From 2001 to 2003, 105 families with clinical features of HLRCC have been identified (Table 41.2-1). In all, 89 families with leiomyomatosis have been found to harbour a germline FH mutation: 7 of 7 studied in Finland (Kiuru et al, 10; Launonen et al, 10) (unpublished data), 35 of 47 in the UK (Alam et al, 1; Alam et al, 2), 31 of 35 in North America (Toro et al, 17), and 16 of 16 from other countries including Tunisia, Ethiopia, Greece, Puerto Rico, Iran, and Australia (Chuang et al, 14; Martinez-Mir et al, 2). At present, there are no population-based prevalence estimates of the disease. Among 299 Finnish patients with tumors not associated with the HLRCC phenotype, no germline mutations were detected (Lehtonen et al, 13). By contrast, among selected patient material of 200 Finnish patients with a tumor associated with HLRCC phenotype, 2 patients with a germline mutation were identified (2 of 200; 1%) (Kiuru et al, 9).
Table 41.2-1Numbers of All Reported HLRCC Families and those with Renal Cell Carcinoma and Uterine Leiomyosarcoma |Favorite Table|Download (.pdf) Table 41.2-1 Numbers of All Reported HLRCC Families and those with Renal Cell Carcinoma and Uterine Leiomyosarcoma
| Reference || Country of Origin || Families || Mutation-Positive Families || Families with Renal Cell Carcinoma (n) || Families with Leiomyosarcoma of the Uterus (n) || Families with Atypical Leiomyoma (n) |
|Launonen et al, 10 ||Finland ||2 ||2 ||2 (6) ||1 (2) ||0 (0) |
|Kiuru et al, 10 ||Finland ||1 ||1 ||1 (1) ||0 (0) ||1 (1) |
|Kiuru et al, 9 ||Finland ||2 ||2 ||0 (0) ||1 (1) ||0 (0) |
|Lehtonen et al, in press ||Finland ||2 ||2 ||2 (5) ||1 (1) ||1 (1) |
|Alam et al, 2 ||UK ||47 ||35 ||1 (1) ||0 (0) ||0 (0) |
|Toro et al, 17 ||North America ||35 ||31 ||5 (13) ||0 (0) ||1-2 (2) |
|Chuang et al, 14; Martinez-Mir et al, 2 ||North America, Tunisia, Ethiopia, Greece, Puerto Rico, Australia, Iran ||16 ||16 ||0 (0) ||0 (0) ||0 (0) |
|Total || ||105 ||89 ||11 (26) ||3 (4) ||3-4 (4) |
The definitive diagnosis of HLRCC relies on FH mutation detection. However, there are certain clinical and histopathologic features suggestive of HLRCC. First, the syndrome typically presents as multiple leiomyomas of the skin with or without positive family history of cutaneous and/or uterine leiomyomas. Toro and colleagues (2003) defined an individual as affected if that individual had >10 skin lesions clinically compatible with leiomyoma and a minimum of 1 lesion histologically confirmed; by this criteria, they identified 35 families. A total of 31 families segregated a germline FH mutation (31 of 35; 86%), and 14 individuals with no skin manifestation were also found to be mutation carriers. Second, the association of the relatively rare papillary type 2 histology with HLRCC has successfully been used in patient identification (Launonen et al, 10). Third, early-onset uterine leiomyosarcoma could be a valuable feature in diagnosis. One HLRCC family was identified based on the presence of early-onset uterine leiomyosarcoma and positive family history of uterine leiomyosarcoma/leiomyoma (Lehtonen et al, 13), and another family was identified among a set of unselected uterine leiomyosarcomas screened for FH mutations (Kiuru et al, 9). Taken together, the above clinical features should be taken into account in clinical practice, and specific diagnostic criteria should be established in the future.
Leiomyomas of the Skin and the Uterus
Leiomyomas of the skin and/or the uterus are the most common feature in HLRCC. Their penetrance is high: approximately 85% (Kiuru, 9; Toro et al, 17). The onset of cutaneous leiomyomas has been reported to range from 10 to 47 years and the onset of uterine leiomyomas from 18 to 52 years (mean, 30 years) (Toro et al, 17). Clinically, cutaneous leiomyomas present as multiple firm, skin-colored nodules ranging in size from 0.5 to 2 cm (Fig. 41.2-1). They are sometimes associated with pain and paresthesias. Uterine leiomyomas in HLRCC patients are often numerous and large and cause typical symptoms such as abnormal uterine bleeding and pain. Onset of uterine leiomyomatosis in HLRCC appears to be earlier than in the general population (30 vs. 40-44 years) and results more often and earlier in hysterectomy due to symptomatic disease (89% had a hysterectomy; 57% were under 30 years of age) (Toro et al, 17).
Multiple cutaneous leiomyomas from a female HLRCC patient.
Cutaneous leiomyomas are composed of interlacing bundles of smooth muscle cells with centrally located blunt-ended nuclei. Uterine leiomyomas are well-circumscribed lesions with a firm and fibrous appearance. Histologically, they are composed of interlacing bundles of elongated, eosinophilic smooth muscle cells surrounded by well-vascularized connective tissue. In addition, four leiomyomas with atypia have been reported (Table 41.2-1). These tumors are variants of leiomyomas and sometimes difficult to discern from malignant tumors.
Leiomyosarcoma of the Uterus
Leiomyosarcoma is a rare malignant mesenchymal tumor of the uterine corpus. In HLRCC, leiomyosarcoma occurs at an earlier age than in the general population, although predisposition to leiomyosarcoma is detected only in a subset of HLRCC families (3 out of 105 families) (Table 41.2-1). The onset of the disease has varied from 30 to 39 years.
As opposed to benign leiomyomas, uterine leiomyosarcomas invade the adjacent myometrium and are not well demarcated from normal tissue. Histologically, the tumors are densely cellular and display spindle cells with blunt-ended nuclei, eosinophilic cytoplasm, and a variable degree of differentiation. Differential diagnosis between benign leiomyomas is sometimes difficult. In those cases, mitotic activity, nuclear atypia, coagulative necrosis, degree of cellularity and differentiation, and invasion to adjacent tissues can be utilized in diagnosis.
At present, 26 patients with renal cancer have been identified in 11 of 105 families (Alam et al, 2; Kiuru et al, 10; Launonen et al, 10; Toro et al, 17) (unpublished data) (Table 41.2-1). Renal carcinomas typically develop earlier than their sporadic counterparts. The median age of onset is 36 years in the Finnish and 44 years in the North American patients, the only UK patient being 18 years old (range of all patients, 18-90 years). Renal tumors in HLRCC are typically solitary and unilateral (Launonen et al, 10; Toro et al, 17), which contrasts with other inherited renal cancer syndromes, such as von Hippel-Lindau syndrome, hereditary papillary renal carcinoma, and Birt-Hogg-Dubé syndrome. Renal carcinomas appear to be associated with an aggressive disease course, as the majority of patients died of metastatic disease within 5 years after diagnosis.
The peculiar and distinct histology of renal cancers in HLRCC originally led to identification of this syndrome (Launonen et al, 10). Typically, HLRCC renal cell carcinomas display papillary type 2 histology and large cells with abundant eosinophilic cytoplasm, large nuclei, and prominent inclusion-like eosinophilic nucleoli (Fig. 41.2-2). The Fuhrman nuclear grade is high: from 3 to 4. Most tumors stain positive for vimentin and negative for cytokeratin 7. Recently, the histologic spectrum of renal cancer in HLRCC has been expanded as three patients were identified as having either collective duct carcinoma or oncocytic tumor (Alam et al, 2; Toro et al, 17).
Renal cell carcinoma from a 50-year-old female patient. A. Papillary architecture (H&E staining; magnification × 10). B Large cells with abundant cytoplasm and inclusion-like nucleoli (× 40).
Renal cancer in HLRCC appears to be associated with an aggressive disease course. Thus, to detect renal tumors at an early stage, regular screening for these lesions is recommended. At present, computed tomography (CT) is the recommended method for screening, as it is more accurate than abdominal ultrasound in detecting papillary renal tumors (Toro et al, 17). By contrast, ultrasound provides a safe and applicable method without risks induced by radiation. However, optimal methods, start age, and frequency of screening need to be evaluated in the future. Moreover, as renal cell carcinoma is present only in a subset of families, there are currently no guidelines regarding whether the surveillance should be carried out in all families.
The following tumor types also have been detected in Finnish patients with a germline FH mutation: breast carcinoma (in two patients also affected with uterine leiomyosarcoma), multiple myeloma, prostate cancer, Hodgkin's lymphoma, and tumors of the salivary and adrenal glands (Kiuru, 9; Launonen et al, 10) (unpublished data). In patients from other countries, only one additional tumor type has been reported: a leiomyosarcoma of the skin in one individual from a North American family (Toro et al, 17).
Gene Structure and Function
The predisposing gene, FH, is located on chromosome 1q42.3-q43. FH consists of 10 exons spanning over 20 kb of genomic DNA. The length of the cDNA is approximately 1.8 kb and it is predicted to encode a 511-amino-acid peptide. The first exon encodes a mitochondrial signal peptide (Edwards, Hopkinson, 1979a; Edwards, Hopkinson, 1979b; Tolley, Craig, 15), but processed FH (without the signal peptide) is present also in the cytosol. Mitochondrial FH acts in the tricarboxylic acid cycle (TCAC, Krebs cycle), catalyzing conversion of fumarate to malate. FH is also known to be involved in the urea cycle. However, the role of cytosolic FH is still somewhat unclear.
The enzymatically active form of the protein functions as a homotetramer. In addition to the signal peptide domain, FH contains several other domains including alpha-helical and lyase domains. Alpha-helixes form a superhelical structure in the core of the tetramer, whereas the lyase domain is thought to be involved in catalytic activity of the protein. These domains belong to a superfamily of proteins including fumarase, aspartase, adenylosuccinate lyase, arginosuccinate lyase, and crystalline (Estevez et al, 8).
FH appears to act as a tumor suppressor gene. Biallelic inactivation of FH has been detected in almost all HLRCC tumors, including skin and uterine leiomyomas, as well as renal cell cancers (Alam et al, 2; Kiuru et al, 10; Kiuru et al, 9; Launonen et al, 10). In addition, FH enzyme activity is known to be very low or absent in HLRCC tumors. The mechanism by which FH acts as a tumor suppressor gene is not known. Multiple hypotheses have been suggested. First, the function of the cytosolic form of FH is not largely understood, and the presence of currently unknown functions of FH related to tumorigenesis cannot be excluded. Second, defective Krebs cycle due to FH mutations might trigger hypoxia response, enhance generation of reactive oxygen species (ROS), or cause activation of antiapoptotic pathways.
Germline mutations in FH have been found in 89 of 105 HLRCC families (85%). In all, 50 different germline mutations have been identified (Fig. 41.2-3). A great majority of the mutations are single base pair substitutions (69 of 89; 78%), of which missense mutations account for 70% (62 of 89). Deletions (including a whole gene deletion) and insertions, as well as splice-site changes, have also been reported (Fig. 41.2-3). Exons 4 and 6, representing 25% of the coding sequence, seem to be mutation hotspots. Mutation frequencies are 28% (14 of 50) and 28% (14 of 50) in exons 4 and 6, respectively. If calculations are based on mutations reported in the 89 separate families, mutation frequencies are 36% (32 of 89) and 17% (15 of 89) in exons 4 and 6, respectively. Furthermore, approximately 18% of all mutations are found in codon 190 (exon 4), including three different amino acid changes: R190H (n = 14), R190C (n = 1), and R190L (n = 1). Mutations from families with renal cell cancer and uterine leiomyosarcoma are indicated in Fig. 41.2-3. In general, the mutations identified in the families with malignancies seemed not to be located in the particular region of the gene. The same mutations (except H153R) have also been detected in families without malignancies.
FH mutations detected in HLRCC and FH deficiency. Mutated codons identified in the families with RCC and/or uterine leiomyosarcoma are specified.
Two founder mutations have been detected in the Finnish population: a missense mutation (H153R, in 3 of 7 families) and a 2-bp deletion in codon 181 (in 3 of 7 families). Families with the H153R mutation included one family with renal cell cancer, one with uterine leiomyosarcoma, and one with both malignancies. Renal cell cancer cases were also observed in two of the three families with the 2-bp deletion in codon 181. One uterine leiomyosarcoma case also was found in one of these families. No malignancies were detected in the third small family with the same mutation (Kiuru et al, 9). Another founder mutation, a splice-site mutation (IVS4+1G>A), was detected in families of Iranian origin. In addition, a missense mutation (R190H) was reported in 35% of the families from North America.
To date, the role of FH in somatic tumorigenesis has been evaluated in three different studies (Barker et al, 3; Kiuru et al, 9; Lehtonen et al, 13). Five mutations were found among Finnish tumors associated with HLRCC, the study including samples such as leiomyomas of the skin and uterus, renal cell carcinomas, sarcomas, and lobular breast carcinomas. Two unselected and nonsyndromic cases–a cutaneous leiomyoma (1 of 10; 10%) and a uterine leiomyosarcoma (1 of 53; 1.9%)–harbored a germline FH mutation. A somatic inactivating point mutation was seen in both tumors. In one soft tissue sarcoma of the lower limb and in two uterine leiomyomas, a purely somatic biallelic inactivation of FH was observed (Kiuru et al, 9; Lehtonen et al, 13). In contrast, no mutations were found in sporadic uterine leiomyosarcomas and uterine leiomyomas in patients from the UK (Barker et al, 3). In a study of 299 non-HLRCC-associated tumors (prostate, nonlobular breast, colorectal, lung, ovarian, testicular, thyroid, head and neck cancers, sarcomas, pheochromocytomas, gliomas, and melanomas) no FH mutations were identified. In summary, FH mutations seemed to be very rare in other tumor types except those associated with HLRCC syndrome. Biallelic inactivation of FH appears to occur notably in a subset of sporadic uterine leiomyomas.
HLRCC patients were found to have reduced FH enzyme activity in lymphoblastoid cell lines compared with controls (Alam et al, 2; Tomlinson et al, 16). The enzyme activity was lower in individuals with a missense mutation compared with individuals with a truncation change or a whole gene deletion. This result is not unexpected because FH functions as a homotetramer, and a missense mutation in 1 allele results in only 1 fully functional homotetramer out of 16. In HLRCC tumors FH enzyme activity is very low or absent.
FH deficiency is a recessive disease caused by biallelic, homozygous or compound heterozygous, germline mutations in FH. Clinically, the syndrome is characterized by neurologic impairment, growth and developmental delay, and fumaric aciduria. Patients usually die within a few months of birth. FH enzyme activity is absent or significantly reduced in the patients' tissues. Heterozygous parents are known to be neurologically asymptomatic carriers of the mutation with a reduced enzyme activity (approximately 50%). In 1 FH deficiency family, a heterozygous parent had cutaneous leiomyomas (Tomlinson et al, 16). Thus far, 10 different FH mutations have been reported in 14 FH deficiency families. Most of the detected mutations are missense changes (12 of 18; 67%) (Fig. 41.2-3). Exons 4 and 6 seem to be mutation hotspots in FH deficiency, including 39% of all detected mutations.
Approximately 85% of families clinically diagnosed with HLRCC have an FH mutation. There seem not to be clear-cut genotype-phenotype correlations between the detected mutations and their clinical outcomes. When comparing HLRCC and FH deficiency, three mutations–K187R, R190C, and R190H–have been reported in both of the syndromes. In 1 FH deficiency family, a heterozygous parent had cutaneous leiomyomas (Tomlinson et al, 16). Renal cell cancer and uterine leiomyosarcoma are seen in 10% (11 of 105) and 3% (3 of 105) of the HLRCC families, respectively (Table 41.2-1). The same mutations (a 2-bp deletion in codon 181, R190H, and H275Y) have been identified in families with or without malignancies. It has been suggested that a modifying gene or genes could play a role in the development of renal cancer and uterine leiomyosarcoma in HLRCC (Eng et al, 7; Tomlinson et al, 16; Toro et al, 17). According to the present data on FH mutation families with malignancies, penetrance of renal cell cancer and uterine leiomyosarcoma seems to be high, perhaps 50%, in individuals with FH mutations and the putative modifier allele. Based on the strikingly malignant phenotype in Finnish HLRCC families, this population would be expected to have a relatively high frequency of the modifier allele.