In the breast, intraductal papilloma (IDP) is a benign proliferative lesion that consists of a branching fibrovascular core with overlying epithelial and myoepithelial layers. These lesions can occur anywhere in the ductal system, and may be solitary or multiple.
Terminology and classification
Haagensen was one of the first to attempt to subclassify these lesions, labeling them as either multiple papillomas or solitary papillomas.  Using this classification scheme, he found that multiple papillomas were smaller than solitary papillomas, and they tended to be peripheral. He also found that women with multiple papillomas had a greater risk of later developing invasive breast carcinoma compared with women with a solitary papilloma.
Using 3-dimensional (3-D) reconstructions, Ohuchi et al later classified papillomas as either peripheral papillomas, originating at the terminal duct-lobular unit (TDLU), or central papillomas, originating in the larger ducts (ie, ducts not part of the TDLU).  The investigators found that peripheral papillomas had a strong tendency to be multiple, whereas central papillomas were usually solitary. Ohuchi et al also showed that women with peripheral papillomas were more likely to have concomitant intraductal or invasive carcinoma compared with women who had central papillomas. 
In their atlas of subgross pathology of the human breast, Wellings et al distinguished between papillomas involving TDLUs, which they called hyperplastic terminal duct with papilloma, and those involving the larger ducts, which they called ductal papilloma.  The investigators found "an arbitrary continuum of morphologic structure links normal [TDLUs] and/or lobules to the entire group of dysplastic, metaplastic, hyperplastic, anaplastic, and neoplastic human mammary lesions, with the exception of intraductal papillomas and rarely occurring hyperplasias of larger ducts."  That is, Wellings et al identified that small, peripheral papillomas were on a morphologic continuum with other lesions arising from the TDLU, whereas large, central papillomas were not.
Based on these and other studies, the authors of this article believe peripheral papilloma, multiple papilloma, and hyperplastic terminal duct with papilloma are all the same lesion. Although these papillomas arise from the TDLU and would be more accurately described as "TDLU papillomas," the long-held convention is to refer to them as "intraductal papillomas," which is the convention that is kept in this article. Accordingly, small, peripheral papillomas are referred as "IDP, small/peripheral subtype (IDP [S/P ST])."
The authors also believe that central papilloma, solitary papilloma, and ductal papilloma are all the same lesion, and they will be referred to as "IDP, large/central subtype (IDP [L/C ST])" in this article. In addition, the authors view IDP (S/P ST) and IDP (L/C ST) as different lesions, with differing prognostic implications. This article is devoted to IDP (S/P ST), whereas a separate one will discuss IDP (L/C ST).
The authors recommend the following resources for additional information:
O'Malley FP, Pinder SE. Breast Pathology: Foundations in Diagnostic Pathology. Philadelphia, Pa: Elsevier; 2006.
Page DL, Anderson TJ. Diagnostic Histopathology of the Breast. Edinburgh, Scotland: Churchill Livingstone; 1987.
Schnitt SJ, Collins LC, eds. Biopsy Interpretation of the Breast (Biopsy Interpretation Series). Philadelphia, Pa: Lippincott Williams & Wilkins; 2008.
Rosen PP, ed. Rosen's Breast Pathology. 3 rd ed. Philadelphia, Pa: Lippincott Williams & Wilkins; 2009.
Wellings SR, Jensen HM, Marcum RG. An atlas of subgross pathology of the human breast with special reference to possible precancerous lesions. J Natl Cancer Inst. Aug 1975;55(2):231-73.
See also Breast Abscesses and Masses.
Small, peripheral intraductal papilloma (IDP [S/P ST]) is primarily a disease of middle-aged women. In Haagenson's study of 52 women with multiple papilloma, the mean age at diagnosis was 41.2 years, and occurrence decreased significantly after age 55 years. 
Prevalence of these lesions shows some variation across reports in the literature. Bartow et al found "papillomas of lactiferous ducts" in 8 of 519 (1.5%) forensic autopsy cases  ; Sohn et al found papilloma in 215 (3.9%) of 5257 consecutive breast biopsies  ; Schnitt et al found papilloma in 1-2% of consecutive benign breast biopsies in a series of Japanese women  ; and Lewis et al found multiple papillomas in 54 (0.6%) of 9108 consecutive benign open breast biopsies, and solitary papilloma in 426 (4.6%) patients in the same cohort. 
Although the studies evaluating prevalence in breast biopsy specimens likely overestimate the prevalence of papilloma in the general population, the above studies are in relative agreement and indicate the combined prevalence of IDP (S/P ST) and IDP, large/central subtype (IDP [L/C ST]) to be 1-5%. The Lewis et al study also suggested that IDP (S/P ST) is considerably less common than IDP (L/C ST).  However, the experience of the authors of this article indicates that IDP (S/P ST) is more common than IDP (L/C ST). This discrepancy may be because IDP (S/P ST) lies on a histologic spectrum with usual ductal hyperplasia (UDH), and many lesions classified as IDP (S/P ST) by the authors of this article would be classified as UDH by others.
Recent molecular evidence has shown that papillomas frequently have loss of heterozygosity (LOH) at loci on chromosome 16. Specifically, Di Cristofano et al found LOH at telomeric 16p (marker D16S423) in 3 of 11 informative papillomas, as well as at centromeric 16q (marker D16S310) in 1 of 3 informative papillomas.  Similarly, Lininger et al found LOH at 16p13 in 6 of 10 informative intraductal papillomas with florid epithelial hyperplasia. 
Extensive studies of usual ductal hyperplasia (UDH) have shown it to have frequent, although inconsistent, chromosomal imbalances, and it has become accepted that UDH is a truly neoplastic process.  Given that allelic imbalances are also characteristic of breast papillomas, as preliminary studies indicate, it is likely they are also truly neoplastic. It should be recognized, however, that these studies did not distinguish between small, peripheral intraductal papilloma (IDP [S/P ST]) and large, central intraductal papilloma (IDP [L/C ST]), and that genetic difference may exist between these lesions.
In addition, IDP (S/P ST) is frequently present in breasts involved by other benign proliferative lesions, including columnar cell hyperplasia (CCH), sclerosing adenosis, and usual ductal hyperplasia (UDH). Therefore, it stands to reason these breasts are somehow the "fertile soil" from which proliferative lesions of all types arise. However, what makes this "fertile soil" is an enormous question that remains unanswered.
Small, peripheral intraductal papillomas (IDP [S/P ST]) are usually asymptomatic and discovered incidentally on breast biopsies prompted by a separate lesion. However, they may rarely present as a palpable breast mass, or they appear as calcifications on mammography.
In the early 20th century, many breast pathologists considered all papillary lesions in the breast to be malignant.  However, it later became clear that many are benign, and that some are associated with a slightly increased risk of developing invasive breast cancer (IBC).
Carter was one of the first to recognize this association.  In a landmark paper, he studied 64 women with breast papilloma and demonstrated that multiple papillomas increased the risk of developing invasive carcinoma more than solitary lesions. He showed that although 2 of 6 patients diagnosed with multiple papillomas later developed IBC, only 4 of 58 patients with solitary papilloma did so. 
Around the same time, Haagensen studied women with multiple papillomas and those with solitary papilloma as separate cohorts and found that 6 of 52 women with multiple papillomas developed IBC (relative risk [RR] 3.7; multiple papilloma vs normal), whereas 7 of 172 women with solitary papilloma developed IBC (RR 1.3; solitary papilloma vs normal). 
Lewis et al later reported the largest series to date (480 women with papilloma) that also showed multiple papillomas increased the risk of IBC more than solitary papilloma (RR 3.01 for multiple papillomas; RR 2.04 for solitary papilloma; both vs normal).  These reports are represented in the image below.
The question as to whether IDP (S/P ST) is a cancer precursor or merely a risk factor has been a topic of debate. Azzopardi did not consider papilloma to be a direct precursor based on what he believed to be a false association between papilloma and IBC in studies available during his time, which he thought were confounded by an overdiagnosis of "pseudo-infiltration" in benign papilloma and the misdiagnosis of papilloma as papillary carcinoma.  In contrast, Haagensen believed "multiple papilloma" was a precancerous lesion, because all 6 women in his study who developed IBC did so in the ipsilateral breast, in the same area where the initial papilloma had been diagnosed. 
However, in the largest series to date, although woman diagnosed with papilloma had an overall increased risk of developing IBC, they had equal bilateral risk, showing no statistically significant difference in breast cancer risk between the breast ipsilateral to and contralateral to the papilloma. [7, 15, 16] Studies evaluating loss of heterozygosity (LOH) in papilloma and subsequent development of breast cancer have also failed to identify LOH at the same allele in both the papilloma and the invasive carcinoma. [8, 13] On balance, it appears that IDP (S/P ST) is best viewed as a benign proliferative lesion that is a risk factor for developing carcinoma rather than as a direct cancer precursor.
Additionally, considering the known increased carcinoma risk in patients with other benign proliferative lesions, which is similar in magnitude to the increased risk associated with multiple papillomas, it must be recognized that the increased IBC risk seen in studies involving IDP (S/P ST) may not be a factor of IDP (S/P ST) itself; instead, it may be related to fundamental abnormalities in the breasts that develop proliferative disease.
Gross examination is generally not informative in assessing for small, peripheral intraductal papillomas (IDPs [S/P ST]), as the lesions are only appreciable microscopically. For this reason, they have been referred to as microscopic papillomas.  Histopathologically, IDPs (S/P ST) are composed of a prominent, branching fibrovascular core that supports a myoepithelial layer and an overlying layer of epithelium. The epithelium is cuboidal to columnar, with basally oriented nuclei. The myoepithelial cells range from spindled to epithelioid, and they lie between the epithelial layer and the fibrovascular core (see the following images).
Many IDPs (S/P ST) lie within a cystic space (see the images below), and attachment to the cyst wall can frequently be identified. The extent of myoepithelial proliferation differs between IDPs (S/P ST) in that some show a single layer of myoepithelial cells underlying the epithelium, whereas others show more hyperplastic appearing myoepithelium (see the last image below).
IDPs (S/P ST) can also be augmented by an epithelial proliferation that resembles usual ductal hyperplasia (UDH) (see the following images).
In some IDPs (S/P ST), the fibrovascular core proliferates in a manner that gives the lesion a solid appearance. These may look like an IDP (S/P ST) with abundant, tightly packed fibrovascular branches (as shown in the first 2 images below), or they may have an adenomalike appearance (as seen in the last 2 images below). Such papillomas have been referred to as duct adenoma and solid papilloma, as well as adenomyoepithelioma for lesions with a hyperplastic myoepithelial component. 
The primary differential diagnosis for small, peripheral intraductal papillomas (IDPs [S/P ST]) includes usual ductal hyperplasia (UDH); large, central intraductal papillomas (IDPs [L/C ST]); atypical papilloma; papillary ductal carcinoma in situ (DCIS); and papillary apocrine metaplasia. These are briefly discussed below.
UDH vs IDP (S/P ST)
UDH can be difficult to distinguish from IDP (S/P ST). UDH is a benign proliferation of a various cell types that takes on a variety of architectures, including micropapillary. It can occasionally be involved by fibrovascular processes, a feature more commonly associated with IDP (S/P ST). Because UDH can have features of IDP (S/P ST), and IDP (S/P ST) can be involved by a UDH-like epithelial proliferation, there is considerable overlap between these lesions, and the authors view them as opposite poles on a histologic spectrum. This spectrum is represented in the images below. In the authors' practice, IDP (S/P ST) is diagnosed only when a prominent, branching fibrovascular core is identified, and the UDH-like proliferation accounts for less than 50% of the lesion.
IDP (S/P ST) vs IDP (L/C ST)
Differentiating IDP (S/P ST) from IDP (L/C ST) is straightforward in most cases. IDP (L/C ST) is generally much larger than IDP (S/P ST), grossly appearing as a papillary structure with associated duct dilatation; it is also solitary and found in the large, central ducts. In contrast, IDP (S/P ST) is usually only appreciable microscopically, is peripherally located, and is often multiple. In cases that are not easily categorized, the authors diagnose IDP (S/P ST) if the cross-sectional area of the papilloma is the approximately 2 mm or smaller, or if multiple papillomas are identified.
IDP (S/P ST) vs papillary DCIS
In contrast to IDP (S/P ST), papillary DCIS usually has more delicate fibrovascular cores and is frequently involved by a clonal-appearing epithelial proliferation with mild to moderate cytologic atypia (see the following images). Most distinctively, however, papillary DCIS does not have a myoepithelial layer (see the 3rd and 6th images below). Immunostaining, therefore, can be useful in distinguishing between these 2 lesions. Many find it helpful to use a cytoplasmic marker, such as calponin or smooth muscle actin (SMA), along with the nuclear marker p63 (see Immunohistochemistry). It should be noted that occasional, scant expression of myoepithelial markers may be identified in papillary DCIS. This should not be taken as strong evidence of a benign papilloma if other features of the lesion favor papillary DCIS.
Cytokeratin (CK) expression also differs between IDP (S/P ST) and papillary DCIS in that CK5/6 and CK14 show stronger staining in IDP (S/P ST) than in papillary DCIS .
In a series by Tan et al, using a predetermined immunohistochemical cutoff score, the investigators were able to differentiate papillary DCIS from benign papilloma using CK5/6 and CK14 with a sensitivity of 95% and 85%, respectively, and a specificity of 86.8% and 94.3%, respectively.  The immunohistochemical score utilized percentage of cells stained multiplied by staining intensity, and both myoepithelial cell and epithelial cell staining was included in the evaluation. 
Tan et al's findings are similar to CK5/6 expression differences found between UDH and DCIS. [19, 20, 21, 22] However, it must be recognized that myoepithelial cells frequently express CK5/6 and CK14, and the staining differences between papilloma and papillary DCIS may largely be a result of the myoepithelial component of papilloma, and thus, they may not reflect differences in epithelial cytokeratin expression between IDP (S/P ST) and papillary DCIS.
IDP (S/P ST) vs papillary apocrine metaplasia
Papillary apocrine metaplasia is encountered in the epithelium of cysts and is quite distinctive. The key to making the diagnosis lies in identifying the diffuse apocrine change in the epithelium, which has a characteristic intense eosinophilia (see the images below). Nucleoli are also often prominent. Fibrovascular cores in papillary apocrine metaplasia are delicate or absent, and the cell apices may show classic apocrine decapitate secretions.
IDP (S/P ST) vs atypical papilloma
Atypical papilloma is a controversial diagnosis. As with IDP (S/P ST), it has a fibrovascular core with overlying epithelial and myoepithelial layers, but, in contrast to IDP, atypical papilloma is involved by an epithelial proliferation similar to that seen in atypical ductal hyperplasia (ADH) or DCIS. For this reason, it has also been referred to as "DCIS involving a papilloma." Although atypical papilloma more commonly presents in the differential diagnosis with IDP (L/C ST), it occasionally presents as a smaller, IDP (S/P ST)–sized lesion.
The diagnostic criteria for atypical papilloma are not well established, and there is considerable interobserver variability in diagnosis between pathologists. This will be discussed in more detail in a separate article.
The primary clinical use of immunohistochemistry in small, peripheral intraductal papilloma (IDP [S/P ST]) is to highlight the myoepithelial cell layer, which is helpful in distinguishing it from papillary ductal carcinoma in situ (DCIS) (see Differential Diagnosis). The myoepithelial cells in IDP (S/P ST) express typical myoepithelial cell markers, including p63 (see the first image below), CD10 (see the second image below), calponin, and smooth muscle actin (SMA) (see the third image below), as well as mammary basal cell markers, such as cytokeratin (CK) 5/6, CK14, and CK17. The epithelial cells usually express keratins typical of mammary luminal epithelium, including CK8/18 (see the last image below) and CK7.