Table 2.
Associations between breast cancer treatment and heart failure risk
| Total | Median value (IQR) | Cases (n = 102) | Controls (n = 306) | RR | Floating 95% CI | P‐value |
|---|---|---|---|---|---|---|
| Model I: Model for all treatments considered simultaneously | ||||||
| Radiotherapy | ||||||
| Median [IQR] mean heart dosea (Gy) | 6.8 [0.9–13.7] | 3.9 [0.9–13.4] | ||||
| Mean heart dose 0–1 Gy | 0.4 [0.2–0.9] | 28 (27.5) | 96 (31.4) | 1.0c | 0.53–1.9 | |
| Mean heart dose 2–9 Gy | 4.3 [3.8–6.6] | 26 (25.5) | 83 (27.1) | 0.8 | 0.49–1.3 | 0.53 |
| Mean heart dose ≥10 Gy | 14.6 [13.6–17.0] | 33 (32.4) | 83( 27.1) | 1.2 | 0.72–1.9 | 0.74 |
| No radiotherapy | 14 (13.7) | 38 (12.4) | 1.4 | 0.73–2.8 | 0.51 | |
| Mean heart dose unknownb | 1 (0.9) | 6 (2.0) | – | – | ||
| Chemotherapy | ||||||
| No chemotherapy | 44 (43.1) | 185 (60.5) | 1.0c | 0.60–1.7 | ||
| CMF‐like | 9 (8.8) | 48 (15.7) | 0.7 | 0.30–1.5 | 0.35 | |
| Anthracyclinese | 33 (32.4) | 60 (19.6) | 6.9 | 3.5–13.6 | <0.001 | |
| Anthracyclines and trastuzumabe | 14 (13.7) | 7 (2.3) | 34.9f | 11.1–110.1 | <0.001 | |
| Other type of chemotherapy or type unknown | 2 (2.0) | 6 (2.0) | –g | – | ||
| Endocrine therapy | ||||||
| No endocrine therapy | 65 (63.7) | 241 (80.4) | 1.0c | 0.67–1.5 | ||
| Tamoxifen | 20 (19.6) | 41 (11.6) | 1.5 | 0.80–2.8 | 0.29 | |
| Tamoxifen and aromatase inhibitors | 8 (7.8) | 17 (6.0) | 1.6 | 0.57–4.5 | 0.42 | |
| Aromatase inhibitors | 8 (7.8) | 5 (1.3) | 4.0 | 1.0–16.3 | 0.06 | |
| Type of endocrine therapy unknown | 1 (0.9) | 2 (0.7) | –g | – | ||
| Model II h: Cumulative anthracycline dose by treatment with radiotherapy and trastuzumab i | ||||||
| No radiotherapy | ||||||
| Median [IQR] cumulative anthracycline dosej (mg/m2) | 242 [230–302] | 252 [241–302] | ||||
| Total | 14 | 38 | ||||
| No trastuzumab | ||||||
| 0 mg/m2 | 11 (78.6) | 33 (86.8) | – | – | ||
| ≤240 mg/m2 | 231 [231–231]k | 1 (7.1) | 0 (0) | –g | – | |
| >240 mg/m2 | 252 [241–302]k | 1 (7.1) | 5 (13.2) | –g | – | |
| With trastuzumab | ||||||
| >240 mg/m2 | 302 [302–302]k | 1 (7.1) | 0 (0) | –g | – | |
| Radiotherapy | ||||||
| Median [IQR] cumulative anthracycline dosej (mg/m2) | 247 [240–319] | 240 [240–300] | ||||
| Total | 88 | 268 | ||||
| No trastuzumab | ||||||
| 0 mg/m2 | 43 (48.9) | 203 (75.8) | 1.0c | 0.48–2.1 | ||
| ≤240 mg/m2 | 240 [221–240]k | 9 (10.2) | 26 (9.7) | 3.3 | 1.5–7.2 | 0.02 |
| >240 mg/m2 | 300 [252–360]k | 23 (26.1) | 29 (10.8) | 8.6 | 4.7–15.6 | <0.001 |
| With trastuzumab | ||||||
| ≤240 mg/m2 | 240 [240–240]k | 13 (14.8) | 10 (3.7) | 25.3 | 9.7–65.9 | <0.001 |
| Model III e : Estimated mean heart dose by anthracycline and trastuzumab | ||||||
| No anthracyclines or trastuzumab | ||||||
| Median [IQR] mean heart dosea , b (Gy) | 3.8 [0.2–14.2] | 3.8 [0.4–14.2] | ||||
| Total | 54 | 236 | ||||
| 0–1 Gy | 0.4 [0.2–0.9] | 9 (16.7) | 60 (25.4) | 1.0c , d | 0.43–2.3 | |
| 2–9 Gy | 3.8 [3.8–5.4] | 12 (22.2) | 64 (27.1) | 0.6 | 0.31–1.1 | 0.30 |
| ≥10 Gy | 14.7 [14.2–18.0] | 21 (38.9) | 77 (32.6) | 0.7 | 0.46–1.2 | 0.57 |
| No radiotherapy | 11 (20.4) | 33 (14.0) | 1.1 | 0.52–2.2 | 0.87 | |
| Mean heart dose unknownb | 1 (1.9) | 2 (0.9) | –g | – | ||
| Anthracyclines but not trastuzumab | ||||||
| Median [IQR] mean heart dosea , b (Gy) | 6.9 [0.9–12.0] | 0.9 [0.2–6.3] | ||||
| Total | 34 | 60 | ||||
| 0–1 Gy | 0.9 [0.2–0.9] | 9 (26.5) | 29 (48.3) | 1.0c | 0.42–2.4 | |
| 2–9 Gy | 6.4 [4.5–6.9] | 13 (38.2) | 17 (28.3) | 1.2 | 0.55–2.8 | 0.70 |
| ≥10 Gy | 15.2 [12.0–16.9] | 10 (29.4) | 5 (8.3) | 2.8 | 0.89–9.0 | 0.15 |
| No radiotherapy | 2 (5.9) | 5 (8.3) | 1.2 | 0.27–5.6 | 0.82 | |
| Mean heart dose unknownb | 0 (0) | 4 (6.7) | –g | – | ||
| Trastuzumab | ||||||
| Median [IQR] mean heart dosea , b (Gy) | 0.9 [0.2–0.9] | 0.9 [0.9–6.4] | ||||
| Total | 14 | 10 | ||||
| 0–1 Gy | 10 (71.4) | 6 (60.0) | – | – | ||
| 2–9 Gy | 2 (14.3) | 3 (30.0) | –g | – | ||
| ≥10 Gy | 1 (7.2) | 1 (10.0) | –g | – | ||
| No radiotherapy | 1 (7.2) | 0 (0) | –g | – | ||
| Model IV l : Joint effects of mean heart dose and anthracyclines | ||||||
| Mean heart dose <10 Gy, no anthracyclines | 3.3 [0.4–3.9] | 32 (31.4) | 160 (52.3) | 1.0c | 0.66–1.5 | |
| Mean heart dose ≥10 Gy, no anthracyclines | 14.7 [14.2–18.0] | 21 (20.6) | 77 (25.2) | 1.1 | 0.59–1.9 | 0.85 |
| Mean heart dose <10 Gy, anthracyclines | 0.9 [0.4–5.0] | 37 (36.3) | 57 (18.6) | 6.3 | 3.0–13.2 | <0.001 |
| Mean heart dose ≥10 Gy, anthracyclines | 14.9 [12.0–16.8] | 11 (10.8) | 6 (2.0) | 12.4 | 4.0–39.2 | <0.001 |
| Mean heart dose unknownb | 1 (1.0) | 6 (2.0) | –g | – | – | |
CI, confidence interval; CMF, cyclophosphamide, methotrexate, 5‐fluorouracil; IQR, interquartile range; RR, rate ratio.
RRs for heart failure were estimated using logistic regression conditional on the matching variables.
In patients treated with radiotherapy.
Heart doses were unknown for seven patients (one case, six controls) because their radiotherapy charts were unavailable.
Reference category. CI for categorical exposure variables were estimated for each category, including the reference category, from the amount of information in that category.
P for trend across categories 0.48.
Anthracycline treatment consisted of an epirubicin‐containing regimen for 14/48 cases and 16/67 controls, and of a doxorubicin‐containing regimen for 34/48 cases and 51/67 controls. Trastuzumab was mostly given in combination with anthracyclines.
RR for anthracyclines plus trastuzumab vs. anthracyclines without trastuzumab is 5.5 (95% CI 1.9–16.0).
Insufficient numbers to produce reliable estimates for this category.
Model additionally included a dichotomous variable for endocrine therapy (no/yes).
Range cumulative anthracycline dose was 90–612 mg/m2 doxorubicin equivalent, the commonest dose was four times 60 mg/m2. Two patients were treated with an anthracycline for breast cancer and then later retreated with an anthracycline for a recurrence or second malignancy. Cumulative anthracycline range for all other patients was 90–366 mg/m2. The commonest regimens were: AC (doxorubicin and cyclophosphamide), FAC (5‐fluorouracil, doxorubicin, cyclophosphamide), TAC (docetaxel, doxorubicin, cyclophosphamide), and FEC (5‐fluorouracil, epirubicin, cyclophosphamide).
In patients treated with anthracyclines.
Median anthracycline dose and IQR in patients treated with and without trastuzumab.
Model also included dichotomous variables for radiotherapy (no/yes), trastuzumab (no/yes), and endocrine therapy (no/yes).