Special Types of Invasive Breast Carcinoma: Tubular Carcinoma, Cribriform Carcinoma, Mucinous Carcinoma, Micropapillary Carcinoma

Clinical Presentation

The majority of patients are diagnosed in their late 50s and early 60s, although the age at diagnosis ranges from 27 to 92 years. Rarely, TCs have been reported in men. Most (60%–70%) TCs are diagnosed as nonpalpable mammographic abnormalities. Occasionally, very small TCs are discovered incidentally in biopsies performed for unrelated reasons.

The exact frequency of multifocality and multicentricity in TC is difficult to ascertain because of varying definitions and methods of sampling used by different investigators. In one study, multifocality was reported in 20% (18/90) of patients with pure TC. Another reported multicentricity in 10 of 17 (56%) mastectomy specimens with TC. This incidence was significantly greater than in a control group composed of mastectomy specimens containing breast cancers of other types. The incidence of contralateral breast cancers in patients with TCs ranges from 13% to 26%.

Clinical Imaging

The structural abnormality associated with TC allows for its easy detection by mammography. In one study, the average mammographic size of nonpalpable TC was 0.8 cm compared with 1.2 cm when the lesion was palpable. Small TCs may not be apparent on mammography but in the majority of patients, it presents as a mass lesion with central density, occasionally associated with microcalcifications. The mass may appear round, oval, or lobulated, with irregular or spiculated margins ( Fig. 29.1 ). Due to similar architectural patterns, TCs cannot be distinguished reliably from invasive carcinoma of no special type (IC-NST) or from radial scars on imaging studies.

Ultrasonography is particularly helpful in detecting some mammographically occult tumors. On sonography, most TCs present as an ill-defined hypoechoic mass with posterior acoustic shadowing, a feature typically associated with malignant tumors ( Fig. 29.1 ).

Gross Pathology

The average size of pure TCs reported in most series is 1.0 cm or less; the median size in two studies was 0.8 cm and 1.2 cm. In general, tumors that are relatively larger are most likely examples of mixed tumors (IC-NST with tubular features) or a coalescence of multifocal TCs.

Grossly, TCs are indistinguishable from IC-NST and appear as firm gray-white lesions with spiculated margins that retract from the cut surface. The cut surface of tumors with extensive elastosis may appear tan or pale yellow.

Microscopic Pathology

Tubular carcinoma has a distinctive microscopic appearance characterized by a haphazard proliferation of well-formed glands or tubules distributed in a stellate configuration ( Fig. 29.2 ). The neoplastic tubules are round to oval with well-formed lumens; many are also sharply angulated with tapering ends. The tubules are lined by a single layer of monotonous cuboidal to low columnar epithelial cells with basally oriented low-grade nuclei and inconspicuous nucleoli ( Fig. 29.3 ). Mitoses, when present, are occasional. The cytoplasm is usually eosinophilic to amphophilic, often exhibiting prominent apical snouts ( Fig. 29.3 ).

The altered stroma of TC makes it easily recognizable from the surrounding normal breast tissue on low-power view. The common stromal alterations are desmoplasia and fibroelastosis ( Fig. 29.4 ). Some consider stromal elastosis as a characteristic finding in TC ( Fig. 29.5 ), but it may not be present in every tumor ( Fig. 29.6 ). Furthermore, stromal elastosis can be present in other invasive carcinomas and in benign lesions. Approximately half of TCs have calcifications identified microscopically. The calcifications can be seen in the stroma, within the neoplastic glands or in the in situ component ( Fig. 29.7 ). Lymphatic and vascular invasion are extremely rare in TC, although metastasis to lymph nodes is not unusual.

A diagnosis of TC should be reserved for cases in which more than 90% of the tumor exhibits this characteristic morphology. Tumors comprising 10% to 90% tubular elements or high-grade nuclei should be classified as IC-NST.

The majority of TCs are associated with a range of atypical epithelial changes including atypical ductal hyperplasia (ADH), atypical lobular hyperplasia (ALH), ductal and/or lobular carcinoma in situ (DCIS and/or LCIS), and columnar cell lesions with or without flat epithelial atypia (FEA). DCIS associated with TC typically has cribriform, micropapillary, papillary, or mixed patterns, is usually of low nuclear grade, and overall may constitute a minor component of the tumor mass ( Fig. 29.8 ).

The frequent coexistence of columnar cell lesions with FEA, lobular neoplasia, and TC has been termed a triad (Rosen Triad) ( Figs. 29.9 and 29.10 ). In fact, some of the incidental TCs are detected in breast biopsies performed for microcalcifications associated with columnar cell lesions ( Fig. 29.11 ). Columnar cell lesions are reported in 93% to almost 100% of TCs, whereas coexistent lobular neoplasia is found in approximately 50% of cases.

Tubular carcinoma retains its well-formed tubular architecture in lymph node metastases ( Fig. 29.12 ). As some of the metastases involve the lymph node capsule, these well-formed glandular metastases can be mistaken for benign glandular inclusions, such as endosalpingiosis.

Prognosis and Treatment

The biological behavior of TC is very favorable. The reported local recurrence rate in patients with TC ranges from 4% to 7%. In a comparative study of TC ( n = 102) versus low-grade invasive ductal carcinoma (IDC) ( n = 212), Rakha and colleagues reported a 6.9% local recurrence rate for TCs compared with 25% for IDC. All seven patients with TC who experienced local recurrence were treated with wide local excision, five patients did not receive postoperative radiotherapy, and none of the seven patients received adjuvant systemic therapy. Min and colleagues compared outcomes of TC with those of DCIS and found no significant difference in disease-free survival, with a 1.4% recurrence rate for TC versus 2.9% for DCIS.

The reported incidence of axillary lymph node metastases in patients with TC varies widely, mostly because of variation in the pathological criteria used in the past. Some of the studies have shown a direct relationship between the degree of differentiation of the tumor and the incidence of lymph node metastases. Overall, tumor size appears to be the greater risk factor for lymph node metastasis. In one study, the median tumor size of the tumors with lymph node metastasis was 1.6 cm, versus 0.9 cm for those without metastasis. In most studies of pure TC, the incidence of lymph node metastases is less than 15% (6.8%–12.9%). In cases where TC does metastasize to axillary lymph nodes, it is usually limited to only one to three level I lymph nodes. Furthermore, the presence of nodal disease does not appear to affect disease-free or overall survival. In view of the low frequency of nodal disease, most patients undergoing sentinel lymph node biopsy are spared an axillary lymph node dissection.

Patients with TC have a longer disease-free survival compared with other invasive breast carcinomas. In the randomized prospective National Surgical Adjuvant Breast and Bowel Project (NSABP)-B06 trial, 120 patients with TC were included among 1,090 node-negative and 651 node-positive patients. In a univariate analysis, both node-negative and node-positive patients with TC experienced significantly greater overall survival at 10 years compared with other patients, and this favorable histology proved to be an independent predictor of survival in node-negative patients by multivariate analysis. In a study by Rakha et al, longer disease-free survival and breast cancer–specific survival was reported in 102 patients with pure TC when compared with 212 patients with grade 1 IDC with a median follow-up of 127 months. There were no cancer-specific deaths in patients with a pure TC diagnosis compared with 9% cancer-specific deaths in patients with grade 1 IDC. Distant metastasis is rare in TC. Rosen et al and Sullivan et al found no distant metastases in 24 and 73 patients with TC, respectively. In three large studies, the overall survival rates (94.1% 5-year and 81.7% 10-year) of patients with TC were not significantly different from those of aged-matched women in the general population (91.3% 5-year and 77.6% 10-year).

Most patients with unifocal TC are excellent candidates for breast conserving surgery (BCS). Although lymph node metastasis is relatively uncommon, sentinel lymph node (SLN) biopsy should be performed at the time of definitive surgery in patients with a prior diagnosis on core biopsy. If an incidental small TC is found in an excisional biopsy performed for a nonmalignant lesion, the need for a subsequent SLN biopsy is questionable considering the low frequency of lymph node metastases and the excellent prognosis of patients with nodal metastasis.

While some studies have questioned the need for radiotherapy in patients with TC considering it's excellent prognosis, several studies have shown benefits of radiotherapy in these patients. Hansen et al demonstrated increased relapse-free survival after adjuvant radiation therapy in patients who had BCS. A retrospective analysis by Fritz and colleagues of two German series found an improved survival rate after postoperative radiation therapy, yet antihormonal and chemotherapy regimens showed no benefit. In another study of younger patients (65 years or younger), adjuvant radiation therapy was a significant predictor of survival.

While TCs are HR+ tumors, adjuvant hormonal therapy may be omitted. The 2022 National Comprehensive Cancer Network (NCCN) Guidelines for Breast Cancer state that adjuvant endocrine therapy should be considered for patients with TCs less than 30 mm that are also pN0 or pN1mi while it is recommended for those that are pN1 of any tumor size. Adjuvant chemotherapy is usually not indicated for TC, although it is considered optional in patients with macrometastasis in one or more nodes.

Prognostic/Predictive Factors

The various biological markers expressed in TCs generally reflect the well-differentiated nature and good prognosis of these tumors. Estrogen receptor (ER) positivity has been reported in 80% to 98% of TCs, and progesterone receptor (PgR) positivity in 72% to 92%. Tubular carcinomas typically show strong and diffuse (more than 90%) staining with ER ( Fig. 29.13 ), and they are almost always diploid, have a low proliferation rate, and do not show HER2 gene amplification or p53 protein accumulation. Occasionally, a prominent basolateral membrane staining with HER2 immunohistochemistry (IHC) ( Fig. 29.14 ) leads to reflex HER2 fluorescence in situ hybridization (FISH) testing, but these tumors typically do not show HER2 gene amplification. If HER2 positivity is found, one should reconsider the diagnosis of TC.

Molecular and genetic studies of TCs reveal a low frequency of cytogenetic abnormalities compared with the complex abnormalities present in most breast cancers of no special type. Tubular carcinomas show similar genetic abnormalities (a higher frequency of 16q loss and 1q gain but a lower frequency of 17p loss), transcriptome, and IHC expression profiles as seen in other low-grade luminal-type breast carcinomas.

Differential Diagnosis

Benign entities such as sclerosing adenosis, complex sclerosing lesions/radial scars (RSs), tubular adenosis, and microglandular adenosis may mimic TC and vice versa.

Adenosis and Sclerosing Lesions

As angulated and tubular glands are commonly encountered in sclerosing adenosis, tubular adenosis, and RSs, these entities are often considered in the differential diagnosis.

Sclerosing adenosis is a common benign lesion that often manifests as microcalcifications on mammogram or an enhancing mass lesion on magnetic resonance imaging (MRI). Sclerosing adenosis sampled in a core biopsy can pose a diagnostic dilemma, particularly if associated with atypia. Because the proliferation pattern in sclerosing adenosis is lobulocentric, it is helpful to examine these lesions at low power to avoid misinterpretation ( Fig. 29.15 ).

Tubular adenosis is an uncommon microscopic finding, may not be lobulocentric, but usually lacks desmoplastic reaction. Examination on higher magnification usually reveals the presence of myoepithelium ( Fig. 29.16 ). In difficult cases, IHC for myoepithelial cells such as p63 and smooth muscle myosin heavy chain (SMM-HC) is helpful. Smooth muscle actin (SMA) stains adjacent myofibroblasts and therefore is not recommended for detection of myoepithelium in breast lesions.

Radial Scars

The center of a RS is comprised of angulated glands and stroma that exhibits fibroelastosis, sometimes making the distinction from TC very difficult. In excisional biopsy specimens, radial sclerosing lesions are easy to recognize because of the characteristic appearance of a sclerotic nidus with trapped angulated tubules from which dilated ducts admixed with ductal hyperplasia and papillary proliferation radiate. Complex sclerosing lesions/RS are most problematic if the center of the lesion is sampled in a core biopsy ( Fig. 29.17 ). In a core biopsy, IHC for myoepithelial cells will help to identify them in the angulated glands of a RS ( Fig. 29.18 ). It is important to note that in highly sclerotic lesions, immunoreactivity for myoepithelial cells may be focally attenuated or absent in a few of the entrapped glands ( Fig. 29.19 ). Thus, caution should be used in the interpretation of IHC stains.

Tubulolobular Carcinomas

Tubulolobular carcinomas are low-grade invasive tumors that may not be readily classifiable as either ductal or lobular carcinomas for various reasons. Some of these tumors may represent a collision tumor—a truly mixed ductal and lobular carcinoma—whereas others may have cytological features that are more typical of TC but invade the stroma in a single-file pattern or have cytological features of invasive lobular carcinoma ( Fig. 29.20 ). This has led some to consider these tumors to be a variant of TC. Although DCIS is an integral part of TC, tubulolobular carcinomas are more likely to be associated with LCIS. Tubulolobular carcinomas are also more likely to be multifocal and metastasize to axillary lymph nodes when compared with pure TCs. IHC staining for E-Cadherin (ECAD) may be useful in making the distinction between ductal and lobular carcinomas in problematic cases. One study reported common expression of ECAD, cytokeratin (CK) 8, and CK34βE12 in eight tubulolobular carcinomas. It should not be surprising that pathologists may find it difficult to categorize a given lesion as ductal or lobular given the fact that some low-grade IDCs share similar chromosomal abnormalities with invasive lobular carcinomas.

Invasive Carcinoma of No Special Type

Distinguishing grade 1 IC-NST from TC can be challenging ( Fig. 29.21 ). Well-differentiated IC-NST in which the glands are more complex in architecture or the tumor cells are higher grade should not be confused with TC. If one adheres to the criteria of TCs as tumors showing at least 90% tubule formation of mostly single-layered cells of low nuclear grade - the distinction between the two becomes attainable.