CLINICAL IDENTIFICATION AND LONG-TERM SURVEILLANCE IN 22 HEREDITARY NON-POLYP- OSIS COLON CANCER (HNPCC) ITALIAN FAMILIES

Background and aim. HNPCC is an autosomal dominant cancer susceptibility syndrome characterised by an increased incidence of colorectal (CRC) and extracolonic cancer (EC). The aim of this study was to assess the efficacy of a HNPCC identification and surveillance policy over 8 years. Material and methods. Familial clustering of CRC and other neoplasms and correspondence to Amsterdam criteria for HNPCC was investigated in 1520 consecutive CRC patients with a family history of CRC. HNPCC families were identified and individuals at risk were advised to take part in a surveillance programme, based on colonoscopy, oesophagogastroduodenoscopy, ultrasonography and other examinations performed at 2-year intervals, starting from 25 years. Results. Twenty-two HNPCC families, with 435 at risk individuals (104 deceased and 331 alive), were identified. None of these families was previously classified as HNPCC. The overall prevalence of CRC was 27.8% (121/435), decreasing from 59.4% in the I evaluable generation, to 24.4 and 8% in the II and III generations, respectively. Multiple CRC were reported in 29 patients. Thirty-four patients with EC were recorded in 12 out of 22 families (Lynch II families). One hundred and ninety-nine patients out of 331 at risk individuals accepted surveillance, the remaining ones were traced at the end of the study. The mean follow-up duration was 48+32 months (range 12–84 months). The overall incidence of CRC was 6/199 (3%) in surveyed patients and 5/132 (3.7%) in unsurveyed patients (P = ns). The mean annual incidence was 0.89% subjects at risk. Four surveyed individuals had CRC detected at first surveillance colonoscopy; in two of them CRC were metacronous appearing 7 and 11 years after a previous CRC. Two individuals developed CRC during follow-up, after a previous negative surveillance colonscopy. CRC detected in surveyed patients were significantly less advanced than those diagnosed in symptomatic, unsurveyed patients. Seven surveyed and six unsurveyed individuals, all belonging to the 12 Lynch II families, had EC during the study period (endometrium, small bowel, thyroid, central nervous system). Moreover, anterior pituitary adenoma was found in two unrelated patients. Twenty-two adenomas in 11 surveyed individuals were found. Tubulo-villous adenomas with high grade dysplasia was removed in one patient. All patients with CRC detected by surveillance are alive. One of the CRC patient in the group of those who refused surveillance died 18 months after the diagnosis. Conclusions. Family history collection in each patient with colorectal cancer leads to identification of often misdiagnosed HNPCC families which, otherwise, could escape an adequate surveillance policy. CRC incidence in at risk subjects was 20 times greater than the incidence reported in our region. Data support the efficacy of repeated colonoscopies for early diagnosis and prevention of CRC in at risk members. Longer follow-up is required to assess the efficacy of surveillance for EC.