Background
Hereditary Non-Polyposis Colorectal Cancer (HNPCC) or Lynch Syndrome is an inherited cancer susceptibility syndrome that predisposes individuals to colorectal cancer as well as extra-colonic cancers, such as endometrial, gastric, ovarian, urothelial and others, with an onset usually before 50 years old [
1]. It is characterized by germ-line mutations in one of the several DNA mismatch repair (MMR) genes, namely
MLH1,
MSH2,
MSH6,
PMS2, and
EPCAM. Mutations in
MLH1 and
MSH2 are the most common, encompassing more than 90% of all cases [
2].
The mutations in MMR genes lead to defects in DNA repair. Tumors with MMR-deficiency status exhibit a high frequency of microsatellite instability (MSI-H), causing particularly unstable regions of the genome to be susceptible to errors that do not get corrected. The National Cancer Institute (NCI) recommends a microsatellite panel of five markers (NCI panel) for determining MSI status [
3], and MSI-H tumors are defined as having instability in two or more markers. Tumor MMR status can also be determined by immunohistochemical (IHC) analysis of the protein products of
MLH1, MSH2, MSH6 and
PMS2. While the gold standard for HNPCC diagnosis is genomic sequencing, screening strategies testing for loss of protein expression by IHC versus polymerase chain reaction detection of the MSI panel are considered to be equivalent [
4,
5].
Hypermethylation of the promoter of the
MLH1 gene, a somatic epigenetic change, can also lead to MSI-H, and contributes to about 15% of colon cancer cases [
4,
6,
7]. The
BRAF c.1799 T > A (V600E) mutation is present in 40% of MSI-H tumors [
8], and almost all are associated with hypermethylation of the
MLH1 promoter [
9].
Tumors with MSI-H (somatic or germline) status have unique clinical pathological features [
10]. Histologically, it can present with mucin-rich, signet ring features, and have an increased number of tumor-infiltrating lymphocytes [
11]. Prognostically, patients with HNPCC have a more favorable stage-matched overall survival, but do not benefit from 5-Fluorouracil-based adjuvant chemotherapy [
12]. More recently, MSI-H, associated with high mutation burden, has become a new biomarker for success of checkpoint inhibitor immunotherapy in various types of metastatic tumors [
13].
Based on the Hampel et al. study, which is predominantly composed of Caucasian patients, as well as the study conducted by Aaltonen et al. in Finland, the prevalence of HNPCC is about 2.2–2.7% in colorectal cancer patients [
7,
14]. In the study by Hampel et al., 23 probands were found to have HNPCC; 10 were more than fifty years of age and only 3 fulfilled the Amsterdam criteria for the syndrome [
7]. Current National Comprehensive Cancer Network (NCCN) guidelines recommend screening for HNPCC for all colon cancer patients diagnosed at an age younger than 70 years old [
15]. Clinical screening criteria, the Amsterdam criteria, and the Bethesda criteria [
16] are not considered sensitive enough to capture all cases of HNPCC.
The clinical pathological features of HNPCC in Asian colorectal cancer patients have been studied and various features have been reported, such as early age onset, left side dominance, and the most common extra-colonic cancer being gastric cancer [
17‐
22]. Asian American immigrants may encounter changes in diet after arriving to the United States, including variations in the quantities of meat and vegetables consumed. They may also benefit from the U.S. Preventive Services Task Force recommendation to start colonoscopy screening at fifty years old. Our hospital, Maimonides Medical Center, serves a large body of the Asian immigrant population. In this study, we evaluated the prevalence, clinical characteristics, and mutation preferences of HNPCC in immigrant Asian patients newly diagnosed of colorectal cancer.
Methods
Patients
This is a single-center study launched in 2009 as a retrospective analysis of patients diagnosed with colorectal cancer between 1/2002 and 12/2009. A subsequent modification was made to extend the study to include patients diagnosed between 1/2009 and 6/2015 and data was collected prospectively. The original study protocol and subsequent modifications were all approved by the Institutional Review Board (IRB) of Maimonides Medical Center (MMC). Eligibility criteria included: (1) patients of Asian origin, (2) who had a diagnoses of colon or rectal cancer, and (3) who had tumor specimens available at MMC for testing, or who had test results available from an outside clinical or reference lab. Clinical characteristics and family history were obtained from medical records. A positive family history, as defined in the revised Bethesda criteria, requires one of the patient’s first-degree relatives to be diagnosed with an HNPCC-related tumor less than 50 years of age, or two or more first- or second-degree relatives, regardless of age. The HNPCC-related cancers were defined as colorectal, endometrial, stomach, ovarian, pancreas, ureter and renal pelvis, biliary tract, brain, sebaceous gland adenomas, and small bowel cancers according to the revised Bethesda criteria [
16].
Testing for HNPCC
The screening tests for HNPCC were carried out at Maimonides Medical Center and utilized immunohistochemistry staining (IHC). The IHC was carried out in the research lab on the first 78 patients. After the revision of the NCCN guidelines to test all colon cancer patients diagnosed younger than seventy years old, the IHC test was later routinely performed in the hospital’s pathology department laboratory or outside commercial labs for 59 other patients. Three patients were tested with germline blood test first.
IHC performed in the research lab followed manufacturer’s instructions. The primary antibodies were obtained from the following commercial sources: MLH1 [Clone ES05, Dako Inc., Carpinteria, CA), MSH2 (Clone FE11, CalBioChem Inc., San Deigo, CA), MSH6 (Clone 44, BD Biosciences, San Jose CA), and PMS2 (Clone A16–4, BD Biosciences, San Jose CA). The secondary antibody was MACH 3 Mouse Probe (Biocare Medical Inc., Concord CA), and the IHC slides were developed following incubation with MACH 3 mouse HRP-polymer, and addition of DAB substrate (Dako Inc., Carpiteria, CA). Hematoxylin was then applied to the slides for counterstaining.
The tumor specimen slides were examined by the pathologist to ensure inclusion of tumor tissue. Each staining procedure included two negative controls: (1) no first antibody, and (2) a specimen of known deficiency of the particular MMR protein. The nuclear staining in the adjacent normal mucosa or the stromal tissue or lymphoid cells was used as the internal positive control.
At the setup of IHC in the research lab, a concordance study was also done by performing IHC on tumor samples known for negative or positive expressions of each of the MLH1, MSH2, MSH6 and PMS2 proteins based on germline test results, and on supplemental slides provided by the clinical laboratory at Ohio State University.
For IHC testing performed in the hospital’s pathology department laboratory, the primary antibodies used are as follows: MLH1 (Clone ESO5, Novocastra, Buffalo Grove, IL), MSH2 (Clone G12–1129, Cell Marque, Rocklin CA), MSH6 (Clone 44, Cell Marque, Rocklin CA), and PMS2 (Clone A16–4, Novocastra, Buffalo Grove, IL).
In selected patients, testing for hypermethylation of the MLH1 gene promoter was performed at ARUP Laboratories (Salt Lake City, UT). Furthermore, all cases of HNPCC were confirmed by genomic sequencing performed at Myriad Genetics (Salt Lake City, UT) or Ambry Genetics (Aliso Viejo, CA). In selected patients with strong clinical characteristics of HNPCC, genomic testing was used as a first and definitive test per treating physicians.
Discussion
In this study, we screened 243 cases of colorectal cancer in Asian immigrant patients managed in a community hospital in the South Brooklyn area of New York City, and compiled data on 143 patients with available information for HNPCC screening and confirmation. We have shown here an HNPCC prevalence of 4.19%, and a number of features that are not entirely consistent with literature. The prevalence of HNPCC in the younger population was as high as 16%, while its prevalence in patients of age 50 years and older was extremely low. The tumor locations were predominantly in the left side of the colon. The hypermethylation rate in the entire cohort and even in the older patient group was remarkably low.
To assess whether there is a difference in the clinical presentation of Asian patients with HNPCC versus Caucasian patients, we reviewed our data in the context of several published studies on Asian patients including patients from the Northern or Southern part of Mainland China [
17‐
19,
21‐
23]. Strikingly, one similar result is the predominance of tumors located in the left side of the colon. Our study showed the dominant location to be in the sigmoid and descending colon (66.7%), which is a finding that was revealed in multiple other studies. Jin et al. analyzed data from hospital colorectal surgery registry and 8 patients were diagnosed of HNPCC among 158 consecutive patients. Two patients had sigmoid cancer and five had rectal cancer, while one had cancer in the hepatic flexure [
17]. Liu et al. drew data from the Fudan Colorectal Registry and reported left versus right colon locations to be 64.7% versus 35.3% [
23]. Furthermore, observations from the Singapore HNPCC registry reported by Chew et al. also showed that 69% of the cancers originated from the sigmoid colon [
18]. On the contrary, Hampel et al. showed left side location in 43% of cases [
7]. Our results are more similar to studies from Asian patients residing in Asian countries than to that from Caucasian patients in the United States. For Asian immigrants, the possible diet change or access to early screening in the U.S. may also make a difference in the clinical pathological features of HNPCC in our cohort. Although all the studies on Asian patients may suffer from the bias of small sample size, the emerging concept of left side preference of colon cancer in Asian HNPCC carriers warrants further investigation. The studies on tumor locations is especially important as recent data emerges to indicate that the location of tumors is a prognostic marker for survival, as well as a predictive marker for response to EGFR inhibitor therapy [
24,
25].
Eighty-three percent of the HNPCC cases (5 out of 6) in this study were attributed to patients younger than 50 years old, and the prevalence of HNPCC was 16.1% in this age group. These results confirmed HNPCC as an early-onset, highly penetrative genetic disease. It is consistent with previous reports of relatively early onset of colorectal cancer in Asian HNPCC patients in some studies [
21,
23] but not in others [
18]. The average onset was 36 years old in the study by Liu et al. [
23], and 37.5 years old in the Hong Kong Registry [
21]. Conversely, in our study, the representation of HNPCC from the older patient group was 16.7% (1 out of 6), and the prevalence of HNPCC in the older patient group was 0.89% only. This result is different from that reported in Hampel et al., composed of a majority of Caucasian patients [
7], where 43% of the total HNPCC patients were diagnosed at an age older than 50 years old. In the current revision of the NCCN guidelines, screening of HNPCC is extended to patients up to 70 years old. Additional studies will be required to further evaluate the yield of screening for HNPCC in Asian colorectal cancer patients older than 50 years old.
Ten patients (7%) were diagnosed at an age younger than 40 years old, including two diagnosed of HNPCC. Others were screened extensively for other genetic syndromes, which showed variants of unknown significance or negative findings. There is also recent publication on increasing colon cancer incidence rates in patients 20–39 years of age, from 0.5 to 1.3% annually in the United States [
26]. Colorectal cancer in the very young patients is a health care issue which will warrant further epidemiology studies considering environmental factors such as diet, obesity, and lack of exercise. From this study, all of the very young Chinese cancer patients (<40 years old) were born in China and are not obese. They are presumably more likely to adopt Western diets than older patients are, but may also partially consume Eastern diets with older family members and still have relatively easy access to Eastern foods. Further studies on this subpopulation may generate useful information on the epidemiology of environmental factors associated with colorectal cancers.
Another notable finding is that all patients diagnosed of HNPCC in the study reported family history of HNPCC-related cancers. Patients of young age without family history did not have HNPCC and similarly, patients with hypermethylation of the
MLH1 promoter did not have family history. Therefore, family history is a sensitive screening tool. However, the revised Bethesda criteria would only capture three of the six HNPCC patients based on family history alone. If further relaxation is allowed on the family history criterion including age limit and the number of diagnosed relatives, then all of the patients can be recommended for screening based on family history alone. Using fulfillment of any of the revised Bethesda guidelines for screening, the sensitivity is 100% and the specificity is 76.6%, similar to that reported in other studies [
27]. Among the revised Bethesda criteria, criterion #1 performed best as a single criterion with a sensitivity of 83.3%.
Mutations on
MLH1 and
MSH2 represent 90% of identifiable HNPCC cases [
28]. Our study showed predominant mutations in the
MSH2 and
MSH6 genes. Similarly, Jin et al. also reported 37.5% of the mutations to be in
MSH6, whereas
MSH6 was found in only 10% of HNPCC patients in an Australian study [
20].
MSH6 gene mutations may give rise to later age of onset and an excess of endometrial cancer, according to Wagner et al. [
29]; in the family that carries
MSH6 mutations in our study, one patient developed cancer at age 71, consistent with this previous report.
Although this is a small study, our result raises the following issues to be studied in the future: (1) It will be interesting to perform epidemiological studies to analyze the potential unique interactions of gene mutations and environmental effects regarding the predominant early onset of colorectal cancers in Asian HNPCC carriers. (2) It will be highly relevant to study the cost-effectiveness of HNPCC screening in Asian patients; should laboratory testing be obtained only in those with early age of presentation and carry positive family history? In that regard, the revised Bethesda criteria performed quite well.
The limitations of this study are its retrospective nature and small sample size. We had to exclude 100 patients due to the lack of tumor specimen and clinical information or patient refusal. We were able to determine the age distribution of those patients and found a higher proportion of patients in the older patient group, which can potentially skew the results on the prevalence of colorectal cancer in the elderly.