The gastrointestinal manifestation of constitutional mismatch repair deficiency syndrome: from a single adenoma to polyposis-like phenotype and early onset cancer
Abstract
Data on the clinical presentation of constitutional mismatch repair deficiency syndrome (CMMRD) is accumulating. However, as the extraintestinal manifestations are often fatal and occur at early age, data on the systematic evaluation of the gastrointestinal tract is scarce. Here we describe 11 subjects with verified biallelic carriage and who underwent colonoscopy, upper endoscopy and small bowel evaluation. Five subjects were symptomatic and in six subjects the findings were screen detected. Two subjects had colorectal cancer and few adenomatous polyps (19, 20 years), three subjects had polyposis-like phenotype (13, 14, 16 years), four subjects had few adenomatous polyps (8, 12–14 years) and two subjects had no polyps (both at age 6). Of the three subjects in the polyposis-like group, two subjects had already developed high-grade dysplasia or cancer and one subject had atypical juvenile polyps suggesting juvenile polyposis. Three out of the five subjects that underwent repeated exams had significant findings during short interval. The gastrointestinal manifestations of CMMRD are highly dependent upon age of examination and highly variable. The polyps may also resemble juvenile polyposis. Intensive surveillance according to current guidelines is mandatory.
Lynch syndrome (LS) is an autosomal dominant condition caused by a defect in one of the mismatch repair (MMR) genes 1, 2. Constitutional (Biallelic) mismatch repair deficiency syndrome (CMMRD) 3-5 is a rare syndrome characterized by childhood malignancies. These malignancies are primarily hematological malignancies, brain tumors and childhood colon cancer, and are often accompanied by Café au lait spots (CAL) 4, 6-11. Recently, Durno et al. 12 have published a clinical surveillance protocol for CMMRD patients including the beneficial results of the screening, based on 10-years of experience observed in one family. However, as the extraintestinal manifestations are often fatal and occur at early age, most of the data on the gastrointestinal evaluation of these patients is obtained from limited case reports and summaries of these case reports 3, 5, 11, 13, 14. CMMRD is an extremely rare syndrome 11, 15, yet the high percentage of consanguinity among the Palestine Arabs, the occurrence of founder mutations for LS among the Ashkenazi Jews 16 enabled us to collect data on a large number of CMMRD patients.
Methods
The cases were recruited from five large gastrointestinal (GI) cancer prevention units in five tertiary referral centers in Israel: Tel Aviv Medical Center, Rabin Medi-cal Center, Hadassah Medical Center, Rambam Medical Center and Soroka Medical Center that provide medical care to most of the Arab Bedouin population that lives in the south of Israel.
This report includes CMMRD subjects with a demonstrated pathogenic mutation, who underwent colonoscopy, upper endoscopy and examination of the small bowel.
The study has been approved by Rabin Medical Center's ethics committee.
Results
Demographics of the families studied
A total of 11 subjects from seven unrelated families were included (Table 1). Four families are of Arab origin (two families are Palestinian Arabs from Gaza, one family is Palestinian Arab from East of Jerusalem and one family is Bedouin Arab from the south of Israel) and three families of Jewish origin (two families are of Ashkenazi origin and the third family is Iranian Jewish family). Four families (57.1%) had PMS2 mutation, two families (28.6%) had MSH6 mutation and one family (14.3%) had MSH2 mutation. Consanguinity was reported among all the four Arab families and also among the Iranian Jewish family. The other two families were of Ashkenazi Jewish origin, in one family both parents harbored the Ashkenazi founder mutation 1906G>C in MSH2 [this family has been described previously by Toledano et al. 17] and in the second family, both parents harbored two different Ashkenazi founder mutations in MSH6 [c3984_3987dup, c.3987_3988insGTCA; this family has been described by Bakry et al. 11]. At the time of diagnosis, LS-related cancer among first-degree relatives was reported in only one family.
Family | Number of included subjects | Origin | Consanguinity | Gene | Mutation | LS-related cancer in first-degree relatives | LS-related cancer in second-degree relatives |
---|---|---|---|---|---|---|---|
I | 1 | Iranian Jew | Yes | PMS2 |
c.3516del4 c.3516del4 |
No | Grandmother: colon cancer, 40 years |
II | 1 | Arab East Jerusalem | Yes | PMS2 |
c.1787A>G c.1787A>G |
No | No |
III | 1 | Askenazi Jew | No | MSH2 |
c.1906G>Ca c.1906G>Ca |
Mother: endometrial cancer, 50 years | Grandfather: pancreatic cancer, 61 years |
IV | 3 | Askenazi Jew | No | MSH6 |
c3984_3987dupb c.3987_3988insGTCAb |
No | No |
V | 1 | Arab Bedouin | Yes | PMS2 |
Exon9-11del Exon9-11del |
No | No |
VI | 3 | Arab Gaza | Yes | PMS2 |
c.245dupA c.245dupA |
No | No |
VII | 1 | Arab Gaza | Yes | MSH6 |
c.2314C>T c.2314C>T |
No | No |
- LS, Lynch syndrome.
- a Founder Ashkenazi Jew mutation in MSH2.
- b Founder Ashkenazi Jew mutation in MSH6 (compound heterozygote).
Patients' information
Six patients were of Arab origin and five patients were Jews (Table 2). All patients had CAL and five subjects have brain lesions compatible with high-grade glioma. At the time of this report, nine subjects are alive, with the oldest patient aged 25 years.
Characteristics | N (%) |
---|---|
Subjects | 11 (100) |
Gender | |
Male | 7 (63.6) |
Female | 4 (36.4) |
Ethnicity | |
Jew | 5 (45.5) |
Arab | 6 (54.5) |
Current alive | 9 (81.8) |
Death | 2 (18.2) |
Neurofibromatosis-like features | |
CAL | 11 (100) |
Freckling | 2 (18.2) |
Plexiform neurofibroma | 1 (9.1) |
Neurofibroma, face | 1 (9.1) |
Brain involvement | |
High-grade glioma | 5 (45.5) |
Undefined bright objects | 1 (9.1) |
Retinal hyperpigmentation | 1 (9.1) |
Hematologic malignancy | |
T-cell lymphoma | 1 (9.1) |
Gastrointestinal involvement | |
Colonic findings | |
CRC and few polyps | 2 (18.2) |
Multiple (>15) polyps with or without cancer | 3 (27.3) |
Few (1–6) polyps | 4 (36.4) |
No polyps | 2 (18.2) |
Type of colonic polyps | |
Adenomatous polyps | 8 (72.7) |
Juvenile polyps with dysplasia | 1 (9.1) |
Colorectal cancer | 3 (27.3) |
Upper gastrointestinal involvement | |
Carcinoma of papilla and few 2 mm adenomatous polyps | 1 (9.1) |
Small bowel | |
Small bowel tumor | 1 (12.5) |
- CAL, Café au lait spots.
The gastrointestinal manifestations
The summarized information is given in Table 2 and the detailed information is given in Table 3. Five subjects were symptomatic and in six subjects the findings were diagnosed during routine surveillance. Two subjects had colorectal cancer and few adenomatous polyps (19, 20 years), three subjects had polyposis-like phenotype (13, 14, 16 years), four subjects had few adenomatous polyps (8, 12, 13, 14 years) and two subjects had no polyps (both at age 6).
Patient | Gene | Presentation (age) | Colonoscopy (age) | Upper endoscopy (age) | Small bowel evaluation (age) |
---|---|---|---|---|---|
1 | PMS2 |
Iron deficiency anemia and rectal bleeding (20 years) Upper GI bleeding (21 years) |
Initial: five polyps, 2–5 mm, LGD Rectal cancer TPC-IPAA (20 years) Normal pouch on repeat exam (2 years) |
Initial normal endoscopy Five polyps, 2–5 mm, LGD Carcinoma of papilla on repeat exam (2 years). Whipple operation (21 years) |
Normal MRE Single balloon: 3 cm carpeting villous adenomatous polyp, operated (24 years) |
2 | PMS2 | Rectal bleeding (16 years) |
Initial: ≥30 polyps, 2–10 mm, atypical juvenile polyps with dysplasia, TPC-IPAA (16 years) No repeat exam |
Normal upper endoscopy No repeat exam |
Normal CT enterography |
3 | MSH2 | Iron deficiency anemia (14 years) |
Initial: ≥30 polyps, 2–5 mm, LGD Rectal cancer (14 years) No repeat exam |
Normal upper endoscopy No repeat exam |
Normal CT enterography |
4 | MSH6 | Rectal bleeding (13 years) |
Initial: ≥15 polyps, 2–10 mm LGD Two polyps 10, 15 mm with HGD on repeat exam (6 m) TPC-APAA (13 years) |
Initial: normal Two polyps, 2–3 mm LGD on repeat screen (6 m) |
Normal CT enterography Normal VCE |
5 | MSH6 | Asymptomatic-systematic screening |
Initial: four polyps, 2–5 mm, LGD (13 years) Two polyps, 2–5 mm on repeat exam (6 m) |
Normal upper endoscopy Normal repeat exam (6 m) |
Normal CT enterography Normal VCE |
6 | MSH6 | Asymptomatic-screening |
One polyp, 2 mm, LGD (12 years) Normal repeat exam |
Normal upper endoscopy Normal repeat exam (6 m) |
Normal CT enterography Normal VCE |
7 | PMS2 | Rectal bleeding (19 years) |
Initial: a 15 mm sessile malignant polyp (19 years) that deserved further Low anterior resection (19 years) Two polyps 10 mm, LGD on repeat exam (1 year) |
Normal upper endoscopy Normal repeat exam (6 m) |
Normal CT enterography Normal VCE |
8 | PMS2 | Asymptomatic-systematic screening |
Two polyps, 2–3 mm, LGD (14 years) No repeat exam |
Normal upper endoscopy No repeat exam |
Normal CT enterography |
9 | PMS2 | Asymptomatic-systematic screening |
Two polyps, 2 mm, LGD (8 years) No repeat exam |
Normal upper endoscopy No repeat exam |
Normal CT enterography |
10 | PMS2 | Asymptomatic-systematic screening |
No polyps (6 years) No repeat exam |
Normal upper endoscopy No repeat exam |
Normal CT enterography |
11 | MSH6 | Asymptomatic-systematic screening |
No polyps (6 years) No repeat exam |
Normal upper endoscopy No repeat exam |
Normal CT enterography |
- CT, computerized tomography; HGD, high-grade dysplasia; LGD, low-grade dysplasia; MRE, magnetic resonance imaging (MRI) enterography; TPC-IPAA, total proctocolectectomy with ileo-anal pouch anastomosis; VCE, video capsule endoscopy.
Five subjects underwent gastrointestinal surveillance by repeated exams. The first subject (patient number 1) had rectal cancer and few adenomatous polyps at age 20. She underwent total proctocolectomy with ileo-anal pouch anastomosis. Upper endoscopy was normal at that time. Two years later, she was presented with upper gastrointestinal bleeding and was diagnosed with high-grade dysplasia of Papilla of Vater, which was found only by side-view duodenoscopy. As a consequence, she then underwent Whipple operation. At age 24, she was diagnosed with a large jejunal villous adenoma that was detected by single balloon enteroscopy performed after retention capsule has failed to pass. The second subject (patient number 4) had initially about 15 adenomatous polyps, 2–10 mm with low-grade dysplasia at age 13. Six months later, she had also two 10–15 mm adenomatous polyps with high-grade dysplasia. These findings directed her toward total proctocolectomy. The third subject (patient number 5) had initially four polyps, 2–5 mm, low-grade dysplasia and two more 2–5 mm polyps with low-grade dysplasia 6 months later. The fourth subject (patient number 6) had initially two polyps, 2–3 mm, low-grade dysplasia at age 14 and no other findings on repeated exams. The fifth subject (patient number 7) initially had a 15-mm sessile, rectal polyp that was resected endoscopically. As the polyp contained invasive cancer, she subsequently underwent further low anterior resection. A year later, two 10 mm polyps with low-grade dysplasia were removed.
Three subjects had polyposis-like phenotype. The first subject (patient number 2) was presented at age 16 with rectal bleeding. Then, colonoscopy revealed thirty 2–10 mm polyps. The histological features were suggestive of juvenile polyposis. An extensive pathologic review was performed and the final conclusion was atypical juvenile polyposis with dysplasia. She underwent TPC-IPAA. Three years later high-grade glioma appeared. This was followed by a genetic evaluation that revealed Café au late spots, Plexiform Neurofibroma and biallelic a PMS2 mutation. The patient died as a result of brain tumor. The second subject had rectal cancer and over 30 adenomatous polyps and the third subject has been described earlier (patient number 6).
Discussion
The first interesting observation is the distribution of the clinical phenotype: from no findings at age 6, few adenomatous polyps at age 11, polyposis-like phenotype with advanced pathology at age 16 to cancer and few polyps at early twenties. About one third of the subjects in this study had polyposis-like phenotype with more than 15 polyps and two of them already had invasive cancer or high-grade dysplasia. The presence of high-grade dysplasia and cancer distinguishes this syndrome from the classic familial adenomatous polyposis (FAP) where cancer appears much later in life 14, 15, 18, 19. As suggested by Durno 3, 14, 18, cases of FAP-like and non-pathogenic APC mutation should be suspected to be CMMRD cases.
The second interesting observation is that the polyps of CMMRD subjects can give the impressions of juvenile polyposis with dysplasia even after meticulous review by expert gastrointestinal pathologist; hence, one should be alert to this possibility, and cases of suspected juvenile polyposis without pathogenic mutation might also be tested for CMMRD.
The third observation is the importance of surveillance. Of the five subjects that underwent surveillance, three subjects had very significant findings during within very short interval. As suggested by the European consortium for the care of CMMRD subjects, in case of initial significant findings, the interval for surveillance might be even shorter than 1 year 20. The surveillance should be performed in advanced endoscopic units, using enhanced imaging techniques (such as narrow band imaging) 20. As in our series, we did not diagnosed significant gastrointestinal findings below the age of 8, and we support the recommendation of the European consortium 20 to start the gastrointestinal evaluation at this age.
The fourth point is the extent of operation. We had two subjects with rectal cancer (without polyposis): one was directed toward total proctocolectomy and the other toward low anterior resection. The first subject had a very complicated course of operation and suffered from long-term morbidity and the second subject had recovered perfectly but had advanced adenomas during 1 year at follow-up colonoscopy. Hence, although the current guidelines recommend extensive surgery 20, the benefits and costs are not so straightforward.
The fifth observation is the presentation of carcinoma of the duodenal papilla at an early age. Duodenal adenomas and cancer have been described previously 4, 13, 15, 21, 22. Yet, in our series this lesion was recognized only with side-view duodenoscopy, suggesting that side-view examination is advised in CMMRD patients. This finding might also resembles FAP, but at much earlier age 19.
The sixth observation is the diagnosis of jejunal tumor by enteroscopy in an asymptomatic subject with normal magnetic resonance imaging (MRI) enterography (MRE). The clinical guidelines advise that CMMRD patients should undergo annual video capsule endoscopy video capsule endoscopy (VCE) 12, 20. However, in this study, patient number 1 could not perform VCE as the retention capsule was retained. Enteroscopy was insisted upon, and consequently a small bowel tumor was detected, which was not detected by the MRE and was not present several months before (as it involved deep anastomosis in the small bowel).
A limitation of the study is the fact that the recruitment of the patients was at GI cancer prevention units, hence the gastrointestinal manifestations are somewhat overrepresented and the hematologic malignancies are underrepresented.
In this unique case series where all CMMRD subjects had comprehensive gastrointestinal evaluation, we have showed that the gastrointestinal manifestations are highly dependent on age of examination and are highly variable from a single adenoma to a polyposis phenotype and early age colorectal cancer. We have witnessed important findings in short interval of surveillance, highlighting the importance of intensive surveillance according to currently published guidelines 20. We also report that the colonic phenotype can resemble histologically to juvenile polyposis and that the upper gastrointestinal phenotype can also mimic FAP with carcinoma of the duodenal papilla at an early age. Also, based on our follow-up, we presented the pros and cons of extensive surgery for subjects with colorectal cancer.
Acknowledgements
We thank Prof Pikarsky Eli from the Department of Pathology, Hadassah-Hebrew University Medical Center, Jerusalem, Israel, for his contribution to the paper as well as Miss Miri Roth for her technical assistance. No grant support in this study.
Author contributions: Prof Z. L., Dr R. K. and Dr S. C.: study concept and drafting. Dr B. H., Dr E. E H., Dr Y. G. and Dr N. A. F.: data collection and helped in drafting. Dr R. G. and Prof Y. N., Dr R. E. and R. D.: data management. Dr S. M., Dr A. V. and Dr E. B.: pathology review. I. B. K.: genetic data. Prof Z. L. is the guarantor of the article. Miss C. C. helped in English editing.