Multidisciplinary approach to a case of Lynch syndrome with colorectal, ovarian, and metastatic liver carcinomas
Lynch syndrome is an autosomal dominant disorder with an estimated prevalence of 3 % of all colorectal cancers. It is attributed to germline mutations in DNA mismatch repair (MMR) genes, which confer increased susceptibility to cancers of the colorectum, endometrium, stomach, small intestine, hepatobiliary system, kidney, urinary bladder, brain, and ovary. We report a thought-provoking Lynch syndrome case with a family history and simultaneous tumors in the colon, pelvis, and liver. These findings made diagnosis and treatment complicated. However, the multidisciplinary approaches followed by a medical oncologist, gynecologist, surgeon, radiologist, and pathologist led to a favorable outcome. This patient had two primary cancers of the colon and ovary, and systemic metastases of colon cancer. The loss of MSH6 protein expression was proven by immunohistochemical examination, but the germline MSH6 mutation was not detected by DNA sequence analysis. Regarding this discrepancy, some possibilities, e.g., genomic rearrangements and epigenetic modifications, which can be missed by conventional sequence analysis, were considered. Theoretically, Lynch syndrome cases with MSH6 impairment exhibit late onset and low penetrance compared to other major cases with MLH1 or MSH6 mutations. Irinotecan hydrochloride (CPT-11) has favorable effects on MMR-deficient tumor cells with high microsatellite instability, although its clinical benefit remains controversial. In this case, the first-line chemotherapy bevacizumab + FOLFIRI regimen has been effective for over a year in the partial response state. We discuss the diagnostic, therapeutic, pathological, and molecular biological characteristics of this intriguing case, indicating the importance of family history, histological assessment, and molecular biological etiology in Lynch syndrome cases presenting a complicated phenotype.
KeywordsLynch syndrome Family history MSH6 Irinotecan (CPT-11) Multiple cancers
Dr. Ishioka (medical oncologist, chairperson of the conference): Good evening, everyone. Today, we would like to discuss a thought-provoking Lynch syndrome case. (A brief summary of the case is given in the Abstract). Dr. Shiono, please begin the case presentation.
Dr. Shiono (medical oncologist, physician in charge of this case): A 51-year-old woman, diagnosed with advanced colon cancer with multiple liver metastases, was referred to our outpatient department by her primary practitioner for systemic chemotherapy.
To identify potential genes for Lynch syndrome, we submitted a colon cancer biopsy specimen for immunohistochemical (IHC) examination of the DNA mismatch repair (MMR) gene products, i.e., MLH1, MSH2, MSH6, and PMS2. Because Lynch syndrome was diagnosed, we could not completely rule out the possibility of ovarian cancer. An effective chemotherapy regimen should be selected based on the origin of the liver metastases. Hence, we consulted a radiologist and gynecologist for differential diagnoses of the pelvic tumor.
Dr. Ito (gynecologist): Because few solid parts were present, which is often the case with primary ovarian cancer, in addition to normal CA-125 levels, a borderline tumor was conceivable in this case. However, laparotomy and histopathological assessment were necessary for the definitive diagnosis.
Initial treatment plan and its course
Dr. Shiono: Given the patient’s history, we decided to prioritize chemotherapy for colorectal cancer, which had already been diagnosed as malignant. Considering her Lynch syndrome background, we selected an irinotecan hydrochloride (CPT-11)-based bevacizumab + FOLFIRI regimen. After confirming the uridine-5′-diphosphate-glucuronosyltransferase 1A1 (UGT1A1) *6 and *28 status as wild type for CPT-11 use, she was admitted for central venous port implantation for outpatient chemotherapy.
Because we used the biopsy specimens for MMR IHC, we performed colonoscopy to obtain biopsy samples for KRAS gene mutation analysis.
Dr. Takase: The upper colon from the stenosis site at the sigmoid with massive tumor was enlarged. Compared with that in the CT images obtained at the former hospital, the ovarian tumor was also extremely enlarged (Fig. 4b).
Dr. Shiono: We consulted a surgeon for palliative surgery, planned elective operation, excluded bevacizumab to avoid interference with postoperative wound healing, and administered FOLFIRI chemotherapy (l-LV 275 mg, CPT-11 220 mg, 5-FU bolus i.v. 570 mg, 5-FU c.i.v. 3500 mg) once during the preoperative waiting period.
Preoperative clinical diagnosis
Ovarian tumor, borderline tumor suspected
Metastatic liver tumor
Subileus due to mechanical obstruction by colorectal cancer
Dr. Ishioka: Please tell us the operative findings, Dr. Miura.
Dr. Miura (surgeon): First, an infant head-sized multilocular and partially villous right ovarian tumor was seen. The left ovary had shrunk. In the abdominal cavity, the disseminated lesion and a small amount of pale yellow ascites were observed at vesicouterine and Douglas pouches, which were considered to be derived from right ovarian cancer. On the other hand, there was a near circumferential 50-mm tumor in the middle portion of the descending colon. However, serous surface invasion was not recognized macroscopically. Multiple metastatic tumors were observed on the bilateral liver lobe, presenting the so-called state of “tumor liver.” Because of diffuse intra-abdominal adhesions due to peritonitis carcinomatosa (PC) and definitive prognostic factors such as tumor liver or PC, we performed minimally invasive, palliative, and debulking surgery, i.e., descending colectomy, oophorectomy, and liver biopsy.
Dr. Ishioka: Dr. Watanabe, please explain the pathological findings.
Immunohistochemistry of colon and ovarian carcinomas
The features of Lynch syndrome-related ovarian cancer are as follows: young onset (mean age 48 years), early stage (FIGO stage I, 47 %), comparatively frequent serous-type histology (endometrioid 35 %, serous 28 %, clear cell 17 %, mucinous 5 %, undifferentiated 15 %), and high attribution rate of MSH6 deficiency among underlying MMR gene mutations (MSH2 49 %, MSH6 33 %, MLH1 17 %) .
Dr. Ishioka: Please describe the MMR IHC results.
Dr. Ishioka: What were the results of sequence analysis?
Dr. Shiono: Despite the above-mentioned IHC results of the colon cancer specimen, we could not detect any pathogenic germline mutations in MLH1, MSH2, and MSH6 by direct sequence analyses. The mechanism of that divergence was unclear and will be discussed later.
Moreover, the genetic status of KRAS was wild type with regard to inspected codons 12 and 13.
Dr. Miura: That was very interesting. Concerning the differential diagnosis of the pelvic tumor, many organs could be the candidate origin, e.g., colon, bladder, prostate, ovary, and uterus. Lastly, we proposed IHC marker sets as the screening criteria . Although these sets seemed unnecessary in this case because detailed molecular analyses had already been performed, they might be useful in other cases depending on the situation.
Lynch syndrome with MSH6 deficiency
Descending colon cancer (tub2, pT4a, pN1a, pM1b, stage IVB) with multiple metastases to the lymph nodes, liver, and peritoneum
Right ovarian cancer (serous adenocarcinoma, pT1c, cN0, cM0, G1, FIGO stage IC)
Dr. Ishioka: Well, tell us the clinical course after that, please.
Dr. Shiono: The postoperative course was favorable. The first visit day after discharge to restart bevacizumab + FOLFIRI therapy was 11 March 2011. While in the waiting room of the Tohoku University Hospital Cancer Center, the Great East Japan Earthquake occurred. Because she resided in the coastal area, she lost her house in the tsunami. Therefore, she moved to Fukuoka prefecture on Kyushu Island with relatives. Fortunately, she restarted the same chemotherapy at the National Hospital Organization Kyushu Medical Center (Bev 230 mg, l-LV 275 mg, CPT-11 200 mg, 5-FU bolus 560 mg, 5-FU civ. 3000 mg). She has remained in the partial response (PR) state over a year.
Dr. Ishioka: We have a comment from Dr. Takami, who is in charge at Kyushu Medical Center. Please read it for us.
Dr. Ishioka: Let’s move on to the discussion.
Dr. Shiono: First, it was challenging to determine whether the liver metastases originated from the colon or ovary because of the difficulty in the differential diagnosis of the pelvic tumor. In this case, the clinical response to the regimen was favorable, which was consequently in line with the histopathological assessment obtained via palliative surgery. Thus, it is important to select a suitable regimen in accordance with the histology, if possible.
Dr. Ishioka: What about the practical treatment?
Dr. Shiono: In the literature, CPT-11 is clearly effective on MMR-deficient tumor cells in vitro and favorable in some clinical studies, but its clinical evidence remains controversial [9, 10, 11, 12, 13, 14, 15, 16, 17]. Hence, we selected the bevacizumab + FOLFIRI regimen as first-line chemotherapy for the metastatic colorectal cancer. In retrospect, because it has still been effective in the PR state for over a year, the chemotherapy choice seems to be reasonable in this case.
Dr. Ishioka: Please explain the standard first-line chemotherapy for advanced or metastatic colorectal cancer, Dr. Kakudo.
Dr. Kakudo (medical oncologist): There are some options. As first line chemotherapy, we choose FOLFIRI or FOLFOX (5-FU/l-LV/l-OHP) regimens as a combination of cytotoxic agents. Sequential therapy like FOLFIRI or FOLFOX regimen as first line, followed by the alternative regimen as second line, has improved outcome regardless of the order of the regimens . CapeOX, the regimen using the oral prodrug of 5-FU, is another option showing an almost identical outcome compared to FOLFOX . The common adverse events are different: peripheral neuropathy in l-OHP and diarrhea in CPT-11. The last choice is whether to add molecular-targeted agents such as bevacizumab (anti-VEGF antibody drug) or cetuximab/panitumumab (anti-EGFR antibody drugs). You should pay attention to contraindications of these monoclonal antibody drugs, e.g., the comorbid severe vascular problems in bevacizumab use. Patients with the KRAS gene mutation should be excluded from cetuximab/panitumumab administration. We make an optimal decision depending on the circumstances of each case [20, 21].
Dr. Shiono: In this case, CPT-11 was used as a key drug considering the Lynch syndrome background. CPT-11's effectiveness against cancer cells resulting from a MMR deficiency has been demonstrated in in vitro analyses. Although the entire mechanism remains unclear, it is speculated that CPT-11, an topoisomerase-I inhibitor, exerts its cytotoxicity by generating DNA double-strand breaks (DSBs) in the administered cell. Conversely, MMR-deficient tumor cells have a tendency to accumulate mutations within microsatellite repeats of genes associated with DSBs repair, such as MRE11 and RAD50. Taken together, MMR-deficient cells exhibit high sensitivity to CPT-11 [9, 10, 11, 12, 13, 14, 15, 16, 17]. Hence, CPT-11 treatment might be effective in patients with microsatellite instability-high (MSI-H) colorectal cancer.
Dr. M. Takahashi (medical oncologist): In fact, a previous randomized study for adjuvant chemotherapy against stage III MSI-H colon cancers demonstrated the significant advantage of the addition of CPT-11 . However, another subsequent study did not reveal the benefit in a similar population . Thus, the clinical benefit of CPT-11 compared with other agents in MSI-H colorectal cancer remains controversial. Because the mutations in MRE11 or RAD50 are detected in many but not all MSI-H tumors (70–85 %) , the MSI-H phenotype may not always correlate with the hypersensitivity of tumors to CPT-11. Another marker to indicate MRE11 or RAD50 deficiency (e.g., mutational analysis and/or IHC) may help to predict the efficacy of the CPT-11 treatment.
Dr. Ishioka: As both the former trials (CALGB89803 and PETACC-3) were adjuvant trials for stage II/III colorectal cancer [12, 13, 23], the clinical relevance of the outcome was slightly different in this stage IV case. While the evidence level of the advantage of CPT-11 for MSI-H tumor was not sufficient, selecting FOLFIRI among the standard therapies was reasonable according to the concept that the most promising therapy should be given priority. Moreover, when you use CPT-11, you must evaluate the patient for the UGT1A1 gene polymorphism. Dr. Akiyama, please offer a general explanation.
Dr. Akiyama (medical oncologist): CPT-11 is inactivated by UGT1A1. If a specific gene polymorphism exists in UGT1A1, the glucuronidation level of SN-38, the active metabolite of CPT-11, would be attenuated, resulting in drug accumulation and toxicity enhancement; this leads to diarrhea, neutropenia, etc. To be more precise, homozygosity for UGT1A1*28 or UGT1A1*6 and heterozygosity for both UGT1A1*6 and UGT1A1*28 are the polymorphisms mentioned in the package insert of the drug. However, optimal criteria for dosage adjustments have not been established. Moreover, there are some differences among ethnicities. In Asians, UGT1A1*6 is more frequent than UGT1A1*28. Conversely, UGT1A1*28 is much more common than UGT1A1*6, which is quite rare in Caucasians and African-Americans. Such discordance is derived from the different genetic background among the races [24, 25, 26].
Dr. Ishioka: With regard to dose, 150 mg/m2 is defined as the maximal dose in Japan, although 180 mg/m2 is the standard in Europe and the US. Accordingly, data from overseas cannot be used for direct comparisons. Many research groups, including ours, are working on this topic, and an appropriate criterion for the Japanese people needs to be established. Well, let us get back to this case. What about MSI in this case? Would you explain the reason, if you did not check?
Dr. Shiono: We obtained positive IHC results, and therefore, we did not perform an MSI examination. IHC is the best initial examination because it directs the candidate gene for subsequent mutation analysis in families with a high probability of having a mutation (the revised Bethesda guidelines or Amsterdam II criteria) . Moreover, the latest analysis on the accuracy and cost-effectiveness of IHC and/or MSI examination to screen for Lynch syndrome  promotes the following strategies: “IHC and MSI performed simultaneously” and “IHC followed by MSI if IHCs were normal.” The latter was slightly better in terms of cost. Therefore, IHC seems to be sufficient if it is performed first. According to this strategy, if IHC demonstrated the candidate mutated gene, you can skip MSI and proceed to direct sequencing. IHC has an advantage in terms of specifying the putative mutated MMR gene compared with MSI .
Dr. Ishioka: OK, so it is reasonable. However, how do you explain the discrepancy between the results of IHC and sequence analyses?
Dr. Shiono: As seen in Fig. 7b, nuclear staining of MSH6 alone was lost compared with that of the other three MMRs. Some possibilities were considered. For example, it is known that genomic rearrangements such as large deletions cannot be detected by conventional sequence analysis [30, 31, 32], actually in a significant proportion of Lynch syndrome families (5–20 %) [33, 34, 35, 36]. Otherwise, it may be a type of epigenetic modification such as methylation . However, further molecular analyses, e.g., the multiplex ligation-dependent probe amplification (MLPA) test, are needed for elucidation [31, 33, 38].
Dr. Ishioka: What about care for the families because this is a hereditary syndrome?
Dr. Shiono: Complying with the guidelines [39, 40], we performed a genetic counseling series for the patient, and her sister wished to accompany her. Her siblings shared the information, recognized the importance of medical follow-up, and have begun to undergo annual screening examinations, including colonoscopy. You can refer to the surveillance recommended by the international collaborative groups [27, 41]. However, a study indicated that the screening recommendations for MSH6 mutation carriers may slightly differ from those for Lynch syndrome carriers as a whole, reflecting the characteristics of MSH6-mutated Lynch syndrome [42, 43]. The weaker phenotype, which is observed as a result of MSH6 mutations, exhibits a later age of onset and lower penetrance compared with that observed as a result of MLH1 or MSH2 mutations [44, 45]. Many types of cancer should be considered in regard to an increased risk, e.g., cancer of the colorectum, endometrium, stomach, small intestine, hepatobiliary system, kidney, urinary bladder, brain, and ovary. The latest prospective study showed that pancreatic and breast cancers had an elevated risk .
Dr. Ishioka: Finally, what is the discriminative point in this case compared with other Lynch syndrome cases?
Dr. Shiono: In general, approximately 90 % of Lynch syndrome cases with mutations in any MMR genes are attributed to MLH1 or MSH2 mutations with distinct clinical features such as early onset (<50 years) and proximal colon predominance [47, 48, 49, 50, 51, 52]. In contrast to these characteristics, it might have been difficult to diagnose Lynch syndrome in this case without a definitive family history. Moreover, with respect to comorbid cancer, while the frequency of endometrial cancer is as high as 60–70 %, the frequency of ovarian cancer is only 7–10 % . Hence, clinical information on ovarian lesions might be relatively less likely to indicate Lynch syndrome. As mentioned above, although the incidence of Lynch syndrome attributed to MSH6 mutation is as low as approximately 10 %, it is known to show “relatively late onset” and “low penetrance” propensity compared with MLH1 or MSH2 mutations [42, 44, 45]. Thus, judging by only clinical manifestation may lead to a diagnostic pitfall. To avoid misdiagnosis of Lynch syndrome, considering a family history is always critically important.
Dr. Shimodaira: Concerning the unique phenotype of MSH6-deficient Lynch syndrome, the mechanism can be understood when the molecular function of the four MMR proteins is considered. First, they function as heterodimers formed by MSH2 in association with MSH6 (MutSα) or MSH3 (MutSβ) and MLH1 interaction with PMS2 (MutLα), respectively. As seen in these complexes, the contribution of MSH6 is relatively small compared with that of major players such as MLH1 or MSH2, which interact with many gene products. In fact, MSH6 functionally participates only in detection of single-base mismatch or small loop-out mutations, while MLH1 or MSH2 engages in widespread mismatches other than single-base abnormalities. Thus, the loss of MSH6 function involves only a partial deficiency of the MMR system, and subsequently it results in an attenuated clinical phenotype, which is approximated to conventional colorectal cancers with regard to late onset and low penetrance, compared with those caused by MLH1 or MSH2 deficiency [54, 55].
Dr. Shiono: In conclusion, this was a very intriguing discussion on diagnostics, treatment, pathology, and molecular biology. It is also dramatic that she was saved from the tsunami, which deprived her of her house on the seashore, by an occasional visit to our hospital during the Great East Japan Earthquake. She has also been spared from life-threatening disease progression by treatment based on the cooperation of many doctors. I appreciate all your kind collaborative work.
Dr. Ishioka: The first-line bevacizumab + FOLFIRI treatment exerted a pronounced effect on this metastatic case of MSH6-mutated Lynch syndrome. The population of Lynch syndrome cases with metastasis is too small to organize a large-scale randomized prospective trial in order to elucidate CPT-11 effectiveness. However, the prevalence of MSI among all colorectal cancers is approximately 15 % [56, 57]. Therefore, it might be possible to conduct a clinical trial by alternatively targeting similar types of cancers. Thus, further analysis is needed to elucidate the clinical benefit of the drug. Are there any questions? Then, this conference is adjourned. Thank you for your attendance.
What we learned from this case conference
You must always collect detailed information regarding family history in order not to overlook familial tumor syndromes.
You should know that weaker phenotypes, such as “late onset” and “low penetrance,” compared to MLH1- or MSH2-mutated Lynch syndrome, can be observed because of MSH6 deficiency.
A histopathological diagnosis must be obtained as soon as possible before deciding on an optimal regimen for patients with multiple primary cancers.
Although it is controversial at the clinical level and requires further study, a CPT-11-based regimen may have favorable effects on Lynch syndrome cases, depending on MMR deficiency.
Conflict of interest
Honoraria: C. Ishioka, Taiho Pharmaceutical Co. Ltd.; C. Ishioka received research funding form Chugai Pharmaceutical Co., Ltd.