Familial Cancer

, Volume 12, Issue 2, pp 279–284

Surveillance for urinary tract cancer in Lynch syndrome


    • The Danish HNPCC-register, Department of Gastroenterology and Clinical Research CenterCopenhagen University Hospital
    • Department of Surgical GastroenterologyAalborg University Hospital
  • Torben Myrhøj
    • The Danish HNPCC-register, Department of Gastroenterology and Clinical Research CenterCopenhagen University Hospital
Original Article

DOI: 10.1007/s10689-013-9634-y

Cite this article as:
Bernstein, I.T. & Myrhøj, T. Familial Cancer (2013) 12: 279. doi:10.1007/s10689-013-9634-y


Hereditary non-polyposis colorectal cancer (HNPCC) is an inherited multiorgan cancer syndrome, which when caused by a germline mutation in the mismatch repair (MMR) genes is known as Lynch syndrome (LS). Mutation carriers are at risk for developing cancers primarily in the colon, rectum and endometrium, but also other extra-colonic cancers. Urinary tract cancers (UTC) have in many studies been reported increased in LS and it has been discussed among researchers and clinicians whether or not screening for urological tumours should be included in the surveillance programme and if so what screening procedures are justifiable. The aim of this review was to elucidate the present knowledge from the literature on the risk of UTC in LS and highlight the pros and cons of screening for asymptomatic neoplasia in the urinary tract. The review is based on a systematic literature search in PubMed database followed by a reference list of retrieved articles and manual searches of further relevant articles. In conclusion there is a moderate increased risk of UTC in LS, but a tremendous lack of knowledge on which screening programme, if any at all to establish, and if so what procedures and time intervals are appropriate. It is recommended that all eventually screening for UTC in LS, only should be performed in clinical trials or with a systematic reporting to a HNPCC-register for future evaluation.


HNPCCLynch syndromeUrinary tract cancerScreening



Colorectal cancer




Hereditary non-polyposis colorectal cancer


Lynch syndrome


Mismatch repair genes, MLH1, MSH2, MSH6, PMS1 and PMS2


Microsatellite instability


Urine cytology




Urinary tract cancer

Characteristics of Lynch syndrome

Hereditary non-polyposis Colorectal Cancer (HNPCC) is characterised by the development of colorectal cancer (CRC), endometrial cancer and other extra-colonic cancers. The terminology regarding the syndrome has changed over the time and today it is generally accepted to use the name “Lynch Syndrome” (LS) for families with strong evidence of mismatch repair (MMR) deficiency [1].

Carriers of an MMR germline mutation have a high lifetime risk of developing various types of cancers and often at an unusually early age and frequently also multiple cancers. Due to the defect in the MMR-system, DNA in the tumours from Lynch Families often show Microsatellite Instability (MSI), a feature found in >90 % of the CRC compared to 15 % of the sporadic CRC. Immunohistochemical (IHC) analysis of MMR-protein expression is another method to detect a defect in the MMR-system by on tumours using antibodies against the corresponding protein of the MMR genes. Loss of protein expression indicates a possible mutation in the corresponding gene and thereby the mutation analysis could be directed towards the specific gene. There is good correlation between the two tests regarding the efficacy to detect families with LS. In prospective studies a slightly better sensitivity has been found using MSI mostly based on examination of colon cancer tissue. But taking into account that IHC is a must cheaper and easier performed test, which even can be performed on a preoperative biopsy and be a part of the decision-making on operation type (segmental resection or subtotal colectomy) to choose, this method have some advantages as the primary screening test for LS compared to MSI [1, 2]. Identification of families carrying a MMR mutation is important, as colonoscopic surveillance in these individual leads to detection of CRC at an earlier stage and to reduction in frequency and mortality of CRC, due to suggestive removal of adenoma [37].

Urinary tract cancers in Lynch syndrome

Several epidemiological studies have demonstrated that Urinary Tract Cancers (UTC) are integrated in LS with the highest lifetime risk for male MSH2-mutation carriers [6, 816]. The majority of the studies find the increased risk of urothelial cancers in the upper urinary tract i.e. the renal pelvis and the ureter, with a lifetime risk varying from a few percent to twelve and UTC in total rank third (5 %) after colon and endometrial cancer in LS [1, 11, 12, 17]. Upper urinary urothelial cancer are rare in western countries with an annual estimated incidence of one or two new cases per 100.000 inhabitants and according to a French national database a significant proportion of patients (21 %) with newly diagnosed upper urinary tract urothelial cancer may have HNPCC [18]. Both Watson et al. and Vasen et al. found the highest risk of UTC in MSH2-mutation carriers, especially in males over the age 50–70 years.

Not all studies differentiate clearly between renal pelvis cancer (urothelial derivates) and kidney cancer (renal cell cancer) when referring to upper urinary tract cancer and therefore the conclusion to what extent kidney cancer is part of LS is difficult to draw [911]. Disagreements exist regarding whether or not carcinoma of the kidney and bladder cancer is part of LS. From molecular analysis with tumour profiling it has been suggested that urothelial cancers of ureter and to a lesser extent the bladder share many characteristics of MMR deficiency driven tumourigenesis, whereas kidney cancer follow a separate pathway and seldom show MSI [1921]. Data from the Swedish Family-Cancer database have found increased risk of renal pelvis cancer but not kidney cancer in families fulfilling the Bethesda Criteria [22]. This is in concordance with a recent study of Aarnio et al. [15] that did not confirm an increased risk of kidney cancer in 974 Finnish mutation carriers from LS families. Formerly an increased risk has been suggested in other studies including also data from Finland [14, 23].

Bladder cancer is far the most frequent UTC in the general Western population and several of these are due to exposure of external causative agents such as smoking or occupational chemicals. The association of bladder cancer and LS is controversial. It has been estimated that the frequency of bladder cancer in families fulfilling the Bethesda criteria are comparable to the background population in the Swedish family cancer database [22]. Based on data from families with proven MMR deficiency urothelial bladder cancer is increased in LS, in particular in males carrying a MSH2 mutation [9, 23, 24].These findings are in contrast to other studies where no increased risk of bladder cancer for any of the germline MMR mutations was observed [12, 13].

Regarding age of onset for UTC in LS, there seems to be a 10–15 years of earlier onset for ureter and bladder cancer, whereas the corresponding young age of onset for kidney cancer is not confirmed [10, 24]. Table 1 summarises the lifetime risks of UTC as reported in the literature in HNPCC.
Table 1

Lifetime risk of urinary tract cancer in Lynch syndrome or HNPCC

Author, year

No families

No individuals





Type of study

Category of evidence level

Watson P 2008

Lynch Syndrome and HNPCC



Overall lifetime risk of UTC of 8.4 %

Lifetime risk in male MSH2 28 %

Retrospective Multiple HNPCC-registries study


Van der Post 2010

Lynch Syndrome



Not assessed

Not assessed

Not assessed

Relative Risk male 4.2 female 2.2

Cumulative Risk (age 70 years) male 7.5 % female 1.0 %

Retrospective Deutch HNPCC-registry study


Vasen H 2001

Lynch Syndrome



Cumulative Risk (age 70 years)

MLH1 1.3 %

MSH2 12 %

Retrospective Deutch & Norwegian HNPCC-registries study


Win 2012

Lynch Syndrome


446 carriers vs 1.029 non-carriers

Median 5 years follow-up Standardized Incidence Ratio 11.2

Not assessed

Median 5 years follow-up Standardized Incidence Ratio 9.5

Retrospective Multiple HNPCC-registries study


Sijmons 1998




Not increased

Relative Risk 14.04

Cumulative lifetime Risk 2.6 %

Not increased

Retrospective Deutch HNPCC-registry study


Geary 2008

Lynch Syndrome



Relative Risk

MLH1 0

MSH2 7.4

Relative Risk MLH1 0 MSH2 9.8

Relative Risk MSH2 3.6

Retrospective Regional (london) HNPCC-register study


Barrow 2012

Lynch Syndrome


821 male

Not assessed

Relative Risk MLH1 38.3 MSH2 42.7 MSH6 0

Relative Risk MLH1 57.2 MSH2 61.9 MSH6 0

Not increased

Retrospective Regional (Manchester) Lynch Syndrome database study


Aarnio 1999

Lynch Syndrome



Standardized Incidence Ratio 4.7

Standardized Incidence Ratio 7.6

Retrospective Deutch HNPCC-registry study


Screening for urinary tract cancer

According to the WHO several aspects should be considered ahead of introducing screening for cancers. The success depends on having sufficient personnel and facilities to perform the screening tests and the subsequent diagnosis, treatment, and follow-up. The screening aims to insure that as few as possible with disease get through undetected (high sensitivity) and as few as possible without disease should be subject to further investigation (high specificity). The likelihood a positive screening will give a correct result (positive predictive value) is strongly depending on the prevalence of the disease in the population. Even the best screening programme will not be effective if the prevalence of the disease is very low. The disease should be common, with high morbidity and mortality and the treating modalities should be acceptable, safe and relatively inexpensive and agreement on guidelines needs to be applied. http://www.who.int/cancer/detection/variouscancer/en/index.html (17-10-2012).

Before establishing screening for detection of UTC in LS the risk and benefits of the different screening methods, their limitations, costs and eventually lack of efficacy has to be considered with respect to the specific cancer type and corresponding risk.

Currently there is no perfect test for detection of bladder cancer. Cystoscopy remains the “gold standard”, but has its limitation as screening procedure due to the invasive character and costs together with the discomfort for the patients. For screening of bladder cancer patients in general many non-invasive urine analysis have been developed and tested, including analysis for microscopic hematuria (dipstick), urine cytology (UC) and tumour-specific molecular markers. Development and testing of the analysis have not been specific addressed on screening in LS. In a systematic review assessing sensitivities and specificity Rhijn et al. found Microsatellite analysis, ImmunoCyt, NMP22, CYFRA21-1, LewisX and Fish the most promising markers and Gossman et al. [25, 26] found NMP22 to be superior to urine cytology regarding sensitivity, but still both authors concluded the evidence insufficient to substitute cystoscopy in follow-up scheme for bladder cancer.

Screening for neoplasia in ureter, pelvis and kidney will in addition to urine analysis include visualization of the upper urinary tract by either by CT-/MR-scan or ultrasound (US). Both CT- and MR-scan are important as a radiological test for characterizing and staging UTC, but as a screening procedure less suitable as minor lesion easily is overseen and regarding CT-scan the repeated irradiation and risk of introducing neoplasia. The intension of screening is detection of cancer at an early stage. A substantial proportion of lesions under 1 cm in the pelvis/kidney were not detected with neither CT-scan nor US in a study of Jamiss-Dow and regarding small tumours (<3 cm) neither CT nor US was superior in characterization, but with the CT depicting substantially more and smaller masses than US [27].

Based on the knowledge of the higher lifetime risk of UTC in LS compared to the general population, pros and cons regarding recommending the clinicians to establish screening for UTC has been discussed during the last 20 years and several authors had suggested screening [1012, 15, 2831]. Due to the higher risk especially among MSH2 mutations carriers it has been suggested to focus screening on these families in a clinical trial setting [10, 11].

Only two studies have systematically analysed the outcome of an established UTC screening program in HNPCC patients [31, 32]. In Denmark UC has been introduced as a national screening program for all HNPCC-patients including LS patients. The advantage of UC is the non-invasive character of the procedures and the relatively few economic costs, which made the national implementation rather simple. Myrhøj analyzed the outcome of biennial UC in 977 persons from LS, Amsterdam I/II and Amsterdam suspected families in a study period from 1991 to 2005 and found UC useless as screening procedures. The sensitivity of diagnosing asymptomatic UTC was only 29 % and just in 0.1 % of the examinations lead to detection of an asymptomatic UTC. Additionally in ten times as many procedures, there was a false positive screening result leading to unnecessary invasive diagnostic procedures in unaffected, non-symptomatic individuals. Later Zachhau studied the outcome of screening in a single urologic department following 20 HNPCC-patients prospectively with a more intensive and invasive program including one initial contrast CT-scan followed by biannual flexible cystoscopy and UC based on immediate urine lavage. Three patients left the programme prematurely and three died of cancer. In total 26 CTs and 48 flexible cystoscopies with UC were carried out and two patients with asymptomatic cancers in the ureter, both of them had a MSH2-mutation, were found. There were no false positive procedures in this study. The overall conclusions in both studies were that evidence for systematically screening in LS is lacking and they indicated the need for further investigation into the effectiveness of different screening methods to find the optimal screening procedure and population to offer screening.

Very little is known about surveillance and outcome hereof for UTC in LS. Therefore, and probably due to the fact that risk of UTC is not extremely high, no standard screening recommendations have been established internationally and the decision of screening is often based on an empirical approach. A collaborative group of European HNPCC-experts, The Mallorca-group [1], have recommended urine analysis with dipstick and UC together with abdominal US annual to biennial from age 30–35 years, if UTC runs in a LS-family, defined as two or more cases of UTC in the family. Lindor and other American HNPCC-experts recommend annual to biennial screening from age 25–35 years in all LS patients with dipstick and UC and Koornstra annual surveillance for hematuria by urine dipstick [29, 30]. All of the recommendation is given with the precaution, that quality of evidence and strength of recommendation is insufficient even though there is a clear association with LS and UTC. Due to the lack of well defined strategies Acher, being a specialists in urology, has proposed surveillance based on risk stratification of the LS families into risk groups with tailored screening programmes. Low risk with no personal or family history of UTC and no MSH2-mutation should be offered UC, NMP22 and dipstick annually, intermediate group with family history of UTC or an MSH2-mutation should in addition have abdominal US and the high risk group with a personal history of UTC should in addition have contrast abdominal CT-scan and flexible cystoscopy [28]. Problems with all these guidelines are that many are country specific and in the majority positive results of the screening programmes are lacking and therefore the most effective and safe surveillance protocols are still unknown. In Table 2 is an overview of publications regarding pros and cons of screening for UTC in LS.
Table 2

Screening for urinary tract cancer in Lynch—recommended program—evidence level

Author, year


Type of study—category of evidence

Grading of recommendations

Vasen 2007

Urine cytology, dipstick and abdominal ultrasound Interval 1–2 years from age 30 to 35 years

If UTC runs in the family (two or more cases)

Experts opinion—IV

Review by a group of European experts on HNPCC/Lynch


Lindor 2006

Urine cytology

Interval 1–2 years from age 25 to 35 years

Experts opinion—IV

Review by a group of American experts on HNPCC/Lynch


Myrhøj 2008

Urine Cytology is not a proper method of screening for UTC in HNPCC

Non-experimental descriptive study—III

Retrospective analysis of a national screening-program with urine cytology in 977 HNPCC-patients in Denmark


Koornstra 2009


Interval annually from age 45–50 years

Experts opinion—IV

Review by a group of specialist at Groningen Hospital, Netherland


Acher 2010

Risk stratification of surveillance

Low risk: no personal or family history of UTC or MSH2 mutation—annual urine cytology, NMP22 and dipstick Intermediate risk: family history of UTC or MSH2 mutation—annual urine cytology, NMP22, dipstick and ultrasound

High risk: personal history of UTC—annual urine cytology, NMP22, dipstick, CT-scan and flexible cystoscopy

Experts opinion—IV

Review by a specialist in urology


Zachhau 2011

Screening should be performed

No specific program advised—further studies needed

Non-experimental descriptive study—III of screening in 20 HNPCC-patient with contrast CT-scan initially followed by flexible cystoscopy with urine cytology biennial


Aarnio 2012

No specific program advised—further studies needed

Non-experimental descriptive study—III of UTC and the method of diagnosis in 30 Lynch patients


Discussion and conclusion

Even though screening for UTC in LS-families has been included in several guidelines during the last decades, there is a huge gap in the current knowledge whether or not these recommendations are appropriate.

The risk of any UTC in LS is not very high, with the highest occurrence in male MSH2-mutation carriers at age 50–70 years. No evidence of an appropriate screening program has been established and only two studies have evaluated the outcome of a screening programme. None of them have the answer to the question of whether screening for UTC in LS is recommendable.

Urine analysis for microscopic hematuria, urine cytology and tumour-specific molecular markers has the advantages of being non-invasive and relatively inexpensive, but the efficacy of these in LS is unknown and UC alone as screening procedure is not recommendable due to very low sensitivity and relatively many false positive results.

The efficacy of the more invasive procedures like CT and cystoscopy as screening procedures in LS is also undetermined. Given the morbidity of the procedures with risk of irradiation damage and screening introduced infections along with the rather expensive character and patients discomfort, screening programmes using CT and cystoscopy should not be introduced generally, but maybe they are usable as screening in selected cohorts of LS patients in clinical trials with the precautions of limiting the irradiation. As abdominal ultrasound is non-invasive, it has the potential as screening procedure, but the efficacy as screening in LS is unknown.

Before introducing any screening procedures for detection of UTC in LS, it should be considered that this may give the LS-patients false sense of security. A better approach may be to increase the information to the LS patients on optional symptoms on UTC in order for the patients to seek medical attention if such should occur. Another challenge for cancer geneticist’s and policy makers of guidelines for screening of high risk families is to update and educate medical health care professionals on the potential risk of UTC in LS.

It is recommended that all eventually screening for UTC in LS, only should be performed in clinical trials with systematic report to a HNPCC-register for future evaluation.

Copyright information

© Springer Science+Business Media Dordrecht 2013