It is not an easy case especially evaluating based on the photos only.

1. Generally speaking it is a papillary lesion.

2. Features: Sclerosing, ductal epithelial proliferation, necrosis.

 3. Cytomorphologic features are not like DCIS, such as streaming, nuclei varing in size and overlapping, cellular swirling, fenestrations. Of cause I do not think it is invasive tumor.

4. Atypical features: focal micropapillary growth, foally with a little monotanous prolifearion, focal central necrosis (though it is not unusual to see the necrosis in UDH).

5. I favor a dianosis of atypical papilloma based on above features.

Of cause if you want to write more you can diagnos as  atypical papilloma with sclerosis and ductal epithelial hyperplasia with focal necrosis.

6. The only differential dx for me is solid papillary ca. But thinking over I denied the dx.

7. I Often noticed the people suggested to use IHC for UDH, ADH, DCIS. It is useless. It is a morphologic dx for these three, not IHC. If the cases showed IHC results well, they should be easily diagnosed by H&E. If  they are borlerline cases by H&E, they will not be distinguished by IHC also. I as a breast/gyn pathologist rarely know the pathologists still use the IHC for differential dx of UDH, ADH, and DCIS in the USA. In 2004 I once started to do a study in this area, I gave up this research after I got the prelimary result.

Dr. Schnitt , Beth Israel Medical center, Harvard once had a breast IHC review paper mentioning the usage of HMW CK. He also mentioned this part in his 2009 breast biopsy book. I asked him in persion in last uscap meeting about IHC for udh, adh, dcis. He said they did not use IHC also in this situation.

Of cause you can still use IHC stains if you feel   they can help you.

Thank Dr. Jin for sharing the chellange case. Just for your reference. cz

 谢谢金主任出的好病例！

认为这是一个从中间到恶性的结构多样混杂的病例。有的酷似乳头状瘤，但却有肿瘤性坏死；有的酷似导管内癌，但却似有浸润（图7、8），甚至怀疑累及脉管（图24）；有灶区酷似乳头状瘤（硬化型）。就是有的为导管内癌，也表现为不同亚型。导管内癌存在，浸润癌不能除外。需多个肌上皮标志物IHC辨别。也需要标记淋巴管血管等，鉴证浸润否。