Peutz-Jeghers syndrome (PJS) is an autosomal dominant disease that is characterized by hamartomatous polyposis in the gastrointestine and mucocutaneous melanin spots in the oral mucosa, lips, wings of the nose, fingers and toes. The first case was reported in a Dutch family by Peutz in 1921 (1). Jeghers et al described the details of another family in 1948 (2) and the syndrome was defined as an independent disease. The clinical features of PJS include gastrointestinal polyps that increase in number and size and the appearance of abdominal symptoms, including ileus and gastrointestinal hemorrhage. Among the types of hereditary gastrointestinal polyposis, the incidence of PJS is second highest following that of familial adenomatous polyposis. PJS develops at an incidence rate of 1 in 5–25×10⁴ individuals.

The tumor suppressor gene, STK11/LKB1, which is located on chromosome 19p13.3, is responsible for PJS. STK11 is believed to be involved in cellular energy metabolism, cell proliferation, cell polarity, p53-dependent apoptosis, the regulation of vascular endothelial growth factor (VEGF) and Wnt signal transduction. Loss of heterozygosity (LOH) due to a mutation, including a deletion in the normal allele, in addition to a germline STK11 mutation, may cause the clinical features of PJS, including gastrointestinal polyposis and cancerization of other organs. PJS is complicated by benign and malignant tumors of various organs and by rare diseases, including sex cord tumor with annular tubules (SCTAT) and minimal deviation adenocarcinoma (MDA). The development of endometrial carcinoma in patients with PJS has also attracted attention in the field of gynecology. Lobular endocervical glandular hyperplasia (LEGH) in a PJS patient with a germline STK11 mutation has recently been reported, which has led to the hypothesis that LEGH is a prodromal lesion of MDA (3).

A disease resembling PJS was described by Peutz in 1921 and by Jeghers et al in 1949 (1,2). The pathology of PJS is considered to be hamartomatous, since PJS polyps are not cancerous and their histology is similar to the normal mucosal structure. A more accurate description of the lesions is hyperplasia of the lacunar epithelium. However, the morphology differs from that of a general hyperplastic polyp. The gland opening is dilated toward the outer side due to the hyperplastic mucoepithelium and muscular fibers overgrow in a dendritic pattern along the epithelium. These changes are preceded by epithelial hyperplasia. Adjacent lamina muscularis mucosae are believed to bend and fuse to give the characteristic appearance and small lesions undergo hyperplastic changes (4,5).

Clinically, PJS is characterized by the development of abdominal symptoms, including pain, ileus and gastrointestinal hemorrhage, which occur with an increase in the number and size of the gastrointestinal polyps. The incidence of PJS is secondary to that of familial adenomatous polyposis among the types of hereditary gastrointestinal polyposis. Malignant and benign tumor complications may occur in PJS, with studies indicating that malignant tumors develop in approximately half of patients by the age of 57 years. The incidence of digestive organ malignant tumors is highest in the colorectum, followed by the stomach, small intestine, duodenum and pancreas. In regions other than the digestive organs, the incidence is reported to be highest in the uterine cervix, followed by the lung and ovary in female PJS patients (Table I) (3). Associations with ovarian SCTAT and uterine cervical LEGH and MDA have been highlighted in the field of gynecology, and an association with endometrial cancer has recently gained interest based on case reports and genetic studies (6–11).

The gene that is responsible for causing PJS is the tumor suppressor gene, STK11/LKB1, which is located on the short arm of chromosome 19 (19p13.3) (12). The gene is 23 kb in size and is comprised of nine coding exons and one non-coding exon. The gene encodes a serine threonine kinase containing 433 amino acids (12). The mRNA is 3.0–3.3 kb in size and is expressed in almost all human tissues. Germline mutations of STK11/LKB1 are observed in more than half of PJS patients. However, the somatic development of PJS in patients with no familial medical history and cases without a mutation in STK11/LKB1 have also been described.

In 1997, Hemminki et al identified the responsible gene region in PJS to be near the chromosome 19 short arm marker using a linkage analysis of families with PJS (13). Amos et al confirmed this finding (14). The STK11 gene was known to be present on chromosome 19 (15,16) and Yoon et al subsequently identified a number of mutation types in this gene, including missense, frame shift, nonsense and splicing site mutations, in 10 PJS patients using polymerase chain reaction-single strand conformation polymorphism (PCR-SSCP; Fig. 1) (17). Germline mutations were present in five of these patients. Tseng et al analyzed STK11 mRNA expression in twin sisters with PJS with the same allele using PCR (18) and revealed that STK11 gene expression was absent in the two subjects. These results suggest that STK11 on chromosome 19 is responsible for PJS and that a mutation or decreased expression of STK11 is the cause of the disease.

The protein product of STK11 is involved in cellular energy metabolism, cell proliferation, cell polarity, p53-dependent apoptosis, the regulation of VEGF and Wnt signal transduction (3). LOH due to a mutation, including a deletion in the normal allele, in addition to the germline STK11 mutation, results in gastrointestinal polyposis and cancerization of other organs, which are common clinical features of PSJ.

As noted previously, Yoon et al identified germline mutations in five of 10 PJS patients, which suggests other developmental mechanisms in the remaining five patients, based on STK11 gene mutation-induced development (somatic case) and the association with other genes. Several studies have suggested the presence of a gene that is associated with PJS other than STK11. In an investigation of 21 PJS patients from 13 families, Papp et al identified that 8 (62%) of the 13 cases of PJS had familial medical histories and that the remaining five cases (38%) were due to de novo mutations (12). Germline mutations were screened in the 21 patients and 13 pathogenic mutations of STK11 were identified. Three of these were frameshift mutations, three were nonsense mutations, two were mutations of the splicing sequence and five were deletions of exons 1–7. This deletion was noted in five of the 13 families, showing a high frequency, and was also shown to affect two genes that were located upstream of the STK11 gene, SBNO2 and GPX4, which are considered to modify STK11. This finding suggests that an abnormality in the genes that modify STK11 function may promote the development of PJS, even in the absence of a mutation in STK11 itself.

Souza et al initially reported a contiguous genetic syndrome in which a developmental disorder, heart malformation and facial dysplasia appear as phenotypes, in addition to the symptoms of PSJ (19). This syndrome is caused by a 19p13.3 chromosomal deletion of a region of ~1.1 Mb that includes STK11(19). Scollon et al also reported a syndrome with a similar gene deletion at a similar site, but the development of a cleft lip and gastrointestinal polyposis differed between the syndromes, suggesting that the phenotypes may vary, despite a common STK11 deletion between the two syndromes (20).

Histologically, SCTAT is characterized by the annular growth of sex cord cells while hyaline bodies form around the nucleus. SCTAT includes ovarian Sertoli cell tumors, ovarian mucous/serous epithelial tumors and ovarian mature teratoma. Scully et al suggested that SCTAT is derived from ovarian granulosa cells and grows in a pattern that is characteristic of Sertoli cells (21). In another hypothesis, SCTAT has been considered to be comprised of sex cord-derived immature cells, which have the possibility to differentiate into granulosa and Sertoli cells. These tumors are positive for immunohistochemical staining of testosterone and estradiol and are diagnosed based on endocrinological symptoms in the majority of cases, including precocious puberty and irregular menstruation, which may involve amenorrhea, hypermenorrhea and postmenopausal hemorrhage (3).

SCTAT is an ovarian tumor that is most likely to complicate PJS, with ~36% of SCTAT cases reported to be associated with the syndrome (22). The clinical features of SCTAT differ between patients with and without PJS (22). In a comparison between SCTAT in 21 patients with PJS and 47 patients without PJS, Young et al identified that SCTAT with PJS was multifocal, bilateral, small and required microscopy to confirm the diagnosis in the majority of the cases. More than half of the cases were accompanied by calcification and the prognosis was favorable. In contrast, sporadic SCTAT was unilateral, large enough for palpation and calcified only in 12% of cases. In total, 20% of cases progressed to malignancy (22). Granulosa cell, Seltori-Lydig cell, borderline malignant and mucous tumors have been reported as other histological types of PJS-associated ovarian tumors, in addition to SCTAT.

MDA is a novel disease type that has been proposed for a condition that was previously termed adenoma malignum by Silverberg and Hurt in 1975 (23). MDA may be a subtype of adenocarcinoma, but it has a well-differentiated histology that is indistinguishable from that of the normal cervical gland. Certain patients that are diagnosed with MDA have a favorable prognosis, and in 1999, Nucci et al proposed the disease type lobular endocervical glandular hyperplasia (LEGH) for these cases (24). LEGH has been assumed to be a precancerous lesion of MDA in one study; however, a conclusion has not been established until recently (6).

MDA is a well-known gynecological tumor that is likely to complicate PJS. The incidence of MDA in PJS patients is estimated to be 15–30%, while ~10% of MDA cases are complicated by PJS. The mean age of MDA patients with PJS has been reported to be 33 years old, whereas that of MDA patients without PJS is 55 years old, indicating a young onset age for the dual disease (25). MDA complicating PJS is considered to be a malignant tumor with the poorest prognosis among all the gynecological tumors complicating PJS.

A mutation in the gene that is responsible for PJS, STK11, and LOH on chromosome 19p13.3, the location of STK11, have been analyzed in studies of MDA without PJS (26,27). LOH at 19p13.3 was identified in ~50% of these cases. In another study, an allelic deletion of >3.5 Mb was noted on chromosome 19p13.3 in nine MDA patients without PJS characteristics (28). Furthermore, in an investigation of the association between STK11 mutation and LOH on chromosome 19p13.3, Connolly et al identified no somatic mutation in the STK11 coding region in eight MDA patients without PJS, but LOH was present on chromosome 19p13.3 in three of those patients. Therefore, somatic mutations in STK11 are absent in MDA without PJS, but LOH may occur at chromosome 19p13.3 (25). These results demonstrate that LOH may occur at chromosome 19p13.3, the location of the STK11 gene that is responsible for PJS, in certain MDA patients without PJS, suggesting the involvement of an unknown tumor suppressor gene on chromosome 19p13.3. Lee et al also identified LOH on the chromosome on which STK11 is present in six of nine MDA patients without PJS and at a site 190 kb away from STK11 in two patients. These findings also suggest the presence of a tumor suppressor gene that is associated with the development of MDA, other than STK11, on chromosome 19p13.3. Linkage with 19q13.4 has also been reported (29).

Kuragaki et al(27) identified STK11 mutations in six out of 11 MDA patients without PJS and also confirmed LOH in all six of the patients with STK11 mutations, suggesting that STK11 on 19p13.3 is associated with MDA, in addition to PJS. With regard to the presence of the STK11 mutations in only six of the 11 patients with mucous MDA, Kuragaki et al suggested that the development of mucous MDA involves the loss of MDA expression, the post-translational modification of the STK11 protein and a gene other than STK11(27). Further analysis of STK11 at the exon level in the six MDA patients with STK11 mutations revealed that a mutation was present in the exons of one out of three of the patients (27). In other studies, Nakagawa et al identified a germline mutation in exon six of STK11 in five out of ten PJS patients (30), Connolly et al observed mutations in exons 4 and 6 in two SCTAT patients with PJS (25) and Hemminki et al identified a mutation in exon 1 in seven out of 12 PJS patients (29). These results suggest that the significant exons of STK11 for PJS are exons 1, 4 and 6.

Hirasawa et al recently described the first case of LEGH in a PJS patient with a germline STK11 mutation. The mutation was a deletion of four bases at codon 263 of exon 6 (6). This is significant in terms of the development of LEGH as a uterine cervical tumor in a PJS patient and also with regard to LEGH as a potential precancerous lesion of MDA.

The prognosis of MDA has been reported to vary between being extremely poor (31) to being mostly equal to that of well-differentiated uterine cervical adenocarcinoma of the same stage (32). Kuragaki et al investigated the association between STK11 mutations and prognosis in 11 MDA patients without PJS (27). The clinical stages of the patients were IB and IIB in five and one out of six patients with STK11 mutations, respectively, and IB and IIB in four and one out of five patients with no STK11 mutations, respectively, showing no difference in the clinical stage between the two groups. However, four of the six patients with STK11 mutations succumbed within 24 months following surgery and the tumor recurred in one of the two surviving patients. In contrast, all five of the patients without STK11 mutations survived for >50 months following surgery, suggesting that the prognosis of MDA with STK11 mutations is poorer than that of MDA without STK11 mutations (P=0.039). The prognosis of PJS patients whose condition is complicated by MDA has also been recorded to be poor in numerous other studies (3,33). This may have been due to a higher level of STK11 mutations in the MDA patients with PJS than in those without PJS.

PJS is characterized by the development of numerous gastrointestinal hamartomatous polyps and mucocutaneous pigmentation and a higher risk of malignant tumors in the digestive tract and other organs compared with the general population. The risk of gynecological cancer is also high, with the lifetime risk of endometrial carcinoma in females with PJS reported to be 9% (34). However, this risk in PJS patients is lower than that of patients with Lynch syndrome and surveillance for endometrial carcinoma in PJS patients is considered unnecessary, unlike for ovarian, cervical and breast cancers (35).

Benign or malignant tumors readily develop in organs such as the digestive tract, mammary glands, ovaries and uterus in PJS patients, with a risk that is 10–18 times higher than that of the general population (36). In females with PJS, the incidence of two rare gynecological tumors, SCTAT and MDA, also tends to increase. In a few patients, several gynecological tumors may complicate PJS simultaneously. One example of this may be observed in the female PJS patient described by Mangili et al, in whom microscopic ovarian SCTAT and uterine cervical MDA were identified (37).

With regard to the association between MDA and PJS, STK11 on chromosome 19p13.3 is likely to be the gene responsible for the two diseases. Gene analyses have been performed separately for PJS and MDA, but only case reports are available for MDA patients with PJS and no comparative gene analysis has been described. Until recently, there have also been no studies of concomitant PJS and LEGH, which is considered to be a precancerous lesion of MDA (6).

The clarification of the association between PJS and LEGH, MDA, SCTAT and endometrial carcinoma requires the elucidation of the developmental mechanism at the genetic level. However, PJS is a rare disease and patients with SCTAT and MDA complications are rarer, making the number of patients very small. A complete gene analysis requires a large number of patients and will require a multicenter study. However, probing of the genetic mechanisms of SCTAT and MDA complicated by PJS may lead to greater understanding of the developmental mechanisms of uterine cervical adenocarcinoma and simultaneous multiple gynecological tumors that complicate PJS. Further studies are anticipated in this field.