SEARCH

SEARCH BY CITATION

Keywords:

  • endometrial carcinoma;
  • histologic grade;
  • prognosis;
  • interobserver variability

Abstract

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

BACKGROUND

The most widely used histologic grading system for endometrial carcinoma is the three-tiered International Federation of Gynecology and Obstetrics (FIGO) system. Although FIGO grading has significant predictive value, the reproducibility of Grade 2 is limited. Recently, a binary grading system was proposed based on the amount of solid growth, the pattern of myometrial invasion, and the presence of tumor cell necrosis. The authors analyzed and compared the prognostic significance and the interobserver variability of both grading systems and of the three criteria for the binary grading system.

METHODS

Eight hundred patients with Stage I–III endometrioid endometrial carcinoma were reviewed and graded independently by two pathologists according to the three-tiered FIGO grading system and the novel binary grading system.

RESULTS

The interobserver agreement for both systems was moderate, with 70% and 73% agreement rates for the FIGO (κ = 0.41) and binary (κ = 0.39) grading systems, respectively. When converting the FIGO grading system into an artificial, 2-tiered grading system (Grade 3 vs. Grades 1–2), the agreement was much better (agreement rate, 85%; κ = 0.58). Of the 3 criteria for the binary grading system, amount of solid growth (≤ 50% vs. > 50%) had the greatest reproducibility (agreement rate, 80%; κ = 0.50). Both the 2-tiered FIGO grading system and the binary grading system were significant predictors of local recurrence, distant recurrence, and disease-specific survival (hazard ratios [HRs]: 1.7, 2.5, and 2.6, respectively, for FIGO and 2.1, 4.1, and 3.8, respectively, for the binary grading system). The amount of solid growth also was a strong prognostic factor for these three endpoints (HRs: 2.4, 3.9, and 3.8, respectively).

CONCLUSIONS

Both the binary grading system and the FIGO grading system had strong prognostic significance. Their reproducibility, however, was limited. A simple architectural binary grading system that divided tumors into low-grade lesions and high-grade lesions based on the proportion of solid growth (≤ 50% or > 50%) had superior prognostic power and greater reproducibility. Cancer 2004;100:764–72. © 2004 American Cancer Society.

The most widely used grading system for endometrial carcinoma is the three-tiered International Federation of Gynecology and Obstetrics (FIGO) system. This histologic grading system is based on both architectural (proportion of solid growth) and cytonuclear criteria.1 It has been shown that FIGO grading has significant predictive value.2–5 However, in a previous retrospective study of 253 patients with Stage I–III endometrial carcinoma,5 we found that patients with Grade 1 or 2 tumors had equal disease-specific survival rates (92% and 94% 5-year survival, respectively), whereas patients with Grade 3 tumors had a significantly worse outcome (63% 5-year survival), indicating that, from a clinical point of view, the discrimination between Grade 1 and Grade 2 is not very useful. Moreover, it was shown that the reproducibility with which Grade 1 and 2 tumors were identified was poor. Initial pathologic examination identified 21%, 57%, and 22% Grade 1, 2, and 3 tumors, respectively, compared with 67%, 8%, and 25% Grade 1, 2, and 3 tumors, respectively, at review. A similar shift from Grade 2 to Grade 1 was reported in a large, Dutch, multicenter, prospective, randomized study, known as the Post Operative Radiation Therapy in Endometrial Carcinoma (PORTEC) trial,3 which included 714 patients with Stage I endometrial carcinoma. Lax et al.6 have proposed a new binary grading system that discriminates between high-grade and low-grade tumors based on the proportion of solid growth (≤ 50% vs. > 50%), the pattern of myometrial invasion (infiltrating vs. pushing border), and the presence of tumor cell necrosis. Those authors found that their binary system had superior interobserver agreement compared with the traditional, three-tiered FIGO grading system. Furthermore, it proved to have prognostic significance for survival.

We conducted a review of 800 patients with endometrioid adenocarcinoma of the endometrium. All slides were reviewed independently by two pathologists and were graded according to both the three-tiered FIGO grading system and the novel binary grading system proposed by Lax et al.6 The objectives of the current study were to analyze and compare the prognostic significance, as well as the interobserver variability, of both grading systems and, separately, to analyze and compare the three criteria for the binary grading system.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Patients

A total of 800 patients with endometrioid endometrial carcinoma from two different studies were analyzed. The histopathologic slides of hysterectomy specimens from 231 patients (91%) from a retrospective study of 253 patients with Stage I–III endometrial carcinoma who underwent surgery and received radiotherapy at the Leiden University Medical Center5 were obtained. Follow-up data were collected from the patient charts, from the cancer registry, and from the patients' general practitioners. All patients were followed for a minimum of 6 years or until death (median follow-up, 9.8 years).

In addition, histopathologic slides of hysterectomy specimens from 569 of 714 patients (80%) who were included in the PORTEC trial3 were obtained. That trial included patients with FIGO Stage I endometrial carcinoma who were randomized to receive pelvic radiotherapy or no further treatment after undergoing a total abdominal hysterectomy and bilateral salpingo-oophorectomy. Of the 569 patients who were included in the current study, 279 patients received postoperative radiotherapy, and 290 patients received no further treatment. The median follow-up was 6.5 years.

Patient, tumor, and treatment characteristics are shown in Table 1. The histologic diagnosis of endometrial carcinoma was confirmed in all patients.

Table 1. Patient, Tumor, and Treatment Characteristics
CharacteristicNo. of patients (%)
  1. FIGO: International Federation of Gynecology and Obstetrics.

Age (yrs) 
 < 60203 (25)
 ≥ 60597 (75)
 Mean66.3
 Range40–91
Radiotherapy 
 Yes506 (63)
 No294 (37)
FIGO stage 
 I740 (92)
 II 32 (4)
 III 28 (4)
Myometrial invasion 
 < 50%309 (39)
 ≥ 50%491 (61)
Total800 (100)

Grading Systems

The 800 slides were graded independently by 2 pathologists (V.S. and H.B., who were blinded to patient outcomes) using 2 different systems: the 3-tiered FIGO grading system and the recently proposed binary grading system. The FIGO grading system is based primarily on the proportion of nonsquamous solid growth. Grade 1 tumors have ≤ 5% areas of solid growth, Grade 2 tumors have 6–50% areas of solid growth, and Grade 3 tumors have > 50% areas of solid growth. In patients with marked notable nuclear atypia, a Grade of 1 or 2 is raised by one grade. Notable nuclear atypia is defined as the presence of cells with markedly enlarged, pleomorphic nuclei that display irregular, coarse chromatin and prominent, eosinophilic nucleoli.7 The FIGO grading system was analyzed not only as a three-tiered grading system but also when converted into an artificial, two-tiered grading system (Grade 3 vs. Grades 1–2).

The binary grading system classifies a tumor as ‘high grade’ if at least 2 of 3 architectural features are present: 1) > 50% solid growth, without distinction between squamous and nonsquamous differentiation; 2) a diffusely infiltrative growth pattern characterized by irregularly distributed glands, masses, cords, or nests of tumor cells infiltrating the myometrium haphazardly, in contrast to an expansive growth pattern in which the invasive tumor has a lobulated appearance with pushing borders; and 3) tumor cell necrosis, defined as areas of necrotic tumor immediately adjacent to viable tumor, whereas necrotic debris within masses of squamous epithelium is not qualified as tumor cell necrosis.6 The pattern of myometrial invasion was judged as infiltrative if glands or tumor cells haphazardly infiltrated the myometrium, by which there was no longer a more or less smooth border between tumor and myometrium, even when this consisted of only a small part of the tumor. In addition to the analyses of the FIGO and binary grading systems, the three criteria of the binary grading system also were analyzed separately for their prognostic significance and interobserver variability.

For tumors that were graded differently by the two pathologists, a consensus grade was obtained during a joint evaluation session. These consensus grades were used for the recurrence and survival analyses.

Statistics

Analysis of the interobserver variability was based on the percentage of agreement and was assessed by the κ statistic. The κ values, as measurements of agreement, are interpreted as follows: 0.00–0.39, poor; 0.40–0.75, fair to good; and 0.76–1.00, excellent.8 In the outcome analyses, only Stage I tumors (92%) were included, because of the relatively small number of advanced-stage tumors (8%). Recurrence and survival rates were calculated using the Kaplan–Meier method, and differences between curves were assessed with the log-rank test. In instances of ordered variables, the trend test was used. Multivariate analysis of prognostic factors was performed using the Cox proportional hazards model. Several multivariate analyses were conducted. The baseline model included age, radiotherapy use, and depth of myometrial invasion. Subsequently, the following factors were added separately: the FIGO grade, the binary grade, the proportion of solid growth, the pattern of myometrial invasion, and the presence of necrosis. All P values are based on 2-sided tests, with P < 0.05 considered statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Interobserver Variability

FIGO grading system

The results of grading by the two pathologists according to the three-tiered FIGO grading system are shown in Table 2. The interobserver agreement for this grading system was fair (κ = 0.41; 70% agreement). When analyzing the FIGO grading system as a 2-tiered grading system (i.e., Grade 3 vs. Grades 1–2), interobserver agreement was much better (85% agreement; κ = 0.58). After the combined grading session, 522 tumors (65%) were considered consensus FIGO Grade 1 lesions, 135 tumors (17%) were considered FIGO Grade 2 lesions, and 143 tumors (18%) were considered FIGO Grade 3 lesions.

Table 2. Grading Results from the International Federation of Gynecology and Obstetrics Grading System
ResultsPathologist 1: No. of patients (%)
Grade 1Grade 2Grade 3Total
Pathologist 2    
 Grade 1430 (54) 18 (2)  4 (1)452 (57)
 Grade 2102 (13)  7 (1) 12 (1)121 (15)
 Grade 3 77 (10) 27 (3)123 (15)227 (28)
 Total609 (76) 52 (7)139 (17)800 (100)
 κ   0.41
Consensus522 (65)135 (17)143 (18)800 (100)
Binary grading system

The grading results according to the two-tiered grading system are shown in Table 3. The interobserver agreement for this grading system was almost similar to that of the FIGO grading system (κ = 0.39; 73% agreement). After the combined grading session, 570 tumors (71%) were considered binary low-grade lesions, and 230 tumors (29%) were considered binary high-grade lesions. The interobserver variability of the three separate criteria also was analyzed. It can be seen in Table 4 that judgment of the proportion of solid growth had the best reproducibility (κ = 0.50; 80% agreement). Determining the pattern of myometrial invasion (Fig. 1A–D) and the presence of tumor cell necrosis (Fig. 2A,B) appeared to be more difficult (κ = 0.41 and κ = 0.36, respectively; 71% and 74% agreement, respectively).

Table 3. Grading Results from the Binary Grading System
ResultsPathologist 1: No. of patients (%)
Low gradeHigh gradeTotal
Pathologist 2   
 Low grade462 (58) 15 (2)477 (60)
 High grade200 (25)123 (15)323 (40)
 Total662 (83)138 (17)800 (100)
 κ  0.39
Consensus570 (71)230 (29)800 (100)
Table 4. Results Using Individual Grading Criteria with the Binary Grading System
ResultsPathologist 1: No. of patients (%)
Solid growthMyometrial invasionNecrosis
≤ 50% (−)> 50% (+)Pushing (−)Infiltrating (+)Absent (−)Present (+)
  1. (−): negative; (+): positive.

Pathologist 2      
 (−)499 (63) 17 (2)195 (25) 29 (3)488 (61) 16 (2)
 (+)146 (18)138 (17)204 (26)363 (46)194 (24)102 (13)
 κ 0.50 0.41 0.36
Consensus623 (78)177 (22)368 (46)432 (54)554 (69)246 (31)
thumbnail image

Figure 1. Patterns of myometrial invasion. (A) Pushing border. (B) Case of disagreement (consensus, pushing border). (C) Case of disagreement (consensus, infiltrating pattern). (D) Infiltrating pattern.

Download figure to PowerPoint

thumbnail image

Figure 2. Necrosis. (A) Necrosis. (B) Case of disagreement (consensus, necrosis).

Download figure to PowerPoint

When analyzing correlations between the three criteria, we found a strong correlation between the proportion of solid growth and the presence of necrosis. Sixty-seven percent the tumors with > 50% solid growth had necrotic areas, compared with 20% of tumors with < 50% solid growth (P < 0.001). The pattern of myometrial invasion was correlated significantly with the depth of invasion. Seventy-three percent of patients who had tumors with an infiltrating pattern showed deep myometrial invasion (> 50%), compared with 47% of patients who had tumors with pushing borders (P < 0.001). Both the pattern and the depth of myometrial invasion were correlated with patient age. Older patients (age ≥ 60 years) had tumors that more often showed deep myometrial invasion (67% vs. 46% for patients age < 60 years; P < 0.001) and an infiltrating pattern of invasion (56% vs. 47%; P = 0.03).

When comparing the FIGO grading system with the binary grading system, we found a significant correlation between both systems. Among the binary low-grade tumors, 95% were FIGO Grade 1 or 2 lesions (82% and 13%, respectively), whereas only 24% and 27% of the binary high-grade tumors were FIGO Grade 1 or 2 lesions, respectively (P < 0.001).

Outcome

Locoregional recurrences (vaginal, pelvic, or both) were diagnosed in 55 of 740 patients with Stage I tumors (7%). The 5-year and 10-year locoregional recurrence rates were 7% and 9%, respectively. Distant metastases were found in 63 patients (9%), in most instances involving multiple sites. Distant recurrence rates at 5 years and 10 years were 8% and 9%, respectively. During follow-up, 205 patients (28%) died. Sixty-six deaths (32%) were related to endometrial carcinoma; 130 patients (64%) died of intercurrent diseases; and, in 9 patients (4%), the cause of death was unknown. The 5-year and 10-year disease-specific survival rates were 93% and 90%, respectively. The prognostic value of postoperative radiotherapy, age (< 60 years vs. ≥ 60 years), and depth of myometrial invasion (≤ 50% vs. > 50%) was analyzed for locoregional recurrence, distant recurrence, and disease-specific survival. In univariate (Table 5) and multivariate analyses, no postoperative radiotherapy (hazard ratio [HR], 4.5) and age (HR, 3.6) were identified as prognostic parameters for locoregional recurrence. Age (HR, 2.0) was a significant prognostic factor for distant recurrence. Age (HR, 2.7) and the depth of myometrial invasion (HR, 1.8) were significant prognostic factors for disease-specific survival (Table 6).

Table 5. Univariate Analyses of Prognostic Factors for Locoregional Recurrence, Distant Recurrence, and Disease-Specific Survival
FactorNo. of patients (%)Five yr survival
LRaP valueDR (%)aP valueDSS rate (%)aP value
  • LR: locoregional recurrence; DR: distant recurrence; DSS: disease-specific survival; FIGO: International Federation of Gynecology and Obstetrics.

  • a

    Total number of events: LR, 55; DR, 63; DSS, 66.

Age (yrs)       
 < 60179 (24)3 5 97 
 ≥ 60561 (76)80.00580.07920.009
Radiotherapy       
 Yes446 (60)3 8 93 
 No294 (40)13< 0.00170.82930.95
Myometrial invasion       
 < 50%292 (39)6 7 94 
 ≥ 50%448 (61)80.1080.37920.15
FIGO grade       
 1491 (66)5 3 97 
 2126 (17)8 8 94 
 3123 (17)14< 0.00125< 0.00176< 0.001
Binary grade       
 Low537 (73)5 3 96 
 High203 (27)110.0118< 0.00184< 0.001
Solid growth       
 ≤ 50%585 (79)6 4 96 
 > 50%155 (21)110.0219< 0.00183< 0.001
Infiltrating pattern       
 Pushing346 (47)7 6 94 
 Infiltrating394 (53)70.9490.16930.62
Necrosis       
 Absent526 (71)6 5 95 
 Present214 (29)80.3714< 0.00187< 0.001
Table 6. Multivariate Analyses of Prognostic Factors for Locoregional Recurrence, Distant Recurrence, and Disease-Specific Survival
FactorLRDRDSS
HR (95% CI)P valueHR (95% CI)P valueHR (95% CI)P value
  1. LR: locoregional recurrence; DR: distant recurrence; DSS: disease-specific survival; HR: hazard ratio; 95% CI: 95% confidence interval; RT: postoperative radiotherapy; FIGO: International Federation of Gynecology and Obstetrics.

Baseline model      
 Age ≥ 60 yrs3.6 (1.4–9.0)0.0072.0 (1.0–4.0)0.0452.7 (1.3–5.7)0.009
 No RT4.5 (2.4–8.1)< 0.0010.9 (0.5–1.5)0.650.9 (0.6–1.6)0.87
 Myometrial invasion ≥ 50%1.8 (0.9–3.3)0.051.6 (0.9–2.7)0.101.8 (1.0–3.0)0.04
Separate addition of the following factors    
 FIGO Grade      
  2 vs. 12.0 (0.9–4.0)0.062.2 (1.1–4.4)0.031.4 (0.7–3.0)0.37
  3 vs. 13.6 (1.9–6.7)< 0.0017.7 (4.4–13.6)< 0.0017.4 (4.3–12.6)< 0.001
  3 vs. 1–21.7 (1.3–2.3)< 0.0012.5 (1.9–3.2)< 0.0012.6 (2.0–3.3)< 0.001
 Binary high-grade2.1 (1.2–3.7)0.0064.1 (2.5–6.8)< 0.0013.8 (2.3–6.1)< 0.001
 Solid growth > 50%2.4 (1.3–4.3)0.0043.9 (2.4–6.5)< 0.0013.8 (2.3–6.2)< 0.001
 Infiltrating pattern0.8 (0.5–1.4)0.511.4 (0.8–2.3)0.231.0 (0.6–1.7)0.99
 Necrosis1.3 (0.8–2.3)0.332.9 (1.7–4.7)< 0.0012.6 (1.6–4.3)< 0.001
FIGO grading system

The consensus FIGO grade was identified as a significant predictor of locoregional recurrence, distant recurrence, and overall survival. The 5-year locoregional recurrence rates were 5%, 8%, and 14% for Grades 1, 2, and 3, respectively (P < 0.001). The 5-year distant recurrence rates were 3%, 8%, and 25% for Grades 1, 2, and 3, respectively (P < 0.001); and the 5-year disease-specific survival rates were 97%, 94%, and 76%, respectively (P < 0.001) (Fig. 3A). In multivariate analysis, FIGO grade proved to be of independent prognostic significance for locoregional recurrence (HR: 2.0 for Grade 2, 3.6 for Grade 3; P = 0.06 and P < 0.001, respectively) (Table 6). FIGO grade also proved to be of independent prognostic significance for distant recurrence (HR: 2.2 for Grade 2, 7.7 for Grade 3; P = 0.03 and P < 0.001, respectively) and disease-specific survival (HR, 7.4 for Grade 3; P < 0.001). When converting the FIGO grading system into a 2-tiered grading system (Grade 3 vs. combined Grades 1 and 2), FIGO grading still proved to be of independent prognostic significance for locoregional recurrence, distant recurrence, and disease-specific survival (HR: 1.7, 2.5, and 2.6, respectively; all P ≤ 0.001).

thumbnail image

Figure 3. (A) Disease-specific survival according to International Federation of Gynecology and Obstetrics grade (P < 0.001). (B) Disease-specific survival according to binary grade (P < 0.001). (C) Disease-specific survival according to proportion of solid growth (P < 0.001).

Download figure to PowerPoint

Binary grading system

The consensus grade according to the binary grading system had prognostic significance for locoregional recurrence, distant recurrence, and disease-specific survival. The 5-year locoregional recurrence rates were 5% and 11% for low-grade tumors and high-grade tumors, respectively (P < 0.001); the 5-year distant recurrence rates were 3% and 18%, respectively (P < 0.001); and the disease-specific survival rates at 5 years were 96% and 84%, respectively (P < 0.001) (Fig. 3B). In multivariate analysis, the binary grade had independent prognostic significance for locoregional and distant recurrences and for disease-specific survival (HR: 2.1, 4.1, and 3.8, respectively; P ≤ 0.006). Of the three separate criteria for the binary grading system, both > 50% solid growth and the presence of necrosis proved to be significant independent adverse prognostic factors for distant recurrence and disease-specific survival. The proportion of solid growth also was identified as a strong, independent prognostic factor for locoregional recurrence. The HRs for each endpoint were comparable to those of the binary grading system and were better than the HRs for the artificially created, two-tiered FIGO grading system. For tumors with > 50% or < 50% solid growth, the 5-year locoregional recurrence rates were 11% and 6%, respectively (P = 0.02), and the disease-specific survival rates at 5 years were 83% and 96%, respectively (P < 0.001) (Fig. 3C). Combining the proportion of solid growth and the presence of necrosis did not improve prognostic power (data not shown). The pattern of myometrial invasion did not have prognostic power.

DISCUSSION

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES

Determining histologic tumor grade is an essential part of pathologic diagnosis, because the histologic grade has prognostic and therapeutic implications. In general, a grading system has to be practical, reproducible, and clinically relevant, with clear prognostic power. For endometrial carcinoma, the most widely used grading system is the three-tiered FIGO grading system. Several previous studies have questioned the reproducibility of this grading system. Reported percentages of interobserver agreement vary between 63% and 81%, with corresponding κ values of 0.49–0.65.6, 9–11 This is comparable to the reproducibility that we observed for FIGO grading (70% agreement; κ = 0.41). In a previous study, we also showed that the reproducibility associated with identification of Grade 2 tumors was poor.5 At pathology review, a shift from Grade 2 to Grade 1 was seen in 78% of the original Grade 2 tumors. Moreover, the outcome of patients with Grade 2 and Grade 1 tumors was practically identical (92% and 94% 5-year survival, respectively), which brings into question the clinical value of Grade 2.

There are several features of the FIGO grading system that may be responsible for this lack of reproducibility. First, distinguishing between squamous and nonsquamous solid growth can be difficult, especially in patients with immature squamous metaplasia. Second, determining whether the percentage of nonsquamous solid growth is > 5% or < 5%, which distinguishes architecturally whether a tumor is a Grade 1 or Grade 2 lesion, is somewhat arbitrary. Finally, the determination of the degree of nuclear atypia is highly subjective, as illustrated by the 35% agreement (κ = 0.22) for nuclear grading reported by Lax et al.6 Several alternative grading systems for endometrial carcinoma have been suggested to improve the reproducibility without loosing prognostic power. Lax et al.6 proposed a binary grading system based on the proportion of solid growth, the pattern of myometrial invasion, and the presence of tumor cell necrosis. Taylor et al.11 divided tumors into low-grade and high-grade lesions based solely on the proportion of solid growth.

We conducted the current study compare the clinical value of the FIGO grading system, the binary grading system proposed by Lax et al.,6 and the three separate factors of that binary grading system (solid growth, pattern of myometrial invasion, and necrosis) by analyzing their reproducibility and their prognostic value in 800 patients with of endometrioid endometrial carcinoma. The interobserver agreement of the binary grading system was almost similar to that of the 3-tiered FIGO grading system (73% agreement; κ = 0.39) and was comparable to the 79% agreement reported by Lax et al.6 When we analyzed the FIGO grading system as a 2-tiered grading system (Grade 3 vs. Grades 1–2), interobserver agreement proved to be much better (85% agreement; κ = 0.58). With regard to the interobserver variability among the three criteria of the binary grading system separately, judgment of the proportion of solid growth showed 80% agreement (κ = 0.50). Determining the pattern of myometrial invasion was more difficult (71% agreement; κ = 0.41). This difficulty probably was caused by disagreement of the two pathologists on a considerable number of tumors with an essentially expansive growth pattern but with some masses or cords of tumor cells infiltrating the myometrium superficially. More detailed definitions of the pattern of myometrial invasion may improve the uniformity in differentiating between both patterns of invasion. It was found that the reproducibility of the presence of tumor cell necrosis was poor (74% agreement; κ = 0.36). Lax et al.6 stated that necrotic debris within masses of squamous epithelium did not qualify as tumor cell necrosis. Judging a necrotic mass as either debris or necrosis often was a matter of debate between the two pathologists. Furthermore, tumor cell necrosis was defined as areas of necrotic tumor immediately adjacent to viable tumor, and that criterion also proved to be somewhat subjective.

In the current study, it was found that the prognostic value of both the FIGO and the binary grading systems was good. In multivariate analyses, both FIGO Grade 3 and the binary high grade were identified as independent adverse prognostic factors for locoregional recurrence (HR: 3.6 and 2.1, respectively), distant recurrence (HR: 7.7 and 4.1) and disease-specific survival (HR: 7.4 and 3.8). Although the HR for patients with FIGO Grade 3 tumors versus patients with Grade 1 tumors was greater than the HR for patients with binary high-grade tumors versus patients with low-grade tumors, the binary grading system was superior to the FIGO grading system when the FIGO grading system was converted into a two-tiered grading system. HRs for FIGO Grade 3 versus FIGO Grades 1–2 for locoregional and distant recurrences and for disease-specific survival were 1.7, 2.5, and 2.6, respectively. Lax et al.6 stated that FIGO Grade 1 and Grade 2 tumors are not equivalent to low-grade tumors in the binary system; the low-grade tumor group was smaller than the FIGO Grade 1 and Grade 2 tumor groups combined (97 patients and 119 patients, respectively). The current results confirm this conclusion, because we found less low-grade tumors than FIGO Grade 1 or Grade 2 tumors (570 patients and 657 patients, respectively). Still, we found a strong correlation between both grading systems (P < 0.001), explaining why the prognostic value of both grading systems was so similar.

Taylor et al.11 compared the three-tiered FIGO grading system with another two-grade system. For 82 patients with Stage I or II endometrioid adenocarcinoma, the percentage of nonsquamous solid areas (in increments of 10%) was scored. The 20% nonsquamous solid growth pattern was chosen to delineate high-grade and low-grade tumors. It was shown that the reproducibility of this grading system was very good (99% agreement; κ = 0.97), and the authors reported that their system had good prognostic significance, because disease recurred only in patients with high-grade tumors. We also analyzed the reproducibility and prognostic value of the proportion of solid growth (defined as > 50% or < 50% squamous and nonsquamous solid growth). In fact, this is a simplification of the FIGO grading system that leaves out the subdivision of > 5% and < 5% solid growth and the differentiation between squamous and nonsquamous solid growth and excluding the influence of nuclear atypia on the final FIGO grade. Although some have reported that nuclear grading is a good prognosticator and has good reproducibility,12 others have questioned its prognostic value and its reproducibility.6, 9, 10 Zaino et al.,9 in their study of 88 patients with endometrial carcinoma, found that the prognosis for patients who had architectural Grade 2 tumors with either high or low nuclear grade was similar, bringing into question the need for the inclusion of nuclear atypia within the FIGO guidelines for grading.

Simply dividing patients into groups with high-grade and low-grade tumors based only on the proportion of solid growth (≤ 50% or > 50%) proved to have an interobserver reproducibility that was better compared with both the binary grading system and the 3-tiered FIGO grading system (80% agreement, κ = 0.50), and it was comparable to the 85% agreement (κ = 0.58) of the artificial, 2-tiered FIGO grading system. Furthermore, its prognostic value was superior to the two-tiered FIGO grading system and was comparable to the binary grading system.

In conclusion, although it was shown that the binary grading system proposed by Lax et al.6 had strong prognostic significance for locoregional and distant recurrence and disease-specific survival, its reproducibility was poor. The traditional, three-tiered FIGO grading system also exhibited only fair reproducibility. However, the reproducibility of the combined FIGO Grades 1 and 2 versus Grade 3 was much better, but the prognostic value of this artificial, two-tiered FIGO grading system was less than the value of the binary grading system. Although a two-tiered grading system has advantages in terms of practical clinical utility, the reproducibility of the proposed binary system was too low to support its use instead of the traditional, three-tiered FIGO grading system. Converting the FIGO grading system into an artificial, two-tiered system improved its reproducibility and affected its prognostic power only moderately; therefore, it is a good alternative. However, a simple architectural grading system dividing tumors into low and high grade, based on the proportion of solid growth (≤ 50% or > 50%), was superior to the 2-tiered FIGO grading system in prognostic power and was equally reproducible.

REFERENCES

  1. Top of page
  2. Abstract
  3. MATERIALS AND METHODS
  4. RESULTS
  5. DISCUSSION
  6. REFERENCES
  • 1
    Shepherd JH. Revised FIGO staging for gynaecological cancer. Br J Obstet Gynaecol. 1989; 96: 889892.
  • 2
    Grigsby PW, Perez CA, Kuten A, et al. Clinical Stage I endometrial cancer: prognostic factors for local control and distant metastasis and implications of the new FIGO surgical staging system. Int J Radiat Oncol Biol Phys. 1992; 22: 905911.
  • 3
    Creutzberg CL, van Putten WL, Koper PC, et al. Surgery and postoperative radiotherapy versus surgery alone for patients with Stage-1 endometrial carcinoma: multicentre randomised trial. PORTEC Study Group. Post Operative Radiation Therapy in Endometrial Carcinoma. Lancet. 2000; 355: 14041411.
  • 4
    Carey MS, O'Connell GJ, Johanson CR, et al. Good outcome associated with a standardized treatment protocol using selective postoperative radiation in patients with clinical Stage I adenocarcinoma of the endometrium. Gynecol Oncol. 1995; 57: 138144.
  • 5
    Scholten AN, Creutzberg CL, Noordijk EM, Smit VT. Long-term outcome in endometrial carcinoma favors a two- instead of a three-tiered grading system. Int J Radiat Oncol Biol Phys. 2002; 52: 10671074.
  • 6
    Lax SF, Kurman RJ, Pizer ES, Wu L, Ronnett BM. A binary architectural grading system for uterine endometrial endometrioid carcinoma has superior reproducibility compared with FIGO grading and identifies subsets of advance-stage tumors with favorable and unfavorable prognosis. Am J Surg Pathol. 2000; 24: 12011208.
  • 7
    Kurman RJ. Endometrial carcinoma. In: KurmanRJ, editor. Blaustein's pathology of the female genital tract (4th edition). New York: Springer-Verlag, 1994: 439486.
  • 8
    Fleiss JL. The measurement of interrater agreement. In: FleissJL, editor. Statistical methods for rates and proportions (2nd edition). New York: John Wiley & Sons, 1981: 212236.
  • 9
    Zaino RJ, Silverberg SG, Norris HJ, Bundy BN, Morrow CP, Okagaki T. The prognostic value of nuclear versus architectural grading in endometrial adenocarcinoma: a Gynecologic Oncology Group study. Int J Gynecol Pathol. 1994; 13: 2936.
  • 10
    Nielsen AL, Thomsen HK, Nyholm HC. Evaluation of the reproducibility of the revised 1988 International Federation of Gynecology and Obstetrics grading system of endometrial cancers with special emphasis on nuclear grading. Cancer. 1991; 68: 23032309.
  • 11
    Taylor RR, Zeller J, Lieberman RW, O'Connor DM. An analysis of two versus three grades for endometrial carcinoma. Gynecol Oncol. 1999; 74: 36.
  • 12
    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; 51: 17051709.