World J Oncol

World Journal of Oncology, ISSN 1920-4531 print, 1920-454X online, Open Access

Article copyright, the authors; Journal compilation copyright, World J Oncol and Elmer Press Inc

Journal website https://wjon.elmerpub.com

Original Article

Volume 16, Number 6, December 2025, pages 555-564

A Simplified Novel Algorithm to Predict the 21-Gene Recurrence Score

↓ Figure 1. Proposed PR% and grade in relation to RS. PR: progesterone receptor; RS: recurrence score.

↓ Figure 2. Proposed algorithm to predict Oncotype DX risk category using histologic grade and PR percentage. PR: progesterone receptor.

↓ **Table 1.** Pathological Features for All Patients
Feature	N (%)
ER: estrogen receptor; FISH: fluorescence in situ hybridization; HER2: human epidermal growth factor receptor-2; ODx: Oncotype DX; PR: progesterone receptor; RS: recurrence score.
ER (%)
Positive	528 (100%)
PR (%)
Negative	30 (5.7%)
Positive	498 (94.3%)
HER2
Equivocal (FISH-negative)	59 (11.2%)
Negative	469 (88.8%)
HER2 score
0	263 (49.8%)
1+	206 (39.0%)
2+ (FISH negative)	59 (11.2%)
ER (ODx value)
Negative	11 (2.1%)
Positive	517 (97.9%)
PR (ODx value)
Negative	72 (13.6%)
Positive	456 (86.4%)
HER2 (ODx value)
Equivocal	3 (0.6%)
Negative	523 (99.0%)
Positive	2 (0.4%)
Histological type
Invasive ductal carcinoma	438 (82.9%)
Invasive lobular carcinoma	68 (12.9%)
Invasive mucinous carcinoma	10 (1.9%)
Invasive tubular carcinoma	7 (1.3%)
Invasive cribriform carcinoma	1 (0.2%)
Invasive solid papillary carcinoma	3 (0.6%)
Invasive breast carcinoma with apocrine features	1 (0.2%)
Group
Learning set	377 (71.3%)
Validation set	151 (28.7%)
Grade
1	99 (18.8%)
2	323 (61.2%)
3	106 (20.1%)
RS
High	96 (18.2%)
Low	432 (81.8%)

↓ Table 1. Pathological Features for All Patients

Feature

N (%)

ER: estrogen receptor; FISH: fluorescence in situ hybridization; HER2: human epidermal growth factor receptor-2; ODx: Oncotype DX; PR: progesterone receptor; RS: recurrence score.

ER (%)

Positive

528 (100%)

PR (%)

Negative

30 (5.7%)

Positive

498 (94.3%)

HER2

Equivocal (FISH-negative)

59 (11.2%)

Negative

469 (88.8%)

HER2 score

263 (49.8%)

206 (39.0%)

2+ (FISH negative)

59 (11.2%)

ER (ODx value)

Negative

11 (2.1%)

Positive

517 (97.9%)

PR (ODx value)

Negative

72 (13.6%)

Positive

456 (86.4%)

HER2 (ODx value)

Equivocal

3 (0.6%)

Negative

523 (99.0%)

Positive

2 (0.4%)

Histological type

Invasive ductal carcinoma

438 (82.9%)

Invasive lobular carcinoma

68 (12.9%)

Invasive mucinous carcinoma

10 (1.9%)

Invasive tubular carcinoma

7 (1.3%)

Invasive cribriform carcinoma

1 (0.2%)

Invasive solid papillary carcinoma

3 (0.6%)

Invasive breast carcinoma with apocrine features

1 (0.2%)

Group

Learning set

377 (71.3%)

Validation set

151 (28.7%)

Grade

99 (18.8%)

323 (61.2%)

106 (20.1%)

High

96 (18.2%)

Low

432 (81.8%)

↓ **Table 2.** Comparison of Various Variables Between the Learning Set and the Validation Set Showing No Significant Differences
Feature	Total, N (%)	Groups	P-value
CI: confidence interval; ER: estrogen receptor; PR: progesterone receptor; RS: recurrence score; SD: standard deviation.
Grade				0.88
1	99 (19.2%)	68 (18.0%)	31 (20.5%)
2	323 (61.7%)	231 (61.2%)	92 (60.9%)
3	106 (19.0%)	78 (20.7%)	28 (18.5%)
Actual RS				0.54
High	96 (18.2%)	72 (19.1%)	24 (15.9%)
Low	432 (82.5%)	305 (81.9%)	127 (84.1%)
Age (years)				0.46
Mean (95% CI)		52.8 (51.9, 53.7)	53.6 (52.1, 55.2)
Median (Min, Max)		52.0 (21.0, 76.0)	53.0 (24.0, 79.0)
SD		10.3	11.4
ER (%)				0.34
Mean (95% CI)		89.8 (88.7 ,91.0)	88.5 (86.5, 90.5)
Median (Min, Max)		95.0 (1.0, 100)	90.0 (5.0, 100)
SD		13.7	15.0
PR (%)				0.87
Mean (95% CI)		66.4 (63.5, 69.3)	67.6 (63.0, 72.2)
Median (Min, Max)		85.0 (0.0, 100)	90.0 (0.0, 100)
SD		34.2	34.1

↓ Table 2. Comparison of Various Variables Between the Learning Set and the Validation Set Showing No Significant Differences

Feature

Total, N (%)

Groups

P-value

Learning set

Validation set

CI: confidence interval; ER: estrogen receptor; PR: progesterone receptor; RS: recurrence score; SD: standard deviation.

Grade

0.88

99 (19.2%)

68 (18.0%)

31 (20.5%)

323 (61.7%)

231 (61.2%)

92 (60.9%)

106 (19.0%)

78 (20.7%)

28 (18.5%)

Actual RS

0.54

High

96 (18.2%)

72 (19.1%)

24 (15.9%)

Low

432 (82.5%)

305 (81.9%)

127 (84.1%)

Age (years)

0.46

Mean (95% CI)

52.8 (51.9, 53.7)

53.6 (52.1, 55.2)

Median (Min, Max)

52.0 (21.0, 76.0)

53.0 (24.0, 79.0)

10.3

11.4

ER (%)

0.34

Mean (95% CI)

89.8 (88.7 ,91.0)

88.5 (86.5, 90.5)

Median (Min, Max)

95.0 (1.0, 100)

90.0 (5.0, 100)

13.7

15.0

PR (%)

0.87

Mean (95% CI)

66.4 (63.5, 69.3)

67.6 (63.0, 72.2)

Median (Min, Max)

85.0 (0.0, 100)

90.0 (0.0, 100)

34.2

34.1

↓ **Table 3.** Relationship Between RS and Grade and RS and PR% in the Learning Set
Variable	Total	Actual RS	P-value
PR: progesterone receptor; RS: recurrence score; SD: standard deviation.
Grade
Grade 1	68 (18.7%)	0	68 (22.9%)	0.00
Grade 2	231 (62.3%)	25 (10.8%)	206 (89.2%)
Grade 3	78 (19.0%)	47 (61.8%)	31 (9.5%)
PR%, mean (SD)		33.7 (35.2)	73.7 (29.5)	< 0.0001

↓ Table 3. Relationship Between RS and Grade and RS and PR% in the Learning Set

Variable

Total

Actual RS

P-value

High, N (%)

Low, N (%)

PR: progesterone receptor; RS: recurrence score; SD: standard deviation.

Grade

Grade 1

68 (18.7%)

68 (22.9%)

0.00

Grade 2

231 (62.3%)

25 (10.8%)

206 (89.2%)

Grade 3

78 (19.0%)

47 (61.8%)

31 (9.5%)

PR%, mean (SD)

33.7 (35.2)

73.7 (29.5)

< 0.0001

↓ **Table 4.** PR% Cut-Offs Combined With Grade as Determined by ROC Curves and Those Yielding a 100% Sensitivity
Variable	PR cut-off (%)	Total	RS	P-value
PR: progesterone receptor; ROC: receiver operating characteristics; RS: recurrence score.
Learning set - all grade (PR cut-off ROC 50)	< 50	99 (26.4%)	43 (63.2%)	56 (18.2%)	0.000
	≥ 50	276 (73.6%)	25 (36.8%)	251 (81.8%)
Learning set - grade 2 (PR cut-off ROC 30)	< 30	58 (24.9%)	22 (84.6%)	36 (17.4%)	0.000
	≥ 30	175 (75.1%)	4 (15.4%)	171 (82.6%)
Learning set - grade 3 (PR cut-off ROC 70)	< 70	28 (39.4%)	24 (57.1%)	4 (13.8%)	0.000
	≥ 70	43 (60.6%)	18 (42.9%)	25 (86.2%)
Learning set - grade 2 (PR at 60) The sensitivity for this cut-off 60 = 100% The specificity for this cut-off 60 = 69.1%	< 60	76 (35.6%)	25 (100%)	51 (27.5%)	0.000
	≥ 60	155 (64.4%)	0	155 (72.5%)
Learning set - grade 3 (PR at 40) The sensitivity for this cut-off 40 = 100% The specificity for this cut-off 40 = 65%	< 40	22 (23.9%)	22 (40.5%)	0	0.000
	≥ 40	56 (76.1%)	25 (59.5%)	31 (100%)

↓ Table 4. PR% Cut-Offs Combined With Grade as Determined by ROC Curves and Those Yielding a 100% Sensitivity

Variable

PR cut-off (%)

Total

P-value

High

Low

PR: progesterone receptor; ROC: receiver operating characteristics; RS: recurrence score.

Learning set - all grade (PR cut-off ROC 50)

< 50

99 (26.4%)

43 (63.2%)

56 (18.2%)

0.000

≥ 50

276 (73.6%)

25 (36.8%)

251 (81.8%)

Learning set - grade 2 (PR cut-off ROC 30)

< 30

58 (24.9%)

22 (84.6%)

36 (17.4%)

0.000

≥ 30

175 (75.1%)

4 (15.4%)

171 (82.6%)

Learning set - grade 3 (PR cut-off ROC 70)

< 70

28 (39.4%)

24 (57.1%)

4 (13.8%)

0.000

≥ 70

43 (60.6%)

18 (42.9%)

25 (86.2%)

Learning set - grade 2 (PR at 60)
The sensitivity for this cut-off 60 = 100%
The specificity for this cut-off 60 = 69.1%

< 60

76 (35.6%)

25 (100%)

51 (27.5%)

0.000

≥ 60

155 (64.4%)

155 (72.5%)

Learning set - grade 3 (PR at 40)
The sensitivity for this cut-off 40 = 100%
The specificity for this cut-off 40 = 65%

< 40

22 (23.9%)

22 (40.5%)

0.000

≥ 40

56 (76.1%)

25 (59.5%)

31 (100%)