A first study of cytokine genomic polymorphisms in CFS: Positive association of TNF-857 and IFNgamma 874 rare alleles, 2006, Carlo-Stella et al

[forestglip]

A first study of cytokine genomic polymorphisms in CFS: Positive association of TNF-857 and IFNgamma 874 rare alleles

N Carlo-Stella, C Badulli, A De Silvestri, L Bazzichi, M Martinetti, L Lorusso, S Bombardieri, L Salvaneschi, M Cuccia

Published: 2006

[Line breaks added]

Objective
In the past two years we have developed a biological bank of genomic DNA, cDNA, serum and red blood cells of Italian patients with certified CFS from the two Italian referral centers for the syndrome. Recent studies have shown an imbalance in cytokine production in disease states similar to Chronic Fatigue Syndrome (CFS), such as sickness behavior, both in animals and in humans. However we notice that serum cytokine concentrations are often inconstant and degrade rapidly.

With this in mind, we investigated cytokine gene polymorphisms in 80 Italian patients with CFS in order to ascertain whether in this group of patients it is possible to describe a genetic predisposition to an inflammatory response.

Methods
We analyzed the promoter polymorphisms of IL-10, IL-6 and the IFNgamma 874 T/A polymorphism in intron 1 with a PCR-SSP method (Cytogen One Lambda Inc. Canoga Park, CA, U.S.A) in 54 patients and TNF-308 G/A and -857 C/T promoter polymorphisms with a PCR-RFLP method (in 54 and 80 patients respectively).

Results
There is a highly significant increase of TNF -857 TT and CT genotypes (p = 0.002) among patients with respect to controls and a significant decrease of IFN gamma low producers (A/A) (p = 0.04) among patients with respect to controls.

Conclusions
We hypothesize that CFS patients can have a genetic predisposition to an immunomodulatory response of an inflammatory nature probably secondary to one or more environmental insults of unknown nature.

Web | PDF | Clinical and Experimental Rheumatology | Open Access

 

[forestglip]

TLDR: They didn't do the replication with a second cohort properly, and the significant SNPs don't replicate in DecodeME.

---

Fukuda criteria. They found significantly more participants had the T allele in the TNF-857 SNP. Also fewer had the A allele in the IFN-γ-874 SNP, but this was much less significant. From the paper:

Our results demonstrate a statistically significant difference (p = 0.0049) between case subjects and controls for the TNF-857 gene polymorphism in which a greater presence of homozygous and heterozygous subjects for the rare allele can be seen.

Second study
To confirm the data, especially on the TNF-857 gene polymorphism we recruited 26 more patients (for a total of 80 patients) and 224 healthy blood donors (a number of them having been used for the first study) to be used as controls and recruited from the same Transfusion Center ward of the Policlinico San Matteo Hospital in Pavia with the same criteria. The analysis by PCR-RFLP was repeated as cited above. The results (shown in Table III) repeat the genotype frequency distribu- tion described in Table II for the TNF-857 gene polymorphism with a p value equal to 0.002.

The TNF-857T allele seems to be associated with an increased production of this cytokine (12). This polymorphism has been studied in a few rheumatological diseases (13, 14).

Spoiler: References

12: Higuchi T, Seki N, Kamizono S, Yamada A, Kimura A, Kato H, Itoh K. Polymorphism of the 5'-flanking region of the human tumor necrosis factor (TNF)-alpha gene in Japanese. Tissue Antigens. 1998 Jun;51(6):605-12. doi: 10.1111/j.1399-0039.1998.tb03002.x.

13: Date Y, Seki N, Kamizono S, Higuchi T, Hirata T, Miyata K, Ohkuni M, Tatsuzawa O, Yokota S, Joo K, Ueda K, Sasazuki T, Kimura A, Itoh K, Kato H. Identification of a genetic risk factor for systemic juvenile rheumatoid arthritis in the 5'-flanking region of the TNFalpha gene and HLA genes. Arthritis Rheum. 1999 Dec;42(12):2577-82. doi: 10.1002/1529-0131(199912)42:12<2577::AID-ANR10>3.0.CO;2-O.

14. Tsuchiya N, Kawasaki A, Tsao BP, Komata T, Grossman JM, Tokunaga K. Analysis of the association of HLA-DRB1, TNFalpha promoter and TNFR2 (TNFRSF1B) polymorphisms with SLE using transmission disequilibrium test. Genes Immun. 2001 Oct;2(6):317-22. doi: 10.1038/sj.gene.6363783.

The trouble is with their attempt at replication. Here are the results for the two cohorts:




They say "To confirm the data, especially on the TNF-857 gene polymorphism we recruited 26 more patients (for a total of 80 patients) and 224 healthy blood donors (a number of them having been used for the first study) to be used as controls".

For patients, looking at the total numbers for the cohorts, I see 54 patients in the first comparison, and 80 in the second. So for their replication, they added 26, but included the original 54, so not actually a replication on a new cohort. And for controls, it looks like 224 in both comparisons of the TNF SNP, which probably means they used exactly the same control group.

That said, it could still be considered one study finding this association, but without replication or multiple test correction. A study I found gives the ID of the TNF-857 SNP as rs1799724. This SNP has a p-value of 0.27 in DecodeME, and 0.43 in the UK BioBank, so doesn't appear to be replicated.

For the IFN-γ-874 SNP (rs2430561), DecodeME has a p-value of .07, but it looks to be the opposite effect direction of this study. (And it's not on Gene Atlas for the BioBank.)

DecodeME results for these two SNPs:

 

[Simon M]

I remember talking to Chris Ponting this study years ago, and the possibility of a study focused on immune genes. These are much more affordable.

His comment was that these kind of targeted genetic studies used to be very popular, but have a poor track record on replication. And that's when replication is done properly.

 

[Jonathan Edwards]

His comment was that these kind of targeted genetic studies used to be very popular, but have a poor track record on replication.

I wonder if in retrospect that was non-sequitur? Maybe they had a poor track record because they often were not done well. I cannot see how a poor track record in replication can be the fault of intelligent targeting of a smaller number of genes. Targeted studies worked very well in rheumatic disease, with excellent replication.

I am sure Chris was right to go for a broad GWAS in ME/CFS because in a sense we were looking for something and we had no idea what it was going to be.

 

[forestglip]

I wonder if in retrospect that was non-sequitur? Maybe they had a poor track record because they often were not done well. I cannot see how a poor track record in replication can be the fault of intelligent targeting of a smaller number of genes. Targeted studies worked very well in rheumatic disease, with excellent replication.

I am sure Chris was right to go for a broad GWAS in ME/CFS because in a sense we were looking for something and we had no idea what it was going to be.

Just adding the context from Wikipedia for why they say candidate gene studies don't replicate:

Candidate genes hypothesized to be associated with complex traits have generally not been replicated by subsequent GWASs[3][4][5][6] or highly powered replication attempts.[7][8] The failure of candidate gene studies to shed light on the specific genes underlying such traits has been ascribed to insufficient statistical power, low prior probability that scientists can correctly guess a specific allele within a specific gene that is related to a trait, poor methodological practices, and data dredging.[9][6][10]

 

[Jonathan Edwards]

...has been ascribed to insufficient statistical power, low prior probability that scientists can correctly guess a specific allele within a specific gene that is related to a trait, poor methodological practices, and data dredging.

i.e. not done well!

 

[Utsikt]

ascribed to insufficient statistical power, low prior probability that scientists can correctly guess a specific allele within a specific gene that is related to a trait, poor methodological practices, and data dredging.

What if you get rid of the last two?

 

[Jonathan Edwards]

not done well