Hereditary nonpolyposis colorectal cancer (HNPCC) is a dominantly inherited syndrome that is estimated to be responsible for between 0.5% to 5% of all colorectal cancers.[1] The syndrome is caused by germline mutations in any of at least four mismatch repair genes.
In one study of HNPCC, mutations in the MSH2 gene were found in 44% of the affected families, MLH1 gene mutations in 23% of families, and PMS1 and PMS2 gene mutations in 6% of families.[2] In 27% of the affected families, mutations could not be ascertained. [See Oncology News International, April 1997, for a review of HNPCC genetics.]
In carriers of the mutated genes, colo-rectal carcinoma develops typically before the age of 50. The tumors appear in the proximal colon more often than in patients with sporadic colon cancer. The lifetime risk of cancer in carriers is close to 100%.[3]
Since the 1970s, there have been reports of a possible survival advantage for colorectal cancer in HNPCC patients.[4-6] In 1981, Henry T. Lynch and colleagues hypothesized "that the genotype in hereditary cancer determines both susceptibility and natural history, such that increased survival is a manifestation of the natural history."[5]
A recent study from Finland provides strong evidence in favor of this hypothesis.[7] The study included 175 patients with colorectal carcinoma from 39 families fulfilling the Amsterdam criteria of HNPCC.[8] Of these patients, 120 came from 22 families with three different germline mutations in the MLH1 gene.
The patients were those diagnosed before the age of 65. Older patients were excluded because of the possibility that they could be sporadic cases among family members not carrying the germline mutation.
The patients were diagnosed between 1953 and 1993. The cancer diagnoses were ascertained from the files of the nationwide, population-based Finnish Cancer Registry. All the rest of the 14,261 colorectal cancer patients diagnosed in Finland in the same age group and calendar years were used as controls.
Utilizing Finland's Central Population Register, the patients were followed until death, emigration, or the common closing date of the study, December 31, 1993, whichever came first. Since everyone in Finland has a unique personal identification number, none of the patients were lost to follow-up.
Avoiding Potential Biases
To avoid confounding by competing causes of death, cumulative relative survival rates were calculated by dividing observed rates by expected rates. The expected rates were derived from sex, age, and calendar-period-specific life tables of the general Finnish population.
All available prognostic factors, ie, disease stage, primary site, age and sex of the patient, and calendar year, were simultaneously adjusted for by calculating relative excess risk of death using a life table proportional hazards model based on annual relative survival rates for the first five follow-up years.
Using these methods, most of the potential biases present in previous studies could be avoided. The HNPCC patients were collected retrospectively from the whole country as they had entered the health care system. In the 1950s through 1970s, HNPCC was not recognized in clinical practice, and thus there was no detection bias. Even in the 1980s through 1993, the proportion of HNPCC patients from all colorectal cancer patients did not increase.
Since the follow-up was complete, all deaths, as well as healthy person-years, were taken into account for both HNPCC and sporadic colon cancer patients.
The expected survival rates were based on the same homogeneous Finnish population from which the patients came. Finally, for the majority of patients, the HNPCC diagnosis was made on a molecular basis at the family level so that results were much more reliable than if diagnosis had been based on the Amsterdam criteria alone.
The cumulative relative survival rates were uniformly better for the HNPCC patients in all strata that were analyzed. The five-year relative survival rate was 65% for all HNPCC patients, compared with 44% for all control colorectal cancer patients.
For the 120 patients from families with germline MLH1 mutations, the relative survival rate was better than for the 55 purported HNPCC patients with no data on germline mutations who were identified using the Amsterdam criteria alone.
The relative survival rates among patients with localized tumors were 85% and 68%, respectively, for HNPCC and sporadic patients. The corresponding figures among patients with nonlocalized tumors were 40% and 18%, respectively.
The relative survival rates were better among HNPCC patients than among controls in age groups 0 to 44 years and 45 to 64 years, in men and women, in patients with tumors of the colon and the rectum, and in all four calendar periods.
Several Hypotheses
Recent advances in molecular genetics have revealed many details of the underlying mechanisms and pathways involved in HNPCC (for a review, see references 1 and 2). However, the reasons for the evident survival advantage remain obscure. Several interesting hypothesis have been put forward.
Despite aggressive histologic tumor features in HNPCC patients,[9] the carcinomas are diploid predominant, indicating better survival.[10,11] The numerous mutations occurring in the tumor cells might trigger the host immune system defenses.[12] Rhyu discusses the possibility that due to the multitude of mutations, the loss of vital functions would drive the tumor cells toward premature death rather than immortality.[2]
Additional DNA repair mechanisms might also be involved.[1] One such alternative could be the p53-related apoptosis pathway.[13]
Until further research provides more detailed understanding of the complex mechanisms and pathways in the tumorigenesis and natural history of HNPCC tumors, the original hypothesis by Lynch from 15 years ago is still valid.[5,14] The same genetic defect that is responsible for tumor initiation and progression may also reduce the viability of cancer cells.[7]
At the moment, there are no reliable clinical indicators of good prognosis in HNPCC-related colorectal carcinomas. Thus, clinical screening and treatment protocols must be based on the strictest criterion: To improve the survival rates of all patients.
