- Email: [email protected]
Noncompliance with adjuvant radiation, chemotherapy, or hormonal therapy in breast cancer patients
The American Journal of Surgery (2008) 196, 500 –504
The American Society of Breast Surgeons
Noncompliance with adjuvant radiation, chemotherapy, or hormonal therapy in breast cancer patients Aye Moe Thu Ma, M.D.*, Julie Barone, D.O., Ashley E. Wallis, Nancy Jade Wu, Luiza Baez Garcia, Alison Estabrook, M.D., Sharon M. Rosenbaum-Smith, M.D., Paul Ian Tartter, M.D. Department of Surgery, St. Luke’s-Roosevelt Hospital Center, 1111 Amsterdam Ave, Mu 2, Section D, New York, NY 10025, USA KEYWORDS: Breast cancer; Breast neoplasm; Chemotherapy; Compliance; Hormonal therapy; Radiation therapy; Survival rate; Tamoxifen
Abstract INTRODUCTION: There is little information on patient-driven noncompliance of adjuvant therapies and its consequences. METHODS: This retrospective study compares clinical, pathological features and outcomes of breast cancer patients who were compliant to recommended radiation, chemotherapy, and hormonal therapies to those who were noncompliant. RESULTS: Noncompliance rates for chemotherapy, radiation, and tamoxifen were 31 of 421 (7%), 30 of 855 (4%), and 294 of (37%) respectively. Old age was associated with noncompliance to chemotherapy and radiation, but younger women tend to be more often noncompliant to tamoxifen. Noncompliance with chemotherapy or radiation did not significantly affect 5-year local and distant disease-free survival rates. Noncompliance with tamoxifen was associated with decreased 5-year local and distant disease-free survivals (87% versus 96%, 76% versus 87%, respectively, P ⬍ .001). CONCLUSION: Noncompliance with tamoxifen is the most common, resulting in significantly increased risk of local and distant disease recurrence. © 2008 Published by Elsevier Inc.
Noncompliance with treatment guidelines can result in increased morbidity, mortality, and medical costs. Recent studies indicate that a significant fraction of women with breast cancer receive less than the recommended or are noncompliant with adjuvant therapies.1–5 However, few have shown whether noncompliance is physician or patient driven.1– 6 We studied breast cancer patients to identify factors influencing their decision not to have adjuvant chemotherapy, raPresented at the 9th Annual Meeting of the American Society of Breast Surgeons, April 30-May 4th, 2008, New York, NY. * Corresponding author: Tel.: ⫹1-212-523-4799; fax: ⫹1-212-523-2814. E-mail address: [email protected] Manuscript received April 17, 2008; revised manuscript June 3, 2008
0002-9610/$ - see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.amjsurg.2008.06.027
diation, or hormonal therapy and to assess the impact of noncompliance on local and distant disease control. The outcomes of patients who did receive adjuvant therapy were compared with those who refused despite physicians’ recommendations.
Patients and Methods The Institutional Review Board approved the protocol for this study. Patients were identified from the breast cancer database of the senior author (P.I.T.) from 1988 to 2002. Data from the registry as well as the electronic medical data system were used in a retrospective chart review that included general demographics, race, histol-
A. M. T. Ma et al.
Noncompliance of adjuvant therapies
ogy, grade, nodal involvement, tumor size, pathology, tumor differentiation, axillary node involvement, treatment with adjuvant radiation, chemotherapy and hormonal therapy, and outcome.
Radiation compliance Breast cancer patients treated with breast conservation with at least 5 years of follow-up were identified. Medical records of patients not receiving radiation were reviewed to determine the reason for omitting radiation therapy. Patients were divided into those treated with adjuvant radiation therapy, those not treated with radiation therapy based on physician advice, and those who refused radiation or were noncompliant. Missing information was obtained by direct contact with the patients for radiation.
Chemotherapy compliance Patients who were treated with curative intent were identified. Patients were advised by the surgeon (P.I.T.) and a medical oncologist as to whether or not they should undergo adjuvant chemotherapy. Patients were considered noncompliant if they refused to undergo chemotherapy as recommended by their physician.
Tamoxifen compliance For tamoxifen compliance, patients with estrogen receptor–positive invasive breast cancers were identified from the registry. All patients had been advised to take tamoxifen by the surgeon (P.I.T.). Patients who did not take tamoxifen as recommended by the surgeon, and, in many cases, by their oncologists, were considered noncompliant. Patients who took tamoxifen for less than 1 year were also considered noncompliant if not stopped on the advice of a physician. Whenever possible, noncompliance was confirmed by contact with the patient.
Data analysis The data were analyzed using SPSS software (SPSS Inc, Chicago, IL) run on a personal computer. A univariate analysis was performed using a chi-square test to identify traditional clinical and pathological factors distinguishing compliant and noncompliant patients. The significance of differences in continuous variables was evaluated using a Student t test. Cumulative 5-year local and distant diseasefree survival rates were calculated using the Kaplan–Meier method. The Breslow test was used to evaluate local and distant disease-free survival times and significance of the difference. Cox’s proportional hazards regression model was used to evaluate the relative significance of prognostic variables for both local and distant disease-free survival.
501 Table 1 Comparison of irradiated patients to those who refused radiation Radiation
Yes
Refused
N Age (mean 57) Tumor sze (cm) Pathology (%) Ductal carcinoma in situ Infiltrating ductal Cancer Infiltrating lobular Cancer Positive nodes Tamoxifen Chemotherapy Local recurrence Distant metastases Died after distant metastases
825 57 1.6
30 62 0.9
117 622 86 178 451 340 52 91 56
15 11 4 2 10 1 4 0 1
(14) (75) (10) (22) (55) (41) (6) (11) (7)
(50) (37) (13) (7) (33) (3) (13) (0) (3)
P value ⬍0.001 .003 ⬍0.001
NS NS ⬍0.001 .127 .064 .715
NS ⫽ not significant.
Results Radiation compliance Four percent (30) of the 855 lumpectomy patients refused radiation therapy (Table 1). These patients were significantly older (62 versus 57, P ⬍ .001), had significantly smaller tumors (0.9 cm versus 1.6 cm, P ⫽ .003), which were more frequently noninvasive (50% versus 14%) than compliant patients. As a result, they were also less likely to have received chemotherapy; only 1 of the 30 (3%) patients refusing radiation was treated with chemotherapy compared with 340 of the 825 (41%) patients who received radiation (P ⬍ .001) (Table 1). Stepwise logistic regression was used to rank the importance of age, tumor size, and noninvasive pathology for association with refusal of radiation. Tumor size (P ⫽ .531) was not significantly related to refusal after consideration for noninvasive pathology (P ⬍ .001) and age (P ⫽ .003). There were 204 disease-related events within 10 years of surgery; 56 patients (7%) developed a local recurrence, 91 patients (12%) developed distant metastases, and 57 patients (7%) died of disease. The local recurrence rate for patients who refused radiation was higher than the rate for patients who were not recommended radiation (13% versus 6%, P ⫽ .127) (Table 1). The lack of statistical significance is attributable to the small number of patients who refused radiation therapy. There was a suggestion that more compliant patients developed distant metastases (91 [11] versus none, P ⫽ .064). This difference is again attributable to the high frequency of noninvasive disease among the patients who refused radiation (Table 1).
Chemotherapy compliance Seven percent (31) of the 421 patients refused chemotherapy recommended by their physicians (noncompliant).
502
The American Journal of Surgery, Vol 196, No 4, October 2008
Table 2 Comparison of patients who received chemotherapy to those who refused
N Age (y) Race White Black Hispanic Other Tumor size (cm) Pathology Iinfiltrating ductal Infiltrating lobular Tumor differentiation Well Moderate Poor Unknown Nodal involvement Negative Positive Surgical procedure Mastectomy Breast conservation Radiation therapy Yes No Tamoxifen therapy Yes No
Received chemotherapy
Refused
390 52
31 57
256 (66) 63 (16) 56 (14) 14 (4) 2.9
23 (74) 4 (13) 4 (13) 0 (0) 2.6
348 (89) 42 (11)
27 (87) 4 (13)
P value .011 0.970
.485 .466 .612
16 153 174 47
(4) (39) (45) (12)
3 13 11 4
(10) (42) (35) (13) .062
152 (39) 238 (61)
17 (57) 14 (43)
135 (34) 253 (66)
11 (37) 20 (63)
270 (69) 120 (31)
14 (43) 17 (57)
169 (43) 221 (57)
11 (33) 20 (67)
.540 .007 .256
These patients were significantly older (57 versus 52, P ⫽ .011) and less likely to have received radiation therapy (43% versus 69%, P ⫽ .007) than patients who received chemotherapy (Table 2). Of the 31 noncompliant patients, 3 (10%) also refused radiation. Stepwise logistic regression was used to rank the importance of variables related to refusal of chemotherapy. Old age (P ⫽ .01) was the most significant variable followed by negative axillary lymph nodes (P ⫽ .016). After a median follow-up of 122 months, patients refusing chemotherapy had better local and distant 5-year dis-
Table 3 Five-year cumulative local and distant disease-free survival for patients who received chemotherapy and those who refused Chemotherapy Local All patients Conservation No radiation Received radiation Distant All patients
Compliant (%) Noncompliant (%) P value 71 74 63
88 94 75
0.0477 0.0766 0.3095
74
100
0.0295
78
80
0.2807
ease-free survival than patients who received chemotherapy (88% versus 71% compared with 80% versus 78%) (Table 3). This seemingly paradoxic result is caused by differences in distributions of nodal involvement and tumor grade among compliant and noncompliant patients (P ⫽ .012 and P ⫽ .026, respectively) (Table 2). In Cox proportional hazards, after consideration for these 2 factors, chemotherapy refusal was not significantly related to local recurrence. Chemotherapy noncompliance did not significantly impact distant disease-free survival (P ⫽ .2807) (Table 3).
Tamoxifen compliance Noncompliance with tamoxifen was common; 37% (294) of the 788 patients with estrogen receptor–positive cancers advised to take tamoxifen either refused or stopped taking the medication within a year. Noncompliant patients were significantly younger (54 versus 59, P ⬍ .001), were more frequently white, had larger ductal cancers (2.4 cm versus 1.8 cm, P ⫽ 0.007), and were more often treated with mastectomy rather than lumpectomy and radiation (38% versus 21%, P ⬍ .001) (Table 4). The cancers of tamoxifen noncompliant patients were more frequently of ductal pathology (39% versus 31, P ⫽ .039). Tamoxifen-noncompliant patients more often did not receive adjuvant radiation therapy. However, this relationship was entirely because of the larger number of patients who have had mastectomies in the noncompliant group. There were no significant relationships between tamoxifen use, tumor differentiation, nodal involvement, and receipt of adjuvant chemotherapy. Stepwise logistic regression was used to rank the importance of variables related to omission of tamoxifen. Young age (P ⬍ .001) was the most significant followed by ethnicity (P ⫽ .005), treatment with mastectomy (P ⫽ .006), tumor size (P ⫽ .009), and pathology (P ⫽ .027). All 5 variables were statistically significant in stepwise regression. Tamoxifen noncompliance impacted local and distant disease-free survival. Five-year distant disease-free survival for tamoxifen users was 87% compared with 76% for nonusers (P ⬍ .001) (Table 4). Five-year local disease-free survival was 96% for tamoxifen users compared with 87% for patients who did not take tamoxifen (P ⬍ .001) (Table 4). This difference cannot be attributed to the higher rate of radiation in the compliant group because even in the subgroup of patients who were treated with breast conservation followed by radiation, 97% of tamoxifen compliant patients were free of local recurrence at 5 years compared with 87% of noncompliant patients (P ⬍ .001). A Cox proportional hazards model was used to evaluate the significance of omission of tamoxifen after consideration for other prognostic factors. After taking into account tumor size (P ⬍ .001) and the number of involved axillary nodes (P ⬍ .001), the use of tamoxifen continued to be a significant prognostic factor (P ⫽ .028).
A. M. T. Ma et al.
Noncompliance of adjuvant therapies
Table 4 Comparison of tamoxifen-compliant and -noncompliant patients Tamoxifen Compliant Noncompliant P value N Age (y) Race (%) White Black Hispanic Other Tumor size (cm) Pathology (%) Infiltrating ductal Infiltrating lobular Tumor differentiation (%) Well Moderate Poor Nodal involvement (%) No Yes Surgery (%) Mastectomy Breast conservation Chemotherapy (%) No Yes Radiation therapy (%) No Yes Five years local recurrence (%) All patients Breast conservation 5-year distant recurrence
494 59
287 54
330 (67) 69 (14) 75 (15) 20 (4) 1.8
219 (76) 41 (14) 24 (8) 3 (1) 2.4
416 (84) 78 (16)
263 (91) 31 (11)
60 (12) 312 (63) 122 (25)
22 (7) 197 (67) 75 (26)
343 (69) 151 (31)
218 (74) 76 (26)
104 (21) 390 (79)
112 (38) 182 (62)
301 (61) 193 (39)
178 (61) 116 (39)
97 (20) 397 (80)
140 (48) 154 (52)
16 (4) 12 (3) 51 (13)
4 (13) 4 (13) 7 (24)
⬍0.001 0.029
0.007 0.039 0.115
0.157 ⬍0.001 0.914 ⬍0.001
⬍0.001 ⬍0.001 ⬍0.001
We also found that tumor differentiation (P ⫽ .034) was significantly related to outcome after consideration for size, nodes, and tamoxifen use. Tumor pathology, race, and age were not significant prognostic factors. Cox proportional hazards model identified tamoxifen noncompliance (P ⫽ .001) and involved axillary nodes (P ⫽ .004) as risk factors for local recurrence.
Discussion The noncompliance rate for adjuvant therapies of breast cancer was highest for tamoxifen therapy at 37%, followed by 7% for chemotherapy, and 4% for radiation therapy. The noncompliance rate of 4% for radiation therapy was slightly higher than our previously reported rate of 1.6% in 487 patients.7 Although several studies have shown poor outcome with an omission of adjuvant radiation treatment, the consequences of patient-driven nonadherence is rarely reported.3–7 In our study, the local recurrence rate of noncompliant patient was 13% compared with 6% for patients
503 who received radiation (P ⫽ .127). The only comparable data are from Chang et al6 who reported that 2 (20%) of 10 of their patients who refused radiation therapy developed local recurrences within 18 months.6 The chemotherapy noncompliance rate in this study is comparable to previous studies that showed rates between 5% and 19%.8,9 Landercasper et al9 evaluated compliance with adjuvant therapy guidelines in 200 patients and found that 30 patients (19%) refused chemotherapy and were not compliant with the National Comprehensive Cancer Network guidelines for systemic adjuvant therapy. In our study, chemotherapy noncompliance did not result in an increased risk of local or distant disease recurrence. In fact, patients with invasive cancers who refused adjuvant chemotherapy were significantly more likely to be local disease free 5 years after surgery. However, this finding was not significant after controlling for nodal involvement and tumor grade. Noncompliance with adjuvant hormonal therapy is an underappreciated problem. Similar to our study, several published reports have also found a high rate of tamoxifen noncompliance or exclusion in estrogen receptor–positive breast cancers. Bickell et al10 reported that 20% of 421 patients with estrogen receptor–positive cancers at 6 institutions in New York City treated in 1999 and 2000 did not receive appropriate adjuvant tamoxifen. Gajdos et al11 studied elderly undertreated patients and found that 35% of women over age 70 with estrogen receptor–positive breast cancers did not receive tamoxifen. Enger et al12 found that 34% of women over 75 years of age were not prescribed tamoxifen. Partridge et al13 measured tamoxifen prescription fill rate among 2,378 New Jersey women on public assistance between 1990 and 1996. They found that 13% of the patients filled their prescription less than 80% of the time in the first year, increasing to 50% by the fourth year of follow-up. Compared with the adherence rates for many chronic drugs therapies of 40% to 50%, the compliance rate for tamoxifen is slightly better.13 The present study did not investigate the specific reasons for refusal, but we were able to identify several factors related to noncompliance. Age has been consistently associated with patient noncompliance whether one measures compliance with blood pressure medications, adjuvant radiation therapy, or adjuvant chemotherapy.14 Our study found that older breast cancer patients were significantly more likely to refuse chemotherapy and radiation therapy. Among many possible reasons, one may be because of the patient’s perception of decreased quality of life during these treatments. Age, on the other hand, has a reverse relation to the use of tamoxifen where younger women were less compliant. Partridge et al13 also found that 1 reliable predictor of nonadherence to tamoxifen was patients younger than 45 years old. Possible explanations may be that younger people feel less of a sense of vulnerability, fear of menopausal symptoms, and the fact that they are also not used to taking
504 daily medication unlike the elderly patients. However, this needs to be explored further because hormonal therapy is the most effective adjuvant therapy for women with estrogen receptor–positive breast cancer, and is shown to improve both survival and the risk of local recurrence.15 Race was reported as a major factor in undertreatment and nonadherence in prior studies for tamoxifen and chemotherapy.10,13 In the present study, whites were more likely to refuse tamoxifen. However, we did not find significant racial differences between chemotherapy compliant and noncompliant patients. This may be because all the patients in the present study were treated by a single surgeon and received surgery, with minimum clearance of 1 mm, radiation, and tamoxifen in the same facilities irrespective of race or ethnicity. One limitation of this study is that there were very few patients who were noncompliant with radiation therapy and chemotherapy. Therefore, the outcomes analysis of noncompliance with radiation and chemotherapy must be interpreted with caution. On the other hand, there was a large number of patients who were noncompliant with tamoxifen therapy, and this study shows that tamoxifen noncompliance can adversely impact both local and distant recurrences. The sad irony is that this is also the most efficacious adjuvant therapy with the least side effects for estrogen receptor–positive breast cancer. Therefore, further research should be encouraged to identify reasons for such noncompliance.
The American Journal of Surgery, Vol 196, No 4, October 2008
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
References 14. 1. Hand R, Sener S, Imperato J, et al. Hospital variables associated with quality of care for breast cancer patients. JAMA 1991;266:3429 –32. 2. Foley KL, Kimmick G, Camacho F, et al. Survival disadvantage among MedicAid-insured breast cancer patients treated with breast
15.
conserving surgery without radiation therapy. Breast Cancer Res Treat 2007;101:207–14. Truong PT, Bernstein V, Lesperance M, et al. Radiation therapy omission after breast-conserving surgery is associated with reduced breast-specific survival in elderly women with breast cancer. Am J Surg 2006;191:749 –55. Vinh-Hung V, Verschraegen C. Breast conserving surgery with or without radiotherapy: Pooled analysis for risks of ipsilateral breast breast tumor recurrence and mortality. J Natl Cancer Inst 2004;96: 15–21. Vinh-Hung V, Voordeckers M, Van de Steene J, et al. Omission of radiotherapy after breast conserving surgery: survival impact and time trends. Radiother Oncol 2003;671:147–58. Chang E, Glissmeyer M, Tonnes S, et al. Outcomes of breast cancer in patients who use alternative therapies as primary treatment. Am J Surg 2006;192:471–3. De Csepel J, Tartter PI, Gajdos C. When not to give radiation therapy after breast conservation surgery for breast cancer. J Surg Oncol 2000;74:273–7. Balasubramanian SP, Murrow S, Holt S, et al. Audit of compliance to adjuvant chemotherapy and radiotherapy guidelines in breast cancer in a cancer network. Breast 2003;12:136 – 41. Landercasper J, Dietrich LL, Johnson JM. A breast cancer review of compliance with National Comprehensive Cancer Network breast cancer guidelines. Am J Surg 2006;192:525–7. Bickell NA, Wang JJ, Oluwole S, et al. Missed opportunities: racial disparities in adjuvant breast cancer treatment. J Clin Oncol 2006;24: 1357– 62. Gajdos C, Tartter PI, Bleiweiss I, Bernstein J. The consequence of undertreating breast cancer in the elderly. J Am Coll Surg 2001;192: 698 –707. Enger SM, Thwin SS, Buist DSM, et al. Breast cancer treatment of older women in integrated health care settings. J Clin Oncol 2006;24: 4377– 83. Partridge AH, Wang PS, Winer EP, Avorn J. Nonadherence to adjuvant tamoxifen therapy in women with primary breast cancer. J Clin Oncol 2003;21:602– 6. Rosenow EC, III. Patients’ understanding of and compliance with medications: the sixth vital sign. Mayo Clin Proc 2005;80:983–7. Early Breast Cacner Trialists’ Collaborative Group. Tamoxifen for early breast cancer: An overview of the randomised trials. Lancet 1998;351:1451– 67.
The American Society of Breast Surgeons
Noncompliance with adjuvant radiation, chemotherapy, or hormonal therapy in breast cancer patients Aye Moe Thu Ma, M.D.*, Julie Barone, D.O., Ashley E. Wallis, Nancy Jade Wu, Luiza Baez Garcia, Alison Estabrook, M.D., Sharon M. Rosenbaum-Smith, M.D., Paul Ian Tartter, M.D. Department of Surgery, St. Luke’s-Roosevelt Hospital Center, 1111 Amsterdam Ave, Mu 2, Section D, New York, NY 10025, USA KEYWORDS: Breast cancer; Breast neoplasm; Chemotherapy; Compliance; Hormonal therapy; Radiation therapy; Survival rate; Tamoxifen
Abstract INTRODUCTION: There is little information on patient-driven noncompliance of adjuvant therapies and its consequences. METHODS: This retrospective study compares clinical, pathological features and outcomes of breast cancer patients who were compliant to recommended radiation, chemotherapy, and hormonal therapies to those who were noncompliant. RESULTS: Noncompliance rates for chemotherapy, radiation, and tamoxifen were 31 of 421 (7%), 30 of 855 (4%), and 294 of (37%) respectively. Old age was associated with noncompliance to chemotherapy and radiation, but younger women tend to be more often noncompliant to tamoxifen. Noncompliance with chemotherapy or radiation did not significantly affect 5-year local and distant disease-free survival rates. Noncompliance with tamoxifen was associated with decreased 5-year local and distant disease-free survivals (87% versus 96%, 76% versus 87%, respectively, P ⬍ .001). CONCLUSION: Noncompliance with tamoxifen is the most common, resulting in significantly increased risk of local and distant disease recurrence. © 2008 Published by Elsevier Inc.
Noncompliance with treatment guidelines can result in increased morbidity, mortality, and medical costs. Recent studies indicate that a significant fraction of women with breast cancer receive less than the recommended or are noncompliant with adjuvant therapies.1–5 However, few have shown whether noncompliance is physician or patient driven.1– 6 We studied breast cancer patients to identify factors influencing their decision not to have adjuvant chemotherapy, raPresented at the 9th Annual Meeting of the American Society of Breast Surgeons, April 30-May 4th, 2008, New York, NY. * Corresponding author: Tel.: ⫹1-212-523-4799; fax: ⫹1-212-523-2814. E-mail address: [email protected] Manuscript received April 17, 2008; revised manuscript June 3, 2008
0002-9610/$ - see front matter © 2008 Published by Elsevier Inc. doi:10.1016/j.amjsurg.2008.06.027
diation, or hormonal therapy and to assess the impact of noncompliance on local and distant disease control. The outcomes of patients who did receive adjuvant therapy were compared with those who refused despite physicians’ recommendations.
Patients and Methods The Institutional Review Board approved the protocol for this study. Patients were identified from the breast cancer database of the senior author (P.I.T.) from 1988 to 2002. Data from the registry as well as the electronic medical data system were used in a retrospective chart review that included general demographics, race, histol-
A. M. T. Ma et al.
Noncompliance of adjuvant therapies
ogy, grade, nodal involvement, tumor size, pathology, tumor differentiation, axillary node involvement, treatment with adjuvant radiation, chemotherapy and hormonal therapy, and outcome.
Radiation compliance Breast cancer patients treated with breast conservation with at least 5 years of follow-up were identified. Medical records of patients not receiving radiation were reviewed to determine the reason for omitting radiation therapy. Patients were divided into those treated with adjuvant radiation therapy, those not treated with radiation therapy based on physician advice, and those who refused radiation or were noncompliant. Missing information was obtained by direct contact with the patients for radiation.
Chemotherapy compliance Patients who were treated with curative intent were identified. Patients were advised by the surgeon (P.I.T.) and a medical oncologist as to whether or not they should undergo adjuvant chemotherapy. Patients were considered noncompliant if they refused to undergo chemotherapy as recommended by their physician.
Tamoxifen compliance For tamoxifen compliance, patients with estrogen receptor–positive invasive breast cancers were identified from the registry. All patients had been advised to take tamoxifen by the surgeon (P.I.T.). Patients who did not take tamoxifen as recommended by the surgeon, and, in many cases, by their oncologists, were considered noncompliant. Patients who took tamoxifen for less than 1 year were also considered noncompliant if not stopped on the advice of a physician. Whenever possible, noncompliance was confirmed by contact with the patient.
Data analysis The data were analyzed using SPSS software (SPSS Inc, Chicago, IL) run on a personal computer. A univariate analysis was performed using a chi-square test to identify traditional clinical and pathological factors distinguishing compliant and noncompliant patients. The significance of differences in continuous variables was evaluated using a Student t test. Cumulative 5-year local and distant diseasefree survival rates were calculated using the Kaplan–Meier method. The Breslow test was used to evaluate local and distant disease-free survival times and significance of the difference. Cox’s proportional hazards regression model was used to evaluate the relative significance of prognostic variables for both local and distant disease-free survival.
501 Table 1 Comparison of irradiated patients to those who refused radiation Radiation
Yes
Refused
N Age (mean 57) Tumor sze (cm) Pathology (%) Ductal carcinoma in situ Infiltrating ductal Cancer Infiltrating lobular Cancer Positive nodes Tamoxifen Chemotherapy Local recurrence Distant metastases Died after distant metastases
825 57 1.6
30 62 0.9
117 622 86 178 451 340 52 91 56
15 11 4 2 10 1 4 0 1
(14) (75) (10) (22) (55) (41) (6) (11) (7)
(50) (37) (13) (7) (33) (3) (13) (0) (3)
P value ⬍0.001 .003 ⬍0.001
NS NS ⬍0.001 .127 .064 .715
NS ⫽ not significant.
Results Radiation compliance Four percent (30) of the 855 lumpectomy patients refused radiation therapy (Table 1). These patients were significantly older (62 versus 57, P ⬍ .001), had significantly smaller tumors (0.9 cm versus 1.6 cm, P ⫽ .003), which were more frequently noninvasive (50% versus 14%) than compliant patients. As a result, they were also less likely to have received chemotherapy; only 1 of the 30 (3%) patients refusing radiation was treated with chemotherapy compared with 340 of the 825 (41%) patients who received radiation (P ⬍ .001) (Table 1). Stepwise logistic regression was used to rank the importance of age, tumor size, and noninvasive pathology for association with refusal of radiation. Tumor size (P ⫽ .531) was not significantly related to refusal after consideration for noninvasive pathology (P ⬍ .001) and age (P ⫽ .003). There were 204 disease-related events within 10 years of surgery; 56 patients (7%) developed a local recurrence, 91 patients (12%) developed distant metastases, and 57 patients (7%) died of disease. The local recurrence rate for patients who refused radiation was higher than the rate for patients who were not recommended radiation (13% versus 6%, P ⫽ .127) (Table 1). The lack of statistical significance is attributable to the small number of patients who refused radiation therapy. There was a suggestion that more compliant patients developed distant metastases (91 [11] versus none, P ⫽ .064). This difference is again attributable to the high frequency of noninvasive disease among the patients who refused radiation (Table 1).
Chemotherapy compliance Seven percent (31) of the 421 patients refused chemotherapy recommended by their physicians (noncompliant).
502
The American Journal of Surgery, Vol 196, No 4, October 2008
Table 2 Comparison of patients who received chemotherapy to those who refused
N Age (y) Race White Black Hispanic Other Tumor size (cm) Pathology Iinfiltrating ductal Infiltrating lobular Tumor differentiation Well Moderate Poor Unknown Nodal involvement Negative Positive Surgical procedure Mastectomy Breast conservation Radiation therapy Yes No Tamoxifen therapy Yes No
Received chemotherapy
Refused
390 52
31 57
256 (66) 63 (16) 56 (14) 14 (4) 2.9
23 (74) 4 (13) 4 (13) 0 (0) 2.6
348 (89) 42 (11)
27 (87) 4 (13)
P value .011 0.970
.485 .466 .612
16 153 174 47
(4) (39) (45) (12)
3 13 11 4
(10) (42) (35) (13) .062
152 (39) 238 (61)
17 (57) 14 (43)
135 (34) 253 (66)
11 (37) 20 (63)
270 (69) 120 (31)
14 (43) 17 (57)
169 (43) 221 (57)
11 (33) 20 (67)
.540 .007 .256
These patients were significantly older (57 versus 52, P ⫽ .011) and less likely to have received radiation therapy (43% versus 69%, P ⫽ .007) than patients who received chemotherapy (Table 2). Of the 31 noncompliant patients, 3 (10%) also refused radiation. Stepwise logistic regression was used to rank the importance of variables related to refusal of chemotherapy. Old age (P ⫽ .01) was the most significant variable followed by negative axillary lymph nodes (P ⫽ .016). After a median follow-up of 122 months, patients refusing chemotherapy had better local and distant 5-year dis-
Table 3 Five-year cumulative local and distant disease-free survival for patients who received chemotherapy and those who refused Chemotherapy Local All patients Conservation No radiation Received radiation Distant All patients
Compliant (%) Noncompliant (%) P value 71 74 63
88 94 75
0.0477 0.0766 0.3095
74
100
0.0295
78
80
0.2807
ease-free survival than patients who received chemotherapy (88% versus 71% compared with 80% versus 78%) (Table 3). This seemingly paradoxic result is caused by differences in distributions of nodal involvement and tumor grade among compliant and noncompliant patients (P ⫽ .012 and P ⫽ .026, respectively) (Table 2). In Cox proportional hazards, after consideration for these 2 factors, chemotherapy refusal was not significantly related to local recurrence. Chemotherapy noncompliance did not significantly impact distant disease-free survival (P ⫽ .2807) (Table 3).
Tamoxifen compliance Noncompliance with tamoxifen was common; 37% (294) of the 788 patients with estrogen receptor–positive cancers advised to take tamoxifen either refused or stopped taking the medication within a year. Noncompliant patients were significantly younger (54 versus 59, P ⬍ .001), were more frequently white, had larger ductal cancers (2.4 cm versus 1.8 cm, P ⫽ 0.007), and were more often treated with mastectomy rather than lumpectomy and radiation (38% versus 21%, P ⬍ .001) (Table 4). The cancers of tamoxifen noncompliant patients were more frequently of ductal pathology (39% versus 31, P ⫽ .039). Tamoxifen-noncompliant patients more often did not receive adjuvant radiation therapy. However, this relationship was entirely because of the larger number of patients who have had mastectomies in the noncompliant group. There were no significant relationships between tamoxifen use, tumor differentiation, nodal involvement, and receipt of adjuvant chemotherapy. Stepwise logistic regression was used to rank the importance of variables related to omission of tamoxifen. Young age (P ⬍ .001) was the most significant followed by ethnicity (P ⫽ .005), treatment with mastectomy (P ⫽ .006), tumor size (P ⫽ .009), and pathology (P ⫽ .027). All 5 variables were statistically significant in stepwise regression. Tamoxifen noncompliance impacted local and distant disease-free survival. Five-year distant disease-free survival for tamoxifen users was 87% compared with 76% for nonusers (P ⬍ .001) (Table 4). Five-year local disease-free survival was 96% for tamoxifen users compared with 87% for patients who did not take tamoxifen (P ⬍ .001) (Table 4). This difference cannot be attributed to the higher rate of radiation in the compliant group because even in the subgroup of patients who were treated with breast conservation followed by radiation, 97% of tamoxifen compliant patients were free of local recurrence at 5 years compared with 87% of noncompliant patients (P ⬍ .001). A Cox proportional hazards model was used to evaluate the significance of omission of tamoxifen after consideration for other prognostic factors. After taking into account tumor size (P ⬍ .001) and the number of involved axillary nodes (P ⬍ .001), the use of tamoxifen continued to be a significant prognostic factor (P ⫽ .028).
A. M. T. Ma et al.
Noncompliance of adjuvant therapies
Table 4 Comparison of tamoxifen-compliant and -noncompliant patients Tamoxifen Compliant Noncompliant P value N Age (y) Race (%) White Black Hispanic Other Tumor size (cm) Pathology (%) Infiltrating ductal Infiltrating lobular Tumor differentiation (%) Well Moderate Poor Nodal involvement (%) No Yes Surgery (%) Mastectomy Breast conservation Chemotherapy (%) No Yes Radiation therapy (%) No Yes Five years local recurrence (%) All patients Breast conservation 5-year distant recurrence
494 59
287 54
330 (67) 69 (14) 75 (15) 20 (4) 1.8
219 (76) 41 (14) 24 (8) 3 (1) 2.4
416 (84) 78 (16)
263 (91) 31 (11)
60 (12) 312 (63) 122 (25)
22 (7) 197 (67) 75 (26)
343 (69) 151 (31)
218 (74) 76 (26)
104 (21) 390 (79)
112 (38) 182 (62)
301 (61) 193 (39)
178 (61) 116 (39)
97 (20) 397 (80)
140 (48) 154 (52)
16 (4) 12 (3) 51 (13)
4 (13) 4 (13) 7 (24)
⬍0.001 0.029
0.007 0.039 0.115
0.157 ⬍0.001 0.914 ⬍0.001
⬍0.001 ⬍0.001 ⬍0.001
We also found that tumor differentiation (P ⫽ .034) was significantly related to outcome after consideration for size, nodes, and tamoxifen use. Tumor pathology, race, and age were not significant prognostic factors. Cox proportional hazards model identified tamoxifen noncompliance (P ⫽ .001) and involved axillary nodes (P ⫽ .004) as risk factors for local recurrence.
Discussion The noncompliance rate for adjuvant therapies of breast cancer was highest for tamoxifen therapy at 37%, followed by 7% for chemotherapy, and 4% for radiation therapy. The noncompliance rate of 4% for radiation therapy was slightly higher than our previously reported rate of 1.6% in 487 patients.7 Although several studies have shown poor outcome with an omission of adjuvant radiation treatment, the consequences of patient-driven nonadherence is rarely reported.3–7 In our study, the local recurrence rate of noncompliant patient was 13% compared with 6% for patients
503 who received radiation (P ⫽ .127). The only comparable data are from Chang et al6 who reported that 2 (20%) of 10 of their patients who refused radiation therapy developed local recurrences within 18 months.6 The chemotherapy noncompliance rate in this study is comparable to previous studies that showed rates between 5% and 19%.8,9 Landercasper et al9 evaluated compliance with adjuvant therapy guidelines in 200 patients and found that 30 patients (19%) refused chemotherapy and were not compliant with the National Comprehensive Cancer Network guidelines for systemic adjuvant therapy. In our study, chemotherapy noncompliance did not result in an increased risk of local or distant disease recurrence. In fact, patients with invasive cancers who refused adjuvant chemotherapy were significantly more likely to be local disease free 5 years after surgery. However, this finding was not significant after controlling for nodal involvement and tumor grade. Noncompliance with adjuvant hormonal therapy is an underappreciated problem. Similar to our study, several published reports have also found a high rate of tamoxifen noncompliance or exclusion in estrogen receptor–positive breast cancers. Bickell et al10 reported that 20% of 421 patients with estrogen receptor–positive cancers at 6 institutions in New York City treated in 1999 and 2000 did not receive appropriate adjuvant tamoxifen. Gajdos et al11 studied elderly undertreated patients and found that 35% of women over age 70 with estrogen receptor–positive breast cancers did not receive tamoxifen. Enger et al12 found that 34% of women over 75 years of age were not prescribed tamoxifen. Partridge et al13 measured tamoxifen prescription fill rate among 2,378 New Jersey women on public assistance between 1990 and 1996. They found that 13% of the patients filled their prescription less than 80% of the time in the first year, increasing to 50% by the fourth year of follow-up. Compared with the adherence rates for many chronic drugs therapies of 40% to 50%, the compliance rate for tamoxifen is slightly better.13 The present study did not investigate the specific reasons for refusal, but we were able to identify several factors related to noncompliance. Age has been consistently associated with patient noncompliance whether one measures compliance with blood pressure medications, adjuvant radiation therapy, or adjuvant chemotherapy.14 Our study found that older breast cancer patients were significantly more likely to refuse chemotherapy and radiation therapy. Among many possible reasons, one may be because of the patient’s perception of decreased quality of life during these treatments. Age, on the other hand, has a reverse relation to the use of tamoxifen where younger women were less compliant. Partridge et al13 also found that 1 reliable predictor of nonadherence to tamoxifen was patients younger than 45 years old. Possible explanations may be that younger people feel less of a sense of vulnerability, fear of menopausal symptoms, and the fact that they are also not used to taking
504 daily medication unlike the elderly patients. However, this needs to be explored further because hormonal therapy is the most effective adjuvant therapy for women with estrogen receptor–positive breast cancer, and is shown to improve both survival and the risk of local recurrence.15 Race was reported as a major factor in undertreatment and nonadherence in prior studies for tamoxifen and chemotherapy.10,13 In the present study, whites were more likely to refuse tamoxifen. However, we did not find significant racial differences between chemotherapy compliant and noncompliant patients. This may be because all the patients in the present study were treated by a single surgeon and received surgery, with minimum clearance of 1 mm, radiation, and tamoxifen in the same facilities irrespective of race or ethnicity. One limitation of this study is that there were very few patients who were noncompliant with radiation therapy and chemotherapy. Therefore, the outcomes analysis of noncompliance with radiation and chemotherapy must be interpreted with caution. On the other hand, there was a large number of patients who were noncompliant with tamoxifen therapy, and this study shows that tamoxifen noncompliance can adversely impact both local and distant recurrences. The sad irony is that this is also the most efficacious adjuvant therapy with the least side effects for estrogen receptor–positive breast cancer. Therefore, further research should be encouraged to identify reasons for such noncompliance.
The American Journal of Surgery, Vol 196, No 4, October 2008
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
References 14. 1. Hand R, Sener S, Imperato J, et al. Hospital variables associated with quality of care for breast cancer patients. JAMA 1991;266:3429 –32. 2. Foley KL, Kimmick G, Camacho F, et al. Survival disadvantage among MedicAid-insured breast cancer patients treated with breast
15.
conserving surgery without radiation therapy. Breast Cancer Res Treat 2007;101:207–14. Truong PT, Bernstein V, Lesperance M, et al. Radiation therapy omission after breast-conserving surgery is associated with reduced breast-specific survival in elderly women with breast cancer. Am J Surg 2006;191:749 –55. Vinh-Hung V, Verschraegen C. Breast conserving surgery with or without radiotherapy: Pooled analysis for risks of ipsilateral breast breast tumor recurrence and mortality. J Natl Cancer Inst 2004;96: 15–21. Vinh-Hung V, Voordeckers M, Van de Steene J, et al. Omission of radiotherapy after breast conserving surgery: survival impact and time trends. Radiother Oncol 2003;671:147–58. Chang E, Glissmeyer M, Tonnes S, et al. Outcomes of breast cancer in patients who use alternative therapies as primary treatment. Am J Surg 2006;192:471–3. De Csepel J, Tartter PI, Gajdos C. When not to give radiation therapy after breast conservation surgery for breast cancer. J Surg Oncol 2000;74:273–7. Balasubramanian SP, Murrow S, Holt S, et al. Audit of compliance to adjuvant chemotherapy and radiotherapy guidelines in breast cancer in a cancer network. Breast 2003;12:136 – 41. Landercasper J, Dietrich LL, Johnson JM. A breast cancer review of compliance with National Comprehensive Cancer Network breast cancer guidelines. Am J Surg 2006;192:525–7. Bickell NA, Wang JJ, Oluwole S, et al. Missed opportunities: racial disparities in adjuvant breast cancer treatment. J Clin Oncol 2006;24: 1357– 62. Gajdos C, Tartter PI, Bleiweiss I, Bernstein J. The consequence of undertreating breast cancer in the elderly. J Am Coll Surg 2001;192: 698 –707. Enger SM, Thwin SS, Buist DSM, et al. Breast cancer treatment of older women in integrated health care settings. J Clin Oncol 2006;24: 4377– 83. Partridge AH, Wang PS, Winer EP, Avorn J. Nonadherence to adjuvant tamoxifen therapy in women with primary breast cancer. J Clin Oncol 2003;21:602– 6. Rosenow EC, III. Patients’ understanding of and compliance with medications: the sixth vital sign. Mayo Clin Proc 2005;80:983–7. Early Breast Cacner Trialists’ Collaborative Group. Tamoxifen for early breast cancer: An overview of the randomised trials. Lancet 1998;351:1451– 67.