Preparation

Harvesting sites are selected only to enhance body contour and avoid iatrogenic donor site deformities, since no overriding clear correlation between donor site location and longevity of the implanted tissue has been demonstrated. ²⁸ , ²⁹ I prefer to harvest fat from the torso, especially the back, love handles, and abdomen, because a reduction in these areas enhances a woman’s body proportion and accentuates the larger breasts with a narrower waist.

Anesthesia

I usually use a general anesthetic or propofol because of the large volumes that must be harvested for fat grafting to the breast. I infiltrate a solution of 0.2% lidocaine with 1:400,000 epinephrine in Ringer’s lactate solution into the donor and recipient sites.

Incisions for Harvesting

Incisions are made with a No. 11 blade scalpel, and a blunt nine-holed Coleman harvesting cannula is used to diffusely inject Ringer’s lactate solution with 1:400,000 epinephrine in a ratio of 1 ml of solution for each 1 cc of fat harvested. For the posterior torso, I usually make one incision in each sacral dimple and one incision in the midline of the back. These may be supplemented by incisions laterally to facilitate harvesting in the hip or flank.

From the anterior approach, I most often place incisions in the pubic hair and umbilicus, with occasional incisions in the flanks or hips to facilitate removal from the more lateral regions.

Cannulas for Harvesting

Fat is harvested in most of the cases with a 10 cc syringe attached to a nine-holed, blunt-tipped Coleman harvesting cannula. The larger-gauge cannula is used here. A combination of slight negative pressure and a curetting action allows parcels of fat to move through the cannula, through the Luer-Lok aperture, and into the barrel of the syringe, resulting in minimal mechanical damage to the fatty tissue.

Harvest

After harvesting, the fat is centrifuged in the 10 cc syringes for 3 minutes at 1300 g. The oil layer is decanted and the aqueous component drained. Codman pads are used to wick off the remaining oil, and the refined, unwashed fat is then transferred to 3 cc Luer-Lok syringes.

Incisions for Infiltration

The incisions are positioned to allow placement from two directions into each area to be grafted. Usually three inframammary incisions, an areolar incision, and an axillary incision on each breast are enough to provide access to the entire breast. Incisions on the sternum, décolleté, or surface of the breast should be avoided.

Cannulas for Placement

Blunt 17-gauge infiltration cannulas, 9 or 15 cm in length, are used to place the fat through 2 mm incisions. Blunt cannulas not only allow more dispersion of the grafted tissue in small aliquots, but also reduce the possibility of an intravascular injection. ³⁰ A type I Coleman cannula is most often used, but if there is scarring or other tissue resistence, a type II or even type III Coleman cannula may facilitate the surgeon’s ability to push the cannula through the denser tissues such as scars and breast parenchyma.

For safety reasons to avoid intravascular emboli, at no time are sharp needles used for injection into the breast. The fat is infiltrated only while the cannula is being withdrawn, not during the advancement of the cannula. Approximately 0.2 cc is placed with each withdrawal of the cannula. This approach maximizes the surface area of contact between the surrounding tissues and the grafted fat. The goal is to allow each parcel of fatty tissue access to a blood supply, thus maximizing the potential for survival and minimizing the possibility of fat necrosis. Also, for safety reasons, cannulas of more than 15 cm in length should be used with extreme caution, because pneumothoraxes have been reported by even the most experienced surgeons while infiltrating fat in the thorax.

Method of Infiltration

The Coleman technique used for placement must be one that facilitates survival of the fatty tissue, not only by minimizing trauma during harvesting and refinement, but also by placing the living tissue in a manner that optimizes the potential for vascularization. When fat is placed into the recipient site in large clumps, some of the fat cells may be too far from a blood supply. This can lead to fat necrosis, causing not only lumps and calcifications, but also the formation of liponecrotic cysts. Bircoll ⁷ advocated the placement of fatty tissue in amounts as small as 1 mm per pass. At the time, the idea of making 130 passes in the breast to place 130 cc of fat was considered “ludicrous.” ³¹ , ³² The technique described in this chapter is based on an even smaller aliquot, with each pass of the cannula depositing from ⅕ to ½ cc per pass, depending on the length of the deposited path. Dramatic volume corrections are possible, but the time requirement for such a procedure should not be underestimated. The time to harvest, refine, and place fat into the breasts in this fashion takes many hours. In general, placement of fatty tissue into the breasts using this method takes approximately 2 hours for the first 100 cc and approximately 45 minutes for each additional 100 cc placed.

Level of Infiltration

The breasts are shaped by layering the fat into different levels until the desired contour is achieved. Although a breast implant augments by expanding the retromammary or retropectoral space, the structural fat grafting technique allows selective augmentation and contouring from the chest wall to the skin. In most of my earlier cases, the largest portion of the fat was infiltrated into the pectoralis major muscle, followed by infiltration into the retropectoral and prepectoral spaces. Those volumes were located close to the chest wall and provided a diffuse increase in the chest size without specific shaping. Shaping of the breast is accomplished by placing aliquots subcutaneously into the superficial breast planes. Placement into the parenchyma of the breast should be limited but is necessary at times to increase projection.

Volume Ranges

With structural fat grafting, corrections with as little as 50 cc to more than 400 cc of fat grafted diffusely in the breast and surrounding tissues produce long-lasting results. The amount placed always depends on the effect desired. As with any surgical procedure, the technique used, the execution of the technique, the experience of the surgeon, and even variations between patients affect the outcome.

Planning

The goal was to create a fuller, natural teardrop shape by placing fat into each quadrant of the breast (focusing on the more deficient medial and inferior areas) as well as into the cleavage, and the retroareolar space (for projection). We marked donor sites (purple marker) in the love handles, posterior hips, sacrum, posterior medial thighs, lateral thighs (all approached from the prone position) as well as the upper and lower abdomen and anterior and medial thighs. The breasts were divided into four quadrants, and the areas for placement were delineated with green marker. The inframammary folds were marked in orange.

Anesthesia

A general anesthetic was used for this patient, with a laryngeal mask airway (LMA). The donor sites were infiltrated with 1:400,000 epinephrine in Ringer’s lactate solution and 0.2% lidocaine (about 500 ml total was used in the donor sites).

Donor Sites

The donor sites were chosen to enhance the patient’s feminine shape by removing most of the harvested fat from the area of the waist down to the hips and up to the breasts.

Purple marks the areas to be removed, red marks possible incision sites, and orange indicates areas in which care was necessary to avoid excessive removal, such as the transition zone between the back and the love handles.

Harvest

When a midline back incision and a curved 23 cm cannula are used, the entire love handles and flanks can be approached without lateral incisions.

The incisions in the sacral dimple can be used to harvest not only from the sacrum and lumbar regions, but also from the hip and flank.

Surgery began with the patient in the prone position, after successful induction of anesthesia and placement of an LMA. The patient was maintained under deep intravenous sedation and breathed on her own. After infiltration with tumescent solution (Ringer’s lactate solution with epinephrine 1:400,000 and lidocaine 0.2%), using blunt Coleman multiholed cannulas, the following volumes were removed with manual suction: right love handle, 170 cc; left love handle, 210 cc; right hip, 100 cc; left hip, 70 cc; right sacral dimple, 70 cc; left sacral dimple, 70 cc; right medial thigh, 50 cc; left medial thigh, 50 cc; right lateral thigh, 100 cc; and left lateral thigh, 120 cc, for a grand total of 1010 cc. The fat was processed in the standard fashion, separating out 2.5 cc aliquots based on density. The fat was noted to be fibrous. (This is a common finding in healthy, fit patients.) The incisions were closed with 5-0 Ethilon interrupted sutures.

The patient was carefully turned to the supine position and reprepped and draped for further suction. A total of 50 cc was removed from the upper abdomen, 50 cc from the lower abdomen, and 40 cc from each medial thigh, yielding a total of 180 cc harvested in the supine position. (We did not suction the preoperatively marked anterior thighs, since sufficient volume was obtained without using this area.) The donor incisions were closed with interrupted 4-0 Ethilon sutures. The grand total volume harvested (from both prone and supine positions) was 1190 cc, and processing yielded 615 cc of usable fat.

Volumes to Be Placed

The chest was then prepped and draped in the usual sterile manner. Stab incisions for placement of fat were made in the medial, lateral and central inframammary folds (as indicated by red dots in the preoperative markings), as well as in the upper medial areolas. The breasts were infiltrated with a small volume of the tumescent solution to reduce bleeding. Using blunt Coleman placement cannulas, a total of 300 cc was placed into the right breast/cleavage, and a total of 307.5 cc was placed on the left side.

Placement

The decision about what level of fat density to use in an area is based on whether one is looking to improve texture, or just create volume. Almost any level of density can be used in the breast, unless scars are being treated—high-density fat is preferred for treating scars. In this case, the higher-density fat was used in the lower quadrants and retroareolar areas, where the most significant structural changes were planned.

On almost every patient I begin with placement into the lower pole, but in this particular patient it was especially important, because this was the most important shaping that would take place in her breast. The placement began against the dermis from the medial incision with a 9 cm Coleman type I cannula, switching to a 15 cm cannula on occasion.

After placing 20 cc from the medial incision, I switched to the lateral incision for placement into the same area of the lower pole of the breast.

After another 20 cc of fat was infiltrated from the lateral incision, I approached the same area from the areolar incision. From all three incisions, fat was placed over the entire lower quadrant, with special attention in this patient to placement that would lower the inframammary fold to the new location several centimeters below the current location.

After 60 cc of fat was placed into each lower quadrant, the medial quadrants were infiltrated from the areolar, medial, and middle inframammary incisions. Again, all placement was subcutaneous. The medial quadrants were felt to be remarkably deficient, so larger amounts were infiltrated here (right, 55 cc; left, 50 cc) than was planned for the lateral quadrants.

Next the lateral quadrant was infiltrated through the lateral inframammary incisions with a significantly smaller quantity of fat.

This placement extended up into the axilla to fill in the defect created by the interface of her large pectoralis muscles with her small breast.

The lateral quadrant was also approached through the middle inframammary fold incision. The placement remains superficial, especially in this patient, to attempt to correct the tuberous deformity. Note the dimpling of the skin as the cannula passes superficial to the skin.

Even with the small amount placed into the lateral quadrant, the fat is infiltrated from three different incisions and presumably three different directions. Every attempt should be made to place the fat aliquots in small quantities with each pass and from at least two and even three directions when possible.

Switching to a 15 cm cannula extends the reach so that distant areas can be infiltrated with fatty tissue in several directions.

Finally, the lateral sternum cleavage area was approached. A 15 cm infiltration Coleman type I cannula is always used in these cases to reach the distant site. In this patient, 40 cc was placed on each side. This amount is about the usual for most patients.

The area was approached from the areolar incision with a 15 cm type I cannula to provide another direction of placement.

After placement was complete into all of the sites, some tissue was placed into the retromammary region. Axillary incisions facilitated finding the plane behind the breast. In this patient, 85 cc was placed on the right and 60 cc on the left in the lower half of the breast; into the upper half of the breast in the retromammary region, 70 cc was placed on the right and 43.5 cc on the left.

All of the incisions were used for this deep placement in the retromammary space with 15 cm 14-gauge type I cannulas. The larger cannulas are used for safety, because the stiffer, bigger cannulas allow greater control in placement. Recently I have begun using the old Coleman harvesting cannulas (bucket-handle type) for placement in this plane and around breast implants.

After fat placement into all of the planned areas, the patient was brought to a sitting position. We determined that her breast projection was deficient, and it did not appear that the lower poles had as much curvature or pseudoptosis as had been hoped.

Through the axillary incision and the other four, another 35 cc was placed into the constricted right lower pole and another 30 cc into the left lower pole. This was an extremely difficult placement, because there was not much space, and several times it was debated whether we should release the apparent adhesions in these areas to allow more curvature of the lower pole. I felt that this would destabilize the fat I had already placed and would probably not allow me to place more fat in a vascularized, healthy fashion. Therefore no release of any sort of adhesion or constriction was performed. However, to enhance projection, 30 cc was placed on the right and 10 cc on the left deep in the parenchyma. No tissue was intentionally placed behind the areola at any time.

Several times during the procedure, the patient is placed in a sitting position to evaluate the effect of the fat transplantation. In these views, taken toward the end of the procedure, one can see the beginning of a release of the constricted lower pole. More placement simply was not possible into the stiff lower pole.

The breast incisions were closed with 5-0 Ethilon sutures. Reston foam was placed over the inframammary folds and between the breasts to prevent migration of fat. Compressed Reston foam was also placed over the areolas to prevent herniation, since the patient had a history of tuberous breasts. The Reston foam was secured with compression in all places with Tegaderm. The surgical bra cushioned with Kerlix was placed, as was a lower body compression garment.

The patient was seen for routine follow-up on the third postoperative day. We were able to attach the external vacuum expansion domes at that time. We instructed the patient to begin using the device nightly and to try to work up to 8 hours per day again. We saw her routinely during her initial postoperative course, and she was able to use the expansion device for 8 hours a day for about 4 weeks. Her postoperative course was unremarkable, and the patient was happy with her results.