Lyme disease researchers call for retraction of paper on deer ticks in Texas

A paper suggesting that 45% of deer ticks in Texas have Lyme disease was raked over the coals in a letter to the editor in a recent issue of Parasites and Vectors, though it doesn’t seem like a retraction is forthcoming.

Steven Norris, a Lyme researcher at the University of Texas at Houston, became concerned when he read the study, which overturned the more commonly acknowledged rate of 1% infection in Texas ticks. He sent the paper around to three other researchers; together, they requested the raw sequencing data from the original group, led by Maria Esteve-Gassent at Texas A&M. That’s how they discovered it was nearly identical to the positive control, making them believe the only differences were the results of sequencing errors.

In their letter published in the journal, they write (emphasis ours):

We support continued studies clarifying the etiology and epidemiology of Lyme disease, Southern Tick-Associated Rash Illness (STARI) and other tick-borne illnesses in Texas and the southern United States in general. However, we believe that the technical flaws in the article by Feria-Arroyo et al. obviate the potential usefulness of the reported results in illuminating the picture of tick-borne illnesses in this region. We further believe that this article provides false information that will mislead public health agencies, physicians and the public, and that it must be rectified. The information included in this letter was provided to the authors three months prior to the letter’s publication, but to date they have not made available information that addresses our concerns regarding the accuracy of the data in their article. Therefore, we have recommended retraction of the article to both the authors and the journal, in keeping with the criteria established by the Committee on Publication Ethics (http://publicationethics.org/files/retraction%20guidelines.pdf). BioMed Central, the publisher of Parasites and Vectors, adheres to these guidelines.

We spoke with Norris for more details:

The article said that 45% of their specimens from Ixodes scapularis ticks in Texas and Northeastern Mexico had Borrelia burgdorferi, the organism that causes Lyme disease in North America. What had been reported previously is less than 1% of ticks in texas had Borrelia burgdorferi in them.

…

I contacted Dr. Esteve-Gassent. My first email to her was on May 8th 2014, and I think the paper was published in April. At that time I sent her a fairly extensive email expressing my concerns and asking for the sequences.

…

A group of four [unrelated tick experts] discussed the article and decided there were serious scientific flaws, so we requested the sequence information. After several email exchanges, Dr. Esteve-Gassent agreed to post the sequences on NCBI and Genebank. At that time we analyzed the sequences and determined that they were virtually identical to the positive control used in their PCR reactions…It was also clear the sequences were poor quality…based on that we came to a consensus that the only differences that were present between their positive control and the sequences they were reporting were due to sequence errors.

…

We wrote a letter to Dr. Esteve-Gassent on June 27 that contained all the information in the letter to the editor. At that time we recommended to them that they retract the paper, but they insisted that the paper was primarily on the ecology of ticks…but in reality most of the attention was directed towards the fact that Borrelia burgdorferi was identified in the samples.

There are a number of organizations here in texas that are sort of Lyme disease support groups, and they saw this as being evidence that Lyme disease is much more prevalent in Texas than had been thought previously. So because we thought this had profound implications on perception to the public and agencies such as public health, we decided to go ahead and contact the journal editor and publisher and to express our concerns. So we sent a letter first to them requesting retraction under the criteria put forth by COPE, and then shortly thereafter we made it a letter for publication.

…

Obviously none of us wants to consider retraction, but in the face of what I think is pretty strong scientific evidence that the information they published is inaccurate, I think that would be the best approach.

We also asked the P.I. from the initial study, Esteve-Gassent, about the letter. She told us that her group is writing a response letter, forthcoming in Parasites and Vectors, though she declined to give us a timeline.