Prognostic Indicators in Stage I Endometrial Adenocarcinoma. With Identification of ... cinous carcinoma, papillary serous carcinoma, and mixed Mullerian ...
Architectural (FIGO) Grading, Nuclear Grading, and Other Prognostic Indicators in Stage I Endometrial Adenocarcinoma With Identification of High-Risk and Low-Risk Groups KHUSHBAKHAT R. MITTAL, MB, BS, PETER E. SCHWARTZ, MD,' AND KENNETH W. BARWICK, MD
We studied 164 cases of Stage I endometrial adenocarcinoma to determine the relative prognostic value of International Federation of Gynecology and Obstetrics (FIGO) and nuclear grading systems. Other factors known to be of prognostic value in endometrial carcinoma also were evaluated. Both the FIGO and nuclear grading systems correlated with five-year mortality rate from cancer. Nuclear Grade 3 proved to be a superior predictor of fatal outcome (nine of 13 169761) over FIGO Grade 3 (four of 13 13181). We advocate the combined use of FIG0 and nuclear grading systems, along with other prognostic parameters, for the detection of most patients with fatal cancer. Cancer 61:538-545, 1988.
T
of endometrial carcinoma was shown to influence clinical outcome in a number of clinicopathologic studies.'-" Most of these studies used Broder's'' or the International Federation of Gynecology and Obstetrics (FIGO) grading system^,^*^-*^'^-' I which are based on the proportion of the tumor with a solid component. Nuclear grade was investigated by others and found to be useful prognostically, or superior to the FIGO grade in prognostication of these patients.394 Nuclear grade also was shown to influence the prognosis in adenocarcinoma of other sites including the prostate,I3 kidney,I4 p a n c r e a ~ , 'and ~ breast.I6 We studied cases of Stage I endometrial carcinoma diagnosed from 1970 to 1980 to evaluate the relative prognostic importance of the nuclear and FIGO grading systems. Other factors known to be of prognostic significance in endometrial carcinoma also were evaluated.
I . Follow-up of less than 5 years unless the patient died from endometrial carcinoma. 2. Presence of another primary carcinoma concurrently in the same patient, or appearance of another primary carcinoma within 5 years of followUP. 3. Patients treated with less than a total hysterectomy and bilateral salpingo-oophorectomy. 4. Specific subtypes of adenocarcinoma, namely clear cell carcinoma, adenosquamous carcinoma, mucinous carcinoma, papillary serous carcinoma, and mixed Mullerian tumors. 5. Patients without diagnostic slides available for review.
HE HISTOLOGIC GRADE
Subjects and Methods All patients with clinical and pathologic Stage I endometrial adenocarcinoma seen at Yale-New Haven Hospital between 1970 and 1980 were reviewed. Patients with the following features were excluded from the study. Presented in part at the 1987 meeting of the International Academy of Pathology. Chicago. Illinois, March 8-1 3, 1987. From the Departments of Pathology and *Obstetrics and Gynecology, Yale University School of Medicine. New Haven, Connecticut. Address for reprints: Khushbakhat R. Mittal, MD, Department of Pathology, Bellevue Hospital, Room 1 IN31C, 27th Street and First Avenue, New York. N Y 10016. Accepted for publication August 25, 1987.
Of the 580 patients with Stage I endometrial carcinoma seen during this period, 164 could be studied. Follow-up information on all patients was obtained from hospital and office medical records and tumor registry data.
FIGO Grading All tumors were graded according to the FIGO system.' Lesions without solid areas were classified as Grade 1 (Fig. I), tumors with less than 50% solid area were classified as Grade 2 (Fig. 2), and those with greater than 50% solid area were classified as Grade 3 (Fig. 3). Benign squamous differentiation foci were not counted as solid areas. The criteria used for defining benign squamous differentiation was the presence of keratinization demonstrated in hematoxylin and eosin (H & E) stained slides, intercellular bridges, or a decreased nu-
538
No. 3
-
STAGEI ENDOMETRIALADENOCARCINOMA Mittal el al.
539
FIG. 1. FIG0 Grade 1 adenocarcinoma of the endometrium (H & E, X200)
clear/cytoplasmic (N/C) ratio in an area of sheet-like growth. These areas often were in the form of morules. Nirclear Grading
Nuclear grading was done on a three-grade system: ( 1) Grade 1: minimal to mild variation in nuclear size and shape (Fig. 4), (2) Grade 2: moderate variation in nu-
FIG. 2. FIG0 Grade 2 adenocarcinoma of the endometrium (H & E. X200).
clear size and shape (Fig. 9, and (3) Grade 3: Marked variation in nuclear size and shape (Fig. 6). Each tumor was given the highest nuclear grade expressed in either biopsy or hysterectomy specimen. Mitotic count per 10 high-power field (HPF) (X45 objective and X 10 eyepiece) also was counted in each case. Also, nucleolar grade and N/C ratio grades were evaluated. Nucleolar grades were defined as ( 1 ) Grade 1: nu-
540
CANCERFebruary I 1988
Vol. 61
FIG. 3. FIG0 Grade 3 adenocarcinoma of the endometrium (H & E, X200h
cleoli absent or present but not conspicuous, (2) Grade 2: prominent nucleoli, and (3) Grade 3: very prominent nucleoli. Nuclear/cytoplasmic ratio grades were defined as (1) Grade 1: minimally to mildly increased N/C ratio, (2) Grade 2: moderately increased N/C ratio, and (3) Grade 3: markedly incrased N/C ratio. Other authors included nucleolar and N/C ratio grades in nuclear grade, but we studied them separately.
The slides were given code numbers and were reviewed by one of us (K.R.M.) without knowledge of the clinical outcome. After the cases were graded and studied, the code was broken and clinical information correlated with the morphologic features of the tumors. The depth of myometnal invasion, presence or absence of vascular invasion, and diffuse infiltrating pattern of growth also were recorded in each case. A diffuse infiltrating pattern
FIG.4. Nuclear Grade I adenocarcinoma of the endometrium (H & E. X400).
No. 3
-
STAGE1 ENDOMETRIALADENOCARCINOMA Mittal et al.
54 I
FIG.5. Nuclear Grade 2 adenocarcinoma of the endometrium (H& E. X400).
of growth was defined as the presence of an invasive tumor predominantly as single glands (Fig. 7). All the patients of this study received preoperative intracavitary radium (ICR) and postoperative vaginal plaque (VP) with the following exceptions: four patients with FIGO Grade 1 tumors did not receive either ICR or VP, three received only ICR. four received only VP, and seven received preoperative external beam radiation
FIG.6 . Nuclear Grade 3 adenocarcinoma of the endometrium ( H & E, X400).
therapy in addition to ICR and VP. Three FIGO Grade 2 patients received only ICR, seven received only VP, five received preoperative external beam radiation therapy in addition to ICR and VP, and three received postoperative pelvic radiation in addition to ICR and VP. Two FIGO Grade 3 patients received only ICR and two received postoperative external beam radiation therapy in addition to ICR and VP.
CANCERFebruary I 1988
542
Vol. 61
FIG.7. Diffuse infiltrating patte'rnof growth (H& E. X40).
All patients that died of endometrial adenocarcinoma within 5 years were treated with ICR and VP with the following exceptions: one FIGO Grade 1 patient also received preoperative external beam radiation therapy, one FIGO Grade 2 patient received only VP, and two received additional postoperative external beam radia-
TABLEl . Distribution of Cases by FIGO and Nuclear Grades Nuclear grade
FIGO grade
I
I 2
2
3
FIGO total
52
14
22
5 18
71
3
38 3
3
9
Nuclear total
93
39
32
78 15
FIGO International Federation of Gynecology and Obstetrics.
TABLE2. Five-Year Mortality Rates by FIGO and Nuclear Grading
DOD at 5 yrs
No. of patients
(%I
-
71 78 15
4 (5.6%)
I
93 39 32
2(2.1%) 2 (5. I % ) 9 (28.1%)
FlGO grade 1
7
3 Nuclear grade
2 3
5 (6.4%) 4 (26.6%)
FIGO: International Federation of Gynecology and Obstetrics; DOD: dead of disease.
tion therapy. Also, one FIGO Grade 3 patient received only ICR and one received additional postoperative external beam radiation therapy. Results
Table 1 shows the distribution of all cases in FIGO and nuclear grading systems. There were approximately twice as many patients in nuclear Grade 3 (32)compared with FIGO Grade 3 ( 15). There were less patients in nuclear Grade 2 (39) compared with FIGO Grade 2 (78). More patients were in nuclear Grade 1 (93)than FIGO Grade 1 (71). Most FIGO Grade 1 tumors also were nuclear Grade I and vice versa. The same was true of R G O Grade 3. However, there was considerable spread in the distnbution of FIGO Grade 2 tumors. FIGO Grade 2 tumors usually had only Grade 1 nuclei. A considerable number of FIGO Grade 2 tumors ( 1 8 of 78) were nuclear Grade 3. Table 2 shows the 5-year mortality rate in various FIGO and nuclear grades. The 5-year mortality rates for FIGO Grades 1, 2, and 3 were 5.6%. 6.4%, and 26.6%, respectively. Nuclear Grades 1,2,and 3 had 2.1%, 5.I%, and 28.1% mortality rates, respectively, at 5 years. More patients with fatal cancer were detected by nuclear Grade 3 (nine of 13 [69%])compared with FlGO Grade 3 (four of 13 [31'701). Nucleolar Grades I, 2, and 3 had 5-year mortality rates of 6.376,9.l%, and 15%, respectively (Table 3). N/C ratio Grades I , 2,and 3 had 5-year mortality rates of 3.3%, 15.4%,and 3 1.270,respectively (Table 3).
-
STAGEI ENDOMETRIAL ADENOCARCINOMA Mittal et al.
No. 3
TABLE3. Five-Year Mortality Rates by Nucleolar and N/C Ratio Grades ~
~
~
No. of patients
~
TABLE4. Five-Year Mortality Rates by Combined Nuclear Grade Combined nuclear grade
No. of patients
DOD at 5 yr (%)
I
11 51 36
2 (2.6%) 2 (3.9%) 9 (25%)
DOD at 5 yr (%)
2 3
Nucleolar grade 1 2 3
543
Ill 33 20
7 (6.3%) 3 (9.1%) 3 (15%)
122 26 16
4 (3.3%) 4 (1 5.4%) 5 (3 1.2%)
DOD: dead of disease.
N/C ratio grade 1
2 3
N/C: nuclear/cytoplasmic; DOD: dead of disease.
Both the nucleolar Grade 3 and N/C ratio Grade 3 detected only a minority of fatal cases (three of 13 and five of 13, respectively) as did FIGO Grade 3 (four of 13). Table 4 shows the results with “combined nuclear” grade consisting of patients given the highest grade achieved on nuclear, nucleolar, or N/C ratio grades. The results are similar to those obtained by using the nuclear pleomorphism grading alone. Table 5 shows the effect of combining the nuclear (nuclear pleomorphism) and FIGO grading systems and giving each patient the higher grade achieved. The 5year mortality rates with this combined grade were 1.9%,2.6%, and 28.696, respectively, for Grade 1,2, and 3 tumors. Ten of 13 fatal cases were detected by the combined FIGO and nuclear Grade 3. The number of mitoses were found to be related to the 5-year mortality rate (Table 6). Tumors with less than 10 mitoses/lO HPF had a 6.2% 5-year mortality rate compared with a mortality rate of 20% in patients with equal to or greater than 10 mitoses/lO HPF. A minority of patients (four of 13) had equal to or greater than 10 mitoses/ 10 HPF. The mortality rate also was related to the percentage of tumor with a solid appearance (Table 7). The 5-year mortality rate was 6.2%, 5%, 18.1% and 5Wo in tumors with 0% to lo%, 1 1% to 50%, 51% to go%, and greater than 80% solid area, respectively. A minority of patients with fatal cancer (four of 13) had greater than 50% solid areas in their tumors. Six of 164 tumors showed vascular invasion. The 5year mortality rate in these patients was 66% (four of six) (Table 8). The 5-year mortality rate in patients without vascular invasive tumor was 5.6% (nine of 158). All fatal cases of carcinoma with vascular invasion, except one, had greater than 90% invasion of the myometrium. Only one case with less than 90% (1 5%) invasion of the myometrium had an FIGO and nuclear Grade 3 tumor. Two patients who had vascular invasion but did not die of their cancer had 40% and 90% invasion of the myometrium, respectively. The FIGO Grade in both of these
TABLE5. Five-Year Mortality Rates by Combined FIGO and Nuclear Grades Combined FIGO and nuclear grade 1
2 3
No. of patients
DOD at 5 yr (%)
52 77 35
I (1.9%) 2 (2.6%) 10 (28.6%)
FIGO: International Federation of Gynecology and Obstetrics; DOD: dead of disease.
cases was 2. The nuclear Grades in these two patients were 2 and 1, respectively. The nuclear Grade in all patients with vascular invasion and fatal cancer was 3. FIGO Grades in these same patients were 1,2, 3, and 3, respectively. The tumor depth was related to the 5-year mortality rate (Table 9). Only 4.1% of the patients with no evidence of residual tumor at hysterectomy died of their cancer. The 5-year mortality rates were 14.9% and 33.3%, respectively, in patients with less than 50% and patients with greater than 50% invasion of the myometrium. Only three of 13 patients with fatal cancer had deep (>50%)myornetrial invasion. A diffuse infiltrating pattern of growth, with individual glands infiltrating the myometrium (Fig. 7), was correlated with poor survival. Two of the five patients with this growth pattern died from their cancer. TABLE6. Five-Year Mortality Rates by Mitotic Activity No. of mitoses ~~
No. of patients
DOD at 5 yr (9%)
144 20
9 (6.2%) 4 (20%)
~
10/10 HPF
DOD: dead of disease. TABLE7. Five-Year Mortality Rates by Percentage of Solid Area in Tumor Percent of solid area
No. of patients
DOD at 5 yr (%)
0% to 10% 11% to 50% 5 1To to 80% 81% to
I29
8 (6.2%) 1 (5%) 2 (18.1%) 2 (50%)
DOD: dead of disease.
20 11 4
CANCERFebruary 1 1988
544
TABLE8. Five-Year Mortality Rates by Vascular Invasion Vascular invasion
No. of patients
DOD at 5 yr (%)
Present Absent
6 158
4 (66.6%) 9 (5.6%)
DOD dead of disease. TABLE 9. Five-Year Mortality Rates by Depth of Myometrial Invasion Tumor depth
No. of patients
DOD (%I
No residual tumor ~50% >50%
121
5 (4.1%) 5 (14.9%) 3 (33.3%)
34 9 ~~~~~
DOD: dead of disease.
Discussion
The results of this study show that both FIGO and nuclear grades correlate well with prognosis in Stage I endometrial adenocarcinoma (Table 2). Our findings on the FIGO grading system agree with those of othen1.5-8.10 who have reported correlation of prognosis with the architectural grade of the tumor. Typically, all of these architectural grading studies detected a minority of patients with fatal cancer as Grade 3. The percentage of patients detected as Grade 3 were 19% (seven of 37),' 32%( I7 of 52),8 44%(43 of 98): and 28%( 1 1 of 40).1° Nuclear grading detected many more cases of fatal cancer as Grades 3 and 4 in one study reported by Austin and MacMohang (1 56 of 207 [75%]).Others3 found nuclear grading to be superior to FIG0 grading, but still detected only a minority of fatal cases in Grade 3 (seven of 26 [28%]).The definition of nuclear grade in the latter study3 was different from ours, and apparently was not based on nuclear pleomorphism. Instead, the grading was based, as far as the published material indicates, on nuclear shape, nuclear size, nucleolar size, nucleolar irregularity, and number of mitoses3 Results of the latter study3 may have been influenced by the relatively poor prognostic ability of nucleolar grade. In our study, only
Vol. 61
23%(three of 13) of the patients with fatal cancer were detected by nucleolar Grade 3. Our findings on the effect of vascular invasion and depth of myometrial invasion agree with the findings of others.l.h4.6-8.10.1I We also found a relationship between mitotic activity and prognosis. Patients having greater than 10 mitosis/ 10 HPF had a 20%5-year mortality rate compared with a 6.2%mortality rate in those with less than 10 mitoses/ I0 HPF. A combination of FIGO and nuclear grading systems was found to be useful (Table 5 ) in detecting more cases (ten of 13) of fatal cancer without any reduction in percentage mortality rate in Grade 3 tumors. A number of studies based on the cytometric analysis of DNA content have found poor prognosis in patients with endometrial cancer showing aneuploid DNA c ~ n t e n t . ' ~We -'~ believe that high nuclear grade may be a histologic marker of aneuploid tumors. It is unclear if an advantage is gained by doing DNA measurements after a tumor has been graded by the nuclear grade. AtkinI7 reported additional benefits of using DNA measurement after endometrial carcinoma grading. The grading system used was not defined but was apparently based on architectural rather than nuclear features. Others found no statistically significant relationship between DNA levels of endometrial carcinoma and prognosis. I 9 Because the mortality rates in FIGO and nuclear Grade 1 and 2 tumors as defined in this study were low and similar, the FIGO and nuclear Grade 1 and 2 tumors may be combined into a low-risk group. Similarly, patients with FIGO Grade 3, and/or nuclear Grade 3, and/or deep (>50%)myornetrial invasion, and/or vascular invasion, and/or diffuse infiltrating pattern of growth may be classified as a high-risk group. This high-risk group would have a cancer-related 5-year mortality rate of approximately 30%(28.2%),and would include approximately 85% of the patients with fatal cancer. The remaining patients would belong to a lowrisk group with a 5-year cancer-related mortality rate of approximately 2%( 1.6%)(Table 10). REFERENCES
TABLE10. Low-Risk and High-Risk Categories of Endometrial Carcinoma Histologic features
Low risk
High risk
FIGO grade Nuclear grade Myometrial invasion Diffuse infiltrating pattern Vascular invasion
I and 2 I and 2 50% Present Present
1.6% (2/ 125)
28.2% ( I 1/39)
Five-yr mortality rate
FIGO: International Federation of Gynecology and Obstetrics.
I. Frick HC, Munnell EW, Richart RM, Berger AP, Lawry MF. Carcinoma of the endometrium. Am JObster Gynecoll973; 1 15:663676. 2. F.I.G.O. News. In1 J Cynecol Obster 197 I ;9: 172- 180. 3. Christopherson WM, Connelly PJ, Alberhasky RC. Carcinoma of the endometrium: V. An analysis of prognosticators in patients with favorable subtypes and stage I disease. Cancer 1983; 5 I :1705- 1709. 4. Creasman WT, Soper JT, M e a r t y KS Jr, McCarty KS Sr. Hinshaw W, Clarke-Pearson DL. Influence of cytoplasmic steroid receptor content on prognosis of early stage endometrial carcinoma. Am J O b stet Gynecol 1985; 15 1 :922-932. 5. Mahle AE. The morphologic histology of adenocarcinoma of the body of the uterus in relation to longevity. Surg Gynecor Obsrer 1923; 361385-395.
No. 3
STAGEI ENDOMETRIAL ADENOCARCINOMA * Mitral et al.
6. Cheon HK. Prognosis of endometrial carcinoma. Obstet Gynecol 1969; 34680-684. 7. Ng ABP, Reagan JM. Incidence and prognosis of endometrial carcinoma by histologic grade and extent. Obstet Gynecol 1970; 35:437-443. 8. Nahhas WA, Lund U, Rudolph JH. Carcinoma of the corpus uteri: A 10-year review of 225 patients. Obstet Gynecol 197 I ; 38564570. 9. Austin JH, MacMahon B. Indicators of prognosis in carcinoma of the corpus uteri. Surg Gynecol Obster 1969; 126:1247-1252. 10. Barber KW, Dockerty MB, Pratt JM, Hunt AB. Prognosis in endometrial carcinoma by modified Duke’s typing. Surg Gynecol Obstet 1962; 114:155-162. 11. Baltzer J, Lohe KJ, Kuenzl R, Scheer KP, Zander J. Prognostic criteria in patients with endometrial cancer. Arch Gynecoi 1983; 234: 121- 129. 12. Broders AC. Carcinoma: Grading and practical applications. Arch Puthol 1926; 2:376-381. 13. Gaeta FJ, Asinvatham JE, Miller GE, Murphy GP. Histologic
grading of primary prostatic cancer: A new approach to an old problem. J Urol 1980; 123:689-693.
545
14. Fuhrman SA, Lasky LC, Limas C. Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Puthol 1982; 6:655-663. 15. Kloppel A, Lingenthal A, Von Bulow M, Kern HF. Histological
and fine structural features of pancreatic ductal adenocarcinomas in relation to growth and prognosis: Studies in xenografted tumors and clinico-histopathological correlation in a series of 75 cases. Histoputhology 1985; 9:841-856. 16. Lash RH, Bauer TW, Hermann RE, Esselstyn CB. Partial mastectomy: Pathologic findings and prognosis. Hum Puthol 1986; 17:813-822. 17. Atkin NB. Prognostic significance of ploidy level in human tumors: 1. Carcinoma of the uterus. J Nut1 Cancer Inst 1976; 56909910. 18. Naus GJ, DeVere R, Richart RM, Deitch AD. Predictive value
of flow cytometric analysis of paraffin embedded tissue in endometrial carcinoma (Abstr). Lab Invest 1985; 52:48. 19. Geisinger KR, Homesley MD, Morgan TM, Kute TE.Marshall RB. Endometrial adenocarcinoma: A multiparametric clinicopathologic analysis including the DNA profile and the sex steroid hormone receptors. Cancer 1986; 58:1518-1525.
