Introduction
Ampullary adenocarcinoma (AMPAC) is a rare malignancy that originates within the ampulla of Vater. Its incidence is approximately 0.59 cases per 100,000, but has been steadily increasing over the past few decades [
1]. Although AMPAC only accounts for 0.2% of gastrointestinal malignancies, it carries significant clinical importance due to its pathological variations and associated prognosis. The 5-year overall survival rate for AMPAC patients who undergo resection ranges from 35 to 50%, with outcomes heavily influenced by various clinical and histological factors, particularly tumor stage and treatment [
2]. Despite nearly 80% of AMPAC cases being resectable at diagnosis, approximately half of these cases may experience disease recurrence [
3]. For patients with advanced AMPAC, first-line systemic therapy typically involves chemotherapy, with the specific regimen selected based on the subtype [
4]. However, the antitumor efficacy of chemotherapy, regardless of the treatment regimen, remains unsatisfactory [
5]. Therefore, gaining a better understanding of the biological features of AMPAC may provide valuable insights for the development of more effective treatment options.
Biomarker-driven cancer therapy is at the forefront of precision medicine in oncology, and the advent of novel technologies, particularly next-generation sequencing (NGS), has exponentially expanded our knowledge of the genetic characteristics and actionable alterations in various cancer types [
6]. However, due to its low incidence, current understanding of the genomic profile of AMPAC lags behind that of other gastrointestinal malignancies [
7,
8]. Recent two major studies have shed important light on the genetic characteristics of AMPAC. Gingras et al. investigated the genomic features of 98 AMPAC cases and identified genomic similarities with bile duct cancer and duodenal adenocarcinomas [
9]. AMPAC was characterized by a high prevalence of genomic alterations (GAs) in genes associated with the WNT pathway, as well as inactivating GAs in
ELF3. Another study by Yachida et al. identified common GAs in
KRAS,
TP53,
CTNNB1,
SMAD4,
APC, and
ELF3 in a cohort of 172 Japanese and American AMPAC patients [
10]. However, it remains unknown whether there are genomic differences between Eastern and Western patients, and little is known about the genomic features of Chinese patients with AMPAC. The National Comprehensive Cancer Network (NCCN) guidelines recommend genetic testing for AMPAC patients, including genes such as
ALK,
NRG1,
NTRK,
ROS1,
FGFR2,
RET,
BRAF,
BRCA1/2,
KRAS, and
PALB2 [
2]. However, it is worth noting that this recommendation is directly adapted from that for pancreatic cancer, and the actual prevalence of these actionable alterations in AMPAC remains uncertain.
To better understanding of the genomic feature of Chinese AMPAC patients, we have enrolled 145 patients in our cohort to perform genomic profiling. We aimed to access (1) the germline and somatic genomic feature of Chinese AMPAC patients; (2) Genomic difference with the Western AMPAC patients; (3) actionable genomic alterations that may have targeted therapy choice.
Discussion
In the present study, we perform a comprehensive genomic study of 145 Chinese AMPAC patients from a single center.
The NCCN guideline recommend patients diagnosed with AMPAC or with a positive cancer history to take genetic testing for detecting genes involved for hereditary cancer syndromes [
2]. The recommend genes, for instance,
ATM,
BRCA1,
BRCA2,
CDKN2A,
MLH1,
MSH2,
MSH6,
PALB2,
PMS2,
STK11, and
TP53, are mainly involved with the hereditary pancreatic cancers. In current study, we found 8.27% of the Chinese AMPAC patients harbored P/LP germline variants, which were all in the DDR pathways. In a small cohort of mainly Caucasian AMPAC patients, the incidence of P/LP germline variants was 18% (8/44), and these variants were in
BRCA2,
ATM,
RAD50 and
MUTYH [
21]. However, this high ratio may be biased because the limited sample size. Intriguingly, we identified a AMPCA patient harbored
MSH3 P/LP germline with notable cancer family history, that his father was diagnosed with bladder cancer and his brother had rectal cancer. Unlike the canonical genes involved in the DNA mismatch pathway (
MLH1,
MSH2,
MSH6, and
PMS2), there is no direct evidence suggesting an association between MSH3 and Lynch syndrome. Previous studies have shown that loss of
MSH3 leads to an enrichment of ≥ 2 bp indel alterations and longer indels in malignant cells, which increases the risk of colorectal cancer [
24,
25]. This may be due to the dysfunction of the MSH2-MSH3 (MutSβ) complex in cells, which is primarily responsible for repairing indel loops instead of base-base mismatches [
26].
We identified several genomic differences between Chinese and Western AMPAC patients, particularly the consistent differences observed in
PIK3CA and
ARID2. However, to date, there have been no studies investigating the biological function of
PIK3CA and
ARID2 specifically in AMPAC.
ARID2, which encodes a subunit of the PBAF complex involved in chromatin remodeling, has been detected as a novel tumor suppressor in other cancers such as PDAC, CRC, and BTC [
27‐
29]. Alterations in
ARID2 in bowel cancer have been associated with worse clinical outcomes and promoted proliferation and metastasis of malignant cells [
28]. Additionally,
ARID2 is implicated in tumor immunology, particularly T cell cytotoxicity. Loss of
ARID2, along with other genes in the SWI/SNF chromatin remodeling complex, may sensitize tumor cells to immunotherapy [
30]. The PI3K-AKT signaling pathway has been shown to limit T cell recognition and the killing of cancer cells in PDAC, while downregulation of PIK3CA has been found to enhance T cell infiltration and promote tumor regression [
31]. The difference in the prevalence of
ARID2 alterations in Chinese AMPAC patients may have implications for prognosis and therapeutic efficacy. Further research is needed to elucidate the specific roles of
PIK3CA and
ARID2 in AMPAC and to explore how their alterations may influence disease progression and response to treatment in this particular patient population.
Previous evidence on the benefit of immunotherapy in treating AMPAC is poor. Cardin et al. has observed clinical benefit in a small cohort of AMPAC patients, but has to close the study because of the slow enrollment [
32]. The MSI, TMB and PD-L1 level are the widely-acknowledged biomarkers associated with the response of immunotherapy [
33]. In western AMPAC patients, the incidence of MMR-deficiency (MMRd) was higher than those with colorectal cancer. Previous studies found that 18% (23/127) of western patients with AMPAC had MMRd as evaluated by IHC [
34]. However, in current study, we only identified two AMPAC patients (1.38%) with MSI-H, which was significantly lower than the incidence in the western patients (8.75%,
p < 0.01) [
35]. In a meanwhile, in a European cohort compromised of 59 AMPAC cases, only 8.2% (4/49) of them were MSI-H, which all had a loss of MLH1 an or PMS2 protein in immunohistochemical analysis [
4]. Wong et al. found only 4.55% (2/44) of AMPAC patients were considered as MSI-H which were caused by somatic GAs in MMR genes [
21]. A previous study also found a low incidence of MSI-H in Chinese with duodenum cancer was 3.2% (8/243) [
36]. The PD-L1 expression level of patients with AMPAC has not been comprehensively evaluated, especially for Chinese patients. Previous study found 26.9% of invasive APC and 6.0% of ampullary dysplastic samples were PD-L1 positive, and PD-L1 expression samples were mainly intestinal-type and poorly differentiated [
37]. Most of the PD-L1-positive tumors (seven of 10) were intestinal-type and poorly differentiated (G3). PD-L1 expression was associated with the MMR status. In 22 patients with MMRd, only 18.18% of them had positive PD-L1 expression in tumor cells (TPS
\( \ge \)1%); however, 10 of them had combined positive score (CPS) over 1%, which indicated a correlation between MMRd and PD-L1 expression in immune cells [
34]. In concordance with our result, previous study has found dysfunction of WNT pathway, especially frequently-altered
CTNNB1 in patients with AMPAC [
9]. Even though multiple targeted drugs have been developed, there still lacks effective personalized medicine targeted at
CTNNB1 and/or WNT pathway currently. However, multiple studies have suggested the
CTNNB1 GAs as a negative predictive biomarker for immunotherapy, which shaped an immunosuppressive tumor microenvironment with depleting of T cells [
38]. In our study, we identified for the first time, a negative correlation between
CTNNB1 GAs and PD-L1 expression in AMPAC, which was in consistent with previously-noted feature in other cancer types [
39].
Conducting of an umbrella study to prove the reasonability of biomarker-matched therapy in AMPAC is problematic due to the rarity of the disease and even lower incidence of actionable GAs in it. However, the real world study, such as “Know your tumor” in pancreatic cancer has proved the survival benefits in biomarker-matched therapy [
40]. In this study, we found over 75% of Chinese AMPAC patients with actionable alterations and over 20% of them had actionable GAs in Level 3, which were standard of care or investigational biomarkers predictive of response to an FDA-approved or investigational drug in another indications. Another study has found a similar frequency (70%) of actionable GAs in 97 Indian AMPAC patients [
41]. Adoption of anti-HER2 therapy, ado-trastuzumab emtansine, has achieved partial response for approximately 6 months for a AMPAC patient with
ERBB2 amplification [
21]. For patients with DDR GAs which may confer sensitivity to PARP inhibitors, a phase II, umbrella clinical trial that evaluate the efficacy of the combination of ATR inhibitor (AZD6738) with PARP inhibitor or immunotherapy for advanced biliary tract cancer, including AMPAC, is ongoing [
42].
There are several limitations to our study that should be acknowledged. Firstly, we utilized a targeted gene panel for genetic testing instead of WES, which resulted in some missing genomic information due to the limited number of genes included. Secondly, we did not test for
NTRK1,
NTRK2, and
NTRK3 fusions, which are recommended pan-cancer druggable biomarkers according to the NCCN guidelines. The exclusion of
NTRK genes was due to the limitations of DNA-based sequencing methods in detecting such fusions [
43]. Additionally, despite our efforts to increase the sample size in our cohort, the study still had a limited number of samples due to the rarity of AMPAC. Moreover, although we identified actionable genetic alterations, these patients did not receive matched therapy based on these biomarkers prior to the finish of the current study. Therefore, it remains unknown whether these identified biomarkers can truly serve as predictive biomarkers for guiding matched targeted therapy in AMPAC. Lastly, we did not collect comprehensive follow-up data to correlate genomic findings with the prognosis of AMPAC. Hence, further studies with larger sample sizes are necessary to investigate the genomic characteristics of Chinese patients with AMPAC, particularly those that may differ from Western patients.
In summary, we conducted a genomic analysis on the largest cohort of Chinese AMPAC patients to date and have made the data available for future drug development and applications.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.