Exploring Gene Profiles in Small Bowel Adenocarcinoma
Small bowel adenocarcinoma (SBA) is one of the rarest tumors found in patients. In fact, SBA only represents about 1-3% of all gastrointestinal malignancies. Due to its rarity, genomic analyses and prognostic biomarkers are limited.
About 55-82% of small bowel adenocarcinomas can be found in the duodenum, 11-25% are found in the jejunum, and 7-17% in the ileum. The data collected from the few studies of this tumor are not homogenous which causes difficulty with early tumor diagnosis. Each year, the USA reports around 5,300 new cases of this disease resulting in an average of 1,100 deaths. Europe reports around 3,600 cases. The median age for this disease is 60 to 70 years old.
Risk factors for SBA include hereditary colorectal cancer syndrome and inflammatory bowel disease. Since the disease is so rare, it’s not as studied as other gastrointestinal cancers, making it hard to find data on clinical trials and treatment.
Although the small intestine is around 80% of the length and 90% of the surface area of the alimentary tract, the occurrence of adenocarcinoma in the small intestine is so rare that it is around fifty-fold less common than colorectal cancer. Many doctors believe this is because the small intestines tend to transmit food quicker which allows for less exposure to carcinogens and bacteria, unlike the large intestines.
The most common location of small intestine adenocarcinoma is in the proximal region. Currently, the cause of occurrences in the duodenum rather than in the jejunum or the ileum is unknown and is debated. Most SBAs occur from existing adenomas which evolve into invasive carcinomas. Difficulty in performing small bowel examinations is thought to be the cause of delayed diagnosis, which is usually 6 to 9 months after symptom onset.
The Study
The Nationale d’ADEnocarcinome de l’intestin GrêlE (NADEGE) cohort recruited 347 patients from 74 french institutions from January 2009 to December 2012. All consecutive Stage 1 to 4 patients with histologically proven, diagnosed, or with recurrent SBA were enrolled into the cohort. Ampullary and non-adenocarcinoma tumors were excluded.
TNM staging was performed following the criteria of UICC and AJCC at diagnosis by CT scan or magnetic resonance imaging. The researchers collected several study data points including patient’s demographics, tumor stage, cancer treatment history, tumor differentiation, lymph node invasion, and initial treatment and survival. The disease or genetic syndrome was assessed by investigator declaration. Researchers also collected tumor biopsy data either from metastasis or surgical resection.
Results
Among the 347 patients included in the analysis, 196 tumor blocks were collected for molecular analysis and immunohistochemistry. The quantity or quality of the extracted DNA wasn’t adequate for molecular analysis in 71 tumors. The researchers also obtained the mutation status of the 125 patients.
The analyse BIOlogique de la cohorte Nationale d’ADEnocarcinome de l’intestin GrèlE (BIONADEGE) study is an secondary study of the NADEGE cohort aimed at assessing the genomic profile due to predisposing factors for SBA, stage, or location of the SBA. Its purpose is to assess the prognostic value of these genomic alterations. The clinical and tumor characteristics of patients were similar in both the NADEGE and BIONADEGE cohorts except for the metastatic stage at a diagnosis was underrepresented in the BIONADEGE cohort (36% in NADEGE vs 18% in BIONADEGE, P < .0001).
Data revealed at least one genomic alteration in almost all tumors (90.4%). There was no difference in the frequency of one genomic alteration according to tumor stage: 89.2% and 95.4% for localized or metastatic tumor at diagnosis, respectively.
Overall, the results showed the number of mutations per tumor was 0 in 9.6% of patients, 1 in 32.0%, 2 in 26.4%, and >3 in 32.0% of the patients. The proportion of tumors with >3 mutations was also similar according to the stage: 30.4% and 40.9% for localized and metastatic tumors at diagnosis, respectively, and according to primary location: 33.3%, 21.4%, and 40.9% for duodenum, jejunum, or ileum, respectively.
The most frequent genomic alterations observed were KRAS (44.0%), TP53 (38.4%), PIK3CA (20.0%), APC (18.4%), SMAD4 (14.4%) and ERBB2 (7.2%). A KRAS mutation was more frequent in metastatic tumors at diagnosis than in localised tumors (72.7% vs 38.2%, P = .003).
The researchers observed a BRAF mutation in 4% of the total cases. Only one of them was a V600E mutation. The researchers didn’t see any significant difference in mutation rate according to the location of the frequently altered genes.
The difference between gene mutation frequency between patients with Lynch syndrome and those with no disease revealed different scenarios. They also found a pattern for less frequent KRAS mutations in Lynch syndrome and more frequent TP53 and PIK3CA mutations in Crohn’s disease compared to patients with no predisposing disease. There was a trend that showed ERBB2 mutation in Lynch syndrome. There were no ERBB2 mutations in Crohn’s disease. There were several rare mutations that are also more frequent in tumors with Lynch syndrome than in other syndromes such as FGFR1. FGFR3, and ATM. The DH1 mutations are always found in tumors with Crohn’s disease.
Here are the results according to the MMR status determined with immunohistochemical analysis of MMR proteins in the 180 subjects.
- A deficient MMR (dMMR) tumor was observed in 50 (28%) patients.
- Negative staining was observed for both MLH1 and PMS2 in 21 (42%) patients.
- MSH2 and MSH6 in 18 (36%) patients.
- PMS2 with MLH1 inconclusive test in 4 (8%) patients.
- MSH6 with inconclusive MSH2 in 2 (4%) patients.
- PMS2 alone in 2 (4%) patients.
- MSH6 alone in 2 (4%) and MSH2 with inconclusive MSH6 in one (2%) patient.
dMMR tumors are found to be associated with young age. They were less metastatic during diagnosis and had fewer KRAS mutations but they had more SMARCB1 mutations. Another pattern was a reduction in TP53 mutations and an increase in ERBB2 mutations in dMMR tumors.
Discussions
The study reveals a large number of patients with SBA have different tumor profiles based on their respective diseases (Chrons, Lynch, or no known disease) or MMR status. The results in this study are largely consistent with the three previous studies which reported a mutation rate for KRAS from 43.4% to 53.6%, SMAD4 from 9.6% to 17.4%, and ERBB2 from 8.4% to 14%. The study also found that KRAS mutations as pMMR status were often associated with growth and metastasis.
Currently, this study is the first to show that genomic alteration is associated with an advanced stage in SBA. The researchers found no remarkable association with prognosis and the mutations observed at primary tumor site.
Previous studies reported associations with ERBB2 mutation and the duodenal location. This study revealed that the mutation rate was higher in tumors of the adjacent small bowel but without reaching significance. Other rare mutations are found to have contrasting distribution according to the small bowel segment. For example, the IDH1 mutations were only found in the ileum tumor which explains its link to Crohn’s disease. The FBXW7 mutation was observed mostly in the jejunum tumor. This is consistent with other studies that report a trend of more FBXW7 mutations in non-duodenal SBAs. The profile of genomic alterations was also found to be distinct from its predisposing diseases or the MMR status.
Crohn’s disease was linked to tumor genomic alterations of IDH1. 1DH1 mutations were increasingly high in TP53 and were often associated with Crohn’s disease in colorectal cancer. Recent studies report the association of 1DH1 and SMAD4 mutations with Crohn’s disease in SBA. In the study, the researchers did not find any link of SMAD4 mutation and Crohn’s disease.
Overall, previous studies and this study support the hypothesis that SBA is associated with Crohn’s disease and has distinct carcinogenesis from sporadic cancer. This is similar to what has been found in colorectal cancer research.
In SBA linked with Lynch syndrome, the study found a trend of fewer KRAS mutations and more ERBB2 mutations compared to other tumors with no linked diseases. Several other rare mutations such as the FGFR1, FGFR3, and ATM are all associated with Lynch syndrome. However, the risk of developing cancer for patients with Lynch syndrome if the patient had an ATM mutant allele is still up for discussion. The researchers couldn’t decipher if ATM mutation was inherited or acquired. Since both R248C and FGFR3 hotspot mutations are already linked to Lynch syndrome, it’s not linked with SBA.
The study found a particularity in the subgroup of dMMR tumors compared to pMMR tumors. The researchers found that patients with dMMR tumors are often younger than patients with pMMR tumors. This is directly opposite in patients with colorectal cancer.
In this study, Lynch syndrome among dMMR tumors typically reached 34%. Nevertheless, it’s observed in colorectal cancer that dMMR tumors are rarely metastasized at diagnosis. In the study, the researchers found that KRAS mutations are less frequent in dMMR tumors compared to pMMR tumors. This hasn’t been reported in any SBA study and needs therefore a confirmatory study.
The researchers found a trend for fewer TP53 alterations but more ERBB2 alterations in dMMR tumors than in pMMR tumors. The association of ERBB2 mutations and dMMR has previously been reported. The study finds a high frequency of SMARCB1 mutations in dMMR tumors.
The study didn’t find any association between prognosis and genomic alteration. A previous study reports poor prognosis associated with genomic alteration of the ERBB but the study found that ERBB2 mutations simply had no prognostic value. Furthermore, the dMMR phenotype was already reported before as a good prognostic factor for disease-free survival.
Like the previous studies, the researchers found a trend of better prognosis for patients with a dMMR tumor. The prognostic effect also shows only to patients with resected and localized tumors.
No patient in the study with a dMMR tumor received immunotherapy. Most TP53 mutations accompanied poor survival rates in previous studies but had no prognostic value in this study. KRAS mutations were reported poor prognostic predictors in a subgroup of patients with a PT1-T3 tumor but were also linked to better survival rates in patients with metastatic tumors. The study shows no significant effect of KRAS mutation but a pattern of worse prognosis in localized tumors.
The prognostic value of KRAS mutation deserves further evaluation in SBA. BRAF mutations only showed in 4% of the tumors. The majority of BRAF mutations were not V600E mutations. It’s also worth knowing that in patients with metastatic colorectal, non-V600E BRAF mutations were not associated with poor prognosis in contrast to the V600E BRAF mutations.
Other genomic alterations mentioned in the study should be targeted. Another study has found that treatment with an immune checkpoint inhibitor gives prolonged survival in patients with metastatic dMMR SBA. This study encourages the further evaluation of other rare mutations.
Limitations of the Study
The study has some limitations. First, even though it was one of the largest genomic profiling studies of SBA, the sample size was underpowered to accurately show rare mutation impact. Second, the gene panel used is limited and contains mostly altered genes in SBA. Third, the constitutional gene mutations were not assessed in cases of Lynch syndrome in the study. Fourth, the researchers did not perform MSI testing because a previous study reported no discordance between MMR IHC and MSI testing. Study researchers believe results are exploratory and should be taken with caution for rare mutations. The researchers did not perform a Bonferroni correction in their analysis. The clinical characteristics were comparable in the NADEGE and BIONADEGE cohorts, except for the metastatic stage at diagnostic underrepresented in the BIONADEGE cohort due to missing tumor samples suitable for genomic analysis. Because of this, the results for the metastatic tumors are limited. Additional studies are needed to specify the association of genomic profile, clinical data, and prognosis.
Conclusion
The study shows that there are several distinct genomic alteration profiles in SBA which depend on the presence or lack of predisposing disease. The study researchers recommend a separate analysis of sporadic SBA and SBA related to other diseases in the future studies. A collection of similar studies and experiences for off-label targeted therapy is thus needed for this rare disease.
Oncology Related Tools
Other
Latest Research
