K. Bilek

Detection of micrometastases in pelvic lymph nodes in patients with carcinoma of the cervix uteri using step sectioning: Frequency, topographic distribution and prognostic impact

OBJECTIVES Limited information exist about the frequency of micrometastases, their topographic distribution and prognostic impact in patients with cervical carcinoma (CX). METHODS Lymph nodes of patients with surgically treated CX, FIGO IB to IIB, with pelvic lymph node involvement, were re-examined regarding the size of metastatic deposits, their topographic distribution within the pelvis. Lymph node status (pN0 vs. pN1mic=metastasis<0.2 cm vs. pN1=metastasis>0.2 cm) was correlated to recurrence free (RFS) and overall survival (OS). RESULTS 31.4% of all patients (281/894) represented pelvic lymph node involvement. 22.2.% of the node positive ones showed micrometastases (pN1mic). Most commonly, obturator and internal nodes were affected by pN1mic, without any side differences. Patients with macrometastases (pN1) and micrometastases (pN1mic) represented significant reduced RFS-rate at 5-years (62% [95% CI: 54.2 to 69.8] for pN1 and 68.9% [95% CI: 55.5 to 82.4] for pN1mic) when compared to patients without metastatic disease (91.4% [95% CI: 89.0 to 93.8]; p<0.001) The 5-years OS-rate was decreased in patients with metastatic disease (pN0: 86.6% [95% CI: 83.7 to 89.5], pN1mic: 63.8% [95% CI: 50.9 to 76.7], pN1: 48.2% [95% CI: 40.4 to 56.0]; p<0.0001). These differences persisted in detailed analysis within these subgroups. In multivariate analysis, tumor stage, pelvic lymph node involvement and micrometastases were independent prognostic factors. CONCLUSIONS A remarkable number of patients with CX show micrometastases within pelvic nodes. Micrometastatic disease represents an independent prognostic factor. So, all patients with pelvic lymph node involvement, including micrometastatic deposits, might be candidates for adjuvant treatment.

p16, p14, p53, and cyclin D1 expression and HPV analysis in small cell carcinomas of the uterine cervix

Summary: Small cell carcinomas (SmCCs) of the uterine cervix are rare tumors. The knowledge regarding protein expression of several checkpoint candidates of cell cycle regulation is limited. Surgically treated SmCCs were selected from our files for immunohistochemical staining (neuroendocrine markers, p53, p16, p14, and cyclin D1). Polymerase chain reaction analysis, using general primers, was performed for human papillomavirus analysis. Nine of 677 tumors (1.3%) were classified as SmCCs after Grimelius staining (8/9 tumors positive) and immunohistochemical reaction against neurone-specific enolase, chromogranin A, synaptophysin (7/9 positive tumors), and CD 56 (8/9 positive tumors). All specimens were positive for at least two of the above. Two SmCCs were p53 positive and one case was p14 positive. Cyclin D1 staining was completely negative. All cases showed strong nuclear and/or cytoplasmic p16-immunostaining. Seven tumors represented human papillomavirus positivity for high-risk types. Four patients died of the tumor after a median time of 36.7 months (range, 15-56 months), representing a 5-year survival rate of 56%. The results suggest that p16 is up-regulated or accumulated in the SmCCs of the uterine cervix, probably caused by infection with human papillomavirus. p14 inactivation is of high prevalence in SmCCs and detection rate of p53 is similar to other histologic types of cervical carcinomas.

DIAGNOSIS AND CLINICAL MANAGEMENT IN MALIGNANT MULLERIAN TUMORS OF THE FALLOPIAN TUBE. A REPORT OF FOUR CASES AND REVIEW OF RECENT LITERATURE

Four out of 42 cases of primary tubal malignancy diagnosed in our histopathological laboratory were malignant mixed Müllerian tumors (MMMT). All four patients were postmenopausal with a mean age of 66.5 years at diagnosis. A correct preoperative diagnosis was made only in one case. Tumor staging (FIGO) revealed stage Ila, IIIc and IV. One patient died of postoperative pulmonary embolism, a second patient of an unknown cause five month after surgery and a third patient died of disease after I I months with secondary deposits in pelvic peritoneum, omentum and paraaortic lymph nodes. The fourth patient is still alive. One patient received chemotherapy alone, one by radiation and chemotherapy and two patients by radiation alone. Tumor spread at the time of diagnosis and the residual tumor volume were the most important prognostic factors. All tumors were histologically the homologous type of MMMT (carcinosarcomas). No heterologous elements were found. Metastatic tumors showed only sarcomatous elements.

p53 in Surgically Treated and Pathologically Staged Cervical Cancer: Correlation with Local Tumor Progression, But Not with Lymphatic Spread

There is only limited information about the prognostic value of p53 immunostaining in cervical cancer. The purpose of this study was to assess the clinical significance of p53 and prognosis in operatively treated cervical carcinoma. A hundred and fourteen primary surgically treated cervical carcinomas (CX) were obtained from the so called Wertheim Archive in the Department of Obstetrics and Gynecology at the University of Leipzig. These included 105 squamous cell cancer (SCC) and nine adenocarcinomas (AC). No cases received neoadjuvant therapy. For immunohistochemical analysis, the cases were tested with the monoclonal antibody DO-7 (DAKO Diagnostics, Denmark). Two hundred tumor cell nuclei were counted for positive nuclear immunostaining, regardless of staining intensity. Cases were stated as positive when a minimum of 10% nuclei showed positive staining. Fresh frozen tissue was available from 21 CX for p53-mutation analysis (exons 4-9) using PCR-based amplification and SSCP-analysis. Of the squamous cell cancers (SCC), 63.8% showed positive nuclear p53-immunostaining; adenocarcinomas (AC) were completely negative (P = 0.0000, Chi2-test). Stage-by-stage analysis revealed no differences in p53-expression. However, combining pT1b- and pT2-cases, the difference in positive immunostaining reached statistical significance (44.4% vs. 71.7%; P = 0.007). There were no differences in p53-reactivity regarding the presence of pelvic lymph node metastases, tumor grading, relapse-free survival and tumor recurrence. In addition, only 5% of CX with positive p53-immunostaining showed genomic alterations in mutational analysis. p53-immunoreactivity showed significant correlation with local tumor progression but not with lymphatic spread, lacking any prognostic impact in surgically treated cervical cancer. There is no correlation of p53-immunostaining with the occurrence of p53-gene mutations in cervical cancer.

COX-2 and Her-2/neu are overexpressed in Paget’s disease of the vulva and the breast: results of a preliminary study

PurposePaget’s disease (PD) of the breast as well as the vulva is a rare condition that accounts for about 4% of breast neoplasms and 1% of vulvar malignancies. Recurrent disease after breast and vulvar surgery might be a challenge. To evaluate relevant molecules therapeutically, tissue from mammary and vulvar PD lesions was investigated immunohistochemically.MethodsHistopathologic samples from 11 patients with mammary PD and eight patients with vulvar PD were stained with antibodies against estrogen and progesterone receptors, HER-2/neu and COX-2 followed by semiquantitative evaluation of the staining results.Results All tested mammary lesions as well as seven out of eight vulvar PD were negative for estrogen and progesterone receptors. Strong membranous staining for HER-2/neu (Score 3) was seen in all cases. Six out of 8 vulvar and 10 out of 11 mammary PD showed COX-2 overexpression.Conclusions PD of the breast and vulva are negative for estrogens and progesterons, therefore, anti-hormonal therapy is not indicated. The high frequency of Her-2/neu and COX-2 overexpression, however, suggests that these molecules could be therapeutically relevant in patients with PD. These results require further investigation.

Peritumoral stromal remodeling, pattern of invasion and expression of c-met/HGF in advanced squamous cell carcinoma of the cervix uteri, FIGO stages III and IV.

OBJECTIVE Different patterns of invasion (PIs) have prognostic impact in several types of cancer and are associated with different grades of peritumoral stromal remodeling, characterized by the desmoplastic stromal response (DSR). One key regulator influencing cellular motility and peritumoral stromal response is c-met/HGF. This study evaluates the association between different PI, peritumoral DSR and its correlation to the expression of c-met/HGF in squamous cell carcinomas of the uterine cervix (CX). STUDY DESIGN 131 advanced stage CX (FIGO III/IV) were re-evaluated histologically regarding PI, using a two-level scoring system. The tumor grows in solid cords/trabeculae in finger-like PI and in very small groups or single cells in spray-like PI. DSR was categorized as none/weak and moderate/strong. The tumors were stained with antibodies against c-met and HGF. The staining of >30% of tumor cells was defined as overexpression. The PI was correlated to the prognostic outcome, different categories of DSR and expression status of c-met and HGF. RESULTS 66.4% of the tumors showed a finger-like, and 33.6% a spray-like PI. The spray-like PI showed a reduced two-year overall survival when compared to the finger-like PI (14.0% vs. 29.1%, respectively; p=0.012), and was associated with moderate/strong DSR. The majority of the tumors showed overexpression of c-met (85.4%) and HGF (74.8%). There was no correlation between the expression status of c-met/HGF and the FIGO stage, peritumoral DSR or the prognostic outcome. CONCLUSIONS Spray-like PI is of prognostic impact in cervical carcinoma FIGO III/IV and is associated with strong peritumoral stromal remodeling. There is no prognostic impact of the immunohistochemical expression of c-met/HGF in advanced stage cervical carcinomas.

Staging of Endometrial Carcinoma: Aspects of Ovarian and Cervical Involvement

To the Editor: We read with interest the article by Richard J. Zaino ‘‘FIGO staging of endometrial adenocarcinoma: a critical review and proposal’’ (1). The study reviewed the more problematic aspects of the staging of endometrial cancer (ECX) and comes up with recommendations for changes in a future revision of the FIGO staging of endometrial carcinoma. One issue that was discussed in the study is the adnexal involvement. Within an unpublished study of 251 consecutive cases with ECX at different stages and histologic types with complete adnexal processing, 10.7% represented adnexal (ovarian) metastases. There was a predominant involvement of the left side (44.1% vs. 29.6%); 25.6% represented with bilateral disease. About 48.2% showed only microscopic adnexal metastatic involvement (Fig. 1A). In the majority of the cases with metastatic disease, lymphovascular space involvement (LVSI) of the mesovarian tissue was observed. Most of LVSI was observed in conjunction with serous ECX. But, adnexal LVSI (either mesovarian, within the ovarian stroma, or within the Fallopian tube) was not stated as metastatic disease in our study. As discussed in the study of Prat (2) and that of Richard Zaino (1), the definition of ‘‘adnexal involvement’’ is quite problematic. Within the current issue of FIGO/TNM-classification, stage IIIA disease is defined as ‘‘tumor involves uterine serous and/or adnexa (direct extension or metastasis) and/or cancer cells in ascites or peritoneal washings’’ (3). There is no clear statement in these classifications of how the term metastasis is defined in that context. Although there are no studies dealing with that topic, LVSI of the meso-salpinx, mes-ovarium, the Fallopian tube, and the ovarian stroma (Figs. 1B–D), in our opinion, represents biologically stage IIIA disease. At our institution we classify ECX with the above-mentioned morphologic features of LVSI in accordance with a FIGO/TNM classification system not as stage IIIA disease, but give the clinician detailed information about the localization of LVSI. Therefore, the patients might be followed as they would represent advanced stage disease. Although uncommon, simultaneous endometrioid carcinomas of the Fallopian tubes may occur and are usually located at the distal end of the tube (4). Furthermore, patients with ECX are usually postmenopausal and the ovaries are usually small and atrophic; therefore extensive embedding may not be time and/or cost consuming. As suggested by others (5), and in recognition of our above-mentioned results, we would recommend complete processing of macroscopic inconspicuous adnexae, regardless of the histologic type of ECX. In cutting the ovary we prefer sectioning along the short axis of the ovary embedding the adjacent mesovarian tissue (6) and for handling of the Fallopian tube we adopted the protocol that is currently used for the handling of the tubes in prophylactic adnexectomies from women with BRCA mutations [so-called SEE-FIM-protocol; (7,8)]. Another problematic point in the staging of ECX, which has been the focus of the paper by Richard Zaino (1), is cervical involvement. Cervical involvement in ECX occurs between 6% and 20% and usually results from a direct surface or stromal extension, but superficial implantation might also occur (2). Women with ECX that spreads to the cervix have a diminished probability of survival, compared with those without. But, detailed studies of the patterns of cervical involvement suggest that FIGO stage IIA tumors (involvement of cervical glands by the ECX) is associated with a similar prognosis as ECX is confined to the myometrium [FIGO stage IA-C; (9,10)]. This subject is currently under discussion for a revised FIGO/TNM classification, which should appear in 2009 (Sobin and Wittekind personal communication). The correct morphologic interpretation of atypical endometrial glands within the cervix might appear as another challenge in staging of ECX. Most tumor tissue, which is observed within the cervical channel, is dislocated during the leading symptom of ECX— the irregular (in most cases postmenopausal) bleeding (Figs. 2A–C)—which does not influence the surgical staging . LVSI might also occur within the cervical stroma, especially in cases with serous endometrial

CD 44 exon v6 as a predictor of lymphatic metastases in cervical carcinoma — an immunocytochemical study of 94 cases

The relation between expression of several splicing variants of the CD 44 glycoprotein by tumor cells and the increased risk of metastases was discussed recently. By means of an immunocytochemical study (imprint cytology specimens from 94 invasive cervical carcinomas) we have shown a significant correlation between expression of CD 44 v6 and invasion of lymphatic vessels, lymphangiosis carcinomatosa in the primary tumor and the total number of positive pelvic lymph nodes. Expression of CD 44 v6 was not correlated with staging, grading and histological type. CD 44 v6 could therefore be considered as a predictor of lymphatic metastases in cervical carcinoma.

A cut-off value of 2 cm in tumor size is of prognostic value in surgically treated FIGO stage IB cervical cancer.

OBJECTIVES Limited knowledge exists about the value of tumor size in surgically treated cervical cancer (CX) using a tumor size of 2 cm as cut-off value. METHODS A total of 366 cases of CX FIGO stage IB who received upfront surgery were evaluated regarding tumor size, the prediction of pelvic lymph node involvement, and recurrence-free and overall survival during a median follow-up time of 94 months. Tumors ≤2.0 cm were defined as small, tumors 2.1-4.0 cm as medium sized and those larger than 4 cm as bulky disease. RESULTS Small tumors were seen in 28.7%, medium sized in 52.5% and bulky tumors in 18.9%. There was a significant higher frequency of pelvic lymph node involvement with increasing tumor size (13.3% vs. 23.4% vs. 43.5%, respectively; p<0.001) and an increase of recurrent disease (6.7% vs. 18.8% vs. 29.4%, respectively; p<0.001). The 5-year overall survival rate was significantly reduced with increasing tumor size (94.0% vs. 85.1% vs. 69.9%, respectively; p<0.001). Pelvic lymph node involvement and maximal tumor size were independent prognostic factors for both recurrence-free and overall survival in multivariate analysis. CONCLUSIONS The results support that tumor size is of prognostic impact in FIGO stage IB cervical carcinomas. A further substaging is suggested for tumors up to 4.0 cm maximum dimension using a cut-off value of 2.0 cm as discriminator. Patients with tumors ≤2.0 cm may represent low risk disease.

Morphological Parameters Associated With Perineural Invasion (PNI) in Carcinoma of the Cervix Uteri

The study determines morphological features that are associated with perineural invasion (PNI) in patients with cervical carcinoma (CX). Histological slides from 194 patients from surgically treated squamous cell carcinoma were re-examined for PNI and correlated to morphological factors of tumor growth. Material from 68 patients (35.1%) represented PNI. PNI was significantly correlated with advanced tumor stage (P < .001). Patients with deep cervical stromal invasion (>66%) showed more PNI than those with more superficial invasion (41% vs 16.9%; P = .001). Tumors with spray-like PI showed significantly more PNI (48.4%) when compared with finger-like PI (26.7%) and those with pushing borders (18.8%; P = .007). Strong peritumoral desmoplastic stromal reaction and absence of peritumoral inflammation were associated with a higher frequency of PNI (P < .001). PNI is associated with advanced tumor stage, deep cervical stromal invasion (>66%), high grade of tumor cell dissociation (ie, spray-like pattern of invasion), strong peritumoral desmoplastic stromal reaction, and reduced peritumoral inflammation.