In their recently published study (1), Smits and colleagues reported an association between rs61764370 and colorectal cancer (CRC) survival of early-stage tumors. Similarly, in our study of 678 (441 controls and 237 cases; Table 1), we found that rs61764370 was also associated with a reduced risk of mortality, albeit in late-stage tumors [HRTG/GG vs. TT, 0.38; 95% confidence interval (CI), 0.16–0.88; n = 124] (Fig. 1 and Table 2). Neither study found an association with risk (Table 1).

Table 1.

Risk of colorectal cancer

StratificationGenotypeCases, n (%)Controls, n (%)Univariable [OR (95% CI)]P
All TT 375 (85.0) 202 (85.2) Reference  
 TG/GG 66 (15.0) 35 (14.8) 0.98 (0.63–1.53) 0.945 
African Americans TT 172 (91.0) 88 (94.6) Reference  
 TG/GG 17 (9.0) 5 (5.4) 0.57 (0.21–1.61) 0.292 
European Americans TT 203 (80.6) 114 (79.2) Reference  
 TG/GG 49 (19.4) 30 (20.8) 1.09 (0.66–1.81) 0.739 
StratificationGenotypeCases, n (%)Controls, n (%)Univariable [OR (95% CI)]P
All TT 375 (85.0) 202 (85.2) Reference  
 TG/GG 66 (15.0) 35 (14.8) 0.98 (0.63–1.53) 0.945 
African Americans TT 172 (91.0) 88 (94.6) Reference  
 TG/GG 17 (9.0) 5 (5.4) 0.57 (0.21–1.61) 0.292 
European Americans TT 203 (80.6) 114 (79.2) Reference  
 TG/GG 49 (19.4) 30 (20.8) 1.09 (0.66–1.81) 0.739 
Table 2.

Survival of patients with colorectal cancer

GeneGenotypeHR(95% CI)aPa
Stage I & II TT Reference   
 TG/GG 0.96 (0.40–2.36) 0.938 
Stage III & IV TT Reference   
 TG/GG 0.38 (0.17–0.92) 0.025 
GeneGenotypeHR(95% CI)aPa
Stage I & II TT Reference   
 TG/GG 0.96 (0.40–2.36) 0.938 
Stage III & IV TT Reference   
 TG/GG 0.38 (0.17–0.92) 0.025 

aAdjusted for age and gender (Cox regression model).

Figure 1.

rs61764370 and survival in colorectal cancer stratified by stage.

Figure 1.

rs61764370 and survival in colorectal cancer stratified by stage.

Close modal

There are some key differences between the two studies that could explain the differential findings. The first of these relates to the ethnicity of the cohorts. Our study was based upon both African Americans and European Americans, whereas that of Smits and colleagues focused on a European population. While we lose power upon stratification, we still observe the same association in European Americans and survival of late-stage tumors (HRTG/GG vs. TT, 0.46; 95% CI, 0.18–1.16; n = 78).

We believe a more likely explanation lies in the differences between KRAS mutation status among the two cohorts. In our study, 70% of the cancers with a KRAS mutation were diagnosed as either stage III or IV. In Smits and colleagues, only 33.5% of stage III and 43.5% of stage IV tumors were KRAS-mutant. We also noted that there were no deaths among patients who carried rs61764370 and whose tumors were mutant for KRAS. This could therefore suggest that the presence of KRAS mutation is a more important factor in determining the relationship between rs61764370 and survival than stage. In addition, it helps to explain the differences between the studies' findings and clarify why our association was primarily seen in late-stage tumors.

Another possible difference that could confound these reports is the predictive properties of the rs61764370 single-nucleotide polymorphism (SNP). Zhang and colleagues previously reported that the variant G allele predicts response to single-agent cetuximab monotherapy in refractory CRCs, particularly in patients whose tumors are KRAS wild-type (2). Intriguingly, Graziano and colleagues reported that the same genotype was predictive of poor outcome in patients treated with both cetuximab and irinotecan (3). Collectively, these results suggest that both KRAS mutation status and therapy modulate the impact of this SNP. Smits and colleagues did not describe the treatment details of their cohort, and where data are available, our cohort primarily received 5-fluorouracil. Collectively, we find that this SNP is not predictive of risk. Rather, an association with survival has again emerged. However, larger studies that take KRAS, BRAF, and treatment status (1–5) into account are needed to more accurately and completely define the role of this SNP in CRCs and its potential to contribute to individualized and personalized treatment.

See the Response, p. 3489

No potential conflicts of interest were disclosed.

1.
Smits
KM
,
Paranjape
T
,
Nallur
S
,
Wouters
KA
,
Weijenberg
MP
,
Schouten
LJ
, et al
A Let-7 microRNA SNP in the KRAS 3′UTR is prognostic in early-stage colorectal cancer
.
Clin Cancer Res
2011
;
17
:
7723
31
.
2.
Zhang
W
,
Winder
T
,
Ning
Y
,
Pohl
A
,
Yang
D
,
Kahn
M
, et al
A let-7 microRNA-binding site polymorphism in 3′-untranslated region of KRAS gene predicts response in wild-type KRAS patients with metastatic colorectal cancer treated with cetuximab monotherapy
.
Ann Oncol
2011
;
22
:
104
9
.
3.
Graziano
F
,
Canestrari
E
,
Loupakis
F
,
Ruzzo
A
,
Galluccio
N
,
Santini
D
, et al
Genetic modulation of the Let-7 microRNA binding to KRAS 3′-untranslated region and survival of metastatic colorectal cancer patients treated with salvage cetuximab-irinotecan
.
Pharmacogenomics J
2010
;
10
:
458
64
.
4.
Zhang
W
,
Labonte
MJ
,
Lenz
HJ
. 
KRAS let-7 LCS6 SNP predicts cetuximab efficacy in KRASwt metastatic colorectal cancer patients: does treatment combination partner matter?
Ann Oncol
2011
;
22
:
484
5
.
5.
Ruzzo
A
,
Canestrari
E
,
Galluccio
N
,
Santini
D
,
Vincenzi
B
,
Tonini
G
, et al
Role of KRAS let-7 LCS6 SNP in metastatic colorectal cancer patients
.
Ann Oncol
2011
;
22
:
234
5
.