Micropapillary Breast Carcinoma: From Molecular Pathogenesis to Prognosis.

Invasive micropapillary carcinoma (IMPC) of the breast is an infrequent type of breast cancer often discussed for its potency for lymphovascular invasion and difficulty in accurate imaging estimation. Micropapillary carcinomas are noted to be present as larger tumors, of higher histological grade and a notably higher percentage of disease-positive lymph nodes. Hormonal and HER-2 positivity in IMPC is also commoner when compared to other NST carcinomas. IMPC occurs either as a pure form or more often as a component of mixed Non-Specific Type (NST) carcinoma. The latest data suggest that despite having comparable survival rates to other histological subtypes of breast carcinoma, effective surgical treatment often requires extended surgical margins and vigilant preoperative axillary staging due to an increased incidence of lymph node invasion, and locoregional recurrence. Moreover, the presence of micropapillary in situ components within tumors also seems to alter tumor aggression and influence the nodal disease stage. In this review, we present an overview of the current literature of micropapillary carcinoma of the breast from biology to prognosis, focusing on biological differences and treatment.

Introduction

Invasive micropapillary carcinoma (referred to as IMPC) is a rare, distinct histological subtype of breast carcinoma. First described as an entity by Fisher et al in 1980,1 it was not until 1993 that the term and classification was introduced by Siriaunkgul et al.2 While micropapillary histological architecture is found in 2–8% of all breast cancers, pure micropapillary carcinoma is infrequent and comprises 0.9–2% of breast carcinomas.3 Mean age of diagnosis is 50–60 years, and it is predominantly found in females, with only a few cases for male IMPC reported.4–10 This review aims to provide an overview of the effect of micropapillary histology on lymph node invasion, LVI, and prognosis. Also, the effect of micropapillary component within non-pure IMPC is discussed, and any recorded differences regarding IMPC treatment compared to other histological subtypes are considered.

There is a distinct pathological morphology of IMPC, consisting of hollow cell clusters with granular or eosinophilic cytoplasm,11 arranged in a pseudopapillary manner, devoid of fibrovascular cores and laid out in an “inside-out” manner, with the luminal cellular surface being the outermost.1,12–18 This arrangement is best presented when MUC1/EMA staining is used, so much so that “reversed” staining of these markers is considered a hallmark of IMPC, shared only by mucinous histology.19–21 The distinctive histological features of pure micropapillary carcinoma can be seen in Figure 1A–D, as taken from one of our cases.161

Figure 1

(A–D) In low magnification, through an atrophic mammary gland a neoplastic population is recognized, infiltrating the remaining ducts (A). The cells are organized in clusters, forming small-sized glandular structures and nests, arranged in a micropapillary pattern (B). Occasionally, a small proportion of them acquire a central lumina. Fibrovascular cores are absent. (C) The neoplastic cells have a moderate amount of eosinophilic cytoplasm and small round nuclei with condensed chromatin and intermediate pleomorphism. (D). In another slide of this lesion, a lymphovascular emboli is recognized (D). The morphology is highly suggestive of invasive micropapillary carcinoma, so immunohistochemical markers are performed to establish the diagnosis.

IMPC is emerging as an oncological and surgical challenge, due to a plethora of characteristics that constitute this histological pattern, interestingly, both elusive and aggressive. Namely, its tendency to present as a palpable mass, often of increased size and higher grade compared to the invasive ductal carcinoma (IDC), currently the most diagnosed type of breast cancer. Another especially troublesome aspect of IMPC is the comparatively increased incidence of lymphovascular invasion (LVI) characterized by both carcinomatous emboli,22,23 and clinically positive axillary lymph nodes,23 which naturally alters the surgical and adjuvant treatment regiments to more aggressive ones, with comparative prognosis still being a point of ongoing debate.5,24–28

Review Methodology

Current literature search on micropapillary carcinoma was performed using the PubMed, SCOPUS and Cochrane Library databases. Studies in the fields of Medicine, Biology, Molecular Biology and Genetics were included. Each report was screened independently for relevance, and the Mendeley referencing tool was used for duplicate detection. Keywords used included “micropapillary breast carcinoma”, “micropapillary DCIS”, “micropapillary cancer” “invasive micropapillary breast carcinoma”. The selection process (carried out under the latest PRISMA guidelines for reporting29), can be seen in Figure 2. A total of 155 reports were included in the review: 117 original articles, 9 review articles, 24 case reports, 2 meta-analyses, and 3 opinion letters/editorials.

Figure 2

Report selection flowchart.

Lymphovascular Invasion and Lymph Node Involvement

We have collected results from several published studies with variable sample sizes and characteristics. A brief summary of study findings on tumor size, lymph node involvement, and LVI presence can be seen in Table 1. One of the most studied respects of IMPC thus far is the seemingly increased frequency of lymphovascular invasion and lymph node (often clinically evident) involvement.13,15,22,30–40 A recent study by Lewis et al,41 published in 2019, used a sample of 2660 patients diagnosed with pure IMPC, one of the largest case series to date. The study demonstrated confirmed regional lymph node metastasis in 55.2% of the patients at the time of diagnosis, with other researchers such as Gokce et al reporting percentages up to 79.6%.33 Risk factors associated with nodal involvement in IMPC include tumor size, ER negativity, and advanced age.42

Table 1

Data on Tumor Size, Tumor Grade, Nodal Status and LVI from Included Studies

Study	Study Type	Tumor Size in IMPC Patients	Histological Tumor Grade in IMPC Patients	Nodal Status of IMPC Patients	LVI Status of IMPC Patients	Comparison with Other Histological Subtypes
Stranix et al (2015)4	Literature Review	–	–	LN positivity in 71.2% (1267/1280 patients)	LVI observed in 73.7% (638/866 patients)	–
Vingiani et al (2013)6	Case–Control Study	T3-T4 in 12.2% of the patients	Grade III in 32.7% of the patients	LN positivity in 69.4% of the patients	LVI observed in 61.2% of the patients	Tumor size, nodal positivity and LVI rates was higher in IMPC patients compared to IDC patients (5.3%, 47.3% and 61.2%, respectively)
Tang et al (2017)13	Case–Control Study	T3-T4 in 9.2% of the patients	–	LN positivity in 64.9% of the patients	LVI observed in 14.7% of the patients	Nodal positivity and LVI differed significantly in IMPC patients compared to IDC patients (46.8% and 0.1%, respectively)No difference in tumor size
Cui et al (2014)14	Case Series Study	–	–	LN positivity in 80% of the patients	LVI observed in 44% of the patients	–
Hashmi et al (2018)15	Case–Control Study	T3 in 11.1% of the patients	Grade III in 26.7% of the patients	LN positivity in 55.6% of the patients	LVI observed in 77.8% of the patients	LVI differed significantly in IMPC patients compared to IDC patients (24.8%) No difference in tumor size, tumor grade or nodal invasion
Pettinato et al (2002)16	Case Series Study	–	–	LN positivity in 90% of the patients	LVI in 72% of the patients	–
Chen et al (2017)23	Case-Control Study	T3-T4 in 10.2% of the patients	Grade III in 22.6% of the patients	LN positivity in 51.3% of the patients	–	No observed difference in tumor size, grade or LN positivity when compared with IMPC patients
Yu et al (2015)28	Case–Control Study	T3 in 5.2% of the patients	Grade III in 40.1% of the patients	LN positivity in 69.3% of the patients	LVI in 61.8% of the patients	LVI differed significantly in IMPC patients compared to IDC patients (43.4%) No difference in tumor size, tumor grade or nodal invasion
Zekioglou et al (2004)30	Case–Control Study	–	Grade III in 82% of the patients	LN positivity in 69% of the patients	LVI in 75.5% of the patients	Tumor grade, nodal positivity and LVI differed were significantly higher in IMPC compared to IDC patients.
Chen et al (2014)32	Case–Control Study	T3-T4 in 10% of the patients	Grade III in 40% of the patients	LN positivity in in 52% of the patients	–	Tumor size, tumor grade and LN positivity were significantly higher in IMPC compared to IDC patients.
Gokce et al (2013)33	Case-Control Study	–	Grade III in 40.8% of the patients	LN positivity in 59.3% of the patients	LVI in 94.7% of the patients	Nodal positivity and LVI were seen significantly more frequently in IMPC patients than IDC patients
Akdeniz et al (2020)38	Case Series Study	T3-T4 in 33.3% of the patients	–	LN positivity in 79.2% of the patients	–	–
Lewis et al (2019)41	Case Series Study	–	Grade III in 34.7% of the patients	–	–	–
Ye et al (2018)42	Case Series Study	–	Grade III in 37.17% of the patients	LN positivity in 50.46% of the patients	–	–
Paterakos et al (1999)43	Case-Control Study	–	–	LN positivity in 94% of the patients	–	Nodal positivity and number of infiltrated lymph nodes were higher in IMPC patients compared to IDC patients.
Hao et al (2019)44	Case–Control Study	T3-T4 in 59% of the patients	–	LN positivity in 69.1% of the patients	LVI in 71.3% of the patients	No difference in nodal positivity and LVI after matching, for IMPC patients and IDC patients
De La Cruz et al (2004)46	Case–Control Study	–	Grade III in 81.3% of the patients	LN positivity in 92.9% of the patients	–	Higher grade tumors and LN positivity were higher in IMPC compared to IDC patients
Chen et al (2013)47	Case Series Study	T3-T4 in 12% of the patients	Grade III in 38% of the patients	LN positivity in 53% of the patients	–	–
Kaya et al (2018)48	Case Series Study	T3-T4 in 5.3% of the patients	Grade III in 42.1% of the patients	LN positivity in 68% of the patients	LVI in 84.2% of the patients	Nodal positivity and LVI were higher in tumors with >75% micropapillary component
Walsh et al (2001)52	Case Series Study	T3 in 7.5% of the patients	Grade III in 67.5% of the patients	LN positivity in 72.3% of the patients	LVI in 62.5% of the patients	–
Kim et al (2020)58	Case–Control Study	T3-T4 in 11% of the patients	Grade III in 34% of the patients	LN positivity in 67.4% of the patients	LVI in 67.2% of the patients	Tumor size, nodal positivity, LVI and grade were higher in IMPC patients compared to IDC patients
Lee et al (2011)98	Case Series Study	T3-T4 in 43% of the patients	Grade III in 55% of the patients	–	–	–
Guan et al (2020)101	Case–Control Study	T3-T4 in 1.5% of the patients	–	LN positivity in 60.8% of the patients	LVI in 78.9% of the patients	Nodal positivity and LVI were significantly higher in tumors with IMPC component, compared to DCIS component
Kim et al (2005)102	Case–Control Study	T3-T4 in 18.4% of the patients	Grade III in 44.7% of the patients	LN positivity in 78.9% of the patients	LVI in 60.5% of the patients
Collins et al (2017)108	Case Series Study	T3-T4 in 21% of the patients	Grade III in 21% of the patients	LN positivity in 42.8% of the patients	LVI in 21.4% of the patients	–
Yoon et al (2019)124	Case–Control Study	–	Grade III in 37.7% of the patients	LN positivity in 63.6% of the patients	LVI in 52% of the patients	After propensity matching, nodal status, Histological grade, and LVI rates did not differ between IMPC and IDC patients
Kim et al (2010)128	Case–Control Study	T3 in 8.3% of the patients	Grade III in 41% of the patients	LN positivity in 70.5% of the patients	LVI in 75.4% of the patients	Nodal positivity, histological grade and LVI were significantly higher in tumors with IMPC component, compared to DCIS component
Chen et al (2018)143	Case–Control Study	T3 in 22.1% of the patients	–	LN positivity in 72.6% of the patients	LVI in 51.65 of the patients	Tumor size, nodal positivity and LVI rates were higher in IMPC patients compared to TN-IDC patients
Li et al (2019)144	Case–Control Study	T3-T4 in 11.79% of the patients	Grade III in 62.71% of the patients	LN positivity in 51.5% of the patients	–	Tumor size and nodal positivity rates were higher compared to IDC patients
Li et al (2016)147	Case–Control Study	T3-T4 in 24.2% of the patients	–	LN positivity in 79.8% of the patients	LVI in 18.2% of the patients	Nodal positivity and LVI rates were higher in IMPC than IDC patients
Lewis et al (2019)148	Case Series Study	T3-T4 in 8% of the patients	Grade III in 36.5% of the patients	LN positivity in 53.3% of the patients	–	–
Liu et al (2014)149	Case–Control Study	T3 in 5.88% of the patients	Grade III in 49.02% of the patients	LN positivity in 69.6% of the patients	LVI in 52.94% of the patients	LVI rates were higher compared to IDC patients
Liu et al (2015)151	Case–Control Study	–	Grade III in 16.4% of the patients	LN positivity in 80.8% of the patients	LVI in 82.9% of the patients	Histological grade, nodal positivity, and LVI rates were higher compared to the mucinous carcinoma group
Kuroda et al (2004)153	Case Series Study	T3-T4 in 33.3% of the patients	–	LN positivity in 66.6% of the patients	LVI in 88.8% of the patients	–
Shi et al (2014)154	Case–Control Study	T3-T4 in 9.6% of the patients	–	LN positivity in 73.4% of the patients	LVI in 75.4% of the patients	Tumor size, nodal positivity and LVI rates were higher in IMPC patients compared to IDC patients
Meng et al (2021)155	Case Series Study	T3-T4 in 6.96% of the patients	Grade III in 14.95% of the patients	LN positivity in 30.4% of the patients	LVI in 42.27% of the patients	–

To put things in perspective, a comparison between IMPC and Invasive Ductal Carcinoma (IDC) is often deemed appropriate, since IDC is undoubtedly the most studied type of breast carcinoma. A comparative study by Hashimi et al15 showed that only 49.5% of the patients with IDC had any nodal involvement, and in fact N3 stage occurred in only 15.6% of the patients, as opposed to 33% in the IMPC group. Lymphovascular involvement has also been found to be more common among IMPC patients, as shown in a study by Tang et al,13 with 14.7% versus only 0.1% in the IDC group, and a staggering 94.7% being reported by Gokce et al.33 Both points are of great surgical significance, since radiologically, clinically or biopsy-proven positive lymph nodes have been an indication for more extensive surgery and axillary dissection.13,32 It is indicative that Tang et al reported selection of partial mastectomy in 7.4% of the IDC group, as opposed to 3.0% of the IMPC group.13 A previous study by Paterakos et al43 showcased not only lymphovascular involvement in 95% of the patients but also a relation with higher-grade tumors at presentation and higher scores on the mitotic index.

Tumor size at diagnosis has also been a much-discussed issue regarding IMPC. Hao et al compared the percentage of tumors larger than 5cm at the time of diagnosis, reporting 4.3% in IMPC and 3% in IDC.44 Ye et al demonstrated that IMPC presented at a higher stage tumor at diagnosis also attributed to a larger size, in a meta-analysis.45 It is worth noting that the reported difference in mean tumor size can be attributed to the rapid growth patterns of IMPC, as well as its insidious presentation, leading to larger tumors being diagnosed more often.38,46 However, more basic research on the underlying molecular biology of IMPC is needed. Another point of concern is the lack of specific guidelines regarding the percentage of micropapillary element required to report a tumor as partially or purely micropapillary. This leads to a lack of systematic sample classification and comparison.13,32,47,48

Pathology – HR and HER2

Molecular testing has provided an insight on the correlations of the hormonal status and clinical presentation, treatment, and prognosis of IMPC patients. Authors report higher percentages of estrogen receptor (ER) and progesterone receptor (PgR) positive tumors when comparing IMPC with IDC.1,3,14,49–52 Collected data on the hormonal status of IMPC tumors, and relevant comparisons from included studies can be found in Table 2. Positive ER staining has been commented upon as positively associated with survival duration in a large series of IMPC patients.49,52,53 A large study by Cui et al14 reported 88% ER positivity and 64% PgR positivity when studying IMPC specimens. A study conducted by Lewis et al, including 865 cases, has reported that the IMPC tumors are characterized as Luminal A in 75.3% of the instances, Luminal B in 14.8%, HER2-enriched in 4.7%, and Triple Negative in 5.2%.41 However, most studies have found that micropapillary carcinomas tend to be in the Luminal B category when genomic sequencing is used instead of staining alone.54–56 While the incidence of the triple-negative classification seems to be lower in IMPC, it is associated with higher-grade tumors, higher disease stage at diagnosis, and an increase in total mastectomies performed.7,15,44,57–59

Table 2

Data on Hormone Receptor and HER-2 Status of IMPC Patients from Included Studies

Study	Study Type	HR Status of IMPC Tumors	PR Status of IMPC Tumors	HER-2 Status of IMPC Tumors	Comparison with Other Histological Subtypes
Stranix et al (2015)4	Literature Review	Positive in 73.4% of the patients	Positive in 62.5% of the patients	Positive in 40.5% of the patients	–
Vingiani et al (2013)6	Case–Control Study	Positive in 87.8% of the patients	Positive in 69.4% of the patients	Positive in 18.4% of the patients	No observed differences compared to IDC patients.
Tang et al (2017)13	Case–Control Study	Positive in 83.5% of the patients	Positive in 78.2% of the patients	Positive in 34% of the patients	HR, PR and HER-2 positivity was observed more frequently in IMPC patients compared to IDC patients.
Cui et al (2014)14	Clinicopathological Study	Positive in 88% of the patients	Positive in 64% of the patients	Positive in 84% of the patients	–
Hashmi et al (2018)15	Case–Control Study	Positive in 86.7% of the patients	Positive in 73.3% of the patients	Positive in 60% of the patients	HR and PR positivity were seen more frequently in IMPC, compared to IDC patients
Pettinato et al (2002)16	Case Series Study	Positive in 36% of the patients	Positive in 27% of the patients	Positive in 72% of the patients	–
Yu et al (2015)28	Case–Control Study	Positive in 66.3% of the patients	Positive in 66.3% of the patients	Positive in 28.8% of the patients	HER2 positivity was observed more frequently in IMPC patients compared to IDC patients
Zekioglou et al (2004)30	Case–Control Study	Positive in 68% of the patients	Positive in 61% of the patients	Positive in 54% of the patients	HR and PR positivity were seen more frequently in IMPC compared to IDC patients
Chen et al (2014)32	Case–Control Study	Positive in 84.1% of the patients	Positive in 70.2% of the patients	–	HR and PR positivity were seen more frequently in IMPC compared to IDC patients
Gokce et al (2013)33	Case–Control Study	Positive in 70.3% of the patients	Positive in 77.3% of the patients	Positive in 52.5% of the patients	No observed differences compared to IDC patients
Akdeniz et al (2020)38	Case Series Study	Positive in 66.7% of the patients	Positive in 66.7% of the patients	Positive in 45.8% of the patients	–
Ye et al (2018)42	Case Series Study	Positive in 89.48% of the patients	Positive in 77.83% of the patients	Positive in 12.15% of the patients	–
Paterakos et al (1999)43	Case–Control Study	Positive in 61% of the patients	–	Positive in 77% of the patients	HER2 positivity was observed more frequently in IMPC patients compared to IDC patients
Hao et al (2019)44	Case–Control Study	Positive in 84.3% of the patients	Positive in 84.3% of the patients	Positive in 33% of the patients	No observed differences compared to IDC patients.
De La Cruz et al (2004)46	Case–Control Study	Positive in 50% of the patients	Positive in 31.2% of the patients	Positive in 50% of the patients	HR, PR and HER-2 positivity were seen more frequently in IMPC compared to IDC patients
Chen et al (2013)47	Case Series Study	Positive in 85% of the patients	Positive in 70% of the patients	–	–
Kuroda et al (2004)51	Case–Control Study	Positive in 70.3% of the patients	Positive in 55.5% of the patients	Positive in 25.9% of the patients	No observed differences compared to IDC patients
Walsh et al (2001)52	Case Series Study	Positive in 90.6% of the patients	Positive in 70.3% of the patients	–	–
Kim et al (2020)58	Case–Control Study	Positive in 75.8% of the patients	Positive in 63.2% of the patients	Positive in 33.3% of the patients	HR, PR and HER-2 positivity were seen more frequently in IMPC compared to IDC patients
Perron et al (2021)62	Case Series Study	Positive in 94% of the patients	Positive in 80.5% of the patients	Positive in 22.5% of the patients	–
Lee et al (2011)98	Case Series Study	Positive in 83% of the patients	Positive in 67% of the patients	Positive in 7% of the patients	–
Guan et al (2020)101	Case–Control Study	Positive in 82.3% of the patients	Positive in 56.2% of the patients	Positive in 30% of the patients	HR and HER2 positivity were seen more frequently in IMPC patients, compared to IDC patients. PR positivity was more frequent in IDC patients
Kim et al (2005)102	Case–Control Study	Positive in 19.4% of the patients	Positive in 19.4% of the patients	Positive in 38.9% of the patients	No observed differences compared to non-IMPC patients
Collins et al (2017)108	Case Series Study	Positive in 100% of the patients	Positive in 85.7% of the patients	Positive in 14.2% of the patients	–
Yoon et al (2019)124	Case–Control Study	Positive in 79.2% of the patients	Positive in 60.7% of the patients	Positive in 38% of the patients	After propensity score matching, HER-2 positivity was significantly higher in IMPC patients compared to IDC patients. No observed difference in ER or PR positivity
Kim et al (2010)140	Case–Control Study	Positive in 77% of the patients	Positive in 73.8% of the patients	Positive in 39.3% of the patients	No observed difference between IMPC and IDC patients
Chen et al (2018)143	Case–Control Study	Positive in 83.2% of the patients	Positive in 74.7% of the patients	Positive in 21.1% of the patients	–
Li et al (2019)144	Case–Control Study	Positive in 88.69% of the patients	Positive in 78.75% of the patients	–	ER and PR positivity rates were higher in IMPC patients, compared to IDC patients
Li et al (2016)147	Case–Control Study	Positive in 81.8% of the patients	Positive in 75.8% of the patients	Positive in 18.8% of the patients	ER positivity rates were significantly higher compared to IDC patients
Lewis et al (2019)148	Case Series Study	Positive in 87.5% of the patients	Positive in 79.4% of the patients	Positive in 14.9% of the patients	–
Liu et al (2014)149	Case–Control Study	Positive in 84.31% of the patients	Positive in 72.5% of the patients	Positive in 15.69% of the patients	ER positivity rates were significantly higher compared to IDC patients
Liu et al (2015)151	Case–Control Study	Positive in 83.3% of the patients	Positive in 74% of the patients	Positive in 28.8% of the patients	HR, PR and HER-2 positivity were seen more frequently in tumors with micropapillary histology, compared to pure mucinous histology
Shi et al (2014)154	Case–Control Study	Positive in 85.1% of the patients	Positive in 78.2% of the patients	Positive in 29.9% of the patients	ER and PR positivity rates were higher in IMPC patients compared to IDC patients
Meng et al (2021)155	Case Series Study	Positive in 78.09% of the patients	Positive in 65.46% of the patients	Positive in 33.99% of the patients	–

Overall, in terms of surveillance, hormonal positivity and HER2-positive staining are reported to be higher in IMPC than IBC.55,60 However, no difference in survival rates is reported between HER2-positive and HER2-negative groups. According to the authors,14,28,41,49,61,62 this is largely attributed to the latest HER2 targeting biological therapeutic regimens added to systemic therapy. A noteworthy study, run by Perron et al, provided insight into the expression of HER2 in IMPC. In particular, it is suggested that due to the tumor’s peculiar histological arrays, the interpretation of HER2 staining in IMPC should be updated from the previously known ASCO/CAP recommendations.62 The authors mention that HER2 expression in IMPC by immunohistochemistry (IHC) ranges from 12.5% to 95%, possibly a result of scoring variability before the 2007/2013 guidelines.54 Furthermore, they analysed 1684 IMPC cases by IHC alone and found 11.6% to be positive (3+) and 29.4% to be equivocal (2+). Analysis of further 1272 IMPC cases by in situ hybridization (ISH) alone showed 20.4% of the cases were HER2-amplified and 7.4% were equivocal. Upon dual analysis of 411 cases by both IHC and ISH, 4.4% of the cases were found to be positive (3+) by IHC and of these, 83.3% were HER2-amplified. Interestingly, they showed that 43% of IMPCs with a HER2 staining score of 1+ were found to be HER2-amplified by ISH.54 They also claim that the morphology of the tumor seems to exclude the luminal side of the cells from staining. Therefore, they suggest lowering the “1+” categorization to tumor staining described as “weak to moderate but incomplete”. In fact, further testing of equivocal staining seemed to yield HER2 positivity in 35% of the specimens, indicating that a more inclusive definition would benefit many IMPC patients by encompassing them in HER2 targeted treatment, a finding also reported by more research groups.54

Lymphovascular Tropism

With the emergence of readily available methods of genomic and molecular analysis, a pathogenetic mechanism to explain the increased incidence of vascular, lymphovascular, and lymph node involvement has been proposed. As discussed earlier, IMPC cases appear with higher percentages of nodal involvement15 and lymphovascular involvement was detected in 14.7% to as high as 94.7% of the IMPC cases, compared to IDC cases.13,33

Recent studies have shown an overexpression of metalloproteinases and adherence molecules,6,15,46,50,63–65 as well as several cytotropic molecules, namely TNF-α, TNFreceptor II, E-cadherin, kindlin-2, integrinβ1, plakoglobin and β-catenin overexpression, occurring within pure IMPC cancer cells.50,51,66–70 Interleukin 1-β is associated with high microvascular density in IMPC tumors, as well as nodal metastases.71 N-cadherin, an adhesive protein, was also upregulated in IMPC cells when compared to non-IMPC cells.72 Well-known tumor chemotaxis factors SDF-1/CXCR4 also facilitate nodal invasion in IMPC.73 The findings mentioned above are indicative of the tumor cell’s ability to separate from neighboring cells, and invade the vascular and lymphatic systems, exhibiting a certain tropism towards lymphatic metastasis.15,50,74,75

The upregulation of glucose transporters has also been observed in a small number of patients, with significant differences in genomic expression when compared to non-IMPC tumors.76 The authors hypothesized that the apparent increase in GLUT-1 transporters with the simultaneous expression of hypoxia-inducible transcription factors is another process that enables IMPC cells to adapt, survive, and metastasize more than their non-IMPC counterparts.77

Another molecular-based study target that can give additional insights in the lymphovascular tropism of the tumors has been the observed predominance of CD44-positive and CD24-negative phenotype on IMPC cells. Alterations in the expression of these two molecules are partially responsible for certain stem cell properties that tumor cells exhibit (self-renewal, survivability, proliferation, lack of apoptosis). Among them, CD24 loss was associated with tumor spread and invasion.78,79 Indeed, a study by Li et al demonstrated a higher percentile presence of such cells, in comparison to IDC tumors, namely 48.5% versus 31.9%.78 CD44 loss was also found to be significantly higher in IMPC tumors when compared to NST tumors and was also associated with lymph node metastasis in IMPC patients as well.69,80 CD146 expression is also positively correlated with high microvascular density and was found to be more significant in IMPC rather than NST tumors.81 These findings serve as a plausible explanation of the IMPC invasive lymphotropic properties. A recent study by Kramer et al showed that IMPC tumor cells were in a highly epithelial state and did not use the EMT pathway, but rather form cell clusters during invasion and metastasis.82

The utilization of deep mRNA sequencing has also demonstrated at least 45 different miRNAs thought to be involved in IMPC development,83 and karyotype studies have also shown certain reproducible aberrations, such as gain of chromosomes 1q,8q,17q,20q and loss of chromosomes 1p,8p,13q,16q,20q, involved in the depolarization of IMPC cells.3,84–86 Among them, alterations in chromosome 8 seem to affect known malignancy-associated genes and could be one of the causes for the tumor’s invasive behavior.87 Other common genetic variations encountered specifically in IMPC include ESR1, KDR, ARID1B, ATR genes.88,89 Loss of LTZS1 expression is associated with IMPC development and nodal infiltration.90

The Role of Micropapillary Element or Micropapillary DCIS

A much-discussed topic in the study of IMPC is the significance and impact of micropapillary DCIS, or micropapillary foci, encountered within breast cancer tumors. Presence of micropapillary DCIS was associated with significantly larger tumor size and higher grade,91,92 as well as lymphatic invasion with nodal metastases.93,94 Recurrence rates, when micropapillary DCIS alone is present, also seem to be elevated,91 with a study reporting 29% versus 8% when compared to patients with non-micropapillary DCIS histology.91 All this is thought to be the result of higher histological grade tumors having a distinctly aggressive comedo necrosis96 and micro-invasion profile, thus explaining the local and locoregional recurrence of disease despite treatment.43,91,97 Another characteristic of micropapillary DCIS is the presentation as a large, multifocal, and often under-diagnosed breast tumor, as reported by a study from MD Anderson Cancer Centre.98 Literature indicates unfavorable recurrence profiles whenever such DCIS histology was present. In fact, even incomplete “inside-out” histological patterns, even without being characterized as micropapillary, are associated with LVI, nodal invasion, poorer survival, and larger tumor size when found in NST carcinomas.99,100

Micropapillary DCIS within NST tumors also differs when compared to non-otherwise specified DCIS within NST tumors. Higher incidence of vascular invasion, increased stage at diagnosis, high recurrence rates and increased lymph node infiltration are all well documented.101

A relatively common histological combination is that of mucinous breast carcinoma with micropapillary DCIS.102–107 Approximately 20% of all mucinous carcinomas are classified as “Mucinous Carcinoma with Micropapillary Features (MPMC)”.20,108,109 MPMC demonstrates higher percentages of lymphovascular invasion and lymph node invasion than mucinous breast carcinoma, likely explained by the higher instances of metastasis-associated mutations in genes associated with the PI3K-Akt, mTOR, AMPK signaling pathways,20,110 such as in GATA3 (20%), TP53 (20%) and SF3B1 (20%).20 Comparison with pure mucinous carcinomas has demonstrated lower frequency of HER2-positivity (20% for IMPC versus none of the mucinous carcinoma of breast111) and PR-negativity, lower nuclear grade and overall more aggressive biological behavior,111–115 as well as worse prognosis.105–107 Micropapillary mucinous carcinoma also shows evidence of being from the same lineage as pure IMPC, a finding that would explain their much-observed combination.116

IMPC Imaging

The mammographic appearance of IMPC is thought to be often nonspecific, and most lesions are an irregular, spiculated high-density mass, with scattered microcalcification in about 66.7% of the cases, often resembling IDC or DCIS.117–121 Micropapillary DCIS imaging in simple mammography often has a segmental or scattered microcalcification pattern.98,122,123 In fact, microcalcification patterns in mammography have been associated with worse prognosis in IMPC.124 Mammographic evaluation has a clear trend to underestimate the true disease size when IMPC is concerned.97,98,122 False-negative rates in mammography evaluation have been reported as high as 12% for IMPC patients,125 whereas patients with Invasive Lobular Carcinoma have false-negative rates higher than 14%,126 and up to 19%.127

When utilizing the ultrasound (U/S), the lesions are mainly hypoechoic, and it has been reported that the use of U/S often misses the true depth of the IMPC tumor invasion.26,119 A single hypoechoic lesion with irregular margins is the most encountered finding in U/S evaluation.117,122–129 In one study, micropapillary DCIS evaluation with U/S yielded a false-negative rate of 47%, and in those that were identified, the true extent was underestimated in 81% of the cases.122 Addition of shear wave elastography has been reported as helpful in better estimating IMPC tumors.130,131 Axillary evaluation of IMPC patients often yields suspicious lymph nodes with cortical thickening, and authors report positivity rates of suspicious nodes in 69% of the patients.128

MRI study is the most helpful at IMPC distinction, with the lesions presenting as spiculated, irregular masses with characteristic rapid enhancement and delayed washout patterns.118,132 Patterns of single or multiple irregular mass with rapid washout waveforms are the most well-recognized patterns of IMPC presentation in MRI.125,128 Mass and non-mass enhancement have also been previously described, while not as frequently as a solitary enhancing mass presentation.91,119,122 The probability of a non-mass enhancement of the lesion being found in MRI ranges from 16.7% to 38.9%.133 The non-mass enhancement is attributed to local lymphovascular infiltration, a finding attributed to the lesion pathology. In literature, non-mass enhancement of IMPC has also been attributed to the presence of DCIS within the lesion, an observation that needs larger case series for validation.128,133 Multifocal IMPC lesions are also better diagnosed and more accurately staged with MRI, compared to any other modality.43,91,98,122,125,128,133 While MRI may be the best imaging modality for IMPC, there is still a percentage of lesions that will be missed, especially diffuse multifocal lesions with extensive DCIS or residual disease after PST.117,134 An example of pre- and post-PST MRI imaging of micropapillary carcinoma can be seen in Figure 3.

Figure 3

(A) Preoperative and pre-treatment MRI image of IMPC. White arrow indicates the central mass of the lesion along with mass and non-mass enhancement. (B) MRI findings post-PST consistent with complete pathologic response. White arrow indicates local scarring in the mass area after PST. Although imaging indicated complete pathological response, residual disease was still found in the scarring area when examined under a microscope, and complete mastectomy was deemed appropriate.

PET-CT scans are also utilized, showing FDG (fluorodeoxyglucose) uptake of the primary tumor, with high (FDG) uptake being a prognostic factor for worse outcomes regarding breast cancer.119,135 As discussed earlier, IMPCs are characterized as Luminal A in 75.3% of the cases.41 Recently, Akin et al investigated how accurately PET-CT scan and MRI could detect breast cancer subtypes in 55 tumors.136 They found that although the SUVmax value from PET-CT scan was high for the Luminal A subtype, it was lower than the SUVmax value of the other breast cancer subtypes. PET-CT scan was better at identifying the molecular subtype of the breast cancer; however, MRI was superior at determining the tumor size, thus better for staging.

Treatment Options

Treatment of IMPC remains controversial, especially among breast surgeons. To begin with, there is a lack of guidelines regarding the impact of micropapillary element being present in several histological subtypes, as well as for the pure IMPC subtype itself. The well-known potency for lymphatic spread did influence surgical approaches in the past, since many authors report high percentages of axillary lymph node dissection (ALND) during surgery32,47 without any current evidence showing a need for more radical axillary approaches.137 While surgeons must strive for breast conserving therapy where possible,137,138 the majority of IMPC case reports were treated with modified radical mastectomy, as shown in a 2017 study by Yu et al, with 99% of the IMPC patients undergoing modified radical, or total mastectomy. Until recently, authors suggested a more radical approach towards locoregional management, with some adding larger surgical margin recommendations,28 and even locoregional radiation therapy to avoid extranodal recurrence. Indications for adjuvant and neoadjuvant treatment administration do not seem to be altered in IMPC, except for more cases being HER2 positive, and therefore candidates for biologically targeted treatment.62 Mercogliano et al demonstrated a possible resistance to HER2-directed therapy in IMPC tumors by investigating the mucin 4 (MUC4) molecule.61 Their study showed that MUC4 was overexpressed in IMPC tumors and had the ability to conceal the target epitope of trastuzumab, leading to treatment resistance and lower survival for IMPC patients (hazard ratio = 2.6, P = 0.0340). It is recommended that physicians have a high degree of suspicion, to avoid underdiagnosis, and to be vigilant in the axillary evaluation of such patients.44,139 To the best of our knowledge, the effect of adjuvant chemotherapy on survival or complete pathological response (CPR), or the role of the endocrine reaction in IMPCs has not been studied.

Newer developments in diagnostic markers and cancer therapy are currently being investigated for use in IMPC. One study evaluated the molecular profile of IMPC for potential response in immune-checkpoint inhibition treatments but showcased unfavorable status of the target ligands.140

Regarding the post-operative radiotherapy treatments (PORT), an informative study was published by Wu et al, studying 881 IMPC patients. The study uses a multivariate analysis of several patient factors and determined that both the surgical approach (mastectomy or breast conserving surgery) and the election to undergo PORT or not, did not alter the 5-year BCSS (breast cancer–specific survival) or OS (overall survival), which remained favorable for patients with IMPC. These results are also in line with older, smaller studies.141

Prognosis of IMPC

The comparative prognosis of IMPC has been a long-standing debate among scientists. A summary of studies evaluating the prognosis of IMPC can be seen in Table 3. However, recent studies and meta-analyses seem to suggest that there is no tangible difference in disease-free survival, recurrence-free survival, or overall prognosis.23,58,142–145  One such meta-analysis, that utilizes a great number of previous prognostic comparative studies, is the one by Hao et al.44 After a meticulous process of balancing key characteristics of the two populations (age, lymph nodes, grade, stage), the analysis demonstrated no statistically significant difference in overall survival and disease-free survival between patients with IMPC and those with IDC. Additionally, they demonstrated that the micropapillary subtype did not carry any gravity as an independent prognostic factor. Favorable prognostic factors for patients with IMPC include receipt of radiation treatment, estrogen receptor positivity, age <65 years and <4 positive lymph nodes.147,148 Lymphovascular invasion and negative ER status are among the most recognized negative predictors for IMPC.53 Lymphatic vessel density and VEGF-C expression are associated with lymph node infiltration in IMPC.149 It is worth mentioning that there are several older or with fewer patients comparative analyses,32,37,47,95,150–153 such as the one of Wu et al,7 or Yu et al28 that demonstrated worse recurrence-free survival, despite being in accordance with similar disease-free survival rates. This was attributed to a higher incidence of lymph node recurrence in the IMPC group of patients.7,28 Therefore, a question arose as to whether locoregional recurrence truly influenced the long-term overall survival of patients with IMPC. A study by Chen et al, also notes that it might be useful to compare overall survival in patient groups with similar nodal involvement and it demonstrated better breast cancer–specific survival as well as overall survival rates in the IMPC group of patients when compared to IDC patients.15,23,142 A recently published nomogram predicting the individual risk for locoregional recurrence, specific for micropapillary breast carcinoma, could be of use in risk-stratifying these patients.154

Table 3

Data on Local Recurrence, Distant Metastasis, and Survival from the Included Studies

Study	Study Type	Local Recurrence Rate of IMPC Patients	Rate of Distant Metastases of IMPC Patients	Survival of IMPC Patients	Comparison with Other Histological Subtypes.
Stranix et al (2015)4	Literature Review	6–80% of the patients (study-dependent)	1–49% of the patients	20–95% of the patients	–
Vingiani et al (2013)6	Case–Control Study	6.1% of the patients	8.2% of the patients	89.8% of the patients	Local recurrence rates and 10-year mortality were higher in IMPC patients compared to IDC patients. No observed difference in distant metastases
Wu et al (2017)7	Meta-Analysis	Locoregional relapse-free survival OR compared to IDC was 2.82	Distant metastasis-free survival OR compared to IDC was 0.95.	Overall survival OR compared to IDC was 0.90	IMPC patients have a higher incidence of locoregional relapse compared to IDC patients. Survival of IMPC and IDC patients did not differ.
Tang et al (2017)13	Case–Control Study	Locoregional recurrence in 4.2% of the IMPC patients	Distant metastasis in 8.2% of the IMPC patients	10-year Overall survival of 84.3% for IMPC patients	Regional and distant relapse-free survival was worse, compared to IDC patients
Cui et al (2014)14	Clinicopathological Study	Locoregional recurrence in 4% of the patients	Distant metastasis in 8% of the IMPC patients	92% of the patients with an average of 36.5 months of follow-up	–
Pettinato et al (2002)16	Case Series Study	Locoregional recurrence in 36% of the patients	Distant metastasis in 45% of the IMPC patients	55% of the patients with an average of 28 months of follow-up	–
Chen et al (2017)23	Case–Control Study	–	–	Overall survival HR compared to IDC was 0.67. Breast Cancer Specific Survival compared to IDC was 0.628.	Compared to IDC patients, IMPC patients had more favorable survival. IMPC histology was an indpendent prognostic factor for survival.
Yu et al (2015)28	Case–Control Study	Locoregional recurrence free	–	10-year Overall Survival of 92.4% of the IMPC patients. Survival HR compared to IDC was 2.56	Compared to IDC patients, IMPC patients had more favorable locoregional recurrence free survival. IMPC histology was an independent prognostic factor for survival.
Zekioglou et al (2004)30	Case–Control Study	Locoregional recurrence in 22.2% of the patients	Distant metastasis in 25% of the IMPC patients	72% of the patients with an average of 56.5 months of follow-up	–
Chen et al (2014)32	Case–Control Study	–	–	5-year DSS survival was 91.8% and OS was 82.9% of the patients on average	No difference was found in OS or DSS of IMPC patients compared to IDC patients
Gokce et al (2013)33	Case–Control Study	Locoregional recurrence in 6.9% of the patients	Distant metastasis in 23% of the IMPC patients	75.9% of the patients with an average of 64.7 months of follow-up	No difference was found in OS of IMPC patients compared to IDC patients
Hao et al (2019)44	Case–Control Study	Locoregional recurrence in 15.4% of the patients	Distant metastasis in 13.6% of the IMPC patients	85% of the patients with an average of 80 months of follow-up	No difference in OS or DFS between IMPC and IDC patients.
Ye et al (2020)45	Meta-Analysis	Locoregional recurrence OR was 3.60 compared to IDC	–	Overall survival OR compared to IDC was 0.87	No difference in OS between IMPC and IDC patients. Higher locoregional recurrence rates of IMPC patients compared to IDC patients.
Chen et al (2013)47	Case Series Study	–	Distant metastasis in 4.1% of the IMPC patients	5-year DS survival was at 91.9% and OS at 83.8%	–
Kaya et al (2018)48	Case Series Study	Locoregional recurrence in 15.8% of the patients	Distant metastasis in 5.3% of the patients	95.7% of the patients with an average of 48.87 months of follow-up	No difference in recurrence rates or survival, between patients with low or high percentage of micropapillary pattern
Kim et al (2020)58	Case–Control Study	–	–	–	No difference in overall survival between IMPC and NST patients, in multivariate analysis
Guan et al (2020)101	Case–Control Study	Locoregional recurrence in 3.1% of the patients	Distant metastasis in 20% of the patients	HR for OS for patients with IMPC component was 1.677 when compared to IDC patients	Locoregional recurrence was lower for IMPC patients compared to IDC patients. OS was better for IMPC patients. Distant metastasis rates were higher for IMPC patients
Kim et al (2005)102	Case–Control Study	Locoregional recurrence in 10.5% of the patients	Distant metastasis in 34.2% of the patients	–	Locoregional recurrence and distant metastasis rates did not differ between IMPC and non-IMPC patients.
Yoon et al (2019)124	Case–Control Study	Local recurrence HR was 2.86 compared to IDC patients	Distant metastasis HR was 1.85 compared to IDC patients	HR for death in IMPC patients compared to IDC patients was 1.30	Local and distant recurrence rates were significantly higher in IMPC patients, compared to IDC patients. No observed difference in overall survival.
Kim et al (2010)139	Case–Control Study	Locoregional recurrence in 13.1% of the patients	–	–	No significant difference in recurrence rates between IMPC and IDC patients. IMPC histology was not independently associated with recurrence
Chen et al (2018)142	Case–Control Study	5-year Locoregional recurrence in 28.6% of the patients	5-year Distant metastasis in 20.2% of the patients	5-year OS 81.9% for IMPC patients	Locoregional recurrence was associated with LN positivity, and was more frequent in IMPC patients. No difference in metastasis rates or OS, compared to TN-IDC patients.
Li et al (2019)143	Case–Control Study	–	–	3-year, 5-year and 10-year survival of 95.9%, 92.3% and 82.1% of the patients respectively	IMPC patients had better OS rates when compared to IDC patients, after propensity score matching
Li et al (2016)146	Case–Control Study	Locoregional recurrence in 5.1% of the patients after a mean of 39 months of follow-up	Distant metastasis in 9.1% of the patients after a mean of 39 months of follow-up	97% of the patients after a mean of 39 months of follow-up	No observed difference in survival or recurrence rates, compared to IDC patients
Lewis et al (2019)147	Case Series Study	–	–	5-year OS of 87.5%	–
Liu et al (2015)150	Case–Control Study	Micropapillary features in histological examination had an HR for RFS of 21.23	–	Micropapillary features in histological examination had no efffect on OS	Mucinous carcinoma with micropapillary features had worse RFS, but non-inferior OS compared to pure mucinous carcinoma
Kuroda et al (2004)152	Case Series Study	–	–	6-year OS of 41.2%	–
Shi et al (2014)153	Case–Control Study	5-year RFS of 67.1%	–	5-year BCSS of 75.9%	Recurrence and death rates were higher for IMPC patients compared to IDC patients
Meng et al (2021)154	Case Series Study	Locoregional recurrence in 18.5% of the patients after 59-month average follow-up	–	–	–

Several prognostic indicators are being studied for IMPC. In a recent study, sialyl LewisX (sLex) and mucin 1 (MUC1) expression in tissue specimens were found to be significantly different in IMPC cells when compared to NOS carcinoma cells. Furthermore, high levels of sLex expression, when combined with low levels of MUC1 expression, were also found to be a reliable prognostic factor for IMPC, making these two molecules potential specialized markers or therapeutic targets.155,156 Absence of caveolin-1 expression in stromal fibroblasts of IMPC is a candidate predictor for advanced axillary staging at diagnosis, as well as shortened progression-free survival.157 GATA3 is another IMPC-specific marker that seems to be expressed in tumors with better prognosis lacking however large confirmatory studies121,158 P63 expression was also found to be significantly associated with high Ki-67 index in IMPC cases, indicating another possible aggression marker that needs further study.159 Loss of ARID1A function was also noted to negatively correlate with disease-free survival (DFS) and 10-year overall survival (OS), especially in luminal B IMPC tumors.160

Conclusion

In the past few years, the previously unknown effect of the presence of micropapillary histological elements or pure IMPC on breast cancer has been explored. Due to its rarity as an entity, and the resulting difficulty in patient accumulation, there are not many studies that have produced tangible and statistically significant conclusions regarding all aspects of IMPC.

Micropapillary carcinomas of the breast have a well-recognized lymphovascular tropism that leads to more patients presenting with clinically disease-positive lymph nodes. In fact, the underlying biology of micropapillary histological patterns is detrimental in the lymphatic tropism of tumors, even when they present as a percentage of the malignancy’s histology or as foci of micropapillary DCIS. Basic research has revealed that there is a multitude of adherence molecules and chemotactic factors involved in the histology’s tendency for lymphatic invasion. Future, translational research, perspectives of such findings could include the utilization of said molecules as treatment targets or prognostic predictors for IMPC patients.

This review highlights the importance of approaching a breast cancer patient in accordance with the personalized medicine principles and making prompt therapeutic decisions in an individualized fashion based on the current literature and taking into consideration all aspects of a patient’s ailment. While no specific guidelines exist yet, it is made clear that micropapillary histology has an effect on treatment choices, and breast surgeons should be aware of the possible wider margin excision needed for this type of breast cancer. Further research is needed to confirm the role of chemotherapy and hormone agents, as well as resistance to trastuzumab. Imaging identification of micropapillary breast cancer is often underestimated regarding tumor invasion and size, and among the available options, breast MRI is the best one to perform. Recent research suggests that the – once thought – worse survival prognosis does not hold true; however, the alarming frequency of lymphovascular involvement and disease recurrence makes a more radical surgical approach more appropriate, for both the axillary and breast tumor burden.