Fax +41 61 306 12 34E-Mail karger@karger.chwww.karger.com

Original Paper

Eur Surg Res 2006;38:545–549 DOI: 10.1159/000096775

Microanatomy of Milk Ducts in the Nipple

Ferit Taneri a Osman Kurukahvecioglu a Nalan Akyurek b Ercument H. Tekin a

Mustafa N. İlhan c Cagatay Cifter a Sukru Bozkurt a Ayse Dursun b

Orhan Bayram a Erhan Onuk a

Departments of a General Surgery, b

Pathology, and c Public Health, Gazi University Faculty of Medicine,

Ankara , Turkey

Introduction

The papilla of the nipple is a conical or cylindrical prominence in the center of the areola. The tip of the nip-ple is fissured and contains the identifiable openings of a number of nipple ducts. Duct numbers ranging from 10 to 20 have been reported in the literature [1–3] , but rele-vant primary data are scarce, and information about the anatomy of human milk ducts and the nipple, largely based on studies performed in the 19th century, is far from complete. Understanding the microanatomy of the nipple and milk ducts has become practically important with developments in diagnostic ductoscopy, duct lavage, and nipple aspiration.

The duct anatomy has been evaluated by ultrasonog-raphy, ductography, and histology [4, 5] . Histological evaluation ought to be the gold standard, since milk ducts can be counted and distinguished from other ducts (e.g., sweat ducts or sebaceous ducts), but histological studies are often limited by small sample size and other factors. The aim of this study was to determine number and di-ameter of milk ducts in the papilla of the nipple in a large series of mastectomy specimens and to investigate pos-sible effects of age, breast weight, and diameter of nipple on the number of ducts in the nipple duct bundle. Such data are required for logical planning of nipple duct can-nulation and are relevant to developing new approaches to ductoscopy and other duct-directed diagnostic modal-ities.

Key Words Milk duct microanatomy, nipple � Nipple, milk duct microanatomy � Microanatomy, milk ducts

Abstract The aim of this study was to determine number and diame-ter of milk ducts in the nipple and to investigate the possible influences of age, breast weight, and diameter of the nipple on the number of ducts. Two hundred and twenty-six carci-noma mastectomy specimens were weighed and the nipple diameters measured. The number of ducts was counted in histological cross sections. Mean diameter of the nipple and mean breast weight were 13.9 mm and 844.6 g, respectively. There was a small but statistically significant positive correla-tion between nipple diameter and number of milk ducts (rho = 0.158; p = 0.01), but no correlation with breast weight. The mean number of ducts in the nipple duct bundle was 17.5. This is significantly higher than the number of ducts re-ported to open on the nipple surface. This discrepancy could reflect duct branching within the nipple or the presence of some ducts which do not reach the nipple surface. Smaller breast ducts (diameter ! 0.5 mm) represent nearly 50% of the nipple ducts and could be a challenge to the ductoscopy technology. Copyright © 2006 S. Karger AG, Basel

Received: March 14, 2006 Accepted after revision: September 4, 2006 Published online: November 3, 2006

Osman Kurukahvecioğlu Gazi Üniversitesi Tıp Fakültesi Genel Cerrahi Anabilim Dalı TR–06510 Ankara (Turkey) Tel. +90 312 202 5719, Fax +90 532 601 1918, E-Mail okurukahveci@yahoo.com

© 2006 S. Karger AG, Basel 0014–312X/06/0386–0545$23.50/0

Accessible online at: www.karger.com/esr

Taneri et al. Eur Surg Res 2006;38:545–549

546

Materials and Methods

We studied 226 breasts prospectively from 222 patients having undergone mastectomy for invasive breast carcinoma between 2002 and 2005. Patients with carcinoma invading the nipple or carcinoma in situ involving nipple ducts were excluded. The pa-tients’ age, weight of the mastectomy specimen, diameter of the nipple, number of ducts in nipple, and the cleft length of each duct were noted.

Mastectomy specimens were weighed, and the largest trans-verse nipple diameter was measured immediately after surgery and before formalin fixation. Transverse sections (at right angles to the nipple duct bundle) from the basal portion towards the apex of each nipple were embedded in paraffin, and 4- � m sections were stained with hematoxylin and eosin for histology. Three his-

tological cross sections were taken for each nipple, and the one in which the ducts were largest was evaluated. Ducts were counted, and the cleft length of each duct was measured and recorded. The number of ducts was counted using an eyepiece grid in an Olym-pus BX51 microscope at a magnification of ! 400. To confirm the distinction of breast ducts from sebaceous and sweat ducts, some sections were stained with an antibody against high molecular weight cytokeratin (Clone AE3; Neomarkers, Fremont, Calif., USA) with streptavidin-biotin-peroxidase detection, as described [4] . In addition, the numbers of ducts in one representative sec-tion from a subset of 24 nipples were counted by an outside ob-server (J.J. Going, University of Glasgow, Glasgow, UK).

The Pearson correlation coefficient was used to evaluate the relations between patient age, breast weight, nipple diameter, and duct number using SPSS version 13.0 for Windows.

Results

The mean age of the patients was 52.2 8 12.5 (range 26–82) years . The mean diameter of the nipple was 13.9 8 7 (range 10–18) mm. The mean breast weight was 845 8 311 (range 210–2,150) g. In total, 3,944 ducts were counted. The range of the number of ducts in the nipple bundle was 8–30 [17.5 8 (SD) 4.6] ( table 1 , fig. 1 ). The mean cleft length of the ducts was 0.57 (range 0.10–1.75) mm. The total number of duct openings according to their diameters are shown in table 2 .

There was no statistically significant correlation be-tween patient age and nipple diameter (p = 0.31) or num-ber of duct openings (p = 0.13). There was, however, a statistically significant but weak positive correlation be-tween nipple diameter and number of ducts (r = 0.158, p = 0.017; table 3 ).

The mean age at natural menopause in Turkish wom-en is 50.8 8 0.42 years [6] . When patients were grouped according to this age ( ! 50 vs. 6 50 years), which reflects the border of natural menopause, there was no statisti-cally significant difference between the two age groups with respect to the number of duct openings (p = 0.73).

The outside observer counted a median of 24.5 ducts (range 17–38, interquartile range 20.5–27.5 ducts) in the sections of the representative 24-nipple subset of our cases.

Discussion

The glandular apparatus of the breast is composed of branching lactiferous ducts, roughly organized in a ra-dial pattern, which spread outward and back from the nipple-areola complex. At the summit of each arborizing

Table 1. Patient characteristics

Mean 8 SD (range)

Age, years 52.2812.5 (26–82)Breast weight, g 844.68311.3 (210–2,150)Diameter of nipple, mm 13.987.0 (10–18)Number of duct openings 17.584.6 (8–30)

Table 2. Distribution of cleft lengths of milk ducts by diameter

Cleft length, mm Counted ducts

n %

0.10–0.25 443 11.240.26–0.37 262 6.640.38–0.50 1,280 32.460.51–0.62 431 10.930.63–0.75 919 23.300.76–0.87 236 5.980.88–1.0 270 6.841.01–1.25 92 2.331.26–1.50 10 0.251.51–1.75 1 0.03Total 3,944

Table 3. Correlation between number of duct openings at the nipple and nipple di-ameter

Rho Two-tailed p

0.158 0.017

Microanatomy of Milk Ducts in the Nipple

Eur Surg Res 2006;38:545–549

547

ductal system, the subareolar ducts are reported to widen to form the lactiferous sinuses [ 7 , but see ref. 5 for a recent comment] which communicate with orifices on the nip-ple [2] . The skin of the nipple contains numerous seba-ceous and apocrine glands. Understanding the ductal anatomy has become more important with the increasing use of ductoscopy in many institutions.

Early studies of the duct anatomy and development and involution of the breast were based on dissections of over 200 breasts [8] . Most textbooks of anatomy state that there exist between 15 and 20 ducts, and the glandular tissue is often depicted as 15–20 lobes radiating out from the nipple [9] . Although it has been stated that up to 22

ducts lead to the nipple in some women, some of these ducts may not be functional, and recent studies [4, 5] sug-gest that there are normally fewer than 12 patent ducts opening at the nipple. This issue is still controversial.

In a recent study [4] , six different ways were used to investigate the ductal anatomy in lactating women; 424 nipples were evaluated, and approximately 5–9 (range 1–17) ducts yielding milk were noted. In 13 mastectomy specimens, a transareolar dye injection technique was used, and 3–7 ducts were determined. Ductograms and mammograms were also used to analyze the ductal anat-omy. This study [4] claimed that more than 90% of the nipples contain 5–9 ductal orifices and hypothesized that

Fig. 1. A Panoramic picture of the ducts at the nipple. HE. ! 20. B AE3 immunostaining of the ducts. Strepta-vidin-biotin-peroxidase-diaminobenzidine/H 2 O 2 (SAB-DAB). ! 20. C Major ducts at the nipple are surround-ed by smooth muscle fibers embedded in fibrocollagenous stroma. HE. ! 200. D AE3 immunostaining of duct-ular epithelium. SAB-DAB. ! 200.

Taneri et al. Eur Surg Res 2006;38:545–549

548

the other ducts were sebaceous gland orifices. In another recent study [5] , the ductal anatomy of 21 lactating wom-en was evaluated with ultrasonography. Ultrasound im-aging identified approximately 9 (range 4–18) milk ducts in each lactating breast.

In contrast, Going and Moffat [10] identified a median of 27 ducts (range 11–48, interquartile range 21–30 ducts) in a series of 72 cancer mastectomy nipples. There seems, therefore, to be a discrepancy between the number of ducts which can be found inside the nipple and the num-ber of ducts which can be identified opening on the nip-ple surface. Love and Barsky [4] suggested that the excess ducts may belong to sebaceous glands. Going and Moffat [10] undertook a detailed three-dimensional reconstruc-tion of one nipple in which only 7 of 27 ducts had a lumen which could be traced right up to the surface of the nip-ple. This model could also explain the discrepancy, but more data are needed.

In our study, the duct number per nipple (mean number of ducts = 18) was also higher than that generally reported in the literature. There might be two reasons for this.

Firstly, ducts as small as 0.10 mm in diameter were counted in our study as opposed to other studies which excluded ducts ! 0.50 mm in diameter. In our study, 3,944 duct openings were counted totally, and half of them (50.3%) were ! 0.51 mm in diameter.

Secondly, counting milk ducts by light microscopy may well identify more ducts than dye injection, ductog-raphy, ultrasonography, or even histology, if tissue blocks for histology are not taken carefully, at right angles to the nipple duct bundle [4, 5, 8] . We believe that histological examination is best, because milk ducts can clearly be distinguished from other ducts (e.g., sweat or sebaceous ducts) using hematoxylin-eosin staining and cytokeratin staining in case of doubt. These two staining methods were used in our study, and milk ducts as small as 0.10 mm in diameter were counted.

It is interesting that the numbers of ducts in the 24 nipples from our series counted by J.J. Going are very close to those Going and Moffat [10] have reported from Scottish women. These numbers are if anything higher than the numbers we ourselves recorded, pointing to the fact that duct counting is not entirely straightforward. Their data probably represent an upper limit of nipple duct numbers.

Another possible explanation for the discrepancy be-tween the numbers of ducts opening on the nipple surface and those present inside the nipple is duct branching within the nipple itself. Although three-dimensional techniques confirm the branching architecture of the

duct system in the body of the breast [11] , the situation within the nipple is less clear. If a milk duct bifurcates near the surface of the nipple, it could run back into the breast as two adjacent ducts. But there are little data to show how often this might happen. For this reason, we counted all ducts separately. It is possible, therefore, that the numbers we found could overrepresent the number of duct systems in each breast. We do not think, however, that we have er-roneously counted sebaceous or eccrine sweat ducts as breast ducts: their hematoxylin-eosin appearance and im-munophenotype are distinctively different.

Breast ductoscopy has evolved over the past decade, as scope diameters have decreased. Access to a duct system is through the orifice at the nipple. Usually Bowmann’s lacrimal dilators (with outer diameters between 0.1 and 1.0 mm) are used to dilate the duct, starting with the fin-est one. This is time-consuming and has to be done cau-tiously [12, 13] . The nipple diameter and the anatomy of the ducts and their ostia are clearly becoming important in the development of new devices. In the light of our own experience we have developed a new duct microdilator [12] , and we think our data on the number of nipple ducts, the cleft length of each duct, and the relation between nipple diameter and number of ducts will guide the de-velopment of new devices and techniques for ductoscopy. In particular, the confirmation that almost 50% of the nipple ducts are smaller-caliber ducts ( ! 0.5 mm) is a def-inite challenge to the ductoscopy technology.

This is a relatively large study on the relation between age, breast weight, and nipple diameter and the number of milk ducts in the nipple bundle. There was a significant correlation between number of ducts and nipple diame-ter, but no correlation between patient age and breast weight. The number of ducts is higher than many studies have shown and highlights the discrepancy between the number of ducts which can be cannulated on the nipple surface and the number of ducts which can be found in the nipple. Data to discriminate between competing ex-planations (duct branching within the nipple, ducts which do not reach the nipple surface) are not available. Many small breast ducts (diameter ! 0.5mm) are likely to be a ductoscopic challenge, but their importance for breast disease is undetermined.

Acknowledgment

The authors thank James J. Going, Senior Lecturer (Cancer Sciences and Molecular Pathology, University of Glasgow, Glasgow, UK), for his excellent critical review.

Microanatomy of Milk Ducts in the Nipple

Eur Surg Res 2006;38:545–549

549

References

1 Bland KI, Copeland EM (eds): The Breast, ed 3. Philadelphia, Saunders, 2004, pp 21–42.

2 Townsend CM, Beauchamp RD, Evers BM, Mattox KL (eds): Sabiston Textbook of Sur-gery, ed 17. Philadelphia, Saunders, 2004, pp 867–928.

3 Brunicardi FC (ed): Schwartz’s Principles of Surgery, ed 8. New York, McGraw-Hill, 2005, pp 453–500.

4 Love SM, Barsky SH: Anatomy of the nipple and breast ducts revisited. Cancer 2004; 101: 1947–1957.

5 Ramsey DT, Kent JC, Hartmann RA, Hart-mann PE: Anatomy of the lactating human breast redefined with ultrasound imaging. J Anat 2005; 206: 525–534.

6 Aydin ZD, Erbas B, Karakus N, Aydin O, Oz-kan SK: Sun exposure and age at natural menopause: a cross-sectional study in Turk-ish women. Maturitas 2005: 52: 235–248.

7 Stenberg SS (ed): Histology for Pathologists, ed 2. New York, Lippincott-Raven, 1997, pp 71–84.

8 Logan W (ed): Breast Carcinoma. New York, John Wiley & Sons, 1977, pp 281–300.

9 Bannister LH, Berry MM, Collins P, Dyson M, Dussek JE (eds): Gray’s Anatomy, ed 38. New York, Churchill Livingstone, 1995, pp 417–424.

10 Going JJ, Moffat D: Escaping from flatland: clinical and biological aspects of human mammary duct anatomy in three dimen-sions. J Pathol 2004; 203: 538–544.

11 Moffat DF, Going JJ: Three-dimensional anatomy of complete duct systems in human breast: pathological and developmental im-plications. J Clin Pathol 1996; 49: 48–52.

12 Tekin EH: Microdilator of milk duct ostium: a new tool for the insertion of the scope (ab-stract). 7th Annu Meet Jpn Assoc of Mam-mary Ductoscopy, 2002, p 5.

13 Mokbel K, Escobar PF, Matsunaga T: Mam-mary ductoscopy: current status and future prospects. Eur J Surg Oncol 2005; 31: 3–8.