Endometrial adenocarcinoma - presenting pathology is a poor guide to surgical management

Awl

N Z J Obstpt Cynoecol2ooo;40:2: 191-194

GYNAECOLOGICAL ONCOLOGY

Endometrial adenocarcinorna - presenting pathology is a poor guide to surgical management Rodney W Petersen,' Julie A Q ~ i n l i v a nGabrielle ,~ R Casperl and James L Nicklid Departments of Obstetrics and G y m b g y , University of Queensland,' Brisbane, Queensland,Australia und Flinders University2W o r d Park, South Australia,Australia

SUMMARY We aimed to evaluate the correlation between the histological grade of endometrial cancer diagnosed on endometrial biopsy or curettage, with the definitive grade and stage of lesion as determined by surgery and histopathological examination and to make recommendations about the suitability of conservative surgery based on preoperative determination of the grade of endometrial adenocarcinoma. A retrospective review of all patients with endometrii adenocarcinoma presenting to the Queensland Centre for Gynaecological Cancer from 1 January 1996 to 31 December 1998 was undertaken. Clinical and pathological data was abstracted h m medical records and case notes of 460 patients. All histological specimens were prospectively reviewed by a panel consisting of gynaecologic pathologists, &ynaecologic oncologists and other doctors involved in the treatment of patients with gynaecological malignancies.The percentage of patients whose management would have been optimised by firll surgical staging at the time of initial surgery was calculated.

INTRODUCTION Endometrial cancer is the commonest femde genital tract cancer in the United States and Australia.' Ninety per cent of patients present with abnormal vaginal bleeding, most commonly postmenopausaI. The cancer is usually diagnosed on histopathology of endometrial curettings or biopsy Presenting histopathology is classified as grade 1 , 2 or 3 according to established criteria based on the relative proportion of solid to glandular architectural features and the Address for correspondence Dr Rodney W Petersen Queensland Centre for Gynaecotogical Oncology E floor. Clinical Sciences Building Royal Brisbane Hospital Queensland. Australia Rodney W Petersen MRANZCOG MBA Fellow in Gynaecological Oncology, Julie A Quinlivan FRANZCOG Clinical Lecturer, Gabrielle R Casper FRANZCOG Fellow in Gynaecological Oncology, James L Nicklm FRANZCOG CGO Senior LecturerGynaecological Oncologist

Only 60%. ?1%, and 84% of the patients with a presenting diagnosis of grade 1, 2 and 3 endometrial adenocarcinomas respectively had this confirmed on final histopatholow. Furthermore, using established criteria, 30%, 46% and 100% of patients presenting with grade 1, 2 and 3 endometrial adenocarcinoma required full surgical staging at the time of their primary surgery There is poor correlation between the pre-operative grade of endometrial cancer and the grade as determined on analysis of the resected uterus. The correlation is poorest with grade 1 endometrial adenocarcinoma, where strongest consideration is given to conservative surgery and the avoidance of subspecialty referral. There is a strong argument that all patients with a diagnosis of endometrial cancer made on endometrial biopsy or curettage, regardless of grade of malignancy, should be offered surgery where the option to perform concurrent comprehensive surgical staging is available.

extent of nuclear atypia? The grade of tumour at diagnosis may contribute to the decision as to whether subspecialty referral is made. Since 1988 FIG0 staging has been based upon surgico-pathologicalcriteria? The principal aims of surgical staging are to identify a high-risk group of patients that require post-operative radiotherapy or other adjuvant therapies and to provide accurate prognostic data. This allows the majority of patients to avoid unnecessary radiotherapy. Patients with pelvic and para-aortic lymph node metastases constitute a high-risk group and have a poor prognosis, with only 33% five-year survivaI.1.2.3.4 Increasing tumour grade is associated with an increased risk of lymph node metastases. The incidence of pelvic node metastases in grade 1 tumours is reported to be 2.8%,compared with 8.7%and 18.3%in grade 2 and 3 tumours respectively3The reported incidence of distant metastases is 2.2% in grade 1 tumours, compared to 10.2%and 39%for grade 2 and 3 tumours respectively4

192

ANWOG

This data suggests that a high percentage of patients presenting with grade 2 or 3 endometrial adenocarcinoma will require full surgical staging including peritoneal washings, pelvic and/or para-aortic node lymphadenectomy and/or other biopsies. Conversely, the low incidence of pelvic node metastases reported in grade 1 tumours has led to the belief that these tumours can be managed by simple hysterectomy and bilateral salpingo-oophorectomy. However, information on the incidence of lymph node metastases was derived from studies relating final, rather than presenting, tumour grade?.* It has been assumed that the presenting grade of the tumour obtained from endometrial sampling accurately reflects the final or 'true' grade of the tumour. However, pathological study of the entire uterus may upgrade 15-25°i~of women with grade 1t ~ m o u r s . ~ The present study was conducted to determine the correlation between the presenting histopathology of endometrial curettings o r biopsy of the endometrium and final histopathology of the uterine specimen with respect to tumour grade and stage. Furthermore, we examined how presenting tumour grade correlated with the percentage of patients who could be optimally treated by simple hysterectomy, without the need for lymphadenectomy or other advanced surgical skills.

METHODS The records of all patients with uterine cancer registered with the Queensland Centre for Gynaecological Cancer between 1 Janum, 1996 and 31 December, 1998 were reviewed and the following data retrieved: (i) the pathological grade and tumour cell type from the endometrial curettings or biopsy that lead to the initial diagnosis of malignancy; (ii) the operative record and intra-operative pathology report on tumour grade and depth of myometrial invasion on frozen section; and (iii) final tumour classification code, stage, cell type, depth of myometrial invasion, and grade. Tumour grade was classified according to established criteria.2 Those adenocarcinomas with squamous differentiation were graded according to the nuclear grade of the glandular component. All presenting histopathology was reviewed prior to surgery at a joint gynaecological oncology clinicopathology meeting to determine tumour grade. A panel of gynaecologic pathologists, gynaecologic oncologists and other doctors involved in the treatment of patients with gyaecological malignancies were in attendance. All cases were further evaluated to determine whether the findings at surgery and/or on intra-operative frozen section indicated that formal staging with pelvic lymph node dissection and para-aortic node sampling was indicated. The protocol to perform full surgical staging is outlined in Table 1. In the absence of these criteria, less than 1% of patients will have occult para-aortic lymph node metasta~es.'.~.~.~

Table 1 The criteria for full surgical 0

0 0

a

Grade 1 or 2 turnour with greater than 50% myornetrial invasion on intra-operative frozen section examination Grade3tumour Cervical extension of tumour Clear cell or papillary serous differentiation

'Derived from Hacker NF. Uterine cancer. In: Berek JS. Hacker NF, prmricalgymlogironmlogy 2nd edition, 1941. William and Wilkins,

Maryland, 285-326 and M o m w CP. Cunin JP. Management of uterine neoplasia. In: Morrow CP. Cunin JP (eds)Gynecologu- Comer Surge0 New York. Churchill Livingston. 1996. j64626;.

There were three protocol violations in cases of

extreme age and coexisting medical morbidity where medical opinion was that pelvic lymph node dissection posed a life threatening operative risk to the patient. These three cases were staged clinically according to the 1971 FIGO classification. The remaining cases were staged surgically according to the 1988 FIGO surgical staging system?

RESULTS During the study period, 460 patients with uterine can. cer were registered with the Queensland Centre for Gynaecological Cancer. Of these patients, 172 (3790), 171 (37%) and 86 (1990)had presenting pathological diagnoses of grade 1 (well dserentiated). grade 2 (moderately differentiated) and grade 3 (poorly differentiated) adenocarcinoma of the endometrium respectively The remaining 31 patients with endometrial adenocarcinoma initially presented with atypical endometrial hyperplasia. uterine sarcoma o r another variant and were excluded from further analysis. leaving 429 patients as the subject of this report. Table 2 summarises presenting versus final pathe logical grade. Only 60°0, 7lo/o. and Moo of patients presenting with grade 1 . 2 and 3 endometrial adenocarcinomas had this grade confirmed on final histopathob ogy Thus. accuracy impmved as the p m n t l n g tumour grade increa.4. Of note, 40"0and 1 3 O 0 of paticnts prr. senting with grade 1 and 2 endomrtrial carcinomas were upgraded on final pathology. No cases p m n t i n g as grade 1 endometrial caminnma wcw downgmdcd. even if the final pathology demonstrat4 no midual tumour. because review of the Initial cuwttinht dcmon. strated welldifferentiated adenocarclnoma. Thew riiscs were included amongst those classlflrd as stnge l A endometrial adenocarcinoma. Grade 2 tumours wcrp downgraded to grade 1 in 1 6 O 0 of patients.

Table 2 Presenting versus final tumour grade NIUII grade

RODNEYW PETERSEN m AL

Table 3 summarizes presenting grade versus final stage of endometrial adenocarcinoma. Of the 172 patients presenting with grade 1 disease, 145 (84%), 8 (50b), 15 YO), and 4 (2%) were stage 1,2,3 and 4 respectively. Using the defined protocols (Table l),52 patients (30%) who presented with grade I tumours required full surgical staging and 4 patients (2%) required debulking of advanced disease. Five of the 52 patients who were surgically staged had lymph node metastases, a rate of 10%. Table 3 Presenting grade versus final stage. Final

Presenting tumour grade

I

2

n (W

n (%)

n (%t

145 (84)

127 (74)

48 (56)

4

4 (2)

4 12)

15 (17)

Total

I72

171

86

stage

I

3

2

3

Of the 171 patients with grade I1 endometrial adenocarcinoma. 127 (74%), 20 (12%). 20 (12%), and 4 (2%) were stage 1 , 2 , 3 and 4 respectivelp Using the defined protocols (Table 2), 79 patients (46%)required full surgical staging and 4 patients (2%) required debulking of advanced disease. Eleven of the 79 patients who were surgically staged had lymph node metastases, a rate of 1496.

All of the 86 patients who presented with grade 111 tumours underwent full surgical staging with 48 (56%).6 (7%), 17 (20%), and 15 (17%)being classified as stage 1, 2, 3 and 4 respectively Fourteen patients had lymph node metastases, a rate of 20%.

DISCUSSION A key finding of this study was that only 60%, 71 YOand 84% of the patients with a presenting diagnosis of grade 1 , 2 or 3 endometrial cancer respectively,had the grade confirmed on final pathology The discrepancy between the grade of adenocarcinoma diagnosed at presentation and that on final histopathology is most likely due to a sampling error in the original diagnostic procedure. It has been reported previously that pathological study of the entire uterus upgrades 15-25% of all grade 1 tumours.5 However, difficulties can also arise because of poor concordance between pathologists with respect to the FIG0 grading system. One study concluded that the E G O grading system was subjective, with a correlation coeficient of 0.60 between a general pathology resident and a senior gynaecologicalpathologist8 Furthermore, using defined criteria (Table 1),30% and 46% of patients presenting with grade 1 and 2 tumours respectively required surgical staging at the time of their primary operation for optimal management.

193

It has been reported previously that 22%of grade 1 tumours have disease invading the middle or outer third of the myometrium, and retroperitoneal nodal metastases occur in 3 to 4% of women with supeSicially invasive grade 1 tumours. These findings have led to the suggestion that at least 30% of women with a pre-operative diagnosis of a grade 1 tumour are at high risk of retroperitoneal disease6 Our finding that 30% of patients with grade 1 tumours and 46% of patients with grade 2 tumours would have their management optimised by performing fuU.surgical staging at the time of initial surgery is consistent with this suggestion. If these patients underwent simple hysterectomy and salpingo-oophorectomy alone, the gynaecologist would be faced with the option of recommending either adjuvant post-operative radiotherapy or submitting the patient to a further staging operation? For patients with disease greater than stage 1B there is evidence that full surgical staging conveys a survival advantage6and will result in optimal patient management.1,2.7,9,10.11,12 Of all patients with disease apparently confined to the uterus, 15% will have lymph node metastases.3~~ Previously, these patients would all have been irradiated meaning that 85% would receive 'unnecessary' adjuvant treatment. Of the 15% with pelvic node metastases, a half to twothirds will have positive para-aortic lymph nodes. A conventional radiotherapy field encompassing the whole pelvis would therefore undertreat 7-10% of all patients?J3 In the present study, positive lymph nodes were identified in 10% and 14% of patients who presented with grade 1 and 2 tumours respectively and whose disease fulfilled the criteria summarised in Table 1. This confiims that these criteria identify a high-risk sub-group of women with endometrial cancer. Furthermore, the negative lymph node pathology in 90% and 86% of these high-risk patients meant that external beam radiotherapy, with its consequent morbidity, could be safely omitted. If surgical staging has not been performed, the gynaecologist cannot predict which patients have retroperitoneal disease and therefore is required to subject all patients in this high-risk group to either post-operative radiotherapy or repeated surgery for the purpose of formal surgical staging? The combination of surgery and radiotherapy is associated with significant morbidity in up to 12?0 of patients.11-12~14~15 These complications include proctitis, rectal stricture, radiation enteritis, small bowel obstruction, rectovaginal fistulae, haemorrhagic cystitis, bilateral ureteric obstruction, enterovaginal fistulae and death.ll.12 Furthermore, the recovery from the physical and psychological stress of radiotherapy may be slow. Significant levels of nausea, anorexia, diarrhoea and fatigue have been documented in these patients up to 3 months following therapyl4Ja

194

ANZJOG

The alternative choice is to re-operate to perform formal staging procedure. This is associated with significant morbidity resulting from repeating major surgery within a short period of time. Specific risks include febrile morbidity, pulmonary embolus, blood loss and transfusion, small bowel obstruction, fistulae formation and adverse psychological sequelae. Gynaecologicalcancer patients are reported to experience crisis feelings and marked sense of helplessness for up to 2 months after their initial surgery.15 Reoperation within this time frame exaggerates these psychological disruptions. Furthermore, such surgery may be technically more difficult as acute reoperation is associated with loss of surgical planes. If full surgical staging is performed at the time of the primary operation, the mean additional time is only W45 minutes.16It has been reported that staging increases blood loss by 60-150mL and there is no increase in overall rates of blood transfusion or total length of hospital stay16 The additional risks of formal surgical staging, if performed at the time of total abdominal hysterectomy and salpingo-oophorectomy, are reported to be 13%.This is largely dominated by the risk of persistent lymphocyst formation, which may require surgical drainage.' Asymptomatic or spontaneously resolving lymphocysts are more common but do not cause major morbidity7 Rarer complications include acute vascular damage and obturator nerve injw,but both are reported to occur in less than 1% of cases.16 In the present study, 30% of patients presenting with grade 1and 46% of patients with grade 2 tumours had features suggestive of high-risk endometrial disease warranting surgical staging. It is apparent that simple hysterectomy and salpingooophorectomy is suboptimal treatment for these women. There is a strong argument that all patients with a diagnosis of endometrial adenocarcinoma made on endometrial biopsy or curettage, regardless of grade of malignancy, should be offered surgery where the option to perform concurrent comprehensive surgical staging is available.

REFERENCES 1 Hacker NF. Uterine cancer. In: Berek JS. Hacker NF. Practical gynecologic oncology 2"d edition, 1991. William and Wilkins.

Maryland. S 3 3 6 . 2 FIG0 stages - 1988revisions: Vulva, ovary, corpus. Gynecol Oncol 1989:35: 125-127. 3 Creasman WT. Morrow CP. Bundy BN. et al. Surgical pathologic spread patterns of endometrial cancer: A Gynecologic Oncology Group Study. Cancer 1987: 60:2035-21341. 4 DiSaia PJ, Creasman WT, Boronow RC, Blessing J A . Risk factors and recurrent patterns in stage 1endometrial cancer. Am J Obsrez G y m 1 1985; 151: 1009-1015. 5 Daniel AG. Peters WA. Accuracy of office and operating Mom curettage in the grading of endometrial carcinoma. Obsrer G y m J 198a;71: 612-614. 6 Orr JW, Orr PF, Taylor PT.Surgical staging endometrial cancer. Clin Ohm G y m 1 1%; 39:65f&1%8. 7 Kilgore LC, Partridge EE. Alvarez RD. el al. Adenocarcinoma of the endometrium: survival comparisons of patients with and without pelvic node sampling. G y m 1 Oncol1995: 56: 24-33. 8 Nordstrom B, Strang P, Lindgren A. Bergstrom R. Tribukait B. Carcinoma of the endometrium: do the nuclear grade and DNA ploidy provide more prognostic information than do the FlGO and WHO classifications? fnr J G y m 1 farhol 1996: 15: 191 201. 9 Aalders J. A b l e r V. Kolstad P. Onsrud M.Postoperative external irradiation and prognostic parameters in stage 1 endometrial carcinoma: clinical and histopathologic study of 540 patients. UhsM GynacOr1980; 56: 419-427. 10 Homesley HD.Kadar N. Barrett R J , Lentz SS. Selective pelvic and peri-aortic lymphadenectomy does not increase morbidity in sur. gical staging of endometrial carcinoma Am J O W C y m l 1 9 9 2 : 167: 1225-1230. 11 Larson DM, Copeland W.Gallager HS. Kong JP. Wahnon JT. Stringer CA. Stage 11 endometrial carcinoma: results and compli. cations of a combined radiotherapeutic - surgical approach. C a w 1988: 61: 1528-1534. 12 KinseUa TJ.Bloomer WD. Lavin PT.Knapp RS. Stage II endomr. trial carcinoma: 10 year foUow.up of combined radiation and sur. gical treatment. G y m 1 OnrOl1980; 10: 290 297. 13 Boronow RC. M o m w CP.Creasman WT et al. Surgical staging in endometrial cancer: clinimpatholqic findings d a p x m , p r t l v ~ study. Ubstpr G y m I 1W:63: w . 14 Nail 1.M. King KR. Johnson JE. Coping with radiation ttratmmt for gynecologtc cancer: M o d and diaruptlon in usual lunc-thin .I &whomm ObsfprGynanl I=: 5: 271 15 Gottesman 0.Irwts M. DifTwenm In crisis martioms nrnong ran cer and surgery pstlenta. .I Consult Clin h y r h o l 1 % ~:.a'.-I .PIR 16 M o m w CP. Curtin JP Mannwment d u t p r i n m p b h In M o m w CP. Cunln J P (eds)C ; y n r r o l ~ + ~Cnnm k~ .wmm NVW York. Church111 I,ivlnmtnn. 199R. 5BB Rar)