Evaluation of a colorectal cancer screening program composed of successive waves of different tests: The experience of the French Calvados County

1Introduction

Colorectal cancer (CRC) is a worldwide public health problem. Every year, 1.9 million CRCs are diagnosed, and 900,000 persons die from this disease [1] . Screening strategies are known to diagnose CRC at an early stage. Yet, an early diagnosis could artificially increase the overall survival without reducing specific mortality [2] . However, screening strategies have been shown to reduce specific mortality, avoiding the effect of a lead-time bias. These strategies also reduce the incidence of CRC by detecting and removing advanced precancerous lesions 34567 .

Noninvasive CRC screening tests are based on occult blood detection in stools. The gaïac-Faecal Occult Blood Test (gFOBT) (Hemoccult II®, SKD, France), which is based on the detection of peroxidase activity, was the first test employed in an average-risk population [8] . Several countries have replaced gFOBT with automated quantitative faecal immunochemical tests (FITs), which directly detect human haemoglobin through specific monoclonal or polyclonal antibodies [9] . Magstream (Fujirebio, Japan) has been demonstrated to be better than Hemoccult II 101112 . An OC Sensor (Eiken Chemical Co, Japan) has also shown improved sensitivity for CRC detection [1314] . Given its enhanced accuracy, OC Sensor is currently used in the French screening program [1215161718 . An increase in the participation rate was also expected due to the reduction in the number of samples needed (only one with FIT, instead of six with gFOBT). Despite the screening program, most of the CRCs diagnosed in average-risk patients 50–74 years of age are still detected through nonscreening measures [1920] . This problem is mostly due to insufficient screening participation. In France, the screening participation rate is approximately 32 % [21] , which is less than the 45% threshold recommended by European guidelines [2223] . Furthermore, some patients develop interval cancers (IC) between successive screening rounds [24] . Notably, such cancers seem to be less frequent after FIT rather than gFOBT [25] . Furthermore, some patients can develop postcolonoscopy IC [2026] . Finally, a few patients with a positive test result will not undergo any colonoscopy, leading to a delayed CRC diagnosis [20] . All these problems raise the question of screening efficiency in real practice.

This study aimed to compare the performances of several CRC screening tests using real-life data from a French area covered by a cancer registry where different tests have been used.