Abstract
Some invasive breast carcinomas are surrounded by a clear space that separate the tumor cells from the adjacent stroma, similar to invasive micropapillary carcinoma (IMPC), but lack the thin strands of connective tissue that separate the cells and characteristic “inside-out” growth pattern of IMPC on immunohistochemical stain for EMA. We consider the presence of the retraction clefts a common phenomenon that may present as a precursor stage of IMPC (PSIMPC). In this study, a total of 2497 cases of invasive breast carcinomas were prospectively collected. Among 2497 cases of breast cancer, 949 (38.0%) cases were PSIMPC, 200 (8.0%) cases were IMPC and 1348 (54.0%) cases were IBC-NST. LVI was seen in128 of 200 (64%) IMPC and 364 of 948 (38.0%) PSIMPC, in contrast to 246 of 1341 (18%) IBC-NST (P<0.001). Lymph node metastasis was seen in 147 of 200 (73.4%) IMPC and 551 of 949 (58%) PSIMPC, in contrast to 563 of 1345 (42%) IBC-NST (P<0.001). The 5-year disease-free survival (DFS) and overall survival (OS) of PSIMPC were 76.8% and 87.6%, compared to 63.2% and 85.9% in IMPC, 86.2% and 93.7% in IBC-NST. PSIMPC demonstrated a more favorable DFS and OS compared to IMPC, but worse DFS and OS compared to IBC-NST. Cox and logistic regression analysis showed that PSIMPC was an independent predictor of DFS and OS. Our findings suggest that the presence of retraction clefts is a precursor state of IMPC, exhibiting IMPC-like features, such as higher incidence of lymphovascular invasion, lymph node metastasis and more aggressive clinical behavior.
Keywords: breast cancer, retraction clefts, invasive micropapillary carcinoma, lymphovascular invasion, lymph node metastasis
1. Introduction
Invasive micropapillary carcinoma (IMPC) is a rare special histologic subtype of breast cancer, accounting for approximately 3% to 8% of all breast cancers [1, 2]. It consists of small, hollow, or morule-like clusters of cancer cells surrounded by clear stromal space [3]. The cells exhibit reversed polarity with the glandular epithelial cells located on the outer surface of the cell nests, revealed by immunohistochemical stain for epithelial membrane antigen (EMA). IMPC is associated with higher incidence of lymphovascular invasion (LVI), lymph node metastasis and poor prognosis [4, 5].
Some invasive breast cancers have tumor cell clusters surrounded by clear stromal space similar to IMPC, but lack the morphological and immunohistochemical characteristics of IMPC. On formalin-fixed paraffin-embedded tissue sections, there are often clear spaces separating tumor cell clusters from adjacent stroma, which can be difficult to distinguish from LVI [6, 7]. Previous studies have shown that the presence of these so-called retraction clefts (RC), may be important for the histologic diagnosis and overall survival of prostate cancer and oral squamous carcinoma [8, 9,10], in general, RCs are considered artifactual, which may be due to inadequate tissue fixation and tissue contraction. However, characteristic retraction clefts around the tumor cell nests may be found even on frozen section of invasive breast carcinoma [6], and this phenomenon may have special significance, rather than artificial products [11].
In this prospective study, we assessed whether the presence of retraction clefts in tumors without the characteristic histological features of IMPC, is associated with lymphovascular invasion, lymph node metastasis and outcomes in a large consecutive series of invasive breast cancers.
2. Materials and methods
2.1. Clinical Data
A total of 2497 cases of invasive breast carcinomas were prospectively collected from Tianjin Medical University Cancer Institute& Hospital (TMUCIH). Patients were diagnosed and had surgical treatment between January 2005 and December 2006. Hematoxylin and eosin (H&E) stained slides were reviewed by three breast pathologists (HQ Jia, WD Li, and L Fu) to assess the diagnosis according to the WHO Classification of Tumors of the Breast [12], including histologic type, grade, LVI and LN involvement.
Special histologic subtypes of invasive breast cancer other than IMPC were excluded. IMPC showed reversed polarity, and the glandular epithelium was located on the outer surface of tumor cell nests labeled by immunohistochemical stain for EMA (Figure 1A&1B). Cases with clear spaces around tumor cell clusters but lack characteristic features of IMPC were identified as PSIMPC. As long as there is any proportion of RC in the tumor, it is classified as IMPC or PSIMPC according to the morphological features of the tumor cells and their arrangements, cases that difficult to distinguish were further identified as IMPC or PSIMPC by immunohistochemistry with EMA. Cases with both IMPC and PSIMPC components were classified as IMPC. In order to compare the characteristics of IMPC and PSIMPC, a case with both IMPC and PSIMPC components is illustrate in Figure 1 (Figure 1C&1D). In selected cases, if LVI and RC were difficult to distinguish according to the morphological features, immunohistochemical stain for D2-40 was performed to support the presence and absence of LVI [13, 14]. Retraction clefts result in clear spaces is seen around DCIS, and D2-40 immunohistochemistry can highlight the myoepithelial cells around breast DCIS. In very difficult cases, additional markers for myoepithelial cells and lymphatic vessels (such as p63, SMMHC and CD31) were performed to avoid misinterpretation of LVI.
Figure 1.
Invasive micropapillary carcinoma (IMPC) showing the characteristic separated from the stroma by a clear space (A, H&E, ×100), and immunohistochemical stains for epithelial membrane antigen (B, immunohistochemical staining, ×100), mark the polarity reversal with the luminal aspect of the cell present on outer surface of the cluster. The upper left part (marked by arrow) is IMPC, and the rest is shown as invasive breast carcinoma with retraction cleft (C, H&E, ×100), EMA immunohistochemical staining showed that the upper left part (marked by arrow) was IMPC with reversed polarity, and the rest lack the characteristic of reversed polarity (D, immunohistochemical staining, ×100). Invasive breast carcinoma with lymphatic vessels (E, H&E, ×100), lymphatic endothelial cells immunohistochemical labeled by D2-40(F, immunohistochemical staining, ×100).
2.2. Immunohistochemistry
The 4 μm sections of formalin fixed and paraffin embedded sections were analyzed by immunohistochemistry. Immunocytochemical stain for EMA (clone E29, Dako, Denmark) was performed on PSIMPC and IMPC which were difficult to distinguish according to the morphological features.
D2-40 (clone D2-40, Zymed) immunohistochemical stain highlights endothelial cells, which was performed to confirm the presence of LVI when necessary (Figure 1E&1F). Estrogen receptor (ER) antibody (clone SP1) and progesterone receptor (PR) antibody (clone SP2) were from Zymed. Immunohistochemical analysis of human epidermal growth factor receptor 2 (HER-2) used DAKO HercepTest TM (Denmark). Assessment of ER, PR and HER2 followed the American Society of Clinical Oncology /College of American Pathologists guidelines [15, 16].
2.3. Statistically analyze
Mann-Whitney U test and Kruskal-Wallis test were used to compare the clinicopathological characteristics between the two groups of variables and the multiple groups of variables. Cox proportional risk model was used to evaluate the effect of tumor variables on survival rate. Univariate and multivariate logistic regression models were applied to analyze the predictors for disease-free survival and overall survival. Kaplan-Meier method calculated the DFS and OS curves, and the differences between the curves were evaluated using a log-rank test. 2-sided P<0.05 was identified statistical significance. SPSS 22.0 software for Windows (IBM, NY, USA) was used for statistical analysis.
3. Results
3.1. Clinicopathologic characteristics
Table 1 shows the clinicopathologic characteristics of PSIMPC and IMPC. Among a total of 2497 cases, 949 (38%) and 200 (8%) showed features of typical PSIMPC and IMPC, respectively. The range of PSIMPC component was 1%-95% of the tumor, with a median proportion of 15% (mean ±standard error of mean, 23.6 ± 0.8). The IMPC component was 1% −100% of the tumor, with a median proportion of 50% (mean ±standard error of mean, 53.7 ± 2.2). In most (68%) PSIMPC cases, the proportion of PSIMPC was less than 25%. In only 14% of PSIMPC cases, PSIMPC constituted more than 50% of the tumor, in contrast to nearly half (49%) of cases of IMPC, the IMPC component constituted more than 50% of the tumor (P<0.001).
Table 1.
Clinicopathologic characteristics of PSIMPC、 IMPC and IBC-NST
| Characteristics | PSIMPC (n,%) | IMPC(n,%) | IBC-NST (n,%) | Pa/Pb |
|---|---|---|---|---|
| 949(38) | 200 (8) | 1348(54) | ||
| Age at diagnosis | 0.482/0.077 | |||
| ≤50 | 453(48) | 90 (45) | 694 (52) | |
| > 50 | 496(52) | 111 (55) | 654 (48) | |
| Tumor size ( mm) | 0.530/<0.001 | |||
| ≤20 | 310(33) | 58 (30) | 591(44) | |
| 20-50 | 562(60) | 124(64) | 680 (51) | |
| > 50 | 65(7) | 12(6) | 60(5) | |
| PSIMPC / IMPC proportion | <0.001 | |||
| ⩽25% | 648(68) | 48 (24) | / | |
| 26-50% | 167(18) | 55(27) | / | |
| 51-75% | 91(10) | 38(19) | / | |
| 76-100% | 43(4) | 59(30) | / | |
| LN | <0.001/<0.001 | |||
| − | 398(42) | 53 (26) | 782(58) | |
| + | 551(58) | 147 (74) | 563 (42) | |
| LVI | <0.001/<0.001 | |||
| − | 584(62) | 72(36) | 1095 (82) | |
| + | 364(38) | 128(64) | 246(18) | |
| TNM stage | <0.001/<0.001 | |||
| 1 | 173(19) | 23(12) | 400 (30) | |
| 2 | 455(49) | 68(34) | 649 (48) | |
| 3/4 | 308(33) | 106(54) | 290 (27) | |
| Histologic Grade | 0.364/<0.001 | |||
| I | 56(6) | 17 (11) | 160(13) | |
| II | 694(79) | 109(73) | 926(75) | |
| III | 133(15) | 24 (16) | 144 (12) | |
| ER | 0.002/0.053 | |||
| − | 239(26) | 30(15) | 319(25) | |
| + | 685(74) | 167(85) | 977(75) | |
| PR | 0.028/0.003 | |||
| − | 285(31) | 45(23) | 325(25) | |
| + | 643(69) | 152(77) | 972(75) | |
| HER-2 | 0.832/<0.001 | |||
| 0/1+ | 696(75) | 150(75) | 1075(82) | |
| 2+ | 159(17) | 34(17) | 168(13) | |
| 3+ | 76(8) | 15(8) | 66(5) | |
| Local regional recurrence/Distant metastasis | 193(26) | 62(39) | 189(19) | <0.001/<0.001 |
| Death of tumor | 93(12) | 35(21) | 57(6) | <0.001/<0.001 |
IMPC invasive micropapillary carcinoma, PSIMPC precursor state of IMPC, LN lymph node, LVI lymphovascular invasion, ER estrogen receptor, PR progesterone receptor, HER-2 human epidermal growth factor receptor 2, Pa and Pb represent the P value of PSIMPC compared with IMPC and IBC-NST respectively.
Patient’s age was not significantly different between PSIMPC and IMPC, IBC-NST group. Compared to IBC-NST, PSIMPC had significantly larger tumor size (P<0.001) and higher histological grade (P<0.001). PSIMPC had significantly higher incidence of LVI (P<0.001) and LN(P<0.001) metastasis comparing to IBC-NST. LVI was seen in 364 (38.0%) of 949 PSIMPC and 128 (64%) of 200 IMPC, in contrast to 246 (18%) of 1341 IBC-NST. LN metastasis was seen in 551 (58%) of 949 PSIMPC and 147 (73.4%) of 200 IMPC, in contrast to 563 (42%) of 1345 IBC-NST (P<0.0001). The rates of LVI and LN metastasis in PSIMPC were significantly higher than those in IBC-NST, but lower than those in IMPC.
IMPC was more frequently ER and PR positive than PSIMPC and (P=0.002 for ER and P=0.028 for PR). There was no significant difference in ER expression between PSIMPC and IBC-NST(P=0.053). While the expression of PR in IBC-NST is higher than that of PSIMPC(P=0.003). The presence of PSIMPC showing significant positive correlation with HER-2 status based on immunohistochemistry, compared to IBC-NST(P<0.001). The rates of locoregional recurrence, distant metastasis and breast cancer related death in PSIMPC were significantly higher than those in IBC-NST (P<0.001).
3.2. Correlation of PSIMPC distribution in tumor with clinicopathologic characteristics
PSIMPC cases were further divided into three groups based on the distribution of PSIMPC in the tumor: both peripheral and internal group, peripheral only group and internal only group. Examples of the different distributions of PSIMPC are illustrated in Figure 2. The proportion of PCIMPC in the both peripheral and internal group was significantly higher than that in the peripheral only and internal only groups (P<0.001).
Figure 2.
PSIMPC located in the internal of tumor (A, H&E, ×100), RC located in the periphery of tumor and cancer cells infiltrate marginal adipose tissue (C, H&E,×100). Note the diffuse EMA immunoreactivity in the tumor cells, lack the polarity reversal architecture (B&D, immunohistochemical staining, ×100).
Comparison of clinicopathological characteristics of the three groups is summarized in Table 2. Internal only group was associated with less frequent LVI (P<0.001) and LN metastasis (P<0.001) than the other two groups. Both peripheral and internal group was less frequently positive for ER and PR (P=0.005 and P=0.006), and decreased expression of HER-2 status based on immunohistochemistry (P=0.024). No significant correlation in patient age, tumor size, histologic grade and regional recurrence/ distant metastasis and death of disease was observed among the three groups.
Table 2.
Correlation of the location PSIMPC with patient clinicopathologic characteristics
| Characteristics | location of PSIMPC (n,%) | P | ||
|---|---|---|---|---|
| both peripheral and internal group |
peripheral only group |
internal only group |
||
| Age at diagnosis | 0.787 | |||
| ≤50 | 275 (47) | 71(50) | 104(47) | |
| >50 | 308 (53) | 70(50) | 116(53) | |
| Tumor size ( mm) | 0.374 | |||
| ≤20 | 188 (33) | 50(36) | 72(33) | |
| 20-50 | 343(59) | 84(60) | 132(61) | |
| > 50 | 46(8) | 5(4) | 12(6) | |
| PSIMPC proportion | <0.001 | |||
| ⩽25% | 364 (62) | 112(79) | 170(77) | |
| 26-50% | 123(22) | 11(8) | 29(13) | |
| 51-75% | 60(11) | 12(8) | 15(7) | |
| 76- 95% | 30(5) | 6(4) | 6(3) | |
| LN | <0.001 | |||
| − | 223 (38) | 56(40) | 118(54) | |
| + | 360 (62) | 85(60) | 102(46) | |
| LVI | <0.001 | |||
| − | 325(56) | 94(67) | 162(74) | |
| + | 258(44) | 47(33) | 58(26) | |
| TNM stage | 0.007 | |||
| 1 | 99(17) | 22(16) | 52 (24) | |
| 2 | 265(46) | 75(54) | 114 (52) | |
| 3/4 | 210(37) | 42(30) | 52(24) | |
| Histologic Grade | 0.881 | |||
| I | 36(7) | 7(6) | 13(6) | |
| II | 424(77) | 104(81) | 162(80) | |
| III | 87(16) | 17(13) | 28(14) | |
| ER | 0.005 | |||
| − | 166(29) | 30(22) | 40(19) | |
| + | 401(71) | 108 (78) | 174(81) | |
| PR | 0.006 | |||
| − | 195(34) | 38(27) | 50(23) | |
| + | 373(66) | 100 (73) | 167(77) | |
| HER-2 | 0.024 | |||
| 0/1+ | 412(72) | 102(73) | 178(83) | |
| 2+ | 103(18) | 27(19) | 28(13) | |
| 3+ | 56(10) | 11(8) | 9(4) | |
| Local regional recurrence/Distant metastasis | 123(27) | 26(22) | 43(25) | 0.467 |
| Death of tumor | 62 (14) | 9(10) | 22(24) | 0.236 |
PSIMPC precursor state of invasive micropapillary carcinoma, LN lymph node, LVI lymphovascular invasion, ER estrogen receptor, PR progesterone receptor, HER-2 human epidermal growth factor receptor 2,
3.3. Correlation of proportion in PSIMPC with clinicopathologic characteristics
According to the proportion of PSIMPC in tumors, we divide it into four categories: ≤25%,26-50%, 51-75%, 76-100% (Table 3). We found no significant correlation between the proportion of PSIMPC and age, tumor size, histological grade, LVI and LN involvement. High PSIMPC proportion was significantly associated with negative ER and PR expression (P<0.001). The proportion of PSIMPC showed a statistically highly significant correlation with HER-2 status based on immunohistochemistry, HER-2 expression is higher in tumors with high PSIMPC proportion than in tumors with low PSIMPC proportion (P<0.001).
Table 3.
The relationship between the proportion of PSIMPC and patient clinicopathological characteristics
| Characteristics | PSIMPC proportion(n,%) | P | |||
|---|---|---|---|---|---|
| ⩽25% | 26-50% | 51-75% | 76-100% | ||
| Age at diagnosis | 0.216 | ||||
| ≤50 | 324 (50.0) | 73(43.7) | 37 (40.7) | 19(44.2) | |
| > 50 | 324 (50.0) | 94(56.3) | 54 59.3) | 24(55.8) | |
| Tumor size ( mm) | 0.668 | ||||
| ≤20 | 214 (33.5) | 57(34.3) | 30(33.3) | 9(21.4) | |
| 20-50 | 382(59.8) | 95(57.2) | 54 (60.0) | 31(73.8) | |
| > 50 | 43(6.7) | 14(8.4) | 6(6.7) | 2(4.8) | |
| LN | 0.911 | ||||
| − | 269 (41.5) | 74(44.3) | 38(41.8) | 17(39.5) | |
| + | 379 (58.5) | 93(55.7) | 53 (58.2) | 26(60.5) | |
| LVI | 0.303 | ||||
| − | 408(63.1) | 92(55.1) | 57 (62.6) | 27(62.8) | |
| + | 239(36.9 | 75(44.9) | 34(37.4) | 16(37.2) | |
| TNM stage | 0.602 | ||||
| 1 | 122(19.1) | 34(20.6) | 12(13.2) | 5(11.9) | |
| 2 | 314(49.2) | 70(42.4) | 49 (53.8) | 22(52.4) | |
| 3/4 | 202(31.7) | 61(37.0) | 30 (33.0) | 15(35.7) | |
| Histologic Grade | 0.756 | ||||
| I | 41 (6.8) | 7(4.5) | 6(7.1) | 2(5.3) | |
| II | 471(77.7) | 127(82.5) | 63(74.1) | 33(86.8) | |
| III | 94 (15.5) | 20(13.0) | 16 (18.8) | 3(7.9) | |
| ER | <0.001 | ||||
| − | 141(22.4 | 43(26.1) | 37(41.6) | 18(43.9) | |
| + | 488(77.6) | 122(73.9) | 52(58.4) | 23(56.1) | |
| PR | <0.001 | ||||
| − | 168(26.6) | 56(33.7) | 39(43.8) | 22(52.4) | |
| + | 463(73.4) | 110(66.3) | 50(56.2) | 20(47.6) | |
| HER-2 | <0.001 | ||||
| 0/1+ | 494(78.0) | 125(75.3) | 55(60.4) | 22(53.7) | |
| 2+ | 94(14.8) | 33(19.9) | 20(2.0) | 12(29.3) | |
| 3+ | 45(7.1) | 8(4.8) | 16(17.6) | 7(17.1) | |
| Local regional recurrence/Distant metastasis | 132(25.8) | 35(28.2) | 19(25.7) | 7(18.9) | 0.731 |
| Death of tumor | 57(11.0) | 17(13.5) | 15(20.3) | 4(10.8) | 0.143 |
PSIMPC precursor state of invasive micropapillary carcinoma, LN lymph node, LVI lymphovascular invasion, ER estrogen receptor, PR progesterone receptor, HER-2 human epidermal growth factor receptor 2,
The rates of regional recurrence, distant metastasis and death of disease were not significantly associated with proportion of PSIMPC (P=0.731 and P=0.143).
3.4. Survival Analysis
The median follow-up was 50 months (range1-180 months) in PSIMPC, 48 months (range 1-157 months) in patients with IMPC and 53 months (range 1-160 months) in IBC-NST. The 5-year DFS and OS for patients with PSIMPC were 76.8% and 87.6%, compared to 63.2% and 85.9% in IMPC (P=0.001 for DFS and P=0.008 for OS), 86.2% and 93.7% in IBC-NST (P<0.001 for DFS and P<0.001 for OS). Patients with PSIMPC demonstrated a more favorable DFS and OS compared to IMPC, but worse DFS and OS compared to IBC-NST (Figure 3A&3B).
Figure 3.
Patients with PSIMPC demonstrated a more favorable DFS and OS compared to IMPC, but unfavorable disease-free survival and disease-free survival compared to IBC-NST without these features (P<0.01 and P<0.01). Kaplan-Meier estimates disease-free survival and overall survival, and compares four different proportions of PSIMPC. No significant difference was found in different proportions of PSIMPC (P=0.645 and P=0.154) (Figure 3C&3D).
On Cox regression multivariate analyses, tumor size (P=0.001), LN metastasis (P<0.001), LVI (P=0.001) and PSIMPC (P=0.029) were independent factors for disease-free survival (Table 4). For overall survival, tumor size (P=0.003), histological grade (P=0.001), LN metastasis (P<0.001), the presence of LVI (P<0.001), and the presence of PSIMPC (P=0.016) were independent factors (Table 4).
Table 4.
Univariate and multivariate analyses of clinicopathologic features for prognosis of PSIMPC and IBC-NST
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| HR | 95 % CI | P | HR | 95 % CI | P | |
| DFS | ||||||
| Age at diagnosis | 1.051 | 0.859-1.25 | 0.631 | / | / | / |
| Tumor size | 1.593 | 1.399-1.895 | <0.001 | 1.366 | 1.141-1.635 | 0.001 |
| Histologic Grade | 1.194 | 0.950-1.501 | 0.129 | / | / | / |
| LN | 2.752 | 2.221-3.428 | <0.001 | 2.286 | 1.809-2.888 | <0.001 |
| LVI | 2.009 | 1.636-2.466 | <0.001 | 1.428 | 1.146-1.778 | 0.001 |
| ER | 0.973 | 0.766-1.238 | 0.827 | / | / | / |
| PR | 0.829 | 0.662-1.036 | 0.100 | / | / | / |
| HER-2 | 1.068 | 0.896-1.274 | 0.463 | / | / | / |
| PSIMPC | 1.458 | 1.193-1.782 | <0.001 | 1.263 | 1.024-1.557 | 0.029 |
| OS | ||||||
| Age at diagnosis | 1.353 | 0.981-1.867 | 0.065 | / | / | / |
| Tumor size | 2.407 | 1.837-30159 | <0.001 | 1.584 | 1.169-2.147 | 0.003 |
| Histologic Grade | 2.248 | 1.582-3.193 | <0.001 | 1.878 | 1.285-2.745 | 0.001 |
| LN | 4.675 | 3.099-7.053 | <0.001 | 3.165 | 2.002-5.002 | <0.001 |
| LVI | 3.022 | 2.190-4.169 | <0.001 | 1.983 | 1.375-2.860 | <0.001 |
| ER | 0.578 | 0.409-0.817 | 0.002 | 0.764 | 0.464-1.257 | 0.289 |
| PR | 0.626 | 0.446-0.880 | 0.007 | 0.843 | 0.526-1.351 | 0.478 |
| HER-2 | 1.318 | 1.020-1.702 | 0.035 | 1.052 | 0.790-1.400 | 0.730 |
| PSIMPC | 2.194 | 1.577-3.053 | <0.001 | 1.577 | 1.089-2.284 | 0.016 |
PSIMPC precursor state of invasive micropapillary carcinoma, LN lymph node, LVI lymphovascular invasion, ER estrogen receptor, PR progesterone receptor, HER-2 human epidermal growth factor receptor 2,
Among patients with PSIMPC, using Cox regression multivariate analyses, we found that tumor size (P<0.001 and P<0.001 respectively), LVI (P<0.001 and P=0.004 respectively), LN metastasis (P=0.006 and P<0.01 respectively) were significantly associated with disease-free survival and overall survival (Table 5).
Table 5.
Univariate and multivariate analyses of clinicopathologic features for prognosis of PSIMPC
| Univariate | Multivariate | |||||
|---|---|---|---|---|---|---|
| HR | 95 % CI | P | HR | 95 % CI | P | |
| DFS | ||||||
| Age at diagnosis | 1.059 | 0.825-1.453 | 0.529 | / | / | / |
| Tumor size | 1.904 | 1.485-2.440 | <0.001 | 1.604 | 1.237-2.080 | <0.001 |
| Histologic Grade | 1.123 | 0.809-1.558 | 0.488 | / | / | / |
| LN | 2.687 | 1.929-3.743 | <0.001 | 2.238 | 1.587-3.156 | <0.001 |
| LVI | 1.787 | 1.347-2.371 | <0.001 | 1.493 | 1.120-1.991 | 0.006 |
| ER | 0.803 | 0.581-1.110 | 0.184 | / | / | / |
| PR | 0.746 | 0.552-1.009 | 0.057 | / | / | / |
| HER-2 | 1.053 | 0.838-1.323 | 0.658 | / | / | / |
| OS | ||||||
| Age at diagnosis | 1.578 | 1.043-2.388 | 0.031 | 1.637 | 1.052-2.549 | 0.029 |
| Tumor size | 2.991 | 2.109-4.241 | <0.001 | 2.272 | 1.548-3.336 | 0.000 |
| Histologic Grade | 1.859 | 1.186-2.915 | 0.007 | 1.577 | 0.987-2.520 | 0.057 |
| LN | 4.733 | 2.631-8.513 | <0.001 | 2.847 | 1.555-5.214 | 0.001 |
| LVI | 2.517 | 1.652-3.835 | <0.001 | 1.950 | 1.237-3.075 | 0.004 |
| ER | 0.479 | 0.311-0.738 | 0.001 | 0.791 | 0.418-1.495 | 0.470 |
| PR | 0.544 | 0.358-0.829 | 0.005 | 0.842 | 0.461-1.537 | 0.575 |
| HER-2 | 1.190 | 0.865-1.639 | 0.285 | / | / | / |
PSIMPC precursor state of invasive micropapillary carcinoma, LN lymph node, LVI lymphovascular invasion, ER estrogen receptor, PR progesterone receptor, HER-2 human epidermal growth factor receptor 2,
Kaplan Meier was used to estimate disease-free survival and overall survival. No significant difference in disease-free survival and overall survival was found among different proportions of PSIMPC (P=0.645 and P=0.154) (Figure 3C&3D) and different distribution of PSIMPC (P=0.279 and P=0.179).
4. Discussion
In this study, we found that the presence of retraction clefts in invasive breast carcinoma, without characteristic IMPC features, even as a small proportion, was significantly associated with lymphovascular invasion and lymph node metastasis, features similar to that of IMPC. Importantly, we found a significant association between the distribution of retraction clefts and LVI and LN metastasis. Retraction clefts located in both peripheral and internal area of the tumor demonstrated more aggressive behavior as measured by LN metastasis and LVI, compared to peripheral only and internal only groups. The proportion of the retraction clefts, however, was not significantly associated with LVI, LN metastasis, disease free survival and overall survival.
The mechanism of retraction clefts is not fully understood. Some studies suggest that this phenomenon is artifactual due to formalin fixation and tissue processing. However, retraction clefts in invasive breast carcinoma also exist on frozen sections [6], indicating that they are real spaces, not just caused by formalin fixation.
The presence of a wide range of retraction clefts in prostate cancer was correlated with tumor characteristics that suggest the tumor is more aggressive and a shorter recurrence free survival [17]. The micropapillary morphology in invasive urothelial carcinoma can predict the presence of infiltration. Even if there is no typical micropapillary morphology, significant retraction in invasive urothelial carcinoma may be associated with more aggressive disease [18].
The presence of retraction clefts in breast cancer may be a morphological reflection of changes in tumor-stromal interactions leading to tumor progression and lymphatic spread. Previous studies have shown that the presence of retraction clefts is associated with LVI and predicts lymph node metastasis and poor prognosis in early stage breast cancer [19]. Essentially all inflammatory breast cancer patients present with lymph node involvement and one-third of patients already have distant metastasis at initial diagnosis [20]. The author found that a tumor extrinsic factor, notably tumor-associated macrophages, promotes and contributes to inflammatory breast cancer’s extreme metastatic phenotype. Retraction clefts and tumor-stromal interactions may lead to lymphatic spread in breast cancer, but the mechanism need further elucidated.
In the current study, we found that the presence of retraction clefts in breast cancer was associated with the presence of LVI and LN metastasis, consistent with previous studies [21]. In addition, we found that the presence of retraction clefts was negatively associated with ER and PR expression.
We propose that invasive breast carcinoma with retraction clefts may represent the precursor state of IMPC (PSIMPC). In addition, retraction clefts correlated with lymphovascular invasion. For tumor with the retraction clefts, it is possible to may benefit from angiogenesis inhibitors. PSIMPC was an independent predictor in this cohort, significantly associated with worse disease-free survival and overall survival than IBC-NST, but better outcomes comparing to IMPC.
In conclusion, we revealed that the presence of retraction clefts in breast carcinomas was significantly correlated with LVI, LN metastases, and poor outcome. Retraction clefts may be part of the early stage of lymphovascular invasion of breast cancer cells. We recommend the presence of retraction clefts should be mentioned in the pathology report to guide precise treatment.
Highlights.
Some invasive breast carcinomas are similar to invasive micropapillary carcinoma (IMPC).
We consider the presence of the retraction clefts a common phenomenon that may present as a precursor stage of IMPC (PSIMPC).
PSIMPC demonstrated a more favorable DFS and OS compared to IMPC, but worse DFS and OS compared to IBC-NST.
Our findings suggest that the presence of retraction clefts is a precursor state of IMPC, exhibiting IMPC-like features and more aggressive clinical behavior.
Funding
This study was funded by National Natural Science Foundation of China (grant number: 81872164, 81672637, 81602340), Program for Chang Jiang Scholars and Innovative Research Team in University of Ministry of Education of China (grant number: IRT0743). Hannah Y Wen was supported in part through the NIH/NCI Cancer Center Support Grant (grant number: P30 CA008748).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflict of Interest The authors declare that they have no conflict of interest.
References
- 1.Yang YL, Liu BB, Zhang X, Fu L. Invasive Micropapillary Carcinoma of the Breast: An Update. Arch Pathol Lab Med 2016;140:799–805. [DOI] [PubMed] [Google Scholar]
- 2.Chen AC, Paulino AC, Schwartz MR, et al. Prognostic markers for invasive micropapillary carcinoma of the breast: a population-based analysis. Clin breast cancer 2013;13:133–139. [DOI] [PubMed] [Google Scholar]
- 3.Siriaunkgul S, Tavassoli FA. Invasive micropapillary carcinoma of the breast. Modern Pathol 1993; 6:660–662. [PubMed] [Google Scholar]
- 4.Yu JI, Choi DH, Huh SJ, et al. Differences in Prognostic Factors and Failure Patterns Between Invasive Micropapillary Carcinoma and Carcinoma With Micropapillary Component Versus Invasive Ductal Carcinoma of the Breast: Retrospective Multicenter Case-Control Study (KROG 13-06). Clin Breast Cancer 2015;15:353–361.e612. [DOI] [PubMed] [Google Scholar]
- 5.Yoon GY, Cha JH, Kim HH, et al. Comparison of invasive micropapillary and invasive ductal carcinoma of the breast: a matched cohort study. Acta Radiol 2019; 60:1405–1413. [DOI] [PubMed] [Google Scholar]
- 6.Acs G, Paragh G, Chuang ST, Laronga C, Zhang PJ.The presence of micropapillary features and retraction artifact in core needle biopsy material predicts lymph node metastasis in breast carcinoma. Am J Surg Pathol 2009; 33:202–210. [DOI] [PubMed] [Google Scholar]
- 7.Acs G, Esposito NN, Rakosy Z, Laronga C, Zhang PJ. Invasive ductal carcinomas of the breast showing partial reversed cell polarity are associated with lymphatic tumor spread and may represent part of a spectrum of invasive micropapillary carcinoma. Am J SurgPathol 2010;34:1637–1646. [DOI] [PubMed] [Google Scholar]
- 8.Kruslin B, Tomas D, Rogatsch H, et al. Periacinar retraction clefting in the prostatic needle core biopsies: an important diagnostic criterion or a simple artifact? Virchows Archiv 2003;443:524–527. [DOI] [PubMed] [Google Scholar]
- 9.Kruslin B, Tomas D, Rogatsch H, et al. Correlation of periacinar retraction clefting in needle core biopsies and corresponding prostatectomy specimens of patients with prostatic adenocarcinoma. Int J Surg Pathol 2005; 13:67–72. [DOI] [PubMed] [Google Scholar]
- 10.Jain D, Tikku G, Bhadana P, Dravid C, Grover RK. The Impact of Peritumoral Retraction Clefting & Intratumoral Eosinophils on Overall Survival in Oral Squamous Carcinoma Patients. Pathol Oncol Res 2019;25:183–189. [DOI] [PubMed] [Google Scholar]
- 11.Irie J, Manucha V, Ioffe OB, Silverberg SG. Artefact as the pathologist's friend: peritumoral retraction in in situ and infiltrating duct carcinoma of the breast. Int J Surg Pathol 2007;15:53–59. [DOI] [PubMed] [Google Scholar]
- 12.Lakhani SR, Ellis O, Schnitt SJ, et al. WHO classification of tumours of the breast tumour. 2019; Lyon: International Agency for Research on Cancer. [Google Scholar]
- 13.Ulamec M, Džombeta T, Čupić H, et al. Periacinar Retraction Clefting and D2-40 Expression in Prostatic Adenocarcinoma. Pathol Oncol Res 2011;18:365–370. [DOI] [PubMed] [Google Scholar]
- 14.Zaorsky NG, Patil N, Freedman GM, Tuluc M . Differentiating lymphovascular invasion from retraction artifact on histological specimen of breast carcinoma and their implications on prognosis. J Breast Cancer 2012; 15:478–480. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Allison KH, Hammond MEH, Dowsett M, McKernin SE, Carey LA, Fitzgibbons PL, Hayes DF, Lakhani SR, Chavez-MacGregor M, Perlmutter J, Perou CM, Regan MM, Rimm DL, Symmans WF, Torlakovic EE, Varella L, Viale G, Weisberg TF, McShane LM, Wolff AC. Estrogen and Progesterone Receptor Testing in Breast Cancer: ASCO/CAP Guideline Update. J Clin Oncol 2020;38:1346–1366. [DOI] [PubMed] [Google Scholar]
- 16.Wolff AC, Hammond MEH, Allison KH, Harvey BE, Mangu PB, Bartlett JMS, Bilous M, Ellis IO, Fitzgibbons P, Hanna W, Jenkins RB, Press MF, Spears PA, Vance GH, Viale G, McShane LM, Dowsett M. Human Epidermal Growth Factor Receptor 2 Testing in Breast Cancer: American Society of Clinical Oncology/College of American Pathologists Clinical Practice Guideline Focused Update. J Clin Oncol 2018;36:2105–2122. [DOI] [PubMed] [Google Scholar]
- 17.Tomas D, Spajić B, Milošević M, Demirović A, Marušić Z, Krušlin B. Extensive retraction artefact predicts biochemical recurrence-free survival in prostatic carcinoma. Histopathology 2011; 58:447–454. [DOI] [PubMed] [Google Scholar]
- 18.Shah TS, Kaag M, Raman JD, et al. Clinical significance of prominent retraction clefts in invasive urothelial carcinoma. Hum Pathol 2017; 61:90–96. [DOI] [PubMed] [Google Scholar]
- 19.Acs G, Paragh G, Rakosy Z, Laronga C, Zhang PJ .The extent of retraction clefts correlates with lymphatic vessel density and VEGF-C expression and predicts nodal metastasis and poor prognosis in early-stage breast carcinoma. Modern Pathol 2012;25:163–177. [DOI] [PubMed] [Google Scholar]
- 20.Allen SG, Chen YC, Madden JM, Fournier CL, Altemus MA, Hiziroglu AB, Cheng YH, Wu ZF, Bao L, Yates JA, Yoon E, Merajver SD. Macrophages Enhance Migration in Inflammatory Breast Cancer Cells via RhoC GTPase Signaling. Sci Rep 2016;6:39190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Acs G, Khakpour N, Kiluk J, Lee MC, Laronga C.The presence of extensive retraction clefts in invasive breast carcinomas correlates with lymphatic invasion and nodal metastasis and predicts poor outcome: a prospective validation study of 2742 consecutive cases. Am J Surg Pathol 2015; 39:325–337. [DOI] [PubMed] [Google Scholar]