VOLUME

23

䡠

NUMBER

19

䡠

JULY

1

2005

JOURNAL OF CLINICAL ONCOLOGY

O R I G I N A L

R E P O R T

Morbidity After Sentinel Lymph Node Biopsy in Primary Breast Cancer: Results From a Randomized Controlled Trial Anand David Purushotham, Sara Upponi, Manfred Borislav Klevesath, Lynda Bobrow, Keith Millar, Jonathan Peter Myles, and Stephen William Duffy From the Cambridge Breast Unit, Addenbrooke’s Hospital, Cambridge University Hospitals, NHS Foundation Trust and University of Cambridge, Cambridge; Section of Psychological Medicine, University of Glasgow; Department of Epidemiology, Mathematics and Statistics, Cancer Research United Kingdom, London, United Kingdom, on behalf of the Cambridge/East Anglia Study Group. Submitted November 12, 2004; accepted March 16, 2005. Supported by a grant from the Eastern Region Research and Development and the National Cancer Research Network. Authors’ disclosures of potential conflicts of interest are found at the end of this article. Address reprint requests to A.D. Purushotham, MD, Cambridge Breast Unit Box 97, Addenbrooke’s Hospital, Cambridge University Hospitals, NHS Foundation Trust, Hills Rd, Cambridge CB2 2QQ, United Kingdom; e-mail: [email protected]. © 2005 by American Society of Clinical Oncology

A

B

S

T

R

A

C

T

Purpose Axillary lymph node dissection (ALND) as part of surgical treatment for patients with breast cancer is associated with significant morbidity. Sentinel lymph node biopsy (SLNB) is a newly developed method of staging the axilla and has the potential to avoid an ALND in lymph node–negative patients, thereby minimizing morbidity. The aim of this study was to investigate physical and psychological morbidity after SLNB in the treatment of early breast cancer in a randomized controlled trial. Patients and Methods Between November 1999 and February 2003, 298 patients with early breast cancer (tumors 3 cm or less on ultrasound examination) who were clinically node negative were randomly allocated to undergo ALND (control group) or SLNB followed by ALND if subsequently found to be lymph node positive (study group). A detailed assessment of physical and psychological morbidity was performed during a 1-year period postoperatively. Results A significant reduction in postoperative arm swelling, rate of seroma formation, numbness, loss of sensitivity to light touch and pinprick was observed in the study group. Although shoulder mobility was less impaired on average in the study group, this was significant only for abduction at 1 month and flexion at 3 months. Scores reflecting quality of life and psychological morbidity were significantly better in the study group in the immediate postoperative period, with fewer long-term differences. Conclusion SLNB in patients undergoing surgery for breast cancer results in a significant reduction in physical and psychological morbidity.

0732-183X/05/2319-4312/$20.00 DOI: 10.1200/JCO.2005.03.228

J Clin Oncol 23:4312-4321. © 2005 by American Society of Clinical Oncology

INTRODUCTION

Breast carcinoma is the leading cause of cancer death in women, with more than 300,000 deaths worldwide annually.1 Currently, surgical management comprises resection of the primary tumor and axillary lymph node dissection (ALND) or four-node sampling of the axilla. There are potentially significant short- and long-term morbidities associated with ALND, such as seroma forma-

tion, impairment of shoulder movement, neuropathy, and arm lymphedema. Estimates for seroma formation vary from 4% to 52%.2-5 Shoulder movement is impaired in the majority of patients in the immediate period after ALND, but usually returns to near-normal values within 12 months of surgery.6-8 Numbness and paresthesia have been reported in up to 70% to 80% of patients after division of the intercostobrachial nerve.6,8,9 The incidence of lymphedema ranges from 6% to 30%.10

4312

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Randomized Trial of SNB in Breast Cancer

Therefore, there is clearly an incentive to avoid ALND in lymph node–negative patients. Sentinel lymph node biopsy (SLNB) is a relatively new method of staging the axilla. The sentinel lymph node (SLN) is the first lymph node(s) to receive lymphatic drainage from the site of a tumor and should be the first site of lymphatic spread. Therefore, a tumor-free SLN indicates the absence of lymph node metastasis in the rest of the lymphatic basin. Initial studies demonstrated the accuracy of sentinel lymph node biopsy (SLNB) in predicting axillary lymph node status in patients undergoing surgery for breast cancer.11-14 In theory, SLNB should result in fewer adverse effects than ALND, and initial reports suggest that morbidity after SLNB is indeed lower than that after ALND. Giuliano et al15 observed an overall complication rate of 3% after SLNB compared with 35% after SLNB plus ALND. A number of studies have demonstrated significantly lower rates of numbness/paresthesia, restriction of shoulder movement, and arm swelling in patients undergoing SLNB when compared with those undergoing ALND.15-24 However, apart from the Milan study,24 these studies were retrospective or prospective observational studies. This article presents the results of a randomized controlled trial with a detailed analysis of the impact of SLNB on postoperative physical and psychological morbidity. The aim of this study was to investigate physical and psychological morbidity after SLNB in the treatment of early breast cancer in a randomized controlled trial comparing standard ALND (control group) with axillary SLNB (study group) in patients undergoing surgery for early breast cancer.

PATIENTS AND METHODS

Patient Selection and Characteristics Patients were included in the study if they had tumors less than 3 cm in diameter (on ultrasound examination) and a proven histopathologic diagnosis of invasive breast cancer. Exclusion criteria included prior treatment for breast cancer (eg, neoadjuvant chemotherapy), pregnancy, clinically involved axillary nodes, multifocal breast cancer, or previous diagnostic excision biopsy. The study was approved by the Local Research Ethics Committees. Between November 1999 and February 2003, after written informed consent was received, 298 patients from Addenbrooke’s Hospital (Cambridge, United Kingdom), West Suffolk Hospital (Bury St Edmunds, United Kingdom), and King’s Lynn Hospital (King’s Lynn, United Kingdom) were randomly assigned to treatment. Allocation was by sealed envelope method. The individual random assignment was performed by a computer random-number generator. The random assignment and preparation of the envelopes was carried out in advance by S.W.D. in the absence of the clinicians involved and blinded to the patients to be allocated thereafter. Surgery The study design is shown in Figure 1 and a flow chart for the study is shown in Figure 2. All patients allocated to the control group underwent a level 2 ALND. In the study group, SLNB was performed using a combined method of blue dye (2.5% Bleu patenté V; Guerbet, Roissy, France) and isotope (40 MBq of Nanocoll, technetium-99 –albumin nanocolloid; Gipharma S.r.l., Saluggia, Italy) with intraoperative

Fig 1. Study design. SLN, surgical lymph node.

4313

www.jco.org

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Purushotham et al

Fig 2. Consolidated Standards of Reporting Trials flow chart.

detection using a gamma probe. Subsequently, mastectomy or wide local excision was carried out as planned. Study group patients underwent ALND as a second procedure if the SLN was positive for macro- or micrometastasis (diameter, 0.2 to 2.0 mm). Division or preservation of the intercostobrachial nerve was recorded in all patients who underwent ALND in both groups. Patients received adjuvant radiotherapy, chemotherapy, and endocrine therapy according to local protocols. Pathology All SLNs measuring less than 5 mm in maximum diameter were bisected and both halves were processed for histologic examination. Nodes that were greater than 5 mm in maximum diameter were cut into three or more slices, all of which were processed for histologic examination. All blocks were sectioned at three levels of 100 ␮m each, and each level was stained with hematoxylin and eosin (HE). If the HE-stained sections showed no evidence of metastasis, serial sections from all of the levels on all of the blocks were stained with low molecular weight cytokeratin antibody CAM5.2 (Becton Dickinson Biosciences, Temse, Belgium) to look for micrometastasis.25 The lymph nodes removed during

conventional axillary dissection were processed according to breast screening pathology guidelines.26 Nodes greater than 5 mm in diameter were cut into two or more slices of approximately 3 mm or less perpendicular to the long axis of the node; those ⱕ 5 mm in diameter were embedded in their entirety. One section from each block was stained with HE. Immunohistochemistry was not performed on these lymph nodes, except under rare circumstances at the histopathologist’s discretion when confirmation of the epithelial nature of cells suggestive of metastasis was sought. Follow-Up All patients were reviewed at the breast clinic at 3-month intervals for the first year and will be observed yearly thereafter for 5 years. Outcome measures were assessed at time intervals indicated. Outcome measures. The key outcome measures are assessment of physical morbidity and psychological morbidity. The following parameters relating to physical morbidity were assessed pre- and postoperatively: numbness and paresthesia, arm swelling, shoulder mobility, and seroma formation.

4314

JOURNAL OF CLINICAL ONCOLOGY

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Randomized Trial of SNB in Breast Cancer

Subjective and objective assessment of sensory loss and paresthesia was performed preoperatively and at 1, 3, 6, and 12 months postoperatively. Objective assessment of arm swelling was performed by circumferential arm measurements at 4-cm intervals from the wrist. This yielded approximately 10 measurements (five each for forearm and upper arm, respectively), which were then used to calculate the limb volume using the formula for the volume of a truncated cone. This technique has been validated previously against the water displacement method to assess lymphedema and has been shown to have good correlation.27-31 Circumferential arm measurements in the ipsi- and contralateral arms were taken preoperatively and at 1, 3, 6, and 12 months postoperatively. Changes in volume in the ipsilateral arm were corrected for changes in the contralateral arm. In addition, the subjective assessment of presence/ absence of arm/hand swelling was recorded. The range of movement in the shoulder joint was measured by recording degrees of flexion, abduction, and internal and external rotation using a goniometer preoperatively and at 1, 3, 6, and 12 months postoperatively. The incidence of postoperative seroma formation was recorded in the two randomly assigned groups. All assessments were performed by S.U. or a member of research staff trained by S.U. (three in total) during the study period. Psychological assessment used a battery of well-validated self-report questionnaires, and was broad-ranging in view of evidence that patients vary widely in their psychological responses to the diagnosis and treatment of cancer.32 The Beck Depression Inventory (BDI)33 assesses the degree of depression and has been used extensively with cancer patients.34 The State-Trait Anxiety Inventory35 assesses stable differences in anxiety proneness (trait) and the degree of anxiety experienced at a given challenge (state), including cancer.36 Assessment of habitual trait anxiety and worry is important because the characteristic predicts poor adjustment in cancer patients.37 The Brief Symptom Inventory (BSI)38 assesses more serious psychiatric disturbance, and its subscales (including somatization and hostility) are sensitive to morbidity in cancer patients.39 The sum of the subscales provides a measure of overall distress: the global severity index (GSI). Cutpoints provided from BSI normative data allow scores on the BSI to be dichotomized to indicate the presence/absence of clinically significant morbidity. The Mental Adjustment to Cancer Scale (MAC) assesses coping responses via four subscales and has been widely applied in cancer patients.40 In addition to the above-described measures of morbidity, the Short Form-36 (SF-36; Medical Outcomes Trust, Waltham, MA)41,42 assesses subjective quality of life (QOL) in terms of physical and social functioning, pain, mental and general health, vitality and role limitations due to physical health, and emotional problems. Patients also completed the simple visual analog QOL scale by grading their perceived state of health on a continuous scale from 0 to

100. Similar to our previous study with breast cancer patients,43 questionnaires were completed immediately after surgery (baseline) and at 3, 6, and 12 months postoperatively. Baseline measures were conducted 7 to 14 days after surgery because, in addition to the ethical issue of sensitivity to the patient’s state before surgery, there is a methodologic advantage in that measures of distress are uncontaminated by the anxiety provoked while awaiting a definite diagnosis of cancer and consequent surgery.44 Statistics and Sample Size We required a 90% power to detect differences in all adverse effects and an 80% power for each individual adverse effect after ALND versus SLNB at a 5% significance level. A total of 300 patients were planned for entry onto the study. Univariate comparisons between groups of continuous variables were made using the t test. These were backed up where necessary by multivariate analyses adjusting for concomitant variables using analysis of variance. Categoric variables were compared using the ␹2 test. These were augmented by multivariate logistic regression analyses where necessary. For the psychosocial and QOL variables, we analyzed standardized T scores, where available, rather than raw scores. In addition, we adjusted for the effects of trait anxiety in the analysis of the BSI in light of evidence that high habitual levels of anxiety predict emotional distress. Introducing trait anxiety as a covariate removes the potentially confounding effect of the individual’s natural level of anxiety. In view of the large number of variables to be compared, adjustment for multiple comparisons was indicated. For the physical morbidity variables and the major psychological and QOL measures, this was not necessary because the measures were designed comparisons, as were the serial analyses at different points of follow-up. For the individual components of the BSI, MAC, and SF-36, however, we applied the Bonferroni principle within each questionnaire. That is, for the 10 components of the BSI and the 10 components of the SF-36, we applied a nominal significance criterion of .005 (.05/10). For the three components of the MAC, we applied a nominal significance criterion of .017 (.05/3). All statistical tests were two sided. The main analysis was performed on an intention-to-treat basis. We also analyzed node-negative and node-positive patients separately, given that these groups should be comparable because of random assignment to treatment. The node-negative patients would represent a comparison of 100% ALND with 100% SLNB only, and the node-positive patients would represent a comparison of immediate ALND with delayed ALND. RESULTS

Physical Morbidity Table 1 lists basic characteristics of the study and control groups in terms of age, histopathologic details, and 4315

www.jco.org

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Purushotham et al

Table 1. Basic Characteristics of Study and Control Patients Study Group (n ⫽ 143) Characteristic

No.

Age, years Mean SD Tumor size, mm Mean SD Gradeⴱ 1 2 3 Node positive LV invasion ER positive Radiotherapy Chemotherapy Endocrine therapy

Control Group (n ⫽ 155)

%

No.

%

57 9.5

58 10.6

16 7.9

16 8.8

25 71 42 49 24 114 132 43 114

18 51 31 34 17 80 92 30 80

38 72 42 40 25 129 137 36 114

25 47 28 26 16 83 88 23 74

Abbreviations: SD, standard deviation; LV, lymphovascular; ER, estrogen receptor. ⴱ Grade missing for five study group and three control group patients.

adjuvant therapy. There were no significant differences between study and control groups. Table 2 lists the odds ratios (ORs) for subjective lymphedema and the objective assessment of arm swelling, expressed as changes in arm volume over time; the latter is corrected for changes in the contralateral arm. That is, for each patient at each time point, the change since preoperative baseline measures in contralateral arm volume is subtracted from the change in ipsilateral arm volume. The study group had significantly smaller increases in arm volume for all time points except at 6 months (mean difference, 20 to 40 mL between the study and control group). For subjective classification of lymphedema as present/absent, the study group was significantly less likely to have

lymphedema at all time points, with a 70% reduction in odds overall (P ⫽ .004). Table 3 lists the results for subjective lymphedema and changes in arm volume stratified by node status. The differences between study and control groups are considerably greater and more significant in the node-negative patients, with typical differences in arm volume increase of 40 to 60 mL and an approximate 80% reduction in the odds of lymphedema in the study group. Interestingly, the node-negative control group patients had greater arm volume increases than the node-positive patients, although this was not significant. No significant differences between study and control groups were observed in arm volume increase or in the odds of lymphedema among the node-positive patients. Table 4 shows seroma incidence by randomization group and node status. Overall, 33 (21%) patients in the control group developed a seroma, 31 of whom required aspiration. In the study group, 20 (14%) patients developed a seroma, 17 of whom required aspiration. This difference was not significant (P ⫽ .1), but the difference in numbers requiring aspiration approached significance (P ⫽ .06). When stratified by nodal status, the proportion of nodenegative patients in the study group who developed a seroma was significantly smaller than the proportion in the control group (11% v 24%; P ⫽ .01), and the difference in proportions requiring aspirations was even more significant (9% v 23%; P ⫽ .008). Table 5 summarizes results for numbness, loss of sensitivity to light touch and pinprick, and any paresthesia, and lists the ORs for ever experiencing each sign/symptom during the year after surgery. There were significantly fewer sensory findings of all kinds (P ⬍ .001) in the study arm, with approximately a 60% reduction in the odds of paresthesia in the study group. For both loss of sensitivity to pinprick and light touch, the differences were also significant at individual follow-up times of 1, 3, 6, and 12 months. The difference between study and control groups was

Table 2. Arm Volume Changes and Subjective Lymphedema ORs by Time Since Surgery Arm Volume Change Corrected for Contralateral Change (mL) Control

Study

Analysis of Presence or Absence of Subjective Lymphedema (OR study/control)

Difference

Time Since Surgery

Mean

SE

Mean

SE

Mean

SE

Significance

ORⴱ

1 month 3 months 6 months 12 months Mean Maximum

49.6 60.0 45.5 56.4 53.1 113.7

9.4 12.5 11.0 10.9 8.1 9.7

8.1 26.1 22.1 18.6 17.7 78.4

8.3 11.0 13.3 13.8 9.2 12.0

41.5 33.9 23.4 37.8 35.4 35.3

12.6 16.7 17.5 17.6 12.2 15.3

.001 .04 .2 .03 .004 .02

0.34 0.40 0.25 0.36 0.30 —

95% CI 0.11 0.16 0.08 0.15 0.18

to 0.90 to 0.94 to 0.66 to 0.86 to 0.68 —

Significance .03 .04 .006 .02 .004 —

Abbreviation: OR, odds ratio. ⴱ For odds ratio estimates of subjective lymphedema, this column refers to the patient ever reporting or the clinician ever diagnosing lymphedema during the follow-up period.

4316

JOURNAL OF CLINICAL ONCOLOGY

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Randomized Trial of SNB in Breast Cancer

Table 3. Arm Volume Changes and Subjective Lymphedema ORs by Time Since Operation Stratified by Nodal Status Arm Volume Change Corrected for Contralateral Change (mL)

Node Status Negative nodes

Time Since Surgery

Mean

SE

Mean

SE

1 month 3 months 6 months 12 months Mean Maximum

59.2 68.0 56.8 59.8 61.3 123.5

10.8 15.7 13.8 12.6 10.0 11.4

⫺1.8 18.1 13.2 20.4 13.1 75.1

10.5 13.1 14.2 18.2 11.4 14.6

1 month 3 months 6 months 12 months Mean Maximum

16.0 35.4 13.3 46.3 29.5 85.5

17.8 16.7 14.6 22.3 11.7 18.1

28.9 43.5 42.7 14.7 27.0 85.1

12.9 20.1 29.8 18.8 15.8 21.1

Control

Study

Analysis of Presence or Absence of Lymphedema (OR study/control)

Difference SE

Significance

ORⴱ

61.0 49.9 43.5 39.4 48.2 48.4

15.2 20.7 19.8 21.8 15.1 18.3

⬍ .001 .02 .03 .07 .002 .009

0.21 0.32 0.18 0.20 0.24 —

0.03 0.08 0.03 0.03 0.09

to 0.80 to 0.99 to 0.71 to 0.75 to 0.59 —

.01 .03 .005 .007 ⬍ .001 —

⫺12.9 ⫺8.1 ⫺29.4 31.5 2.5 0.4

21.4 27.0 36.3 29.0 20.3 28.3

.6 .8 .4 .3 .9 .9

0.65 0.50 0.33 0.71 0.56 —

0.10 0.11 0.06 0.14 0.19

to 3.92 to 2.11 to 1.56 to 3.23 to 1.63 —

.3 .1 .6 .2 —

Mean

95% CI

Significance

Positive nodes

Abbreviation: OR, odds ratio. ⴱ For OR estimates of subjective lymphedema, this column refers to the patient ever reporting or the clinician ever diagnosing lymphedema during the follow-up period.

tive values. Although reduction in mobility was always lower on average in the study group (less impairment), this was significant only for flexion (P ⫽ .04). For all mobility variables, the pattern during the year showed a marked reduction soon after surgery, with a gradual return to preoperative mobility in both groups. The postoperative reduction in mobility tended to be smaller and the return to preoperative values was slightly faster in the study group, but the difference between the reduction compared with preoperative values was only significant for abduction at 1 month (P ⫽ .006) and flexion at 3 months (P ⫽ .02). In both cases, the study group showed a smaller reduction in mobility. When stratified by node status (data not shown) the difference between the study and control groups was greater for flexion (P ⫽ .001) in the node-negative group.

considerably attenuated when adjusted for preservation of the intercostobrachial nerve. When stratified by lymph node status, the differences between groups were greater in the node-negative patients and smaller in the nodepositive patients. For any paresthesia, 23% of nodenegative study group patients and 48% of node-negative control group patients ever had sensory findings (OR, 0.31; 95% CI, 0.16 to 0.60; P ⬍ .001), whereas the corresponding figures for node-positive patients were 45% in the study group and 48% in the control group (OR, 0.89; 95% CI, 0.35 to 2.26; P ⫽ .8), respectively. Similar results were observed at individual time points and for individual sensory findings. Table 6 lists the results for shoulder mobility. These are expressed as average reductions in the range of movement during the year postoperatively, compared with preopera-

Table 4. Incidence of Seroma Formation Control

Study

Patient Group

Seroma

No.

%

No.

%

OR

95% CI

Significance

All (node positive and node negative)

No Yes Aspirated No Yes Aspirated No Yes Aspirated

122 33 31 87 28 26 35 5 5

79 21 20 76 24 23 88 12 12

123 20 17 84 10 8 39 10 9

86 14 12 89 11 9 80 20 18

1.00 0.60 0.54 1.00 0.37 0.32 1.00 1.79 1.58

— 0.33 to 1.11 0.28 to 1.03 — 0.16 to 0.82 0.13 to 0.75 — 0.55 to 5.75 0.48 to 5.17

— .1 .06 — .01 .008 — .3 .4

Node negative

Node positive

Abbreviation: OR, odds ratio.

4317

www.jco.org

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Purushotham et al

Table 5. Sensory Findings in the Control and Study Groups in the 12 Months After Surgery and After Adjustment for Preservation of ICBN

Symptom/Sign Numbness Loss of pinprick Loss of light touch Any paresthesia

Experienced Symptom During Year

No.

%

No.

%

No Yes No Yes No Yes No Yes

40 115 37 118 34 121 25 130 ()

35 65 24 76 22 78 16 84

75 68 63 77 59 81 48 92 ()

52 48 45 55 42 58 34 66

Control

Study Odds Ratio Not Adjusted for ICBN

95% CI

Odds Ratio Adjusted for ICBN

95% CI

0.32

0.19 to 0.51

0.52

0.30 to 0.91

0.38

0.22 to 0.64

0.65

0.36 to 1.15

0.39

0.23 to 0.65

0.69

0.38 to 1.26

0.36

0.20 to 0.66

0.49

0.23 to 0.86

Abbreviation: ICBN, intercostobrachial nerve.

Psychological Morbidity In view of the number of comparisons conducted, we adjusted the significance criteria for the individual components of BSI, MAC, and SF-36 (as described in Patients and Methods) to reduce the risk of type I errors. Adjustment was also made in the analyses of the BDI, BSI, and MAC for patients’ individual levels of trait anxiety as a factor that might confound the other measures of morbidity. Depression and anxiety (BDI and State-Trait Anxiety Inventory). There were no significant differences between study and control groups in depressive symptoms or state anxiety during the 1-year follow-up period, nor were there differences as a function of node-positive or node-negative status. BSI and GSI. Table 7 lists mean and dichotomized GSI scores by time since surgery for the study and control groups overall, and in more detail as a function of nodepositive and node-negative patients. The mean GSI declined (improved) during the 12-month follow-up period for study and control groups, indicating an overall reduction in morbidity. The control group started at a higher (worse) level and their scores declined more rapidly than those of the study group, so that at 12 months the two groups had similar levels. The study group’s global mean scores were consistently lower (better) than those of the controls, but the difference was significant only at the immediate postoperative period (baseline) and ceased to be

significant when adjusted for trait anxiety. However, closer analysis as a function of node status showed that, when adjusted for trait anxiety, the immediate postoperative (baseline) score of node-negative patients in the study group was significantly lower than that of the controls. In contrast, and after adjustment for trait anxiety, the mean score of node-positive study group patients was significantly higher than that of the control group at 3 months. The subscales of the BSI provided additional insight into the elements of psychological disturbance contributing to overall distress. After adjustment for trait anxiety, in the immediate postoperative period (baseline) the study group scored lower (better) than the control group on somatization (P ⬍ .001). When considered more closely as a function of node status, and after adjustment for trait anxiety, the somatization scores of node-negative patients in the study group were found to be significantly lower in the immediate postoperative period (baseline; P ⬍ .001). In node-positive patients, differences between study group and control were found at 3 months: study group scores were significantly higher (worse) for interpersonal sensitivity (P ⫽ .003). After adjustment for trait anxiety, the difference in interpersonal sensitivity remained significant. MAC scale. There were no significant differences between study and control groups in MAC scores during the

Table 6. Average Reductions in Range of Shoulder Movement (in degrees) Compared With Preoperative Values Mean Reduction in Control group

Mean Reduction in Study Group

Measure of Mobility

Mean

No.

SD

Mean

No.

SD

Mean Difference

95% CI

Significance

Flexion Extension Abduction Internal rotation External rotation

13.0 ⫺1.5 6.3 1.7 2.9

141 139 138 139 139

32.9 10.7 11.5 12.7 12.3

6.7 ⫺2.2 3.1 0.3 1.5

134 134 132 134 134

15.6 8.1 15.7 12.0 11.0

6.3 0.7 3.2 1.4 1.4

0.1 to 12.6 ⫺1.5 to 3.3 ⫺0.5 to 6.3 ⫺1.5 to 4.4 ⫺1.5 to 4.4

.04 .6 .08 .3 .3

Abbreviation: SD, standard deviation.

4318

JOURNAL OF CLINICAL ONCOLOGY

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Randomized Trial of SNB in Breast Cancer

Table 7. Mean GSI Levels in Study and Control Groups and Analysis of Dichotomized GSI Scores by Time Since Surgery Mean Standardized GSI by Study Group Control Time Since Surgery All patients (node positive and node negative) Baselineⴱ 3 months 6 months 12 months Node-positive patients Baselineⴱ 3 months 6 months 12 months Node-negative patients Baselineⴱ 3 months 6 months 12 months

Study

OR for Morbid GSI Levels (study/control)

Difference

Mean

SE

Mean

SE

Mean

SE

Significance

OR

95% CI

Significance

52.9 50.9 50.8 49.7

0.9 0.9 1.1 1.1

49.7 50.4 48.9 48.4

0.8 0.9 0.9 0.9

3.2 0.5 1.9 1.3

1.2 1.3 1.4 1.4

.01 .7 .2 .4

0.45 0.54 0.36 0.55

0.13 to 1.35 0.11 to 2.24 0.05 to 1.80 0.08 to 2.94

.2 .5 .3 .6

50.0 46.0 50.7 47.2

1.5 1.5 2.0 2.0

49.0 53.0 50.3 49.8

1.3 1.3 1.8 1.6

1 ⫺6.7 0.39 ⫺2.6

2.1 2.0 1.3 2.6

.6 .01 .88 .3

— 0.88 1.04 0.95

— 0.01 to 71.9 0.1 to 13.8 0.01 to 78.4

— .92 .96 .97

53.9 52.5 50.8 50.7

1.1 1.1 1.3 1.3

50 49.5 48.5 47.8

1.0 1.1 1.0 1.1

3.9 3.0 2.5 2.9

1.5 1.6 1.7 1.7

.009 .05 .13 .08

0.41 0.49 0 0.45

0.10 to 1.34 0.07 to 2.43 0 to 0.87 0.03 to 3.31

.10 .32 .36 .35

Abbreviations: GSI, Global Severity Index; OR, odds ratio. ⴱ Refers to the immediate (7 to 14 days) postoperative period.

1-year follow-up period, nor were there differences as a function of node-positive or node-negative status. SF-36. In the immediate postoperative period, the physical combined score (P ⫽ .001), physical functioning score (P ⫽ .003), and vitality score (P ⫽ .004) were significantly higher (better) in the study group. In node-negative patients, the study group had significantly higher scores for vitality (P ⫽ .001) in the immediate postoperative period (baseline). Mean levels of these components in the study and control groups are listed in Table 8. QOL. The QOL scores were usually higher (better) in the study group, and significantly so immediately postoperatively (P ⫽ .01). Higher scores were accentuated in the node-negative patients, but again the only significant difference was in the immediate postoperative period (baseline; P ⫽ .01). No significant differences between study and control group were observed in the node-positive patients.

DISCUSSION

Several validation studies have indicated that SLNB is a reliable method of staging the axilla in patients with breast cancer, although there is a recognized false-negative rate associated with this procedure.11-14 A recently published randomized study on SLNB in breast cancer has shown a reduction in physical morbidity in the group of patients undergoing SLNB.24 The authors reported less pain and better arm mobility in patients who underwent SLNB compared with those who underwent ALND. In this study with 1-year follow-up, a detailed analysis has shown a significant reduction in physical morbidity in the study group. There was a significant reduction in postoperative arm swelling, rate of subjective lymphedema, and seroma formation. Numbness, loss of sensitivity to light touch and pinprick, and the experience of any paresthesia by the patient were also significantly reduced in the study

Table 8. Mean Immediate Postoperative Levels in Study and Control Groups of Those SF-36 Components Showing a Significant Difference Study Group Patient Group All

Node negative

Control Group

Difference

SF-36 Component

Mean

SD

Mean

SD

Mean

Physical combined Physical functioning Vitality Vitality

42.3 44.5 51.8 51.4

10.4 8.1 9.8 10.3

38.6 41.3 48.2 46.7

8.2 9.0 10.2 9.8

3.7 3.2 3.7 4.6

95% CI 1.2 1.1 1.1 1.6

to to to to

6.1 5.4 6.2 7.6

P .001 .003 .004 .001

Abbreviation: SD, standard deviation.

4319

www.jco.org

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Purushotham et al

group. Although shoulder mobility was less impaired on average in the study group, this was significant only for abduction at 1 month and flexion at 3 months. This reduction in physical morbidity is compatible with the less invasive surgical procedure of SLNB. In general, scores reflecting QOL and psychological morbidity were significantly better in the study group in the immediate postoperative period, with fewer long-term differences. This finding is not unexpected, given the significant reduction in physical morbidity in the study group. The one exception to this was the advantage of the control group in various psychological scores at 3 months when analysis was restricted to node-positive patients. This could be explained by the fact that, at 3 months, node-positive patients in the study group would have recently undergone their second operation (ALND) and commenced adjuvant therapy, which may have adversely affected their psychological state. When assessing psychological morbidity, the analysis confirmed the importance of accounting for habitual or chronic anxiety by applying the trait anxiety score as a covariate. A number of significant effects were lost when the covariate was entered. Our trial has some limitations. One is the impossibility of double blinding in a surgical trial of this nature. Another potential problem is the large number of study versus control comparisons. Some of these are designed contrasts, for which we had good reason a priori to seek a difference between study and control groups. For others, notably the individual components of the psychological and QOL questionnaires, this was not the case. We therefore adjusted the analysis of the components of the BSI, MAC, and SF-36 questionnaires for multiple testing. It should also be noted that we had a relatively small proportion of node-positive patients (20%). However, this is in keeping with the fact that SLNB is a technique designed to benefit patients who are likely to have node-negative disease. In the clinical setting, SLNB is primarily indicated for patients with a low risk of lymph node metastases and it is appropriate to reproduce this situation in research. In addition, as time passes, a higher proportion of tumors are REFERENCES 1. Pisani P, Parkin DM, Bray F, et al: Estimates of the worldwide mortality from 25 cancers in 1990. Int J Cancer 83:18-29, 1999 2. Say CC, Donegan W: A biostatistical evaluation of complications from mastectomy. Surg Gynecol Obstet 138:370-376, 1974 3. Budd DC, Cochran RC, Sturtz DL, et al: Surgical morbidity after mastectomy operations. Am J Surg 135:218-220, 1978 4. Tadych K, Donegan WL: Postmastectomy seromas and wound drainage. Surg Gynecol Obstet 165:483-487, 1987 5. Somers RG, Jablon LK, Kaplan MJ, et al: The use of closed suction drainage after lumpec-

presenting at an early stage, resulting in an increasing number of node-negative patients. An interesting incidental observation is that among the control group, the node-negative patients had greater arm volume increases than the node-positive patients (Table 3). This observation is of borderline statistical significance and is not attributable to outlier effects, given that the variability of arm volume changes is much the same in the node-negative controls as in the node-positive patients. One could speculate that in node-positive patients, metastatic involvement of axillary lymph nodes has led to the development of compensatory mechanisms for lymphatic flow, so that the subsequent effects of ALND are reduced, thereby resulting in less arm swelling postoperatively. With continued attempts at detecting breast cancer at an earlier stage through national breast screening programs, and with the reduction in mortality due to earlier detection and better treatment, it is important to minimize the morbidity associated with breast cancer treatment and to maintain patients’ QOL. This study demonstrates that SLNB in early breast cancer results in a significant reduction of physical and psychological morbidity. ■ ■ ■

Acknowledgment We thank J. Warwick, PhD, for helpful discussion. We thank the patients who participated in this study. Appendix The following are participants in the Cambridge/East Anglia Study Group: Cambridge: A.D. Purushotham, S. Upponi, M.B. Klevesath, G.C. Wishart, D. Barbera, S.J. Pain, L.G. Bobrow, A. Marker, R.M. Warren, P. Britton, A. Freeman, R. Sinnatamby, K.K. Balan, R. Barber, J. Rowley. Bury St Edmunds: E. Coveney. King’s Lynn: P. Gough, M. Robbins, P. Beckett. London: S.W. Duffy, J.P. Myles. Glasgow: K. Millar, T. Musiello. Authors’ Disclosures of Potential Conflicts of Interest The authors indicated no potential conflicts of interest.

tomy and axillary node dissection for breast cancer: A prospective randomized trial. Ann Surg 215:146-149, 1992 6. Hladiuk M, Huchcroft S, Temple W, et al: Arm function after axillary dissection for breast cancer: A pilot study to provide parameter estimates. J Surg Oncol 50:47-52, 1992 7. Gutman H, Kersz T, Barzilai T, et al: Achievements of physical therapy in patients after modified radical mastectomy compared with quadrantectomy, axillary dissection, and radiation for carcinoma of the breast. Arch Surg 125:389-391, 1990 8. Lin PP, Allison DC, Wainstock J, et al: Impact of axillary lymph node dissection on the therapy of breast cancer patients. J Clin Oncol 11:1536-1544, 1993

9. Ivens D, Hoe AL, Podd TJ, et al: Assessment of morbidity from complete axillary dissection. Br J Cancer 66:136-138, 1992 10. Petrek JA, Heelan MC: Incidence of breast carcinoma-related lymphedema. Cancer 83: 2776-2781, 1998 11. Giuliano AE, Kirgan DM, Guenther JM, et al: Lymphatic mapping and sentinel lymphadenectomy for breast cancer. Ann Surg 220:391401, 1994 12. McIntosh SA, Purushotham AD: Lymphatic mapping and sentinel node biopsy in breast cancer. Br J Surg 85:1347-1356, 1998 13. Veronesi U, Paganelli G, Galimberti V, et al: Sentinel-node biopsy to avoid axillary dissection in breast cancer with clinically negative lymph-nodes. Lancet 349:1864-1867, 1997

4320

JOURNAL OF CLINICAL ONCOLOGY

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.

Randomized Trial of SNB in Breast Cancer

14. Veronesi U, Paganelli G, Viale G, et al: Sentinel lymph node biopsy and axillary dissection in breast cancer: Results in a large series. J Natl Cancer Inst 91:368-373, 1999 15. Giuliano AE, Haigh PI, Brennan MB, et al: Prospective observational study of sentinel lymphadenectomy without further axillary dissection in patients with sentinel node-negative breast cancer. J Clin Oncol 18:2553-2559, 2000 16. Burak WE, Hollenbeck ST, Zervos EE, et al: Sentinel lymph node biopsy results in less postoperative morbidity compared with axillary lymph node dissection for breast cancer. Am J Surg 183:23-27, 2002 17. Haid A, Koberle-Wuhrer R, Knauer M, et al: Morbidity of breast cancer patients following complete axillary dissection or sentinel node biopsy only: A comparative evaluation. Breast Cancer Res Treat 73:31-36, 2002 18. Haid A, Kuehn T, Konstantiniuk P, et al: Shoulder-arm morbidity following axillary dissection and sentinel node only biopsy for breast cancer. Eur J Surg Oncol 28:705-710, 2002 19. Peintinger F, Reitsamer R, Stranzl H, et al: Comparison of quality of life and arm complaints after axillary lymph node dissection vs sentinel lymph node biopsy in breast cancer patients. Br J Cancer 89:648-652, 2003 20. Schijven MP, Vingerhoets AJ, Rutten HJ, et al: Comparison of morbidity between axillary lymph node dissection and sentinel node biopsy. Eur J Surg Oncol 29:341-350, 2003 21. Schrenk P, Rieger R, Shamiyeh A, et al: Morbidity following sentinel lymph node biopsy versus axillary lymph node dissection for patients with breast carcinoma. Cancer 88:608614, 2000 22. Swenson KK, Nissen MJ, Ceronsky C, et al: Comparison of side effects between sentinel lymph node and axillary lymph node dissection for breast cancer. Ann Surg Oncol 9:745-753, 2002

23. Temple LK, Baron R, Cody HS III, et al: Sensory morbidity after sentinel lymph node biopsy and axillary dissection: A prospective study of 233 women. Ann Surg Oncol 9:654-662, 2002 24. Veronesi U, Paganelli G, Viale G, et al: A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer. N Engl J Med 349:546-553, 2003 25. Makin CA, Bobrow LG, Bodmer WF: Monoclonal antibody to cytokeratin for use in routine histopathology. J Clin Pathol 37:975-983, 1984 26. National Coordinating Group for Breast Screening Pathology: Pathology Reporting in Breast Cancer Screening (ed 2). Sheffield, National Health Service Breast Screening Programme Publication, No. 3, 1997 27. Engler HS, Sweat RD: Volumetric arm measurements: Technique and results. Am Surg 28:465-468, 1962 28. Casley-Smith JR: Measuring and representing peripheral oedema and its alterations. Lymphology 27:56-70, 1994 29. Pani SP, Vanamail P, Yuvaraj J: Limb circumference measurement for recording edema volume in patients with filarial lymphedema. Lymphology 28:57-63, 1995 30. Stanton AW, Levick JR, Mortimer PS: Current puzzles presented by postmastectomy oedema (breast cancer related lymphoedema). Vasc Med 1:213-225, 1996 31. Stranden E: A comparison between surface measurements and water displacement volumetry for the quantification of leg edema. J Oslo City Hosp 31:153-155, 1981 32. Payne DK, Sullivan MD, Massie MJ: Women’s psychological reactions to breast cancer. Semin Oncol 23:89-97, 1996 33. Beck A, Steer RA, Brown GK: Beck Depression Inventory II Manual. San Antonio, TX, Psychological Corp, Harcourt Brace & Co, 1996

34. Absetz P, Aro AR, Sutton SR: Experience with breast cancer, pre-screening perceived susceptibility and the psychological impact of screening. Psychooncology 12:305-318, 2003 35. Spielberger CD: State-Trait Anxiety Inventory Manual. Palo Alto, CA, Consulting Psychologists Press Inc, 1983 36. Stark D, Kiely M, Smith A, et al: Anxiety disorders in cancer patients: Their nature, associations, and relation to quality of life. J Clin Oncol 20:3137-3148, 2002 37. Tjemsland L, Soreide JA, Malt UF: Posttraumatic distress symptoms in operable breast cancer III: Status one year after surgery. Breast Cancer Res Treat 47:141-151, 1998 38. Derogatis LR: Brief Symptom Inventory Manual (ed 3). Minneapolis, MN, NCS National Computer Systems Inc, 1993 39. Hanson Frost M, Suman VJ, Rummans TA, et al: Physical, psychological and social well-being of women with breast cancer: The influence of disease phase. Psychooncology 9:221-231, 2000 40. Watson M, Greer S, Young J, et al: Development of a questionnaire measure of adjustment to cancer: The MAC scale. Psychol Med 18:203-209, 1988 41. Brazier JE, Harper R, Jones NM, et al: Validating the SF-36 health survey questionnaire: New outcome measure for primary care. BMJ 305:160-164, 1992 42. Wapnir IL, Cody RP, Greco RS: Subtle differences in quality of life after breast cancer surgery. Ann Surg Oncol 6:359-366, 1999 43. Purushotham AD, McLatchie E, Young D, et al: Randomized clinical trial of no wound drains and early discharge in the treatment of women with breast cancer. Br J Surg 89:286-292, 2002 44. Nosarti C, Roberts JV, Crayford T, et al: Early psychological adjustment in breast cancer patients: A prospective study. J Psychosom Res 53:1123-1130, 2002

4321

www.jco.org

Downloaded from ascopubs.org by 37.44.207.190 on January 22, 2017 from 037.044.207.190 Copyright © 2017 American Society of Clinical Oncology. All rights reserved.