Panel gene profiling of small bowel adenocarcinoma: Results from the NADEGE prospective cohort

Small bowel adenocarcinoma (SBA) is a rare tumour. Large genomic analyses with prognostic assessments are lacking. The NADEGE cohort has enrolled 347 patients with all stage SBA from 2009 to 2012. Next‐generation sequencing investigates the presence of 740 hotspot somatic mutations in a panel of 46 genes involved in carcinogenesis. The mismatch repair (MMR) status was assessed by immunochemistry. We have collected 196 tumour samples and 125 had conclusive results for mutation analysis. The number of mutations was 0 in 9.6% of tumours, only 1 in 32.0%, 2 in 26.4% and ≥3 in 32.0%. Overall, at least one genomic alteration was observed in 90.4% of tumour. The most frequent genomic alteration was in KRAS (44.0%), TP53 (38.4%), PIK3CA (20.0%), APC (18.4%), SMAD4 (14.4%) and ERBB2 (7.2%) genes. KRAS mutations were more frequent in synchronous metastatic tumours than in localised tumours (72.7% vs 38.2%, P = .003). There was no significant difference in the mutation rates according to primary location for the most frequently altered gene. ATM, FGFR3 and FGFR1 gene alterations were associated with Lynch syndrome and IDH1 mutations with Crohn disease. dMMR tumours were associated with younger age, localised tumours, less KRAS but more SMARCB1 mutations. No genomic alteration was associated with overall survival. There is a trend for better survival in patient with dMMR tumours. In conclusion, there is a different genomic alteration profile in SBA according to predisposing diseases. No association between genomic alterations and prognoses was observed except for a trend of better prognoses associated with dMMR.

carcinogenesis. The mismatch repair (MMR) status was assessed by immunochemistry. We have collected 196 tumour samples and 125 had conclusive results for mutation analysis. The number of mutations was 0 in 9.6% of tumours, only 1 in 32.0%, 2 in 26.4% and ≥3 in 32.0%. Overall, at least one genomic alteration was observed in 90.4% of tumour. The most frequent genomic alteration was in KRAS (44.0%), TP53 (38.4%), PIK3CA (20.0%), APC (18.4%), SMAD4 (14.4%) and ERBB2 (7.2%) genes. KRAS mutations were more frequent in synchronous metastatic tumours than in localised tumours (72.7% vs 38.2%, P = .003). There was no significant difference in the mutation rates according to primary location for the most frequently altered gene. ATM, FGFR3 and FGFR1 gene alterations were associated with Lynch syndrome and IDH1 mutations with Crohn disease. dMMR tumours were associated with younger age, localised tumours, less KRAS but more SMARCB1 mutations. No genomic alteration was associated with overall survival. There is a trend for better survival in patient with dMMR tumours. In conclusion, there is a different genomic alteration profile in SBA according to predisposing diseases. No association between genomic alterations and prognoses was observed except for a trend of better prognoses associated with dMMR.

K E Y W O R D S
cohort study, Crohn's disease, genomic profiling, Lynch syndrome, MMR status, small intestine adenocarcinoma 1 | BACKGROUND Small bowel adenocarcinoma (SBA) is a rare tumour of poor prognosis. 1 Nevertheless, it is the first aetiology of small bowel cancer in France 2 and second aetiology in the United States. 3 Concordant findings report an increasing incidence of SBA. 2,4,5 Few studies have investigated the molecular phenotype of SBA.
A previous study reports that the genomic profile of SBA is closer to colon adenocarcinoma rather than gastric adenocarcinoma. 6 Recently, a large genomic analysis mainly on Stage IV tumours has reported a distinct profile of SBA compared to gastric or colon adenocarcinoma. 7 Indeed, if RAS mutation prevalence is similar to colon cancer, APC mutations are much less frequent, BRAF rarely involved V600E point mutations and ERBB2 mutations or microsatellite instabilities (MSI) are more frequent than in colon cancer. 7-10 A prognostic value had been inconsistently associated with ERBB2 mutations, 11 MSI 9 or TP53 mutations. 12 Some differences of genetic profile were reported according to the small bowel segment. Indeed, several studies found that ERBB2 mutations were more frequent in duodenum, 7 Lynch syndrome was associated with younger age, poor differentiation, an early stage and duodenum location. Tumour grade and stage were the main prognostic factors. 13 The analyse BIOlogique de la cohorte Nationale d'ADEnocarcinome de l'intestin GrèlE (BIONADEGE) study is an ancillary study of the NADEGE cohort aimed to assess the genomic profile according to a predisposing disease for SBA, to SBA localisation or stage and assess the prognostic value of these genomic alterations.

What's new?
Because small bowel adenocarcinomas (SBAs) are quite rare, genomic analyses and prognostic biomarkers are lacking. In this study, the authors found at least one genomic alteration in 90.4% of SBAs. The most frequent alterations occurred at KRAS, TP53, PIK3CA, APC, SMAD4 and ERBB2. Additional alterations were specific to SBAs from patients with Lynch syndrome, while IDH1 mutations were associated with Crohn disease. No association was found between SBA prognosis and specific genomic alterations, except for a trend toward better prognosis with tumors deficient in mismatch repair (dMMR).

| Study population
The NADEGE cohort has recruited 347 patients in 74 participating French institutions from January 2009 to December 2012. All consecutive Stage I-IV patients with histologically proven, newly diagnosed or with recurrent SBA (local or distant) were enrolled into the NADEGE cohort. Ampullary and non-adenocarcinoma tumours were excluded. TNM staging was done according to the criteria of AJCC and UICC (7th UICC TNM Staging System) performed at diagnosis by computed tomography (CT) scan and/or magnetic resonance imaging. The following clinical data were prospectively collected: demographics, cancer treatment history, tumour stage, lymph node invasion, tumour differentiation, initial treatment and survival.
The predisposing disease or genetic syndrome was assessed by investigator declaration. The tumour blocks of either tumour biopsy from primary or metastasis or tumour surgical resection were collected.

| Immunohistochemistry
Tissue microarrays (TMA) were constructed from 0.6-mm diameter tissue cores obtained from formalin-fixed paraffin embedded tumour specimens. Haematoxylin and eosin (H&E) staining was performed on each TMA slide to confirm the presence of tumour tissue. The expression of MLH1, MSH2, MSH6 and PMS2 was assessed as previously described. 9 Briefly, 4 μm sections were cut onto silane-treated Super Frost slides (CML, Nemours, France) and left to dry at 37 C overnight.
The slides were deparaffinised in xylene and rehydrated in pure ethanol. Endogenous peroxidase was blocked using 3% hydrogen peroxide in methanol for 30 minutes. Before immunostaining, antigen retrieval was performed by immersing sections in citrate buffer (pH 6.0). Sec

| Molecular analysis
The same paraffin blocks were used for DNA extractions and for IHC analyses. DNA was extracted from formalin-fixed, paraffin-embedded neoplastic tissue that had been macrodissected with reference to the H&E stained section.  (Table S1). DNA extraction, NGS and mutation calling were performed as described previously. 10

| Statistical analysis
Descriptive analysis of the initial tumour stage (reference) and variables measured at baseline was performed. Categorical variables were summarised as frequencies and percentages and continuous variables as medians and ranges. The comparison of gene alteration frequencies according the subgroup of patients was assessed with the χ 2 test or Fisher's exact test, as appropriate, for categorical variables.
Patients with metastatic disease were defined as those who had metastasis at the time of the inclusion and those who developed additional metastatic recurrence tumours during follow-up. Therefore, some patients in this trial were analysed twice: first, as cases with localised tumours, and second, as cases with metastases.
Overall survival (OS) was defined as the time from diagnosis of a primary tumour (localised tumour) or of metastasis (synchronous or metachronous) until death due to any cause. Patients who were still alive at the last follow-up were censored. Patients with synchronous resected metastasis were excluded from the analysis of metastatic patient subgroup in order to assess OS of patients with unresectable metastases.
The survival curves for OS were estimated by the Kaplan-Meier and were compared using the log-rank test. The follow-up time was assessed by the reverse Kaplan-Meier method. The medians and 95% confidence intervals (95% CIs) were calculated and 3-year rates with 95% CI were also provided.

| Survival analysis in the 125 patients with mutational status available
Univariate analysis was performed in the subgroup of patients with mutation statuses available to assess the prognostic factors for OS including clinical parameters and the gene mutation with a frequency over 10% (Table 5). No genomic alteration was associated with OS ( The risk of developing a cancer for patient with Lynch syndrome if they had an ATM mutant allele is a matter of debate. 17 We could not determine in our study if the ATM mutation was inherited or acquired.
FGFR3 R248C hotspot mutation has already been associated with the Lynch syndrome in upper tract urothelial carcinoma 18 but not with SBA until our report.
We found some specificity in the subgroup of dMMR tumours compared to pMMR tumours. Patients with dMMR tumours are younger than patients with pMMR tumours. This is the inverse result that it is observed in colorectal cancer. 19 That may be explained by the fact that in our study the proportion of Lynch syndrome among dMMR tumour reach 34%. Nevertheless, as it is observed in colorectal cancer, the dMMR tumours are rarely metastatic at diagnosis. In our study, KRAS mutations are less frequent in dMMR tumours compared to pMMR tumours. This has not been previously reported in SBA and deserves a confirmatory study. There is also a trend for less TP53 alterations but more ERBB2 alterations in dMMR tumours than in pMMR tumours. The association of ERBB2 mutations and dMMR has previously been reported. 10 We report a higher frequency of SMARCB1 mutations in dMMR tumours, SMARCB1 has already been described in dMMR colorectal tumours. 20 We did not find any association between genomic alteration and prognosis. One previous study reports a poor prognosis associated with a genomic alteration of the ERBB signalling cascade, 11 but ERBB2 mutations solely had no prognostic value. The dMMR phenotype was already reported as good prognostic factor for disease-free survival in one study. 8 In our study as in a previous one, 9 there is a trend for better prognosis in patients with a dMMR tumour. The prognostic effect of dMMR phenotypes seems restricted to patients with localised and resected tumours. In patients with metastatic tumours, the MMR status seems to have no effect. It must be pointed out that no patient with a dMMR tumour received immunotherapy. TP53 mutations were reported associated with poor survival in a previous study, 12 but had no significant prognostic value in our study either in localised tumour or metastatic tumour like in another previous study. 9 KRAS mutations were reported as a poor prognostic predictor in the subgroup of patients with a pT1-T3 tumour 21 but also associated with a better survival in patients with metastatic tumour. 9 In our study, there was no significant effect of KRAS mutation but a trend of a worst prognosis in localised tumours and a better prognosis in metastatic tumours. It has been previously reported that KRAS mutations were associated with a poor OS in colorectal Stage III pMMR tumours. 22  Our study had some limitations: first, even if this study is one of the largest genomic profiling of SBA, the sample size does not allow an accurate evaluation of rare mutation impact. Second, the gene panel used is limited but contains the most frequently altered genes in SBA. Third, the constitutional gene mutations were not assessed in case of Lynch syndrome in our study. Fourth, we did not perform MSI testing, nevertheless a previous study has reported no discordance between MMR IHC and MSI testing. 9 Finally, we assume that our results are exploratory and should be taken with caution for the rare mutations as we did not perform a Bonferroni correction in our analysis. Moreover, the clinical characteristics were comparable in the NADEGE 13 and BIONADEGE cohorts, except for metastatic stage at diagnostic underrepresented in the BIONADEGE cohort due to missing tumour samples suitable for genomic analysis. Thus, our results in metastatic tumours are limited. Additional studies pooling several databases are needed to specify the association of genomic profile, clinical data and prognosis.
In conclusion, our study shows that there are different genomic alteration profiles in SBA that depends on the existence, or lack thereof, of a predisposing disease. This advocates to analyse separately sporadic SBA and those related to predisposing disease in future studies. With caution due to sample size, genomic alteration had no prognostic impact except a trend for a favourable prognosis associated with dMMR phenotypes in localised tumour. Nevertheless, some genomic alterations may be targeted. A compilation of worldwide experiences for off-label targeted therapy is urgently needed for this orphan disease.