195例子宫颈病变LEEP术后切缘状态及全子宫切除标本病变残留情况的分析

目的:探讨宫颈病变LEEP术后切缘状态以及全子宫切除标本中病变残留情况。方法:收集2008年11月至2009年6月本院195例LEEP术后行全子宫切除术病例的临床病理资料,对比分析其LEEP术后切缘状态及全子宫切除标本病变残留情况。结果:在本院行阴道镜下多点活检组织学诊断的162例患者,其活检诊断与LEEP术后病理诊断的总体完全符合率为75.3%,LEEP术后病理诊断升级者占19.8%,降级者占4.9%。195例患者中,128例因LEEP切缘阳性行全子宫切除术,67例切缘阴性者因其他高危因素行全子宫切除术。128例LEEP切缘阳性患者中,75.8%未在全子宫切除标本中发现任何上皮内病变;当LEEP切缘为HSIL、CIS、微小浸润癌和浸润癌时,子宫颈中HSIL以上程度病变残留率分别为13.2%、19.6%、21.7%和70.0%,HSIL以上程度的病变残留率随LEEP切缘病变程度增加而升高。67例LEEP切缘阴性者的全子宫切除标本中,4例(6.0%)发现残留病灶。结论:除LEEP切缘为浸润癌阳性,其它阳性切缘状态并非必须行全子宫切除术,而需进一步评估。规范LEEP过程可能有助于减少阳性切缘比例。     

The risk of residual neoplasia in women with microinvasive squamous cervical carcinoma and positive cone margins

The objective was to evaluate the prevalence and factors affecting residual disease in women with cervical microinvasive carcinoma (MIC) with positive cone margins for high-grade lesions and invasive carcinoma. We reviewed histopathology slides of 129 women with MIC who had high-grade lesions or invasive carcinoma at cone margins. These patients underwent hysterectomy following cone biopsy between January 1994 and June 2004. Of the 129 patients, 77 (59.7%) had residual disease in the hysterectomy specimens, in which 57 (44.2%) had residual high-grade lesions. Twenty patients (15.5%) had residual invasive carcinoma: 18 were microinvasive and 2 were invasive. Factors significantly affecting the risk of residual disease included positive postconization endocervical curettage (P= 0.001), positive cone margins for invasive carcinoma (P= 0.003), and depth of stromal invasion >1 mm (P= 0.014). Cox proportional hazards analysis revealed positive cone margins for invasive carcinoma as significant predictor of residual invasive disease (hazard ratio, 3.22; 95% CI 1.21-8.60, P= 0.019) In summary, patients with MIC and positive cone margins for high-grade lesions or invasive carcinoma are at high risk of residual neoplasia. Repeat cone biopsy should be performed to determine exactly the severity of lesion before planning treatment.     

The role of radical parametrectomy in the treatment of occult cervical carcinoma after extrafascial hysterectomy

OBJECTIVES: To assess the morbidity and efficacy of radical parametrectomy (RP) performed following extrafascial hysterectomy in patients with occult cervical carcinoma. METHODS: An IRB approved retrospective chart review identified 23 patients that underwent RP with pelvic and/or para-aortic lymphadenectomy and upper vaginectomy. Data were collected on demographics, tumor stage, grade, histology, indication for hysterectomy, surgical findings, complications, recurrence, and survival. RESULTS: Of the 23 patients, 2 patients had a stage IA(2) lesion while 21 patients had a stage IB(1) lesion. There were 5 patients with a grade 1 tumor, 10 with grade 2, 4 with grade 3, and 4 with unknown grade. Median age was 41 years (range 27-59). The most common indication (48%) for extrafascial hysterectomy was CIS of the cervix. Four patients (17%) had metastasis to pelvic nodes or evidence of tumor at the margin at the time of RP. Three of these 4 patients with a positive specimen received adjuvant radiation and all are alive (mean follow-up 66 months). One patient declined radiation and is alive at 42 months. There were 7 (30%) operative complications: Most notably 4 patients received blood transfusions. Two of 19 patients (11%) with no residual tumor in RP specimen recurred and 1 patient was salvaged with radiation (follow-up 103 months). With a median follow-up of 61 months (range 9-103), overall 5-year survival is 96%. CONCLUSIONS: RP is an acceptable option for patients diagnosed with an occult cervical carcinoma at the time of extrafascial hysterectomy. Careful selection of RP for patients unlikely to have residual tumor will obviate the need for radiation in most instances.     

Correlation of cone biopsy with findings at radical hysterectomy and use of adjuvant radiation therapy

Objective

To determine if pathologic findings in cone biopsy specimens correlate with residual invasive disease in radical hysterectomy specimens and the need for adjuvant chemo-radiation therapy.

Study design

We identified 65 patients who underwent a cone biopsy and subsequent radical hysterectomy. Clinico-pathologic parameters in the cone specimens were correlated with the presence of residual invasive disease in the radical hysterectomy specimens and the need for adjuvant chemo-radiation.

Results

A positive endocervical margin, a positive deep margin, a positive post-cone ECC, and positive LVSI were significantly associated with the presence of residual disease in the radical hysterectomy specimen, while positive LVSI, a positive ECC, a positive deep cone margin, and greater than 1 positive margin were significantly associated with the use of adjuvant chemo-radiation therapy.

Conclusion

Pathologic parameters in cone biopsy specimens can estimate the risk of residual invasive disease in radical hysterectomy specimens and the use of adjuvant chemo-radiation.     

Value of endocervical margin examination of conization specimens. Prospective study conducted on 150 patients

OBJECTIVE: To assess the usefulness of frozen sections (FS) on endocervical margin in surgical conization or loop electrosurgical specimens. MATERIAL AND METHODS: In a prospective study, 150 patients were treated from October 1995 to December 1997: 69 cases without FS, 81 cases with FS. CIN on frozen section resulted in an immediate additional resection. RESULTS: In the group without FS, 13 patients had involved endocervical margin by high-grade CIN (18.8%). Frozen section was impossible in a conization specimen that was too short. FS revealed 64 normal glandular epitheliums, seven squamous metaplasias in which two lesions were under-evaluated (being in fact CIN on permanent sections), eight high-grade CIN followed by additional resection in six cases and two invasive carcinomas. Endocervical margin on additionals section were always free of disease. The rate of failure was 2.6% among 77 cases. This rate corresponded to two under-evaluations. Invasive carcinoma and CIN without additional resection were excluded because frozen section only allowed a peroperative diagnosis. The average height of the cone and the rate of complications were similar. Repeat surgery was necessary in nine cases in the group without frozen section, in which five showed residual lesions, absent in the other group. CONCLUSION: The ultimate histological interpretation was never difficult after frozen section. This method permits reduction of cases with involved cone margin and residual lesions and, despite some limitations, it may be useful for surgical management.     

The value of frozen sectioning for the evaluation of resection margins in cases of conization

Objectives To determine the role of frozen section examination (FSE) of the cone specimen in the evaluation of the resection margin status and to rule out invasion in patients with high-grade cervical intraepithelial neoplasia. Methods Thirty patients with cervical intraepithelial neoplasia underwent conization followed by FSE and planned hysterectomy. The results of the definitive paraffin exam were compared with FSE. Results In evaluation of the margins by FSE, 4 patients (13%) had positive cone margines and 26 (87%) had negative margins. The definitive of paraffin examination of margin status was concordant in all the cases. Intraoperative diagnosis of invasion was made in three cases such that all of them were invasive squamous cell carcinoma. Among the remaining 27 cases, we detected five CIN1, two CIN2 and three CIN3, so the diagnosis of the FSE was concordant with paraffin section in 26 out of 30 cases (87%). Also we detected four additional CIN (one CIN 1, two CIN2 and one CIN3) after paraffin study, whose frozen specimens were reported normal. Conclusion Frozen section examination can provide immediate and precise evaluation of the cone margin status in high-grade cervical intraepithelial neoplasia. It can identify frank invasion and permit adequate treatment in a one-stage procedure and reliably detect clear resection margins.     

Evaluating cervical cone biopsy specimens with frozen sections at hysterectomy

Frozen section evaluations of cervical cone biopsy specimens were performed at the time of hysterectomy to exclude invasive cervical cancer. During a two-year period we prospectively evaluated 43 cone biopsy specimens. We found all the diagnoses made with frozen sections to be accurate when compared with prospective permanent sections, and all patients received appropriate therapy. Thirty-eight cases showed no evidence of invasion. Two patients had invasive squamous cell cervical cancer, one had invasive cervical adenocarcinoma extending to the endometrial cavity, and two had microinvasion. All invasive cancers were diagnosed correctly with frozen sections and confirmed with permanent sections. When hysterectomy immediately followed conization, no complications occurred, and no significant increase in blood loss was noted. We found frozen section evaluation of a cone biopsy specimen at the time of hysterectomy to be a reliable procedure that saves time, eliminates the risk of additional anesthesia and decreases patients' costs.     

Early invasive adenocarcinoma of the uterine cervix: criteria for nonradical surgical treatment

OBJECTIVE: This retrospective study was undertaken to identify selection criteria for nonradical surgery for early invasive adenocarcinoma of the uterine cervix. METHODS: Seventy-nine patients with surgically treated cervical adenocarcinomas (with invasion to 5 mm or less) were examined clinicopathologically. The evaluation of stromal invasion was conducted according to the FIGO (1995) staging system. RESULTS: The mean age was 46 (range: 29-73) years, and the median follow-up was 118 (9-348) months. Definitive treatment modalities included radical hysterectomy in 71 (89.9%) cases, modified radical hysterectomy in 2 (2.5%), and simple extrafascial hysterectomy without pelvic lymphadenectomy in 6 (7.6%). Postoperative adjuvant external radiation therapy was given to 5 (6.3%) patients. The histological subtypes were endocervical in 37 (46.8%) cases, endometrioid in 32 (40.5%), and adenosquamous in 10 (12.7%). Forty-one (51.9%) patients had lesions with up to 3 mm of stromal invasion; of these, 24 (58.5%) had lesions with up to 7 mm of horizontal extension (stage IA1). Thirty-eight (48.1%) patients had lesions with stromal invasion greater than 3 mm and no greater than 5 mm; of these, 4 had lesions with no wider than 7 mm of horizontal extension (stage IA2). Of 73 patients with pelvic lymphadenectomy, one (1.4%) tumor (depth: 5 mm; width: 15 mm) had node metastases. Parametrial involvement was present in one (1.4%) patient (lesion depth: 5 mm; lesion width: 16 mm). None had adnexal metastasis. Eighty-eight percent of the patients with stromal invasion up to 3 mm had well-differentiated adenocarcinoma, compared to 53% of the patients with lesions invading more than 3 mm. Of all of the patients, 5 (6.3%) patients who received curative radical hysterectomies had recurrences and died. Among 5 patients, one patient with central pelvic recurrence had a lesion invading to a depth of 3 mm and width of 7 mm, and the others had lesions with more than 3 mm of invasion and 15 to 36 mm of width. CONCLUSIONS: Patients with early invasive adenocarcinoma to a depth of 3 mm or less stromal invasion, including those who meet the criteria for FIGO stage IA1, may be treated with simple extrafascial hysterectomy without lymphadenectomy and oophorectomy.     

Laparoscopic modified radical hysterectomy: a strategy for a clinical dilemma

OBJECTIVE: To investigate the role of laparoscopic modified radical (type 2) hysterectomy when cervical cancer cannot be excluded or documented preoperatively. METHODS: Between 1996 and 2004, 50 patients with cervical intraepithelial neoplasia (CIN III) or adenocarcinoma in situ (AIS) involvement of cone endocervical margins and/or endocervical curettings, who were not candidates for observation or repeat conization, underwent laparoscopy to perform a modified radical hysterectomy. RESULTS: Forty-nine (98.0%) modified radical hysterectomies were completed laparoscopically and one (2.0%) patient required a laparotomy. Of the overall group, 35 (70.0%) had residual pathology; 26 (52.0%) were precancerous lesions, and 9 (18.0%) had invasive disease (5 adenocarcinomas, 3 squamous lesions, and 1 adenosquamous carcinoma). Of the nine with cancer, one had stage IA1 disease, three had stage IA2 disease, and five had stage IB1 disease. Five (55.6%) invasive lesions were diagnosed intraoperatively (frozen section), and a laparoscopic pelvic and lower aortic lymph node dissection was performed. The median operative time was 96 min (range 58-185), blood loss 100 ml (50-450), and postoperative hospital stay 2.5 days (range 1-14). There were no incidences of prolonged urinary retention fistulas, or other serious complications. All patients with cancer remain disease-free (median follow-up 44.2 months, range 1-88.7 months). CONCLUSIONS: Laparoscopic modified radical hysterectomy is a treatment option for patients for whom cervical cancer cannot be definitively excluded, and can be completed with acceptable operative time, blood loss, and hospitalization.     

Negative predictive value of human papillomavirus test following conization of the cervix uteri.

OBJECTIVE: The goal of this study was to determine/evaluate the negative predictive value of human papillomavirus (HPV) testing following conization of cervix uteri. METHODS: A prospective analysis was undertaken on 79 cone biopsies of women with high-grade lesions (cervical intraepithelial neoplasia (CIN) III). HPV testing was performed on cervical smears before and after conization. We correlated the margin status (defined as positive cone margin or endocervical curettage status) and positive conization HPV status with the residual disease in a hysterectomy specimen. A Digene II kit was used to perform HPV testing. HPV detection was done by Hybrid Capture assay. RESULTS: Of the 79 patients, 47(59.5%) had positive margins after conization. HPV testing was positive in 37 cases (78.7%) and negative in 10 cases (21.3%). Residual disease was found in 31 of 47 (66%) postconization hysterectomy specimens. No residual lesions were found in HPV-negative cases. Of the 32 cases with negative margins following conization, HPV testing was negative in 25 cases (78%) and was positive in 7 cases (22%). Among these 25 cases with negative HPV tests, no residual lesion was detected, and in 7 HPV-positive cases, only one residual lesion was found. CONCLUSION: HPV testing is potentially an effective tool in predicting residual dysplasia after conization and could potentially assist in the decision between hysterectomy and conservative follow-up in women with CIN III.     

Frozen section evaluation of cervical conization specimens

Frozen section evaluation was done on 96 conization specimens. One patient had invasive carcinoma, which was correctly diagnosed on frozen section. No patient received inappropriate therapy on the basis of an erroneous diagnosis made from the evaluation of frozen sections. The mean time from initiation of surgery to the surgeon's receipt of the frozen section diagnosis was 0.9 hours. There was no significant increase in the blood loss or intraoperative complication rate when frozen conization was added to either abdominal or vaginal hysterectomy.     

Extended experience in the use of laparoscopic ultrasound to detect pelvic nodal metastasis in patients with cervical carcinoma

OBJECTIVE: To evaluate the use of laparoscopic ultrasound (USG) to detect pelvic nodal metastasis in patients with early stage cervical carcinoma. METHODS: Laparoscopic USG was used to search for pelvic lymph node metastasis in stage Ia2 to IIa cervical carcinoma patients before radical hysterectomy. Suspicious lymph nodes identified by laparoscopic USG were removed laparoscopically for pathological confirmation by frozen section. If nodal metastasis was diagnosed, radical hysterectomy would be cancelled but enlarged lymph nodes were removed preferably by laparoscopic approach before closing the abdomen. These patients were treated with radiotherapy after recovering from the surgery. By comparing the laparoscopic USG and pathological findings of lymph nodes removed with or without radical hysterectomy, diagnostic accuracy of laparoscopic USG was determined. RESULTS: Ninety-three patients were recruited and the final analysis included 90 patients. Laparoscopic USG found suspicious lymph nodes in 17 patients and nodal metastases were confirmed pathologically in 14 of them. Three patients with macroscopic and five patients with microscopic pelvic nodal metastases were missed by laparoscopic USG. The accuracy, sensitivity, specificity, positive and negative predictive value of laparoscopic USG in detecting pelvic lymph node metastasis were 87.8%, 63.6%, 95.6%, 82.4%, and 89%, respectively. Macroscopic metastatic nodes were successfully removed laparoscopically in 11 out of 14 patients and laparotomy was required for the other three patients. CONCLUSIONS: Laparoscopic USG can be performed with no major morbidity. This technique is sensitive in detecting macroscopic but not microscopic metastatic pelvic lymph nodes. Removal of macroscopic metastatic nodes identified via laparoscopic USG via laparoscopic approach could be accomplished in majority of patients.     

Intravaginal amputation of the uterine cervix with frozen section examination of the endocervical margin: a review of 414 consecutive patients

We report our institutional experience with the accuracy and usefulness of cervical amputations with frozen section evaluation of the endocervical margin in the management of preinvasive squamous epithelial lesions. Four hundred and fourteen consecutive patients, who underwent amputation of the cervix because of a preinvasive epithelial lesion, or discrepancy between cytologic and biopsy findings especially when colposcopic evaluation was unsatisfactory, had frozen section evaluation of the endocervical margin. Medical records were reviewed and pathologic findings were compared with those obtained on paraffin embedded sections. Frozen section analysis of the upper endocervical margin led to the diagnosis of a residual lesion in 90 (21.7%) cases. In 59 (14.2%) of these cases a further excision was performed during the same operative procedure leading to complete resection in 34 (8.2%) cases. In 403 (97.3%) cases the diagnosis based on the frozen section was corroborated by the permanent sections. For the diagnosis of insufficient cervical resection, the sensitivity and specificity of frozen sections were 93.8% and 99.7% respectively. We conclude that frozen section evaluation of the upper endocervical margin at the time of cervical amputation is a reliable procedure that increases the rate of complete resection. The risks associated with additional anesthesia are then reduced, as are inconvenience to the patients and costs.     

Is it necessary to make screening pap smears after hysterectomy?

OBJECTIVE: To precise the risk of cancer of the vagina after hysterectomy. PATIENTS AND METHODS: In our file of cervicovaginal and vulvar pathology, we looked for all VAIN and invasive cancers of the vagina on a 10-year period. RESULTS: Out of 2152 patients, we found but 45 cases, 13 of which only after total or radical hysterectomy: 4 cases of invasive cancer of the vagina (1 after radical hysterectomy for invasive cancer of the cervix, and 3 after total hysterectomy for CIN); 9 cases of VAIN (5 after total hysterectomy for CIN; and 4 VAIN (3 after radical hysterectomy for cervical invasion). DISCUSSION AND CONCLUSIONS: In our series, we did not observe precancerous or invasive lesion of the vagina after hysterectomy for benign lesion. Indeed, the 13 cases of invasive or in situ cancers of the vagina we found had undergone simple or radical hysterectomy for cervical lesion. We think that the cytological follow-up of the vaginal vault after hysterectomy for benign lesion can be, if not stopped, at least quite spaced out. On the other hand, the follow-up must be imperatively maintained in the event of hysterectomy for precancerous lesion or cancer of cervix.     

Clinicopathologic features of early adenocarcinoma of the cervix initially managed with cervical conization

OBJECTIVE: To evaluate the clinicopathologic features of microinvasive adenocarcinoma of the cervix in order to guide the management of patients with this disease. MATERIALS AND METHODS: A retrospective review was conducted of patients diagnosed with early invasive, 1 mm and 2 mm and 3 mm and 4 mm and 相似文献   

A prospective surgical pathological study of stage I squamous carcinoma of the cervix: A Gynecologic Oncology Group study

Thirty-three institutions collaborating in the Gynecologic Oncology Group gathered surgical and pathological data on 1125 patients with primary, previously untreated, histologically confirmed stage I cervical carcinoma with more than 3 mm of invasion who were selected to undergo radical hysterectomy and paraaortic and pelvic lymphadenectomy. Of the 940 eligible, evaluable patients, 732 had squamous carcinoma. Of the study group, 87 (12%) did not undergo radical hysterectomy because of gross disease beyond the uterus or microscopic aortic node involvement documented at exploratory laparotomy. Among the 645 patients undergoing pelvic and paraaortic lymphadenectomy and radical hysterectomy, five risk factors were significantly associated with microscopic pelvic lymph node metastasis: depth of invasion (P = 0.0001), parametrial involvement (P = 0.0001), capillary-lymphatic space invasion (P = 0.0001), tumor grade (P = 0.01), and gross versus occult primary tumor (P = 0.009). The factors identified as independent risk factors for pelvic lymph node metastasis by multivariate analysis were capillary-lymphatic space involvement (P less than 0.0001), depth of invasion (P less than 0.0001), parametrial involvement (P = 0.0005), and age (P = 0.02). The model was used to predict the chance of a patient having nodal metastasis for any combination of risk factors.     

Microinvasive Carcinoma of the Uterine Cervix: Histological Findings on Cone Specimens Related to Residual Neoplasia on Hysterectomy

The treatment of cervical microinvasive carcinoma is controversial. Hysterectomy is performed in almost all cases, associated or not with more radical procedures. Currently, there is a tendency to adopt conservative management to treat patients with early invasion, as long as it can be assured that the whole lesion has been removed. The aim of this study was to establish which histological information should be obtained from the cones that would give the best possible assurance of absence of residual neoplasia in the patient. This was done by comparing cone and hysterectomy specimens from each patient. One hundred sixty-three cases, treated from 1967 to 1994, underwent simple or radical hysterectomy following cone biopsy. We evaluated the following histological features in the cones: (i) invasion depth, (ii) lateral extension of the lesion, (iii) unifocal or extensive lesion, (iv) vascular invasion, (v) morphological signs of HPV infection, and (vi) free or involved cone surgical margins. Residual neoplasia in the histerectomy was more frequent when the margins of the cone were involved by atypical epithelium, and in cases with signs of HPV infection. However, according to statistical analysis, these two variables were not mutually independent, and the only important parameter to predict residual neoplasia in the hysterectomy specimens was involved surgical margins in the cone.     

Lateral microscopic extension of squamous cell carcinoma of the vulva

PURPOSE: The aim of this study was to measure the radial occult microscopic spread of tumor in patients with invasive squamous cell carcinoma of the vulva. MATERIALS AND METHODS: In the operating room the gross tumor border was marked. The pathologist took a radial section in each quadrant and measured the distance of occult lateral spread of the tumor. RESULTS: From 7/01/93 to 6/30/96, 24 tumors from 21 patients were studied. The mean maximum tumor diameter was 3. 2 cm (0.5-7.0) and the mean depth of invasion was 9.1 mm (1.1-28.0). The gross and microscopic extent correlated in 20 tumors. Maximum lateral microscopic extent of the other 4 tumors was 3.5, 5 (to the margin), 10, and 16 mm. These 4 tumors were ulcerative and infiltrative and arose from or involved mucosa. CONCLUSION: The gross and microscopic periphery of most invasive squamous vulvar cancers are approximately the same. Ulcerative tumors with an infiltrative pattern of invasion which involve mucosal epithelium may be more likely to extend beyond what is grossly apparent. Measurement of the tumor-free margin should be included in future studies.     

Benign granular cell tumor mimicking carcinoma of the breast. Report of a case.

A case of granular cell tumor of the breast in a 59 year-old woman is presented. Clinical evaluation suggested carcinoma with infiltration of the skin. The granular cell tumor was diagnosed on frozen section and the lesion was treated by wide local excision including the overlying skin. Granular cell tumors of the breast, which are usually benign, may closely mimic breast carcinoma, both clinically and on frozen section. The possibility of granular cell tumor of the breast with its potential for a false-positive diagnosis on frozen section supports a two-step procedure for the treatment of breast cancer, especially in young women, to prevent inappropriate radical surgery.     

The prognostic significance of radiation dose and residual tumor in the treatment of barrel-shaped endophytic cervical carcinoma

OBJECTIVE: The aim of this study was to evaluate the impact of total radiation dose on residual tumor and the prognostic significance of persistent disease in women with bulky, barrel-shaped cervical carcinoma who received definitive radiation followed by adjuvant hysterectomy. METHODS: The medical records of 57 patients with bulky endophytic cervical carcinoma treated at the University of Washington between 1976 and 1997 were reviewed. All patients received external beam pelvic radiotherapy supplemented by intracavitary brachytherapy, followed by extrafascial hysterectomy 6 to 8 weeks later. RESULTS: The mean pretreatment tumor diameter was 5.9 cm, with a range of 4-9 cm. Total radiation dose to point A ranged from 5040 to 9700 cGy, and the mean for the group was 7966 cGy. Residual disease was present in 35 (61%) of the hysterectomy specimens. The frequency of cervical tumor sterilization correlated significantly with the mean radiation dose to point A (P = 0.016). Patients without histologic residual disease had a significantly improved outcome, with 95% of patients remaining clinically free of disease at last follow-up, versus 31% of those with residual disease (P < 0.001). As expected, the pelvic control rate was excellent (100%) in patients with complete tumor eradication compared to the group with residual tumor (44%). Those with no residual disease enjoyed a significantly improved survival compared to those with residual tumor (P < 0.001). Furthermore, a statistically significant higher survival was realized in patients harboring only microscopic residual compared to those with either macroscopically evident tumor residuum and/or positive surgical margins (P = 0.036). CONCLUSIONS: Higher radiation doses are associated with an improved likelihood of tumor eradication in the treatment of bulky, endophytic cervical cancer and complete tumor sterilization at adjuvant hysterectomy is predictive of significantly enhanced survival and pelvic control. The high rate of histologic tumor persistence in our series emphasizes the need for more efficacious therapies in patients with bulky endophytic cervical cancer and argues for escalation of radiation dose even when adjuvant hysterectomy is planned.