seeking clarification of all this.

Patient Heidi Dunlap-Bauer provided a summary of the FDA meeting on her Facebook page. Here is what she said about the Working Group phase 1: "Next up was the Blood XMRV Scientific Research Working Group report by Graham Simmons. This group has just about all the names we are familiar with and then some: Harvey Alter, Jerry Holmberg, Frank Ruscetti, Roger Dodd (you remember him from the May transcripts referring to the perception of an XMRV emergency rather than a real emergency), Suzanne Vernon, Judy Mikovits, John Coffin, Shyh-Ching Lo, Bill Switzer, etc. Presently, the main thrust of this group seems to be to find agreement on an effective assay. Graham Simmons concluded that the study was "too small to conduct meaningful statistical comparisons" and "more work on analytical panel development will need to be performed." The main labs in this group are WPI, FDA (Lo), FDA (Hewlett), NCI and BSRI."
 

Dr. Yes

Shame on You
Messages
868
Likes
45
Thanks Jennie. This matter concerned me as well.

Without knowing what kind of assays, nor even what kind of samples (how they were screened, strain of XMRV, etc) were being used, this was a meaningless piece of information from the CAA. Given this, and the fact that so few samples appear to have been tested, reporting on a single 'false' positive at this stage was premature, except perhaps to have a 'scoop', but the only effect it could have is to unjustly cast doubt on the WPI (and perhaps make the CDC look better). Not mentioning the apparent 'false' positive at the FDA lab was a strange omission as well. We would need more information than this out-of-context snippet, particularly about the means by which samples are initially determined to be negatives. I think it was inappropriate and irresponsible for the CAA to release that tidbit of information to begin with, even leaving aside the question of whether it was a misinterpretation.

These labs are evaluating their assays on spiked blood, so at this stage they are basically just repeating the calibrations used in the individual studies. More technically 'sensitive' PCR as determined by spiked samples does not necessarily equal effective PCR in detecting XMRV in CFS patients. Essentially the assays are being tested for their sensitivity in detecting a synthetic clone of prostate cancer strain XMRV in whole blood, not on their ability to detect XMRV in vivo in PBMC from CFS patients (who have strains of XMRV that differ genetically from the prostate cancer strain). And how one determines that any blood sample is truly XMRV negative in the first place is a good question.
 

eric_s

Senior Member
Messages
1,925
Likes
75
Location
Switzerland/Spain (Valencia)
Thanks for that Omerbasket. I didn't think the CAA were being as alarmist as Mikovits implied - the CAA didn't really do any interpretation for there to be a misinterpretation (other than not mentioning the apparent FDA positive). I understand it being a touchy issue though.

Still though - if it was a false positive, it does raise concerns about contamination while things are so uncertain.

If there are two independent blinded studies, I'm not too worried - I just want to see this new one!
I think they explain the Lo lab's false positive. In the top right corner of the top cell in the Lo column there seems to be an asterisk and at the bottom of the picture there's probably text eplaining it. Unfortunately it's too small for us to read.
 
Messages
1,471
Likes
5
Location
UK
It may explain why the FDA (Lo) box is green, but the idea of a false negative cannot be testable at this stage.