Multivariate analysis of histopathologic prognostic factors for invasive cervical cancer treated with radical hysterectomy and systematic retroperitoneal lymphadenectomy

BACKGROUND: The aim of this study was to identify the independent histopathologic prognostic factors for patients with cervical carcinoma treated with radical hysterectomy including paraaortic lymphadenectomy. METHODS: A total of 187 patients with stage IB to IIB cervical carcinomas treated with radical hysterectomy and systematic retroperitoneal lymphadenectomy were retrospectively analyzed. The median follow-up period was 83 months. Cox regression analysis was used to select independent prognostic factors. RESULTS: Using multivariate Cox regression analysis, lymph node (LN) status (negative vs. metastasis to pelvic nodes except for common iliac nodes vs. common iliac/paraaortic node metastasis), histopathologic parametrial invasion, lymph-vascular space invasion (LVSI), and histology of pure adenocarcinoma were found to be independently related to patients' poor survival. For patients who had a tumor histologically confined to the uterus and have neither parametrial invasion nor lymph node metastasis, LVSI was the most important prognostic factor, and histologic type, depth of cervical stromal invasion, and tumor size were not related to survival. The survival of patients with a tumor extending to parametrium or pelvic lymph node(s) was adversely affected by histology of pure adenocarcinoma. When the tumor extended to common iliac or paraaortic nodes, patients' survival became quite poor irrespective of LVSI or histologic type of pure adenocarcinoma. Patients' prognosis could be stratified into low risk (patients with a tumor confined to the uterus not associated with LVSI: n = 80), intermediate risk (patients with a tumor confined to the uterus associated with positive LVSI, and patients with squamous/adenosquamous carcinoma associated with pelvic lymph node metastasis or parametrial invasion: n = 86), and high risk (patients with pure adenocarcinoma associated with pelvic lymph node metastasis or parametrial invasion, and patients with common iliac/paraaortic node metastasis: n = 21) with an estimated 5-year survival rate of 100 +/- 0 (mean +/- SE)%, 85.5 +/- 3.9%, and 25.1 +/- 9.7%, respectively. CONCLUSIONS: LN status, parametrial invasion, LVSI, and histology of pure adenocarcinoma are important histopathologic prognostic factors of cervical carcinoma treated with radical hysterectomy and systematic retroperitoneal lymphadenectomy. Prognosis for patients with cervical carcinoma may be stratified by combined analysis of these histopathologic prognostic factors. Postoperative therapy needs to be individualized according to these prognostic factors and validated for its efficacy using randomized clinical trials.     

Prognostic significance of serous and clear cell adenocarcinoma in surgically staged endometrial carcinoma

BACKGROUND: The serous adenocarcinoma (SA) and clear cell adenocarcinoma (CCA) of endometrium have been shown to be associated with high relapse rate and poor survival. It is not clear whether prognostic significance of these specific cell types of tumor is independent of retroperitoneal lymph node metastasis and other histopathologic prognostic factors in endometrial carcinoma. METHODS: We examined 240 consecutive patients with clinical stage I to stage III endometrial carcinoma who were treated prospectively with radical surgery and/or platinum-based chemotherapy. Surgery included extended hysterectomy, bilateral salpingo-oophorectomy, and systematic pelvic and paraaortic lymph node dissection. Prognostic significance of various histopathologic factors was determined by Cox regression analysis. RESULTS: SA/CCA were more frequently associated with deep myometrial invasion, high nuclear grade (G3), lymph-vascular space invasion (LVSI), and pelvic lymph node metastasis when compared to endometrioid adenocarcinoma (EMA). Of 216 clinically staged stage I or II disease, seven of 12 cases of SA/CCA had extrauterine disease. This incidence was much higher than that for EMA (46/204) (p<0.01). A multivariate Cox regression analysis revealed that cell type, grade, LVSI, and paraaortic node metastasis (PANM) were independent prognosticators. CONCLUSIONS: Prognosis of patients with endometrial carcinoma depends on cell type, grade, LVSI, and PANM. Poor prognosis for patients with SA/CCA is independent of lymph node metastasis and other histopathologic prognostic factors. The SA/CCA should be strictly discriminated from EMA.     

Is it possible to predict para-aortic lymph node metastasis in endometrial cancer?

Objective

The purpose of this study was to determine the histopathologic risk factors for pelvic lymph node (PLN) and para-aortic lymph node (PALN) metastasis in endometrial cancer (EC) and to identify in which patients PALN dissection should be performed.

Study design

A total of 204 consecutive patients, with EC and underwent systematic pelvic and para-aortic lymphadenectomy extending to the renal vessels, were studied retrospectively. Statistical significance between risk factors was examined using multivariant logistic regression analysis.

Results

Cell type, depth of myometrial invasion and tumor size were found to be independently related to PLN metastasis. PLN metastasis in any site and lymphovascular invasion (LVSI) were independent prognostic factors for predicting PALN metastasis. The sensitivity, specificity and the NPV of PLN metastasis for detecting PALN metastasis were 80.8%, 89.3% and 97%, respectively. Furthermore, the 204 patients were divided into two groups according to the presence of one of these following factors: (1) non-endometrioid cell type, (2) PLN metastasis, (3) LVSI, (4) adnexal metastasis and (5) serosal involvement. Among these 204 patients, 104 had one or more of these factors (group A), and 100 patients had none of these factors (group B). PALN metastasis was significantly greater in group A, compared to group B. The sensitivity and the NPV of these combined prognostic factors for predicting PALN metastasis were 96.2% and 99%, respectively.

Conclusions

Presence of non-endometrioid cell type, PLN metastasis, LVSI, adnexal metastasis or serosal involvement diagnosed by frozen section (FS) seem to be poor prognostic factor for PALN metastasis in EC. Also, PALN dissection should be extended to the level of the renal vessels in all patients who will undergo PALN dissection, due to frequent involvement of the supramesenterial region.     

Selective radiation therapy in stage IB uterine cervical carcinoma following radical pelvic surgery

Thirty-five patients with squamous cell cancer of the cervix, Stage IB, treated by radical hysterectomy and bilateral pelvic lymph node dissection are analyzed for histopathologic factors including (1) depth of cervical stromal invasion, (2) degree of differentiation of the tumor, (3) vascular tumor invasion, and (4) lymph node involvement. The influence of these factors on the outcome is studied. A correlation is found in that the advanced depth of cervical stromal invasion appears to increase the risk of nodal involvement and also the subsequent recurrence in patients with negative nodes. The Grade 2 and Grade 3 disease is more often present in patients with outer one-third cervical stromal invasion, while the vascular tumorous invasion was only found in patients with middle and mostly with outer one-third cervical stromal invasion. A subgroup of patients is described, who are at high risk for recurrent disease following surgery and who may benefit from adjuvant radiation therapy.     

The prognostic value of histopathologic grading parameters and microvessel density in patients with early squamous cell carcinoma of the uterine cervix.

Abstract. Graflund M, Sorbe B, Hussein A, Bryne M, Karlsson M.
The purpose of this study was to investigate the prognostic importance of clinical and histopathologic factors, including malignancy grading systems (MGS), partial index (PI), invasive front grading (IFG), and microvessel density. A complete geographic series of 172 early stage (FIGO I–II) cervical carcinomas treated by Wertheim-Meigs surgery during the period 1965–1990 was studied. The patients were followed up for at least 10 years. Significant prognostic factors for disease-free survival were lymph node status ( P < 0.0000001), radical surgical margins ( P = 0.00003), and tumor size ( P = 0.008). In a multivariate Cox analysis it was shown that lymph node status was the single most important prognostic factor with regard to disease-free survival. The total MGS and the PI scores were highly significantly ( P = 0.0001) associated with pelvic lymph node metastases and disease-free survival rate in squamous cell carcinomas. The MGS and the PI systems were superior to the IFG system in predicting lymph node metastases. The total IFG score was also a statistically highly significant ( P = 0.003) prognostic factor with regard to disease-free survival in both univariate and multivariate analyses. Microvessel density was a nonsignificant prognostic factor. There was a highly significant ( P = 0.002) association between vascular space invasion of tumor cells and the presence of lymph node metastases. In conclusion, histopathologic malignancy grading systems provide valuable prognostic information in patients with early stage squamous cell carcinomas of the uterine cervix.     

The significance of local cellular immune response of women 50 years of age and younger with endometrial carcinoma

OBJECTIVE: Local cellular immune response and traditional histopathologic parameters in endometrial carcinoma patients 50 years old and younger were compared with those of patients older than 50 years of age. MATERIALS AND METHODS: We retrospectively compared the clinicopathologic factors and outcomes of 24 younger and 82 older women who were operated on for endometrioid type endometrial carcinoma at our institution. RESULTS: No significant difference was determined in survival with respect to age and menopausal status in patients with endometrial carcinoma. Younger and older age groups had a similar distribution of most pathologic features including myometrial invasion, cervical involvement, lymph node metastasis, vascular invasion, perivascular lymphocytic infiltrates and tumor infiltrating lymphocytes. While older women had higher-grade tumors, younger women had more frequent ovarian metastasis. Tumor infilrating lymphocytes were seen more frequently in the postmenopausal group than those of the premenopausal group. CONCLUSION: The distribution of local cellular immune response, most histopathologic parameters and survival were the same for younger and older women.     

Expression of vascular endothelial growth factors (VEGF-A/VEGF-1 and VEGF-C/VEGF-2) in postmenopausal uterine endometrial carcinoma

OBJECTIVE: We investigated the expression of two angiogenic vascular endothelial growth factors, VEGF-A/VEGF-1 and VEGF-C/VEGF-2, in 228 cases of uterine endometrial carcinomas from postmenopausal patients to evaluate the correlation with histopathologic features and clinical outcome. METHODS: Immunohistochemistry was used to assess VEGF-A/VEGF-1 and VEGF-C/VEGF-2 expression in 228 primary surgically treated cases of postmenopausal endometrial carcinomas and the results were statistically analyzed in relation to vascular invasion, depth of invasion (myometrial vs serosal-parametrial invasion), lymphatic vessel invasion, lymph node metastasis, disease-free 5-year survival rate (DF5YR), and disease-free 10 year-survival rate (DF10YR). RESULTS: The results of univariate analysis showed that VEGF-A/VEGF-1 and VEGF-C/VEGF-2 expression correlated with vascular invasion (P < 0.0001, P = 0.0006), depth of invasion (P = 0.0004, P = 0.043), lymphatic vessel invasion (P = 0.021, P < 0.0001), lymph node metastasis (P = 0.0017, P = 0.0008), DF5YR (P = 0.0081, P = 0.0002), and DF10YR (P = 0.0077, P = 0.0001). Multivariate analysis showed that lymph node metastasis (P = 0.0017, P = 0.0008), parametrial-serosal invasion (P < 0.0001, P < 0.0001), and VEGF-C/VEGF-2-positive status (P = 0.03, P = 0.01) were significant factors in DF5YR and DF10YR. CONCLUSIONS: We conclude that VEGF-A/VEGF-1 and VEGF-C/VEGF-2 expression was predictive of these histopathologic features of endometrial carcinoma and clinical outcome.     

Identification of histopathologic risk groups in stage IB squamous cell carcinoma of the cervix

This study was undertaken to identify histopathologic risk factors in 100 women with stage IB squamous cell carcinoma of the cervix treated surgically. Histologic factors included maximum depth of stromal invasion, presence of lymph-vascular invasion, mitotic activity, nature of the tumor-stromal borders, plasma cell-lymphocyte stromal response, histologic grade, and metastases to regional lymph nodes. Using a multifactorial analysis, the maximum depth of stromal invasion was found to be the most important prognostic indicator (P less than .0001). The depth of invasion also correlated significantly with the presence of nodal metastases (P less than .0001), lymph-vascular space invasion (P = .0003), and "spreading" versus "pushing" borders (P = .0315). The number of mitoses, grade of tumor, or plasma cell-lymphocyte stromal response did not correlate significantly with depth of stromal invasion. Lymph-vascular involvement, although present in 59% of the patients, did not significantly affect survival. Depth of stromal invasion and lesion diameter were combined to constitute three risk groups: Patients with small size cervical tumors (less than 2 cm), regardless of depth of stromal invasion, as well as patients with intermediate size lesions (2.1 to 3 cm) with stromal invasion less than or equal to 1.5 cm, constituted a low-risk group; the intermediate-risk group was comprised of those patients with cervical lesions between 2.1 and 3 cm in size and deep stromal invasion (greater than 1.5 cm), as well as those patients with large cervical lesions (greater than 3.0 cm) and stromal invasion less than or equal to 1.5 cm.2+ (greater than 3 cm) and deep stromal invasion (greater than 1.5 cm).(ABSTRACT TRUNCATED AT 250 WORDS)     

Myometrial invasion of endometrium cancer assessed by transrectal ultrasonography.

Transrectal ultrasonographic examinations before surgery were performed on 27 patients with stage I endometrial cancer to assess myometrial invasion. The findings were compared with the histopathologic data obtained by surgery. Sensitivity and specificity of transrectal ultrasonography in myometrial invasion of endometrium cancer were 82.6 and 100%, respectively. Therefore, transrectal ultrasonography may be a useful diagnostic tool to determine the myometrial invasion, which is the most important prognostic factor in endometrial cancer.     

Prognostic significance of proliferation in endometrial adenocarcinomas: a multivariate analysis of clinical and flow cytometric variables

In a prospective study, the prognostic significance of clinical and flow cytometric variables was evaluated in univariate and multivariate analyses of 489 cases of endometrial adenocarcinoma. Age, uterine cavity depth, stage, histopathologic grade, myometrial invasion, ploidy level, and S-phase fraction were of prognostic value in the univariate analysis. In a multivariate analysis of clinical variables, age and uterine cavity depth were insignificant. When adding flow cytometric variables (DNA content and S-phase fraction) separately to the clinical variables, the significance of myometrial invasion was reduced, and histopathologic grade became nonsignificant when S-phase fraction was included. When adding both DNA content and S-phase fraction, the DNA content became prognostically insignificant. S-phase fraction was, next to clinical stages III and IV, found to be the strongest predictor of outcome. It is concluded that S-phase fraction is a new, objective parameter that can be used for the identification of prognostically different subgroups of patients within a given stage and grade.     

Prognostic factors affecting disease-free survival in early-stage cervical cancer patients undergoing radical hysterectomy and pelvic-paraaortic lymphadenectomy

OBJECTIVE: This study was performed to identfy surgical and histopathologic prognostic factors that could predict 5-year disease-free survival (DFS) after patients underwent radical hysterectomy and pelvic-paraaortic lymphadenectomy for FIGO Stage I-II cervical carcinoma. METHODS: A retrospective review was performed for all patients undergoing primary radical hysterectomy and pelvic-paraaortic lymphadenectomy for Stage I-II cervical cancer at Ankara Oncology Hospital from 1995 to 2000. Clinical and pathologic variables including age, tumor size (TS), clinical stage, depth of invasion (DI), lymphovascular space involvement (LVSI), cell type, tumor grade, lymph node metastases (LNM), parametrial involvement, surgical margin involvement and pattern of adjuvant therapy were analyzed using univariate analyses. DFS was performed by the Kaplan-Meier method and the log-rank test. Independent prognostic and predictive factors affecting DFS were assessed by the Cox proportional hazard method. RESULTS: Ninety-three patients underwent primary type III radical hysterectomy and pelvic-paraaortic lymphadenectomy. Five-year DFS was 87.1%. LVSI, parametrial involvement and grade were the prognostic factors that independently affected survival. DFS was not significantly different for age, disease status of the surgical margins, tumor size, depth of invasion, cell type, pelvic lymph node metastases and adjuvant radiotherapy. CONCLUSIONS: LVSI, parametrial invasion and histologic grade 2-3 were independent prognostic factors in early-stage cervical cancer patients. Adjuvant radiotherapy in these patients provides no survival advantage.     

MMPs及其调控因子与滋养层细胞的侵袭

滋养层细胞的侵袭行为受严格的时空调控 ,细胞外基质降解是滋养层细胞侵袭过程的关键步骤 ,也是成功妊娠建立的必要条件。基质金属蛋白酶 ( MMPs)是参与细胞外基质降解的关键酶 ,因此也是滋养层细胞侵袭行为调控的关键因子 ,滋养层细胞的自分泌和子宫的旁分泌因子如 :激素、细胞因子、生长因子、基质蛋白等可调节 MMPs活性 ,进而参与滋养层细胞侵袭过程     

Value of preoperative serum CA125 in early-stage adenocarcinoma of the uterine cervix without pelvic lymph node metastasis

OBJECTIVE: This investigation attempted to clarify the value of preoperative serum CA125 in predicting histopathological prognostic factors for early-stage cervical adenocarcinoma without lymph node metastasis. METHODS: This study initially surveyed 163 patients with clinical stage Ib or IIa cervical adenocarcinoma treated with radical hysterectomy and pelvic lymphadenectomy. Of the 163 patients, 116 had preoperative serum CA125 levels, and 14 had pelvic lymph node metastasis. The investigation group comprised 102 lymph node-negative patients. RESULTS: A cutoff value of 26 U/ml was obtained after the discriminant function analysis for identifying patients with positive lymph vascular space invasion (LVSI) or depth of stromal invasion > or =2/3 thickness. Multivariate analysis revealed that among the preoperative clinicopathological variables, including age, tumor size, parametrial invasion, and CA125 level, raised CA125 most significantly influenced the assessment of the LVSI (P = 0.040) and depth of cervical stromal invasion (P = 0.002). CONCLUSIONS: In early-stage cervical adenocarcinoma with negative pelvic lymph node metastasis, preoperative serum CA125 levels at the cutoff value of 26 U/ml impacted the determination of the poor histopathological prognostic factors.     

Factors regulating trophoblast invasion

Implantation in the human is unique. This uniqueness is characterized on the maternal side by a spontaneous and massive decidualization of the endometrium and on the embryonic side by an almost unlimited invasive potential. Human embryos express an intrinsic invasive potential, which allows them to implant almost anywhere except in the endometrium because it protects itself from implantation. Human implantation is thus only possible during a limited period of time known as the implantation window. This mini review stresses the importance of studying trophoblast invasion into the endometrium as a model for human implantation. Cytotrophoblastic cells (CTB) can easily be isolated from first-trimester legal abortions and retain their invasive behavior when cultured in vitro. This model shows that matrix metalloproteinases (MMPs) are produced by CTB and are instrumental to their invasive behavior. Embryo implantation and tumor invasion use these same biochemical mediators for invasion. However, in contrast to tumor invasion, trophoblast invasion is limited both in time and space: it occurs during the first trimester of pregnancy and invasion does not go beyond the proximal third of the myometrium. Factors regulating MMP expression are of maternal and fetal origin.     

Endometrial cancer: factors affecting survival

PURPOSE: To evaluate the influence in survival of clinical and pathological findings in patients with endometrial cancer. METHODS: In 152 women treated for endometrial cancer from 1982 to 1996, personal, obstetrical and oncological data, histology, grade, myometrial invasion, peritoneal cytology, FIGO stage and treatment were correlated with survival. RESULTS: Mean age was of 60.3 +/- 11.1 years old. Eight patients had a previous history of other neoplasms (seven of them gynecological). The mean clinical complaint was abnormal uterine bleeding. The most common histological type was endometrioid (84.9%), only 51 cases did not show myometrial invasion and 119 women were in Stage I at diagnosis. Peritoneal cytology was negative in 113 patients. Seven patients out of 85 in whom lymphadenectomy was performed showed metastasis. Seventeen of the patients died. The factors influencing survival were age, myometrial invasion and lymph node metastasis. CONCLUSION: Lack of myometrial invasion, absence of lymph node metastasis and age younger than 60 years seem to be the most significant predicting factors of survival.     

An analysis on prognostic significance of histopathologic risk factor in uterine endometrial carcinoma]

Prognostic risk factors in histopathologic findings were analyzed in the data for one hundred and thirteen patients with uterine endometrial carcinoma who were treated surgically. Univariate survival analysis with Kaplan-Meier methods revealed that the nuclear grade (p less than 0.005), lymph-vascular space invasion (p less than 0.005), histologic grade (p less than 0.01) and histologic type (p less than 0.01) correlated with the patient's prognosis. Among surgical FIGO stages II and III, there was similar significance in the correlations in the nuclear grade (p less than 0.05), lymph-vascular space invasion (p less than 0.0001) and histologic type (p less than 0.05), although there was difference (p less than 0.05) only between grade 1 and grade 3 in the histologic grade. Multivariate survival analysis with a proportional hazard regression model showed that the nuclear grade (p less than 0.005) and lymph-vascular space invasion (p less than 0.01) correlated significantly with the prognosis. The hazard ratios with a 95% confidence interval for each of these factors were 19.2 (3.2-115.7) and 16.9 (2.1-135.3), respectively. The 3-year survival rate was 98% for a hazard ratio less than 64.7, and 45% for one more than 67.4, between which there was a significant difference (p less than 0.0001). And univariate survival analysis of this hazard ratio also revealed strong correlations with the patient's prognosis. These results suggest that it is of great importance to adopt the hazard model including such factors as the nuclear grade and lymph-vascular space invasion in estimating the patient's prognosis and proceeding to adequate post-operative therapy for individual patient.     

Risk factors for the development of lymph node metastasis in vulvar squamous cell carcinoma.

One hundred and ten women who underwent vulvectomy and inguinal-femoral lymphadenectomy for stages I-IV vulvar squamous cell carcinoma were studied. The most important factors that affected the inguinal lymph node status in the order of importance were vascular invasion, clinical stage, tumor thickness, depth of stromal invasion, and amount of keratin. Fourteen (88%) of 16 tumors with vascular invasion in the primary tumor metastasized. In the absence of vascular invasion, 18 (19%) of 94 tumors metastasized. Overall, 82% of tumors were correctly classified into lymph node negative and positive groups on the basis of vascular invasion. Tumor thickness and depth of stromal invasion had a similar accuracy in predicting lymph node status. The risk of lymph node metastasis increased from 0% when tumor thickness or depth of stromal invasion was less than 2 mm, to over 20% when depth of stromal invasion was greater than 2 mm, and to over 40% when tumor thickness exceeded 4 mm. A combination of vascular invasion, tumor thickness (or depth of stromal invasion), and the amount of keratin correctly classified 97% (76/78) of the lymph node negative group and 63% (20/32) of the positive group with an overall accuracy of 87%. The probability of having lymph node metastasis was computed for individual patients on the basis of one or more pathologic parameters using a logistic regression model. This feasibility is an important step toward individualized therapy for vulvar carcinoma.     

Recommendations for the oncologic pathology report and morphologic factors associated with prognosis in endometrial carcinoma

Summary.The endometrial carcinoma (EC) is the fourth common malignant tumor in women with a reported incidence for Germany of 24.5. The prognostic evaluation shows an overall five-year survival of more than 76 %. The histopathologic report should include the tumor type in detail with special consideration of serous-papillary, clear cell and endometrioid type. Stage T2 a and T2 b should only be diagnosed with clear-cut invasion of endocervical glands or the cervical stroma. In most cases of positive endocervical curetting, endmotrial carcinoma has been dislocated during fractional curetting and represents no indication for advanced surgery (Wertheim-Meigs-procedure). Special reference should be made on isthmic carcinoma, because there is a combination of lymphatic drainage of the cervix and corpus uteri. Well established prognostic factors are surgical tumor stage (pTNM-system), pelvic lymph node metastases and tumor grading. Molecular biologic and cellkinetic parameters should be used for research purpose only.     

Diagnostic problems on frozen section examination of myometrial invasion in patients with endometrial carcinoma with special emphasis on the pitfalls of deep adenomyosis with carcinomatous involvement

Frozen-section and paraffin section diagnoses were compared in 55 patients with Stage I endometrial adenocarcinoma. In 44 patients (80%), a corresponding depth of myometrial invasion and in 54 (98%) patients the same tumor grade were found. Regarding the depth of myometrial invasion and histologic grade, sensitivity, specificity, positive and negative predictive values were 70%, 86%, 73%, 83% and 92%, 100%, 100%, 94%, respectively. Concerning myometrial invasion 9% false-positivity and 10% false-negativity rates were noted. The histopathologic characteristics of false-positive and false-negative patients are emphasized because carcinomatous involvement of deeply situated adenomyosis and advanced grade tumors are the main diagnostic pitfalls. It is important for pathologists to be able to identify carcinomatous involvement of adenomyosis and adjacent foci of minimal myometrial invasion during frozen-section examination which can prevent aggressive surgery.     

Tumor size, depth of invasion, and grading of the invasive tumor front are the main prognostic factors in early squamous cell cervical carcinoma.

OBJECTIVE: The objective of this study was to evaluate the prognostic significance of clinical and histopathologic factors, including a new grading system focusing on the invasive tumor front. METHOD: A retrospective analysis of 125 surgically treated patients with squamous cell cervical carcinoma FIGO stage IB was conducted. For each tumor, the degree of keratinization, nuclear polymorphism, pattern of invasion, and degree of lymphoid infiltration at the invasive tumor front were graded and given scores between 1 and 4. RESULTS: Clinical tumor size, depth of invasion, and grading of the invasive front had prognostic significance in multivariate analysis, while lymph vascular space involvement, lymph node status, and grade of differentiation did not. Based on clinical tumor size, depth of invasion, and grading of the invasive tumor front, patients could be separated into three groups: One group with minimal risk of recurrence (5-year disease-free survival (DFS) of 100%) consisting of 24% of the patients, an intermediate group with a fairly low risk of recurrence (5-year DFS of about 92%), and a high risk group with a 5-year DFS of 45%. This latter group contained 26% of the patients with 78% of all relapses occurring in the total group of patients. The invasive tumor front grading was reliably reproducible, with inter- and intraobserver agreement of 79 to 87% and kappa values of 0.47 to 0.66. CONCLUSION: Clinical tumor size, depth of invasion, and grading of the invasive tumor front were the main predictors of prognosis in patients with stage IB squamous cell cancer of the cervix.