BREAST CONSERVATION SURGERY IN LOCALLY ADVANCED BREAST CANCERS FOLLOWING NEOADJUVANT CHEMOTHERAPY

Guide:[Link] Co-Guides:[Link] [Link] Dr. Basu

OBJECTIVES
To assess the adequacy and effectiveness of Breast Conservation Surgery (BCS) in locally advanced cancers after downstaging by neoadjuvant chemotherapy by margin analysis.

MATERIALS AND METHODS

It is a prospective interventional study between September 2007 and April 2009. Sample size : 30 Inclusion criteria : Patients with LABC who receive NAC and have residual lesion less than or equal to 4 cm.

Exclusion criteria : 1. Presence of metastasis at the time of diagnosis /during the study period. 2. Any contraindications for radio therapy like collagen vascular disease. 3. Patients not willing for BCS . 4. Presence of multi focal lesions / diffuse microcalcification / chest wall involvement / extensive intraductal component by histopathology. 5. Male breast cancer

Results
BCS done in 30 patients with locally advanced breast cancers . The mean age was 47.27 +/- 9.7 years

 Out of 30 patients 29 had infiltrating ductal carcinoma and 1 had mucinous carcinoma. 15 patients had received cyclophosphamide/adriamycin/5fluorouracil regimen and the other 15 had received paclitaxel/adriamycin for neoadjuvant chemotherapy

TABLE 1A

Clinicopathologic characteristics of patients

CHARACTERISTIC Tumor T3 T4b Nodal Status N0 N1 Stage IIB IIIA IIIB Grade I II III Unknown Number of patients(n=30) 28 ( 93.3%) 2 ( 6.7%) 16 ( 53.3%) 14 (46.7%) 16(53.33%) 12(40%) 2(6.7%)

6 ( 20.0%) 17 ( 56.7%) 1 ( 3.3%) 6 ( 20.0%)

 Table 2 shows the mean size of tumor before and after NAC with mean percentage reduction in size both by clinical and MDCT assessment . The prechemotherapy mean tumor size was significant but the postchemotherapy size was not significant.

TABLE 2: Mean tumor size by clinical assessment and MDCT

Mean tumor size Clinical MDCT P value

Prechemotherapy size (cm)

5.03 1.27

4.00 1.7

0.002

Postchemotherap y size (cm)

2.18 1.45

1.90 1.004

0.116

% reduction in size

56.54 26.58

52.65 23.67

0.239

 The clinical assessment overestimated the degree of response in 5 patients who had clinically complete response but showed some residual enhancement on MDCT. In the other 5 patients, clinical assessment underestimated the degree of response when compared to MDCT. In 2 patients there was an ill-defined indurated area palpable with no enhancement on MDCT. Clinically three patients had no response but had partial response by MDCT.

TABLE 3: Comparison of response to NAC by clinical and MDCT assessment

Clinical assessment MDCT assessment (n=26) Complete response Complete response Partial response No response 0 Partial response 5 13 No response

0 3

 On comparing the residual tumor size after Neoadjuvant Chemotherapy by clinical examination and MDCT assessment with gross appearance by pathological examination, it was found that the difference in size was not significantly different from that of pathological examination i.e. both methods of assessment correlated well with pathological gross assessment (Table 4).

TABLE 4: Correlation of mean tumor size by clinical measurement and MDCT after NAC with gross pathologic assessment
Method of assessment Gross appearance Clinical size Mean ( cm ) P value

1.811.36

Comparison group 0.066

1.971.13

CT size

1.901.064

0.46

 Out of 30 patients, 8 patients had pathological complete response and 22 had incomplete response Out of 8 patients who had complete pathological response, 6 received paclitaxel and adriamycin and 2 received CAF regime. The clinical assessment for tumor response was in agreement with pathologic tumor response in 23 (76.6%) out of 30 cases. In 5 patients response was underestimated and in 2 it was overestimated.

TABLE 5 Comparison of assessment of response to NAC by clinical assessment and histology

Pathological response Clinical response(n=30 )
Complete Incomplete

Complete

Incomplete

20

 MDCT assessment was in agreement with pathologic assessment of tumor response in 20 out of 26 cases (76.9%). In the other 6 patients it underestimated tumor response. All the 6 patients had pathologic complete response but showed evidence of residual enhancement by MDCT

TABLE 6: Comparison of assessment of response to NAC by MDCT and histology MDCT assessment(n=26) Pathology

Complete response Complete response Incomplete response 2

Incomplete response 0

18

 Before chemotherapy the mean tumor size by clinical assessment using vernier calipers was 6 +/- 1.69 cm in complete responders. By MDCT assessment the mean tumor size was 5.46 +/- 3.31 cm in complete responders and 4.16 cm +/- 1.32 cm in incomplete responders. By both methods of assessment, there was no significant difference observed between complete responders and incomplete responders with respect to tumor size before starting chemotherapy

TABLE 7 Comparison of mean Pre-chemotherapy tumor size between complete and incomplete responders by HPE Method of Complete assessment response Incomplete P value response

Mean tumor 5.37 1.25 4.91 1.28 0.382 size ( Clinical in cm ) Mean tumor 4.73 2.44 3.67 1.14 0.273 size ( CT in cm

 On comparing the hormonal receptor status (ER, PR, Her2Neu) in complete responders and incomplete responders there was no significant difference observed

TABLE 8: Comparison of estrogen receptor status between complete and incomplete responders ER status

Pathological response
Incomplete Complete

Total

Positive

12 (80.0%) 10 (71.4%) 22 (75.9%)

3 (20.0%) 4 (28.6%) 7 (24.1%)

15 (100%) 14 (100%) 29 (100%)

Negative

Total

TABLE 9: Comparison of progesterone receptor status between complete and incomplete responders PR status Pathological response Complete 3 (23.1%) 4 (25.0%) 7 (24.1%) Incomplete 10 (76.9%) 12 (75.0%) 22 (75.9%) Total 13 (100.0%) 16 (100.0%) 29 (100.0%)

Positive

Negative

Total

TABLE 10: Comparison of Her2Neu receptor status between complete and incomplete responders Her2Neu receptor Pathological response Complete 4 (40.0%) 3 (17.6%) 7 (25.9%) Incomplete 6 (60.0%) 14 (82.4%) 20 (74.1%) 10 (100.0%) 17 (100.0%) 27 (100.0%) Total

Positive

Negative

Total

 Out of 30 patients, 15 had received cyclophosphamide, adriamycin and 5flurouracil (CAF) and the other 15 had received paclitaxel and adriamycin(AP ) The two groups (CAF and AP) were comparable with respect to age, hormone receptor status, grade of the tumor and the pathological response

TABLE 1 1: Clinicopathologic characteristics of patients in the two chemotherapy groups CAF ( n=15 ) Paclitaxel P value (n=15)

Her2Neu status
Positive Negative 4 (40.0%) 10(58.8%) 6 (60.0%) 7 (41.2%) 0.440

Grade
I II III Unknown 3 (50.0%) 8 (47.1%) 1 (100.0%) 3 (50.0%) 3 (50.0%) 9 (52.9%) 0 (0%) 3 (50.0%) 0.787

TABLE 1 1: Clinicopathologic characteristics of patients in the two chemotherapy groups CAF ( n=15 ) Paclitaxel (n=15) Pathological response 5 (62.5%) 10 (45.5%) P value

Complete Incomplete

3 (37.5%) 12 (54.5%)

0.682

 The mean tumor size before starting chemotherapy was not significantly different between the two groups both by clinical assessment and MDCT assessment. However the mean residual tumor size after 3 cycles of chemotherapy was statistically different by clinical assessment (P<0.04) although not by MDCT.

 The mean percentage reduction in size in CAF group was 37.00% +/- 15.44 and in AP group was 64.14% +/- 22.42 and this difference was statistically significant (P<0.001) (Table 12). The taxol based regime caused significant decrease in the original size of the tumor and thereby increasing breast conservation rates.

TABLE 12: Pre and post chemotherapy tumor size in patients receiving two different regimes.
Method of assessment Pre-chemo (clinical in cm) Post-chemo (clinical in cm) Pre-chemo (CT in cm) Post chemo (CT in cm) Percent reduction in size (CT) CAF ( n=15) Paclitaxel (n=15) P value

4.87 1.06 2.70 1.32

5.20 1.46 1.67 1.43

0.48 0.05

3.66 2.22

1.49 0.75

4.28 1.82 1.58 1.14

0.359 0.08

37.78 % 16.14

65.39 % 21.87

0.001

 A total of 174 margins were analyzed by frozen section. Nine margins were positive (fig 4a), 4 were close (fig 4b), 1 was suspicious (fig 4c) and all others were negative (fig 4d). Margins were revised intraoperatively if they were reported as positive, close or suspicious. In 2 patients, 2 out of 6 margins were positive and in another patient, 3 out of 6 margins were positive. In effect, 10 patients required intraoperative revision of margins.

TABLE 13: Comparison of Frozen section of margins with final Histopathological examination

Frozen section

Final Histopathology examination Positive Negative 12 0 12

Sensitivity: 100% Specificity: 98.8% Positive Predictive value: 85.7% Negative Predictive value: 100%

Total

Positive Negative Total

2 160 162

14 160 174

 Analyzing the patient and tumor characteristics with those who had revision and who did not have revision of the margins, no statistical difference was obtained with respect to age, menopausal status, ER & PR status and grade of the tumor. However, Her2Neu receptor status was found to be statistically significant (P< 0.009) between these two groups (Table 14).

TABLE 14: Clinicopathologic characteristics of patients with and without margin revision.
Revised ( n=10 ) Age (years) 40 > 40 Not Revised (n=20) P value

4 6

4 16 Menopausal status 9 11 ER status 9 10

0.465

Premenopausal Postmenopausal

4 6

1.000

Positive Negative

6 4

0.798

TABLE 14: Clinicopathologic characteristics of patients with and without margin revision.

Revised ( n=10 )

Not Revised P value (n=20) PR status 9 1.000

Positive Negative

4 6

10 Her2Neu status 10 8 0.009

Positive Negative

0 9

TABLE 14: Clinicopathologic characteristics of patients with and without margin revision. Revised ( n=10 ) Not Revised (n=20) P value

Histological grade I II III 1 6 0 5 11 1

0.566

 Comparing the mean pre-chemotherapy size by clinical and CT assessment between margin revised and not revised groups; the difference was not statistically significant. But the mean post chemotherapy size by clinical assessment, CT and gross assessment by pathology was significantly different between these two groups i.e. residual tumor size appears to be a significant predictor of positive or close surgical margin which requires reexcision (Table 15).

TABLE 15: Mean tumor size in patients with and without margin revision. Method of Revised Not revised P Value assessment 5.20 cm 1.84 4.95 cm 0.90 0.619 Pre-chemo size (clinical) 3.76 cm 1.41 4.08cm 1.79 0.667 Pre-chemo size (CT) Post-chemo size (clinical) Post chemo size (CT) 3.2 cm 2.5 cm 0.75 0.85 1.67 cm 1.46 0.005 1.6 cm 0.95 0.018

3.18 cm Pathologic residual tumor 16.14 size

1.39

1.5

0.001

 On gross examination of the tumor, out of 173 margins 31 margins had gross evidence of tumor less than or equal to 0.5 cm from the cut margin. Out of these 31 margins, 9 were positive, 3 were close and 19 were negative by histology (Table 16) and it was statistically significant (p<0.0001).

TABLE 16: Comparison of gross inspection with histologic examination for margin assessment Gross inspection 0.5 Negative Close Positive Total

19 ( 61.3% ) 141 ( 99.3% ) 160 ( 92.5% )

3 ( 9.7% ) 1 ( 0.7% ) 4 ( 2.3% 0

9 ( 29.0% ) 0 ( 0% ) 9 ( 5.2% )

31 ( 100.0% ) 142 ( 100.0% ) 173 ( 100.0% )

> 0.5

Total

SHRINKAGE PATTERN

Shrinkage patterns after neoadjuvant chemotherapy

Pre chemo

Post chemo