A 46-year-old woman underwent radical left mastectomy with axillary lymphadenectomy in October 2005 due to hormone receptor-positive infiltrating ductal carcinoma in the upper outer quadrant and lymph node metastasis in 1 of 13 isolated nodes.
In the same surgical procedure, immediate reconstruction was performed by placing an expander prosthesis (anatomical Becker) with a 300 cc. volume textured silicone gel layer, starting expansion 2 weeks after surgery and reaching the maximum expansion volume 1 month later.
As a complementary treatment, chemotherapy was established according to a sequential scheme of adriamycin-cyclophosphamide and taxane, receiving a total of 24 sessions during the period from November 2005 to May 2007, followed by trastuzumab. Complementary hormonal treatment with letrozole was maintained for a period of approximately 3 years.
The second stage of reconstruction was performed in July 2006, by replacing the expander prosthesis with a 320 cc textured silicone gel breast implant in the same subpectoral location.
In the follow-up clinical controls, a depression was observed at the medial edge of the mastectomy scar, in the form of a hatchet, which displeased both the patient and the surgical team, so it was decided to remodel it in March 2007 by means of a fat graft of adipose tissue obtained from the abdominal area and processed using the Coleman technique, after infiltration of the scar with 3 cc of local anaesthetic (lidocaine). The scar was infiltrated with 3 cc. of local anaesthetic (lidocaine 1% with adrenaline); it was performed without incident and a total of 5 cc. of fatty tissue was infiltrated in the area of the defect.
The patient continued to be monitored by the Gynaeco-Oncology Department, and follow-up imaging tests were carried out in accordance with our hospital's guidelines for the diagnosis and treatment of malignant tumours. Specifically, 4 annual mammograms and 2 breast ultrasound scans were performed between May 2006 and April 2009, with no evidence of alterations suggesting damage to the integrity of the prosthesis.
In March 2010 a new breast ultrasound was performed in which a previously unreported cystic ovoid image was visualised on the anterior surface of the interior of the breast prosthesis, so the radiologist decided to perform an MRI. This procedure was carried out one month later and a nodular image was observed in the prosthesis, located in the internal interquadrant area, about 2.5cm in diameter, which appeared as an image of liquid content, with no inflammatory signs or signs of prosthetic rupture. In different sections it appeared as if an elongated hook-shaped structure was introduced into the prosthesis, similar to the way a finger is introduced into a half-inflated balloon. The patient did not report any symptoms.

In view of these images, we decided to perform a surgical exploration, taking into account the possibility of having to remove the breast prosthesis. At first it was thought that the fatty infiltrate might have entered the prosthetic capsule and pushed the prosthesis, without breaking it, creating a deep concavity inside it, like a tunnel.
During the operation, it was found that the prosthesis appeared to be intact and that there was no tissue inside the capsule that could alter the shape of the implant. There were no alterations in the area of the capsule that was in contact with the site of fat infiltration, but we nevertheless decided to take a biopsy of the area of injected fatty tissue; this biopsy corresponded in the histological study with fibrous tissue with no signs of malignancy.
When we removed the prosthesis for examination, we were struck by the fact that there were several opaque spots scattered in different places inside it, with no linear distribution, which might lead us to believe that these were needle punctures. However, at a point close to the largest of these, after applying strong compression to the prosthesis, we saw that the cohesive silicone of the implant was slightly herniated. We then decided, once the implant had been removed, to open it and send samples of the material inside. The removed implant was replaced with another of the same characteristics.

The anatomopathological study reported the sample analysed from inside the prosthesis as haematic material.