Statistics from Altmetric.com
All papers from the first issue are well before my time, but I had no problem making my choice. Going through the index, my eye was immediately caught by a name (Dukes), and seeing the title of the paper (Familial intestinal polyposis) I realised this had to be my pick. With today's knowledge, it was fun to look back and see what at that time was already known and what may have been wrong or incomplete.
Well, nothing much was wrong. The paper1 gives an accurate description of the first data on families with polyposis and colorectal cancer. It is explained that in some families multiple polyps occur, strictly confined to the colon, and that this predisposes to colonic cancer. Several statements still hold true: the young age at diagnosis of cancer, the “tendency for malignant disease to begin at an earlier age period in each succeeding generation,” the need for “constant medical supervision,” and that “no other method of treatment than surgery has any permanent effect on the course of the disease.” However, the mentioned “waiting game,” where “a radical excision may be undertaken as soon as malignancy is detected,” has been replaced by preventive colectomy at a certain age, which was at that time apparently not an option. Interestingly, although the relation between the polyps and subsequent cancer is made clear, it is never explicitly stated that the cancers in fact usually occur in the polyp.
Several statements are vague or incomplete in the light of today's knowledge, such as those on the pattern of inheritance. It is stated that both males and females are affected, that transmission can be by the female or the male, and that about half the family members are affected, pointing to an autosomal dominant pattern. In this respect, the remark that “polyposis may possibly result from the chance mating of two individuals, each of whom was destined to die of cancer of the intestine” sounds somewhat weird. The observation that in two of 10 St Mark's Hospital families studied the disease seemed to start with both father and mother, who died of cancer of the rectum or colon, can probably be explained by the chance mating of two carriers, leading to an excessive number of affected individuals in this family.
Although all the patients described in the paper are considered to have the same syndrome, knowledge has progressed and we now discern several syndromes. From the description of the patients' histories, it seems that most of them had a disease which we would now designate familial adenomatous polyposis (FAP). However, the four very young patients would now be considered to have juvenile polyposis, and the patients from the original Warthin paper which was cited by Dukes (cancer family “G”) were later, of course, found to have a different disease which we now call hereditary non-polyposis colorectal cancer (HNPCC) syndrome, based on a germline mutation in one of the genes coding for the DNA mismatch repair apparatus.2
It is stated that “members of polyposis families show no special proclivity to the development of tumours in any other organ of the body than the intestine,” which we now know not to be true. Several phenotypical variants of this inherited disease caused by a germline mutation in the APC gene exist. Gardner's syndrome (polyposis, epidermoid cysts, and osteomas), a proportion of the cases with Turcot's syndrome (polyposis and brain tumours), and AFAP (attenuated familial adenomatous polyposis—that is, APC mutations with less than 100 polyps and colorectal cancer at a later age than in “common” FAP) all belong within the spectrum of FAP. In addition, since many FAP patients do not now die from colorectal cancer, other tumours appear, such as duodenal polyps, desmoid tumours, and thyroid and hepatocellular cancers. HNPCC also involves an inherited predisposition to cancer which is certainly not limited to the large intestine. Endometrial, ovarian, gastric, and urothelial cancers often occur in patients with HNPCC, as well as cancers at other sites. The combination of sebaceous gland tumours and an internal malignancy is known as Muir-Torre syndrome. Based on the presence of microsatellite instability, it has been suggested that a proportion of Muir-Torre syndrome cases would be variants of HNPCC.
Nevertheless, there is no doubt that this review from over 50 years ago is in retrospect a seminal paper, which has to a large extent withstood the course of time. Since it was written, St Mark's Hospital has developed into one of the leading institutes for follow up studies of hereditary and sporadic polyps.
I am grateful to Dr Gerrit Meijer for critically reading the manuscript.