Malignant melanoma of the vulva in a nationwide, 25-year study of 219 Swedish females: clinical observations and histopathologic features.

BACKGROUND: Because the clinical and histopathologic features of vulvar melanoma had not been characterized completely in a large, homogeneous population, the authors retrospectively analyzed all such patients recorded in Sweden during a 25-year period. METHODS: The Swedish National Cancer Registry opened its records to the authors for review of all 219 females with primary vulvar melanoma reported from 1960 to 1984. Histopathologic specimens and clinical histories of the 198 patients who qualified for this study were reanalyzed and the tumors rigorously subtyped. RESULTS: Macroscopically amelanotic tumors were observed in 27% of patients, predominantly in glabrous skin; the clitoral area and labia majora were the most common primary sites. Of all melanomas, 46% emerged in glabrous skin, 12% emerged in hairy skin, and 35% extended to both areas. On average, approximately 2.5 times more melanomas appeared in the vulva than on the whole body surface. Overall, 57% were of the mucosal lentiginous (MLM) type, 22% were nodular melanomas (NMs), 12% were unclassified, and only 4% were superficial spreading melanomas (SSMs); this was the reverse of the order observed for cutaneous melanoma. Almost all vulvar melanomas underwent a vertical growth phase; other common features were marked thickness and ulceration, particularly in the glabrous skin. Preexisting nevi occurred in 11 cases, all in hairy skin, and 71% in conjunction with SSM but only 4% with MLM. CONCLUSIONS: Several clinical and histopathologic features indicated that the natural history of vulvar melanomas is at variance with that of cutaneous melanomas. Because preexisting nevi, which are often considered a precursor to melanoma, were significantly linked to SSM and only in the vulvar hairy skin, melanomas in the glabrous skin apparently emerged de novo.

Malignant melanoma of the vulva in a nationwide, 25-year study of 219 Swedish females: predictors of survival.

BACKGROUND: In an epidemiologic study of 219 Swedish females with vulvar melanoma, the authors previously established the incidence of this disease as 0.19 per 100,000 women, with a 3% annual decrease from 1960 to 1984 and a 5-year relative survival rate of 47%. After reviewing the medical histories of all of the 219 patients, the authors documented their precise clinical and histopathologic features, which, along with treatment, are assessed herein as predictors of survival. METHODS: Of 219 consecutive cases of vulvar melanoma collected from the Swedish National Cancer Registry, 21 were excluded because of inadequate data. Clinical and histopathologic materials from the remaining 198 cases were then reexamined. With a clinical three-stage system, lesion types and treatment modalities were assessed as survival factors in univariate and multivariate analyses. RESULTS: In univariate analysis, significant predictors of survival for patients at Stage I were tumor thickness, ulceration, number of mitoses, macroscopic amelanosis, preexisting nevi, extent of tumor invasion (lateral labia majora), and patient age. The mode of treatment was not significant. In multivariate analysis, staging (Stage I vs. II and III) and tumor thickness were independent predictors of survival. For Stage I only, tumor thickness, ulceration, and clinical amelanosis independently predicted survival time. CONCLUSIONS: To the authors' knowledge, this is the largest series of patients with vulvar melanoma ever reviewed, and an ethnically homogeneous and nationwide female population is represented. In this series, clinical stage, macroscopic amelanosis, and tumor characteristics such as tumor thickness and ulceration, rather than treatment mode, were the best factors for predicting the outcome of these patients.

Screening of germline mutations in the CDKN2A and CDKN2B genes in Swedish families with hereditary cutaneous melanoma.

BACKGROUND: Approximately 10% of human cutaneous melanomas occur in families in which several members are affected. The familial predisposition to this disease is often associated with dysplastic nevus syndrome, a condition in which afflicted family members have multiple dysplastic nevi (atypical moles). The chromosome region 9p21 and markers on chromosomes 1p and 6p have been linked to melanoma susceptibility. The tumor suppressor genes CDKN2A and CDKN2B have been mapped to the 9p21 region, and genetic analyses have revealed the presence of germline CDKN2A alterations in melanoma families. The reported frequencies of such alterations, however, vary among these families. PURPOSE: The present investigation was carried out to determine the frequencies of CDKN2A and CDKN2B germline gene mutations among members in a population-based cohort of Swedish melanoma families (i.e., melanoma kindreds). METHODS: DNA was prepared from blood samples obtained from 181 individuals belonging to 100 melanoma kindreds. The polymerase chain reaction (PCR) technique, followed by single-strand conformation polymorphism (SSCP) and nucleotide sequence analyses, were used to identify the types and frequencies of mutations in exons 1, 1beta, 2, and 3 of the CDKN2A gene and in exons 1 and 2 of the CDKN2B gene. RESULTS: CDKN2A gene aberrations were independently identified by both SSCP and nucleotide-sequence analyses. Nucleotide-sequence analysis identified a single point mutation leading to a substitution of leucine for proline in codon 48 of exon 1 in a family with a history of melanoma and several other cancers. A second abnormality, leading to an insertion of an extra arginine residue at codon number 113 of exon 2, was seen in four separate families. The CDKN2A exon-3 coding region had the wild-type sequence in all samples. No germline mutations were found in the alternative exon 1beta of the CDKN2A gene or in exons 1 and 2 of the CDKN2B gene. CONCLUSIONS: The present investigation demonstrates that CDKN2A germline gene mutations were observed in 7.8% of the 64 Swedish melanoma kindreds that each included at least two first-degree relatives with melanoma and dysplastic nevus syndrome. No CDKN2A exon 1beta or CDKN2B mutations were identified. The critical genes responsible for the inheritance of a susceptibility to develop melanoma among family members in this population have yet to be identified.

Genes involved in cell cycle G1 checkpoint control are frequently mutated in human melanoma metastases.

A common characteristic of cancer cells is unrestrained cell division. This may be caused by mutational changes in genes coding for components of cell cycle-controlling networks. Alterations in genes involved in G1 checkpoint control have been registered in many human tumours, and investigations from several laboratories show that such alterations, taken together, are the most frequent changes detected in cancer cells. The present paper describes mutational analysis by polymerase chain reaction-single-strand conformation polymorphism (PCR/SSCP) and nucleotide sequence analysis of the genes coding for the p15, p53 and N-ras proteins in 26 metastases from 25 melanoma patients. The registered mutation frequencies add together with previously registered mutations in p16 in the same patient samples to a substantial total frequency of 44% of patients with mutation in at least one of the investigated genes. These results show the occurrence of heterogeneous defects among components of the cell cycle controlling machinery in a human melanoma tumour sample collection and demonstrate that the total frequency of detected alterations increases with the number of cell cycle controlling genes included in the screening panel.

Predictors of late mortality in cutaneous malignant melanoma--a population-based study in Sweden.

We determined risk factors for late deaths from cutaneous malignant melanoma (CMM) based on clinical characteristics at diagnosis, initial surgical treatment, histopathologic features of the primary tumor and type of eventual recurrences during follow-up. We examined deaths from CMM 8 or more completed years after the initial diagnosis in a case-control study nested in a nationwide cohort comprising all 8,838 patients with CMM diagnosed in Sweden during 1960-1978 with complete follow-up through 1986. There were 285 case patients and 285 control patients, individually matched by sex, age and follow-up time. Conditional logistic regression was used to obtain odds ratios (OR) as estimates of the relative risk. The risk of late mortality increased stepwise, almost 19-fold, with increasing tumor thickness from < or = 0.75 to > or = 7.00 mm. Besides the thickest tumors (> or = 7.00 mm), those of intermediate thickness (1.50-2.49 mm) had the highest risk (OR 8.5). After adjustment for tumor thickness, non-radical primary surgical treatment increased the risk of late mortality almost 3-fold while prophylactic lymph node dissection entailed a significantly reduced risk of late mortality (OR 0.5); the histopathologic features increasing level of invasion and vertical growth phase also remained significantly associated with a poor outcome. In a multivariate model, non-radical primary surgical treatment, prophylactic lymph node dissection, vertical growth phase, level of invasion and lymphocyte reaction were independent predictors of late mortality.

Resection margins of 2 versus 5 cm for cutaneous malignant melanoma with a tumor thickness of 0.8 to 2.0 mm: randomized study by the Swedish Melanoma Study Group.

BACKGROUND: The traditional surgical treatment for primary malignant melanoma has often been a wide excision with a margin of about 5 cm. Since the risk of local recurrences is dependent on tumor thickness, thin tumors (<1 mm) have routinely been excised with a narrow margin. For thick tumors, the optimal resection margin is controversial, and can be determined only by prospective, randomized trials. METHODS: The Swedish Melanoma Study Group performed a prospective, randomized multicenter study to evaluate an excision margin of 2 versus 5 cm for patients with cutaneous malignant melanoma with tumor thickness > 0.8 and < or = 2.0 mm. The trial includes 769 patients. Patients with melanomas of the skin of the head, neck, hands, feet, or vulva were not included in the trial. In the event of an excision biopsy for diagnosis, radical surgery was completed within 6 weeks. The median follow-up time was 5.8 years for estimation of survival and 4.0 years for diagnosis of recurrent disease. RESULTS: No significant differences have been observed between the treatment groups regarding local or regional recurrences or survival. CONCLUSIONS: We recommend an excision with a margin of 2 cm for cutaneous malignant melanoma with a tumor thickness > 0.8 and < or = 2.0 mm.

Mutational analysis of the CDKN2 gene in metastases from patients with cutaneous malignant melanoma.

We analysed 26 metastases from 25 patients with sporadic cutaneous malignant melanoma for alterations in the CDKN2 gene by a combined polymerase chain reaction/single-strand conformation polymorphism (PCR/SSCP)/nucleotide sequencing approach. Eleven alterations (one in exon 1, five in exon 2 and five in the 3' non-coding sequence of the exon 3 region) were concordantly and independently detected by both SSCP and nucleotide sequence analysis. Two of the exon 2 changes and the five changes in the non-coding exon 3 region are likely to represent natural polymorphism. Four (15%) of 26 metastases thus had CDKN2 mutations and belonged to 3 (12%) of 25 patients. Semi-quantitative PCR furthermore revealed no sign of homozygous deletions of the CDKN2 exon 2 region. The results support an involvement of the CDKN2 product in the development of a subgroup of sporadic melanomas and encourage the search for alterations in additional genes of the 9p21 region.

Utility of a proliferation marker in distinguishing between benign naevocellular naevi and naevocellular naevus-like lesions with malignant properties.

In the present study we have investigated the utility of the proliferation marker MIB 1 in distinguishing between benign naevocellular naevi and naevocellular naevus-like lesions with malignant potential. Percentages of MIB 1 immunoreactivity in the intradermal portion of the lesions were determined. In benign congenital and acquired naevi, as well as in dysplastic naevi, there was no or only a slight intradermal melanocytic proliferation (0-2%), whereas vertical growth phase melanomas exhibited a substantial proliferative activity (11-48%). In five cases of naevus-lke lesions, which had all relapsed as unmistakable malignant melanomas (locally or metastatically) after primary surgery, there was also clear proliferative activity (9-67%). Our findings suggest that MIB 1 may be a useful tool in the routine histopathological examination of problematic naevocellular lesions.

Immunohistochemical analysis of the N-ras p21 and the p53 proteins in naevi, primary tumours and metastases of human cutaneous malignant melanoma: increased immunopositivity in hereditary melanoma.

Immunohistochemical analysis of the N-ras p21 and the p53 proteins was carried out on formalin-fixed sections of naevi, primary melanomas and metastases from patients with sporadic melanoma (SCMM) and with hereditary melanoma (HCMM)/dysplastic naevus syndrome (DNS). Seven out of 11 (64%) common naevi and three out of nine (33%) dysplastic naevi showed increased cytoplasmic N-ras expression. No p53 immunopositivity could be recognized in any of the naevus samples. However, strong N-ras expression as well as immunopositivity for p53 was recognized among primary melanomas and metastases with significantly higher frequency among samples from patients with HCMM compared with samples from SCMM cases (for N-ras, 40% vs 10%, P < 0.01; and for p53 43% vs 17%, P < 0.05). We have earlier registered N-ras codon 61 mutations among metastases from 59% of patients with HCMM and from 24% of subjects with SCMM. A comparison of the genetic data with the immunohistochemical results showed occurrence of increased N-ras p21 expression in the presence and absence of detectable N-ras mutant alleles. Increased expression of wildtype N-ras p21 may contribute to tumorigenicity in the absence of mutational activation, at least in a subset of melanomas. Altogether, N-ras p21 alterations are registered at earlier stages than p53 alterations in melanoma development and may be of aetiological importance, whereas p53 alterations may be associated with tumour progression in the late stages.(ABSTRACT TRUNCATED AT 250 WORDS)

Glutathione transferase P1-1 expression in human melanoma metastases: correlation to N-RAS mutations and expression.

Expression of the detoxication enzyme glutathione transferase P1-1 (GST P1-1) at elevated levels has been noted in many types of human tumors, including melanomas. The products of the human H-RAS, K-RAS and N-RAS genes play a key role in intracellular signal transduction leading to transcriptional activation of AP-1 (Fos/Jun) responsive genes. The oncogenic mutated forms of the ras proteins are constitutively active and interfere with normal signal transduction. Mutated RAS genes as well as increased expression of wild-type ras proteins are common features in human tumors including melanoma. We have characterized 30 melanoma metastases from 23 melanoma patients with reference to N-RAS expression and mutation as well as to GST P1 expression (immunohistochemistry and genetic analysis). Twenty-three of 30 samples (70%) had high N-Ras p21 and/or N-RAS codon 61 mutations and 18 of these 23 samples also had high GST P1-1 immunoreactivity. Seven of 30 (23%) samples had low N-Ras p21 immunoreactivity and no detectable N-RAS codon 61 mutations. Six of these 7 samples (86%) also had low GST P1-1 immunoreactivity. The results indicate a statistically significant correlation (Spearman correlation coefficient, r = 0.56, p = 0.001, 2-tailed test) and provide, for the first time, indirect evidence for a possible coregulation of N-RAS and GST P1 in human malignant melanoma which should be further evaluated.

Estramustine binding protein in primary tumours and metastases of malignant melanoma.

The presence of estramustine binding protein (EMBP), a 54 kD cytosolic glycoprotein, which is distinct from the estrogen receptor, was investigated in a pilot study of primary malignant melanomas and their metastases. In 11 primary melanomas EMBP was demonstrated by immunohistochemistry. The percentage of positive melanoma cells ranged between 11-91% with a mean of 57%. Radioimmunoassay on metastatic tissue revealed significant amounts of EMBP, with values ranging between 0.6-4.2 with a mean of 1.6 ng/mg protein. A high number of cells with a positive stain for EMBP in the primary tumours was significantly correlated to a short interval between diagnosis and the occurrence of metastases. Presence of EMBP in malignant melanoma may have prognostic significance and the role of hormone-linked cytostatic drugs in the treatment of this disease needs further investigation.

Elective lymph node dissection in stage I cutaneous malignant melanoma of the head and neck. A report from the Swedish Melanoma Study Group.

The effect of elective lymph node dissection in patients with cutaneous malignant melanoma of the head and neck was investigated in a retrospective study. Of 517 patients in clinical stage I, 84 underwent elective dissection of the ipsilateral neck lymph nodes. In six of these patients, lymph node metastases were demonstrated at histopathological examination. There was a slight reduction in the incidence of recurrent disease in the regional lymph nodes in the group of patients who had undergone elective lymph node dissection, but this difference was not statistically significant. No significant differences were seen between the two groups regarding overall survival of disease-related survival.

Melanoma metastases from patients with hereditary cutaneous malignant melanoma contain a high frequency of N-ras activating mutations.

Mutations in N-ras exon 2 codon 61 were studied in formalin-fixed human melanoma metastases. DNA fragments including codon 61 were amplified by polymerase chain reaction (PCR) and mutational analysis was performed by oligonucleotide hybridization (ODN), allele specific PCR and PCR combined with single strand conformation polymorphism analysis (SSCP). Thirty metastases from 25 patients with 'spontaneous' cutaneous melanoma were compared with 35 metastases from 17 patients with 'hereditary' cutaneous melanoma. The frequency of mutations as measured by PCR/ODN was significantly higher in patients with hereditary melanoma (mutations in 24% versus 59%, p < 0.05). The most frequent mutations were C/A transversions to lysine (AAA). The occurrence of lysine mutations was, in addition, studied by allele specific polymerase chain reaction. Again, the mutation frequency was significantly higher in metastases from patients with hereditary melanoma. PCR/SSCP finally enabled the isolation of lysine mutant alleles and nucleotide sequence analysis which confirmed the presence of the mutated codon 61. The relatively higher frequency of N-ras mutations in tumours from patients with hereditary melanoma may be related to the hypermutability described in hereditary melanoma and dysplastic naevus syndrome. The results support an involvement of N-ras mutations in the molecular pathogenesis of melanoma.

Prognostic factors in thin cutaneous malignant melanoma.

BACKGROUND: Thin melanomas can metastasize and be lethal. The predictive importance of tumor thickness in thin melanomas and the specific features identifying the patients at risk have not been investigated fully. METHODS: Prognostic factors were analyzed in 585 patients with clinical Stage I invasive cutaneous malignant melanoma with a thickness of less than or equal to 0.8 mm. The patients were included in a population-based cancer registry in Stockholm county during 1976-1987. They constituted about 64% of all patients with thin melanomas who were diagnosed in the region during the study period. Information was available on age, sex, anatomic site of the tumor, histologic type of melanoma, level of invasion, tumor thickness, and tumor regression. In a Cox regression analysis, the prognostic importance of each factor was studied. By a case-control technique with individual matching for the identified independent predictors of recurrence, the additional prognostic information given by type and grade of inflammatory response, presence of vertical growth phase, mitotic rate/mm2, and histologic ulceration of the tumor was assessed. RESULTS: After a median follow-up time of 50 months, recurrent disease developed in 26 patients (4%). There was no difference in recurrence rate between patients treated with narrow (1-2 cm) or wide (5 cm) excision. Anatomic site, tumor thickness, level of invasion, and tumor regression were found to be independent prognostic factors in the multivariate analysis. In the case-control study, only grade of inflammatory reaction added significant prognostic information. No subgroup could be identified that was without risk of recurrent disease. CONCLUSIONS: Thin melanomas do not seem to constitute a separate form of melanoma, but compose one end of a continuous spectrum of biologic behavior.

Cutaneous malignant melanoma of the head and neck. Analysis of treatment results and prognostic factors in 581 patients: a report from the Swedish Melanoma Study Group.

BACKGROUND: Results of surgical treatment of cutaneous malignant melanoma (CMM) have been highly variable, probably because of patient selection. Therefore, a study of representative patients with this disease was performed. METHODS: In a defined area of Sweden, 581 patients were analyzed. Clinical records and histopathologic findings were reviewed. The minimum follow-up time was 7 years. Prognostic factors were evaluated by the Cox proportional hazards model. RESULTS: Evaluation of sex distribution, age, and anatomic site of the primary tumor showed that the patients were representative of all Swedish patients with CMM of the head and neck. The mean patient age at diagnosis was 64 years for both sexes. Fifty-three percent of the patients were women. Female patients had more tumors of the face than did male patients, whereas male patients were overrepresented among patients with tumors of the auricle-external ear canal and scalp-neck area. Localization to the face was observed in 68%, which is an overrepresentation of three to four times when skin surface is taken into consideration. Twenty-four percent of the patients had lentigo maligna melanoma. Only 33% of the patients had superficial spreading melanoma. In univariate analyses, sex, anatomic site of the primary tumor, histogenetic type, Clark level of invasion, and tumor thickness had prognostic power. In a multivariate analysis, tumor thickness, anatomic site of the primary tumor, and sex of the patient were independent prognostic factors. CONCLUSIONS: In representative patients with CMM of the head and neck, tumor thickness, anatomic site of the primary tumor, and sex of the patients were independent prognostic factors.

Regional hyperthermic perfusion with melphalan after surgery for recurrent malignant melanoma of the extremities. Swedish Melanoma Study Group.

A prospective randomized trial testing regional hyperthermic perfusion with melphalan has been conducted. Sixty-nine patients with recurrent malignant melanoma of the extremities were randomly allocated to surgery (36 patients) or surgery plus regional perfusion (33 patients). Prognostic variables concerning primary tumor as well as the recurrent disease were evenly distributed in the groups, excluding any bias in the randomization. Median tumor-free survival after randomization was 17 months in the perfusion group and 10 months in the control group. There were 15 locoregional recurrences in the perfusion group and 24 in the control group. The tumor-free survival curve was significantly (P = .044) better for the perfusion group than for the control group. Median survival time after randomization was 57 months in the perfusion group and 35 months in the control group. This difference was not significant. One patient died within 1 month after perfusion of pulmonary embolism. Regional hyperthermic perfusion after surgery of recurrent malignant melanoma should only be recommended in prospective and controlled trials, until its value has been proven in several randomized studies.

Nuclear DNA analyses in archival material from primary malignant melanoma. A pilot study.

In 28 cases of malignant melanoma, paraffin-embedded specimens were analyzed in order to determine the reliability of ploidy results. The material consisted of thin and thick melanomas. The results indicate that useful prognostic information may be obtained in this kind of material by means of DNA measurements, provided that the analysis is performed on morphologically identified tumor cells. The value of DNA measurements in malignant melanomas may, however, not be as clear as has been reported for several other tumors.

Comparison of nuclear DNA content in primary and metastatic malignant melanoma.

The DNA patterns obtained from 23 primary malignant melanomas and 35 corresponding metastases were compared and found to differ in many cases. In eight cases the primary tumors and their metastases had a ploidy type I ("euploid") DNA pattern. One case had a type I primary tumor and both type I and type II metastases. Five cases had type I primary tumors and ploidy type II ("aneuploid") DNA pattern metastases. In five cases the primary tumors and corresponding metastases were type II, and in another four cases the primary tumors were type II, whereas the metastases were type I. We interpret these data as indicating that malignant melanomas (more often than adenocarcinomas) are composed of genetically heterogeneous tumor sublines that frequently give rise to heterogeneously composed metastases. Since we sometimes observed a change in the DNA content in malignant melanomas, it seems to be more difficult to obtain prognostic information from DNA analysis in malignant melanoma as compared to the more stable adenocarcinomas.