Table 4 Multivariable models showing the association between RCB index and class (hazard ratios with 95% confidence intervals and associated p values shown), stratified by histologic group and receptor subtype; models include age, T category, nodal status, and tumor grade

From: A pooled analysis evaluating prognostic significance of Residual Cancer Burden in invasive lobular breast cancer

 

ILC (n = 181)

Non-ILC (n = 4397)

HR+HER2-ILC (n = 129)

HR+HER2- non-ILC (n = 1586)

TN or HER2+ILC (n = 52)

TN or HER2+ non-ILC (n = 2811)

Multivariate model evaluating RCB index

RCB

1.82 (1.23–2.68), p = 0.003

1.74 (1.64–1.85), p < 0.001

1.53 (1.00–2.44), p = 0.077

1.48 (1.31–1.67), p < 0.001

4.75 (0.81–27.79), p = 0.083

1.83 (1.71–1.96), p < 0.001

Age

0.98 (0.95–1.02), p = 0.310

1.00 (0.99–1.00), p = 0.100

0.99 (0.95–1.03), p = 0.720

1.00 (0.99–1.01), p = 0.670

0.90 (0.78–1.04), p = 0.160

0.99 (0.99–1.00), p = 0.110

T category (reference: T2)

T0-1

1.54 (0.45–5.27), p = 0.490

1.07 (0.84–1.38), p = 0.570

1.31 (0.34–5.01), p = 0.690

0.98 (0.67–1.45), p = 0.930

2.91 (0.01–728.34), p = 0.710

1.15 (0.84–1.59), p = 0.380

T3

0.80 (0.23–1.51), p = 0.490

1.36 (1.16–1.59), p < 0.001

0.77 (0.38–1.56), p = 0.460

1.11 (0.85–1.44), p = 0.460

1.56 (0.18–13.31), p = 0.680

1.47 (1.21–1.78), p < 0.001

T4

0.84 (0.18–3.84), p = 0.820

1.95 (1.60–2.39), p < 0.001

0.54 (0.07–4.43), p = 0.560

2.26 (1.63–3.13), p < 0.001

3.15 (0.03–288.43), p = 0.620

1.82 (1.41–2.36), p < 0.001

Nodal status (reference: node negative)

Node positive

1.29 (0.67–2.47), p = 0.440

1.19 (1.03–1.37), p = 0.019

1.05 (0.50–2.19), p = 0.900

1.41 (1.11–1.80), p = 0.006

9.06 (0.36–228.13), p = 0.180

1.11 (0.93–1.33), p = 0.250

Grade (reference: grade 1 or 2)

Grade 3

3.16 (1.37–7.28), p = 0.007

1.17 (1.02–1.35), p = 0.003

3.01 (1.10–8.26), p = 0.032

1.52 (1.23–1.87), p < 0.001

3.70 (0.10–143.58), p = 0.480

0.88 (0.74–1.07), p = 0.200

Multivariable model evaluating RCB class

RCB 0

REF

REF

REF (combined RCB 0/I group)

REF (combined RCB 0/I group)

REF

REF

RCB I

1.43 (0.27–7.62), p = 0.680

1.91 (1.46–2.52), p < 0.001

REF (combined RCB 0/I group)

REF (combined RCB 0/I group)

0.43 (0.01–28.50), p = 0.690

2.08 (1.53–2.83), p < 0.001

RCB II

1.45 (0.30–6.95), p = 0.640

3.78 (3.08–4.64), p < 0.001

0.55 (0.17–1/73), p = 0.300

1.99 (1.42–2.77), p < 0.001

3.10 (0.06–157.98), p = 0.570

4.04 (3.21–5.09), p < 0.001

RCB III

6.89 (1.41–33.64), p = 0.017

7.38 (5.89–9.24), p < 0.001

2.16 (0.64–7.31), p = 0.220

3.16 (2.19–4.54),p < 0.001

222.83 (1.12–44,368.31), p = 0.045

9.77 (7.55–12.65), p < 0.001

Age

0.97 (0.94–1.01), p = 0.130

1.00 (0.99–1.00), p = 0.160

0.98 (0.94–1.02), p = 0.360

1.00 (0.99–1.01), p = 0.800

0.86 (0.71–1.05), p = 0.140

0.99 (0.99–1.00), p = 0.110

T category (reference: T2)

T0-1

1.50 (0.43–5.18), p = 0.520

1.05 (0.82–1.35), p = 0.670

0.70 (0.19–2.58), p = 0.590

0.93 (0.63–1.38), p = 0.730

2.17 (9.67 × 10–6–488,756.31), p = 0.900

1.12 (0.81–1.55), p = 0.480

T3

1.00 (0.54–1.85), p = 1

1.40 (1.20–1.64), p < 0.001

0.60 (0.16-2.24), p = 0.440

1.12 (0.86-1.45), p = 0.420

2.80 (0.23–34.37), p = 0.420

1.52 (1.25–1.84), p < 0.001

T4

1.76 (0.39–8.01), p = 0.460

2.02 (1.65–2.48), p < 0.001

0.53 (0.05–5.49), p = 0.590

2.39 (1.72–3.31), p < 0.001

1.90 (0.03–126.25), p = 0.760

1.83 (1.41–2.38), p < 0.001

Nodal Status (reference: node negative)

Node positive

1.33 (0.70–2.54), p = 0.380

1.36 (1.18–1.57), p < 0.001

1.13 (0.56–2.28), p = 0.730

1.57 (1.24–1.99), p < 0.001

12.83 (0.18–891.08), p = 0.240

1.24 (1.04–1.49), p = 0.018

Grade (reference: grade 1 or 2)

Grade 3

2.60 (1.14-5.94), p = 0.023

1.15 (0.99–1.32), p = 0.063

3.05 (1.12–8.35), p = 0.030

1.45 (1.18–1.79), p = 0.001

5.85 (0.18–193.67), p = 0.320

0.87 (0.72–1.05), p = 0.140

  1. In these adjusted analyses, RCB index is significantly associated with EFS for ILC and non-ILC cases overall. When stratified by receptor subtype, higher RCB index remains significantly associated with EFS in non-ILC cases, and trends towards shorter EFS in ILC subsets but has wide confidence intervals in these smaller groups. For models including RCB class, RCB 0 and RCB I cases were combined as a single reference group (RCB 0/I) for HR+HER2- cases due to only 4 ILC patients having RCB 0 disease.
  2. Bold indicates statistical significance (p < 0.05).
Back to article page