Daniel Satgé

A tumor profile in Down syndrome

We conducted a review of cancers in Down syndrome (DS), because solid tumors are poorly understood in DS. Cancers are in excess in this condition because of the 20-fold excess of leukemias, whereas malignant solid tumors seem to be globally underrepresented as compared with those in the general population. However, among these tumors, some tumors are in excess: lymphomas, gonadal and extragonadal germ cell tumors, and possibly retinoblastomas and pancreatic and bone tumors. Neoplasms in excess are seen earlier, sometimes in fetal life (leukemias and testicular germ cell tumors) or neonatally (leukemias and lymphoma) and affect mainly male subjects. There seems to exist an excess of rare karyotypes. Other tumors are underrepresented, particularly neuroblastomas and nephroblastomas, in young children, and perhaps common epithelial tumors in adults. These observations suggest that DS has a particular tumor profile, with some tissues more affected by malignant diseases (hematopoietic tissue and germ cells) and others that seem to be protected (central and peripheral nervous system, renal tissue, and epithelial tissues). The mechanism is mainly genetic, but differences in exposure to exogenous agents compared with the general population must be kept in mind. These findings are of interest for the management of these patients and early detection of cancers. Better knowledge of this tumor profile could help us to understand the mechanisms of carcinogenesis and should be compared to the current knowledge of genes on chromosome 21.

An excess of testicular germ cell tumors in down's syndrome†

The incidence of specific solid tumors in Downs syndrome (DS) is not well established. Testicular germ cell tumors (TGCT) might be increased in this population.

ABNORMAL CONSTITUTIONAL KARYOTYPES IN PATIENTS WITH NEUROBLASTOMA: A REPORT OF FOUR NEW CASES AND REVIEW OF 47 OTHERS IN THE LITERATURE

Anomalies of constitutional karyotype, which have led to the discovery of oncogenes and tumor-suppressor genes in embryonal tumors such as retinoblastoma and Wilms tumor, have, until recently, rarely been reported until recently in neuroblastoma. We present four new cases of neuroblastoma associated with (a) a mosaicism for monosomy 22; (b) an 11q interstitial deletion; (c) a pericentric inversion of chromosome 9 at band 9p21; and (d) a Robertsonian translocation t(13;14). These anomalies and 47 others in the literature are worthy of interest, because some are recurrent, involving the same chromosome regions (1p36, 2p23, 3q, 11q23, and 15q), and some anomalies are situated on chromosome regions known to contain genes involved in neuroblastoma development (1p, 2p, 9p, 11q, 16q, and 17q). Chromosome regions 3q and 15q, observed several times, may also contain genes significant for neuroblastoma onset or development. Furthermore, the lack of neuroblastoma in patients with Down syndrome and Klinefelter or triple-X syndromes, together with a probable excess of neuroblastoma in patients with Turner syndrome, suggests that genes of importance for neuroblastoma may map to chromosomes X and 21. A search for genes implicated in neuroblastoma biology should use these data.

Down's syndrome protects against breast cancer: is a constitutional cell microenvironment the key?

Dear Sir, A recent review article in International Journal of Cancer underlined the role of tissue microenvironment in developmental regulation of tumor cells and showed that modifications of the stroma may revert tumor cells to a normal phenotype.1 We present preliminary data suggesting that Down’s Syndrome (DS) could be a natural model of cancer protection due to a particular cell microenvironment. An epidemiological study dealing with breast cancer in subjects with DS2 and reports on cancer incidence3–5 and death rates6,7 in DS subjects have clearly showed a strikingly lower rate of breast neoplasms among DS subjects in comparison to non-DS subjects. If we consider the considerable increase in life expectancy among the DS population over the past 5 decades, breast neoplasms are estimated to be 10–25 times less frequent in DS subjects than in the general population.2,7 This makes DS, which is caused by a constitutional supernumerary chromosome 21 (trisomy 21),8 the most powerful natural condition that offers protection against the most common malignant neoplasm in women, breast cancer. It is surprising that a genetic impairment may have such a favourable effect, but we believe that it deserves our undivided attention. From a broader oncological perspective, neoplasms in DS subjects have a very particular distribution pattern in comparison to those in the general population; this pattern is what we refer to as the “Down Syndrome tumour profile.”9 In DS subjects, some neoplasms are considerably or moderately overrepresented, some are represented at a similar frequency, whereas others are slightly or markedly under-represented, when compared to the non-DS population. One could therefore assume that the theoretical 1.5-fold overexpression of genes on chromosome 21 that characterises the DS condition has different positive, neutral and negative effects on tumour onset and progression, depending on the original tissue. Chromosome 21 is the smallest of our chromosomes, harbouring nearly 250 genes, which is just 1% of the human genome.10 It is probable that some of the overexpressed genes resulting from the gene dosage effect and residing on the supernumerary chromosome 21 could have a protective or favourable effect on various tumour cells. To date, however, no single gene has been isolated to explain oncogenic events in DS. Another cause could be an epigenetic process at work due to the host microenvironment in persons with DS. On examining the features common to neoplasms that are rare or lacking in DS subjects, we were forced to seriously consider the role played by stroma. The stroma of a solid epithelial tumour consists of a population of non-malignant tumour cells including fibroblasts, and a main cell component, which produces the fibroblast extracellular matrix (ECM) surrounding the malignant cells. It is well established that the stroma plays a key role in malignant epithelial tumour homeostasis and development,11 and it is particularly interesting that the most common malignancy in DS children and adults is leukaemia, which is devoid of stroma. Germ cell tumours, lymphomas, sarcomas and retinoblastomas,3,6,12 which are all renowned for their very poorly developed stroma, are also more common in DS subjects. Conversely, cancers of the breast, digestive tract, head and neck, bronchus and skin, which are very rare in adults with DS,3,7,9 are neoplasms that are sustained by well-developed stroma. In the absence of an as yet conclusive explanation for this protective effect against cancer, it is worth considering the hypothesis that the stroma plays an important role in inhibiting solid tumours, and particularly breast cancer. Fibroblasts may promote the development of cancer, and many studies indicate that cancer cells change their microenvironment by modifying fibroblasts and their related extra cellular matrix (ECM).11,13,14 Moreover, fibroblasts taken from distal non-tumorigenic sites in patients with a hereditary predisposition to cancer in different sites, including the breast, present altered phenotypes that favour tumour development.14 Conversely, it would be interesting to investigate if DS fibroblasts have an inhibitory effect on the development of breast cancer in DS subjects. If this hypothesis were valid, we could consider that trisomic 21 fibroblasts and their related ECM offer more resistance to cancer development than euploid fibroblasts and their related matrix. In our studies, neither the in vitro co-culturing of a breast cancer cell line and fibroblasts from a patient with DS or in vivo xenografts on nude mice of both cell lines resulted in the inhibition of breast cancer cell growth. In contrast, when cultured onto an ECM (Fig. 1) secreted by confluent DS fibroblasts, the same breast cancer cell line showed significant growth inhibition (30%) compared to a matrix produced by confluent euploid fibroblasts (Fig. 2). The ECM was not fully secreted during the in vivo experiments (it needs 10 days to be secreted by confluent fibroblasts in vitro), which leads us to

Are Solid Tumours Different in Children with Down's Syndrome?

A review of the literature showed a peculiar distributionof neoplasms in persons having Down’s syndrome (DS). Itrevealed a well-known increase in leukaemia combined witha less often described global decrease in solid neoplasms. Italso found a difference in organ site and histology distribu-tion of solid neoplasms.

Age and stage at diagnosis: a hospital series of 11 women with intellectual disability and breast carcinoma

BackgroundBreast cancer has been poorly studied in women with intellectual disability (ID), which makes designing a policy for screening the nearly 70 million women with ID in the world difficult. As no data is available in the literature, we evaluated breast cancer at diagnosis in women with ID.MethodsWomen with ID were searched retrospectively among all women treated for invasive breast cancer in a single hospital over 18 years. Age at diagnosis was compared among the whole group of women. Tumor size, lymph node involvement, SBR grade, TNM classification, and AJCC stage were compared to controls matched for age and period of diagnosis using conditional logistic regression.ResultsAmong 484 women with invasive breast cancer, 11 had ID. The mean age at diagnosis was 55.6 years in women with ID and 62.4 years in the other women. The mean tumor size in women with ID was 3.53 cm, compared to 1.80 cm in 44 random controls from among the 473 women without ID. Lymph node involvement was observed in 9 of the 11 women with ID compared to 12 of the controls (OR = 11.53, p = 0.002), and metastases were found in 3 of the 11 women with ID compared to 1 of the 44 controls (OR = 12.00, p = 0.031). The AJCC stage was higher in women with ID compared to controls (OR = 3.19, p = 0.010).ConclusionsWomen with ID presented at an earlier age with tumors of a higher AJCC stage than controls despite no significant differences in tumor grade and histological type. Thus, delayed diagnosis may be responsible for the differences between disabled and non-disabled women.

Down syndrome patients are less likely to develop some (but not all) malignant solid tumours.

To the Editor : A recent article in Clinical Genetics reports that ‘. . . the mortality rate from solid-tumour cancers is reduced more than 90% from that of the general population’ in patients with Down syndrome (DS) (1). Based on strong epidemiological evidence, we would like to modulate this sentence somewhat. First, studies on cancer mortality, commonly underestimate the cancer frequency if an autopsy is not performed as often in patients with intellectual impairment (2). Secondly, mortality studies using standardized mortality ratios falsely lead to decreased cancer mortality values when other causes of death are increased as observed in Yang et al.’s study (3). Epidemiological studies based on cancer incidence provide a better estimate of its frequency. The three best epidemiological studies conducted in Denmark, Finland, and Australia (4–6) indicated a standardized incidence ratio respectively of 0.50, 0.57, and 0.44. This 50% reduced frequency of solid tumours observed in DS constitutes a very important observation. Interestingly, although are the following observations, some solid tumours such as breast cancers, and neural tumours neuroblastoma and medulloblastoma (4, 7) are very rare in patients with trisomy 21. On the contrary, the frequency of gastric, colon, ovarian cancers, and gliomas does not differ significantly from that in the general population. Finally, gonadal and extragonadal germ cell tumours, retinoblastoma, and liver cancer (4–6, 8) are more frequent in DS than in the general population. In a nutshell, one observes a ‘tumour profile’ in DS patients. This ‘tumour profile’ results from both the genetic dosage imbalance due to the supernumerary chromosome 21 and life conditions experienced by intellectually disabled patients. For instance, lung cancer is rarer and some digestive tract cancers are more frequent both in DS and intellectual disabled patients (9, 10). To recognize the existence of a ‘tumour profile’ (11) has important practical consequences. First at the clinical level, if one considers that all solid tumours are very rare in patients with DS, one might take the risk of depriving these patients from a good medical surveillance particularly for cancers which have the same or an increased frequency in this population. This will result in a decreased opportunity for treatment and cure in DS patients. Secondly, before searching for a protective gene against cancer on chromosome 21, it is necessary to wonder whether the observed variation is restricted to the DS population, and thus linked to the supernumerary 21 chromosome. Conversely, if the reported variation is shared with other causes of intellectual disability, it is probably due to environmental factors. Also, research should focus mainly on tumours occurring less frequently and specifically in DS, particularly medulloblastoma, neuroblastoma, and breast cancer. In addition, if the decreased frequency of solid tumours observed in DS results from an antiangiogenic effect of some gene(s) on chromosome 21, one should be able to document a reduced frequency of all solid tumours in DS. Furthermore, as impaired angiogenesis alters tumour growth, one should observe, many small tumours and a small number of large tumours with a reduced growth rate in patients with DS. As far as we are aware of, this is not the case. In conclusion, taking good care of DS patients is a noble task and discovering factors protecting children and adults with DS from developing some solid tumours is a fascinating scientific adventure with potential benefits for all. Further well-designed epidemiological studies focussing on malignant tumours observed less frequently in children and adults with DS will achieve the answers with greater economy and accuracy.

A 23-year-old woman with Down syndrome, type 1 neurofibromatosis, and breast carcinoma.

It is well known that some genetic conditions predispose to breast cancer [Oesterreich and Fuqua, 1999]. Interestingly, according to recent epidemiological studies, Down syndrome (DS) seems to protect against breast cancer [Scholl et al., 1982; Hasle et al., 2000; Satgé et al., 2001; Yang et al., 2002]. We report a unique case of a woman with DS and type 1 neurofibromatosis (NF1). This patient developed an adenocarcinoma in situ with areas of invasion of her right breast when she was 23-years old. A 23-year-old patient with both DS (47,XX,þ21) and NF1 presented with a 2.5 cm lump in her right breast. Her mother was 38-years old and her father was 44years old at conception. The patient did not have major malformations. She experienced numerous respiratory infections and was treated by tonsillectomy during childhood. The diagnosis of sporadic NF1 was based on the presence of numerous neurofibromas, twelve café-au-lait spots, and lack of a first-degree relative with NF1. She experienced menarche at the age of 14 years. In her large family, breast cancer was present in one of her four paternal aunts at the age of 82 years. There was no history of breast, ovarian, or other BRCA1-associated cancer in either parent’s family. Mammography showed microcalcifications suggesting a comedocarcinoma. Surgical resection revealed a grade 2 intraductal carcinoma (Fig. 1) with central areas of necrosis, comedocarcinoma type, and rings of slightly polymorphic cells with some mitoses. Two areas of stromal invasion (less than 5% of the whole tumor surface) were identified (Fig. 2). Ten axillary lymph nodes were free of neoplastic tissue. Hormone receptors were negative on tumors cells. The patient received post-operative radiotherapy. She is now in good health today, 14 years after her surgical treatment. According to the data in eight French populationbased cancer registries, the age-specific breast cancer incidence for women below the age of 30 years is consistently less than 15 new cases per 100,000 womanyears [Parkin et al., 1997]. The unusual early onset of this malignant tumor in this patient with DS suggests a genetic predisposition to breast cancer even though her familial history was negative. Solid tumors, particularly breast cancer, are rare in DS [Satgé et al., 1998; Hasle et al., 2000]. A recent study conducted in Denmark could not find a single case of breast cancer among 1,278 women with DS, whereas 7.32 were expected [Hasle et al., 2000]. Two additional studies conducted in the USA and in France found only one case (versus 11.65 expected) [Scholl et al., 1982] and five cases (versus 68.98 expected) [Satgé et al., 2001], respectively. A study conducted on 19, 000 persons with DS found a decreased (0.04) standardized mortality odds ratio for breast cancer [Yang et al., 2002]. These data suggest that trisomy 21 protects against breast cancer. Environmental factors peculiar to women with DS are conflicting. Their dietary habits resulting in excess weight and a lack of fruit and vegetable intake are considered as risk factors for breast cancer. On the other hand, the low use of oral contraceptives and substitutive menopausal hormonal therapy reduces the risk of breast cancer [Satgé et al., 2001]. Possible factors in support of a protective effect include hormonal differences, particularly low exposure to estrogens, some environmental exposures such as those mentioned in terms of hormone therapy and diet, and an over-expression of genes from chromosome 21, such as ANA, ETS2, TIAM1, and S100 beta which are involved in cell proliferation and cell differentiation [Satgé et al., 2001]. It has been shown that women with NF1 are at much higher cancer risk than men with NF1 [Airewele et al., 2001]. Breast carcinoma has been reported in patients with NF1. Brasfield and Das Gupta [1972] found five patients with breast cancer among 54 women with NF1, but this is in the context of a potentially biased sampling. Zöller et al. [1997] reported in a recent study that patients with NF1 have an overall cancer incidence approximately four times as high as expected in the general population, and although separate calculations were not produced for specific tumor types, two breast adenocarcinomas (at ages 38 and 56 years) and one breast tumor of borderline malignancy were described. In a Danish long-term follow up study of patients with NF1 [Sorensen et al., 1986], breast cancer ranked Grant sponsor: Fondation Jérôme Lejeune.

A tumor profile in Edwards syndrome (trisomy 18).

Constitutional trisomy 18 causes Edwards syndrome, which is characterized by intellectual disability and a particular set of malformations. Although this condition carries high mortality during prenatal and early postnatal life, some of the rare infants who survive the first months develop benign and malignant tumors. To determine the tumor profile associated with Edwards syndrome, we performed a systematic review of the literature. This review reveals a tumor profile differing from those of Down (trisomy 21) and Patau (trisomy 13) syndromes. The literature covers 45 malignancies: 29 were liver cancers, mainly hepatoblastomas found in Japanese females; 13 were kidney tumors, predominantly nephroblastomas; 1 was neuroblastoma; 1 was a Hodgkin disease; and 1 was acute myeloid leukemia in an infant with both trisomy 18 and type 1 neurofibromatosis. No instances of the most frequent malignancies of early life—cerebral tumors, germ cell tumors, or leukemia—are reported in children with pure trisomy 18. Tumor occurrence does not appear to correlate with body weight, tissue growth, or cancer genes mapping to chromosome 18. Importantly, the most recent clinical histories report successful treatment; this raises ethical concerns about cancer treatment in infants with Edwards syndrome. In conclusion, knowledge of the Edwards’ syndrome tumor profile will enable better clinical surveillance in at‐risk organs (i.e., liver, kidney). This knowledge also provides clues to understanding oncogenesis, including the probably reduced frequency of some neoplasms in infants and children with this genetic condition.

A Lower Risk of Dying from Urological Cancer in Down Syndrome: Clue for Cancer Protecting Genes on Chromosome 21

Objective: It was the aim of this study to evaluate the risk of dying from bladder and kidney cancer in persons with Down syndrome (DS), as compared with the general population. Methods: Using data of the French national mortality statistics (INSERM) during a 21-year period, 1979–1999, we compared the observed number of deaths from renal and bladder cancer in DS subjects with the expected number of deaths from these cancers. The expected number of deaths was calculated taking into account the prevalence of DS at birth and the life expectancy of persons with DS, assuming the risk was identical to the one of the general population. Results: A significant 6-fold decreased risk of dying from urological cancer was found in persons with DS, with 5 cases observed, while 30 were expected. The relative risk of dying was 0.27 for bladder cancer (p = 0.0017) and 0.06 for kidney cancer (p < 0.0001). Other mortality studies provided similar values. Conclusions: Children and adults with DS have a decreased risk of dying from urological neoplasms. Genes on chromosome 21 could play a protective role against urological cancer.