History of fat grafting to the breast

In 1895 Czerny first described transplanting a lipoma to reconstruct a breast defect ( ). In the early part of the 20 ^th century, Lexer openly used resected fat to treat bilateral breast defects ( ). Because of limited success and the significant donor site defect of en bloc fat harvesting, these techniques were not adopted. In 1983, Illouz published his large experience with liposuction confirming our ability to remove fat cells from small port incisions using a cannula connected to vacuum ( ). This offered surgeons a new ability to address body contour. The ease of fat removal by liposuction also created an opportunity for its use as an autologous filler. However, because many of the variables important to fat grafting were not well understood, early results were disappointing. At the same time, the silicone breast implant was experiencing rapid growth over its prototype saline counterpart. The popularity of silicone further weakened interest in fat grafting for breast augmentation.

In 1988, Bircoll published his experience with the autologous grafting of liposuctioned fat for breast augmentation ( ). At the same time a series of opinion letters to the editor culminated in the American Society of Plastic Surgeons issuing a position statement questioning the safety of fat grafting to the breast ( ). The statement suggested that fat grafting would compromise breast cancer detection and should be prohibited. Because of this unprecedented strong ban and because early results were neither impressive nor reproducible, the technique was largely abandoned for more than 20 years.

The controversy of silicone breast implants in the United States in the 1990s and the subsequent moratorium on silicone breast implants until 2007 stimulated the potential for autologous tissue as a means of breast augmentation and reconstruction, and certainly this has been an undercurrent supporting the advance of this technique. Recently, there has been a resurgence of interest in fat grafting to the breast. In 2006, at the Annual Meeting of the American Society for Aesthetic Plastic Surgery, Thomas Baker presented a series of 20 patients augmented with liposuctioned fat with 90% graft survival and 180 mL growth documented by serial MRI and 3D volumetric analysis. None of the women had difficult-to-interpret findings on mammogram ( ). At the latest update of this series, with over 40 women followed up for at least 6 months and for an average of 30 months, there were still no issues with breast imaging and difficult to interpret masses. The average breast volume augmentation increased to 200 mL, while the percentage survival deceased to 80% ( ).

In 2007, Coleman published his review of 17 patients who were grafted using autologous fat and were followed up with serial photography ( ). The results were overall successful with maintenance of volume over 7–12 years of follow-up. This publication helped stimulate renewed interest in this technique. Coleman generally used serial grafting sessions instead of injecting larger volumes in a single session in a pre-expanded recipient breast like Baker et al. Following this renewed enthusiasm for the technique and with the realization that the radiographic arguments behind the ASPS-imposed ban may no longer be valid, many surgeons across the world have started publishing their own experiences with fat grafting to the breast ( ).

Donor graft (‘the seeds’)

The quality of the donor fat graft seems to depend upon donor age, donor site, harvesting technique & processing techniques.

Donor age (and therefore recipient age)

Age is thought to be a factor in the success of fat grafting. Animal studies in nude mice suggest this to be the case ( ). Human donor fat over a range of ages was injected subcutaneously into nude immuno-compromised mice. The data suggested higher volume retention in the recipient with fat from younger donors. In practice, autologous fat grafting does not afford us the opportunity to control for this variable and this may only serve as a pre-operative prognosticator.

Donor site

There have been anecdotal reports that fat harvested from some areas in the body survives better than fat from others. This may be an effect related to the size of adipocytes in various locations of the body. For example, fat harvested from the neck during a facelift may contain smaller adipocytes than larger fat cells harvested from the anterior abdomen. Smaller fat lobules have a better surface contact to volume ratio and therefore a higher likelihood of revascularizing. There also may be less trauma and cell damage when smaller fat cells are harvested. However, in practical terms, especially for large volume grafting, there is no demonstrable evidence that fat harvested from any area is better than another.

Harvesting techniques: the role of vacuum

Pressures utilized vary greatly in liposuction and certainly impact cell survival and graft take. Several studies have demonstrated that for optimal adipocyte viability, vacuum pressures should remain below one half atmosphere (<15 inches mercury, while regular liposuction is >30 inches of mercury) and that the lower the suction pressures, the more viable the adipocytes. Liposuction aspirator machines dialed down to very low pressures may be used to aspirate the fat. This technique, however, has many limitations as it requires: i) a rigid collection containers (lest they collapse under vacuum); ii) a stable stand away from the surgical field to keep the container upright (lest the aspirate continues straight into the suction machine); iii) long sterile tubing resulting in significant dead space loss. Another drawback of using liposuction aspirators is the sudden high airflow gushing through the tubing whenever a side hole of the liposuction cannula loses the vacuum. This causes sudden violent splashing of fat into the collection containers and desiccation of the graft through the long tubing.

Generally, hand-held, smaller size (<10 mL) syringe methods are thought to generate lower vacuum and be less traumatic and are therefore recommended to harvest fat. They have the inconvenience of being much more time, resource, and personnel demanding. A more recent advance is the constant pressure syringe that maintains a constant 280–300 mmHg (11–12 inches of mercury, about one-third atmospheric) vacuum throughout the excursion of the plunger.

The role of the liposuction cannula

When Illouz described liposuction and in the early days of fat grafting, the most commonly used cannulae were 10 mm in diameter. These large bore cannulae harvested large tissue chunks. Because of the resultant low surface to volume ratio when grafting these larger grafts, there is significantly less revascularization potential. The preference now is to harvest with cannulae less than 3 mm in diameter with side holes less than 1 × 2 mm in size. The smaller fat graft droplets thus harvested have much better chance of surviving. We also showed that harvesting efficiency per 10 liposuction strokes increased as we increased the number of side holes, and that increasing the number of openings up to 12 holes allowed us to efficiently harvest fat at pressures well below the ones described in the literature ( Figure 9.1 ) . In addition, the lipo-aspirate obtained by these smaller gauge cannulae with smaller aspiration holes has a much lower tendency to clump and clog and is easier to re-inject.

A 12 gauge multiple small hole cannula used in fat harvesting for grafting.

Ostensibly, one might think that fat surgically resected en bloc, which is then diced with minimal trauma, would maintain cellular integrity better than suctioned fat by any method, and might result in better graft take ( ). Ongoing studies are being performed in this area to better understand the role of minimizing graft trauma ( ) and there is an opportunity to validate this question and to potentially improve instrumentation in this area.

Processing techniques

There have been multiple reports of ‘percent graft take’ by volume ( ). Because of lack of standardization in grafting technique, we must consider re-thinking the results of many of these studies. For example, 60 mL of aspirated fat using the tumescent technique will decant to a variable aliquot of fat and serum, including blood and crystalloid. 60 mL of aspirate may decant to 30–40 mL of fat. When this fat is then centrifuged or rolled on a Telfa pad, two techniques used to further concentrate fat, the resultant fat may reduce to 20 mL by volume. It is, therefore not surprising that when fat is grafted, even if all the fat survives, if one does not effectively process the fat and remove crystalloid effectively, one has already committed to at best a 30–40% volume take, because that is the actual amount of fat that has been inserted by volume.

The controversy of centrifugation vs decantation

Separation by decanting ( Figure 9.2 ) involves simple gravity to separate higher density blood and crystalloid from adipocytes. A high speed centrifuge uses much higher gravitational forces (3–5 g) and separates fat from crystalloid extremely well. These centrifuges also require transfer of fat into multiple individual 5 mL or 10 mL syringes. However, it has been demonstrated that subjecting adipocytes to 3–5 g of centrifugation results in a higher degree of cell death ( ). Therefore, a compromise between these two techniques, used by several practitioners ( ) is manual centrifugation. Prototype devices, similar to the geared concept used in salad spinners, can subject larger volumes of adipocytes to 1.5 g forces to better separate out unwanted crystalloid, without subjecting the fat to excessive (3–5 g forces) trauma or excessive syringe manipulation ( Figure 9.3 ).

Simple decanting of fat from crystalloid.

Manual centrifugation in closed IV collection bag system.

Straining and filtration

Some separation methods advocate filtration of the lipo-aspirate. With this method, smaller fat droplets, pre-adipocytes, and other precious tissue such as platelets pass through the filter and are lost while the larger less useful clumps remain. Straining, as a method of preparation has the same drawbacks as filtration with the added disadvantage of decreasing adipocyte survival by exposing them to air. Air exposure, extensive manipulation and transfer from syringe to syringe increases the chance of contamination and the duration of hypoxia and has a potentially further deleterious effect on adipocyte survival.

In addition to potential cellular damage, centrifuged, strained or filtered lipo-aspirate delivers grafts in coalesced concentrated clumps with less chance to revascularize by interfacing the recipient bed than the loosely dispersed decanted fat. This is important for the large volume grafts required for the breast and is less of an issue in the well vascularized face where the smaller volumes usually grafted still have enough recipient contact and where more accurate control of the contour correction is mandatory.

Concentration of ‘stem cells’

Recently, there has been a great deal of excitement and many claims made over the role of stem cells or the ‘pre-adipocyte’ present in high concentration in the lipo-aspirate. In order to separate out pre-adipocytes, several well-known processes exist. Crucial steps in separation techniques utilize enzymatic digestion of the lipo-aspirate tissue and centrifugation. In the current clinical trial protocols, stem cells isolated from half the lipo-aspirate are mixed with the remaining unprocessed raw aspirate and re-injected as a stem cell enriched lipo-graft. At the time of this writing, it is still unknown whether this practice represents a significant contribution from stem cells versus adult adipocytes alone. This is an exciting area of potential research both in the physiology of the stem cell and in its ability to be separated from viable adipocytes, and hopefully this will be delineated over the next decade.

Such concerns support the argument that fat grafting in large volumes (unlike those performed for lip or nasolabial folds) might best be accomplished with a team approach. Ostensibly, it is recommended that an assistant or several assistants process fat simultaneously while surgical liposuction harvest is performed. On the other hand, a standardized, simple and efficient, closed system for non-traumatic low pressure aspiration, processing of the aspirated material, and easy re-injection of the preserved active slurry with minimal dependence on assistants has the highest potential for use in this type of surgery.

Recipient site size

In the familiar two-dimensional skin grafting of wounds, the surface of the wound determines the maximal limit of skin graft that survives, and over grafting does not help. Similarly, in three-dimensional grafting, the volume of the recipient bed sets the maximal limit to the volume of graft that can survive, and over grafting will lead to necrosis and oil cysts. Geometrical models of optimal dispersion of the individual fat droplets in the three-dimensional recipient matrix dictates that to preserve the three-dimensional vascular stromal interconnections of the matrix, anything more than a two-thirds ratio of donor volume to recipient volume filling (assuming recipient stromal cell islands are approximately same size as the fat droplets) will lead to overcrowding and by worsening the graft-to-recipient contact drastically increase the amount of necrosis.

Using the ‘seeds in a field’ analogy, and with everything else optimized, the size of the field sets the upper limit of the crop size. Overfilling the recipient space with more grafts will lead to crowding, will reduce graft to recipient interface and will result in massive necrosis. This opens up an opportunity for positive intervention. For mega-volume grafting in a tight space, temporarily expansion of the recipient loosens the extracellular matrix and creates many more interstices where many more individual fat droplets can optimize their graft to recipient interface and survive ( Figure 9.4 ).

Pre-expansion theoretically increases interstitial space and blood supply.

Recipient site quality

It is well established that muscle tissue with its high capillary density is an excellent graft recipient and that the more vascular the recipient, the better the graft survival. Furthermore, a number of cytokines and growth factors can accelerate the process of graft neoangiogenesis and improve survival. Thus attempts at priming the recipient site by increasing its vascularity and stimulate angiogenic cytokine production should improve engraftment. Tissue expansion is well known to be angiogenic and to stimulate the production of the potent angiogenic cytokine VEGF.

Recipient site filling pressure

From the general surgery trauma literature and from hand and upper extremity trauma, the importance of compartment pressure and the grave consequences of high interstitial pressure are well understood. If it were possible to pre-stretch the recipient space prior to fat grafting, it would make it more compliant and allow the injection of larger volumes of graft into the recipient site before reaching prohibitively high interstitial pressures.

Advances to improve the recipient site

The Brava ^® bra was initially developed in the 1990s as an external soft tissue Ilizarov expander to enlarge the breast. The device consists of a pair of semi-rigid domes with silicone gel rim cushion that interface with the skin and that are worn around the breasts like a bra. A small pump maintains inside the domes a low negative pressure that does not interrupt capillary flow and that imparts on the breast surface a constant gentle isotropic distraction force. Studies have shown that when the device is worn in an uninterrupted fashion for 10–12 hours per day, true and long lasting breast augmentation ensues over a period of weeks to months ( ). Forty thousand women have used Brava over the past 7 years, but compliance with the cumbersome treatment and the slow pace of real tissue growth has limited its popularity. However, within a few weeks of use, substantial temporary swelling of the breast develops. This was found to be an ideal way to prepare the breast for large volumes of fat grafts. The three to four fold increase in volume that can be attained over 4 weeks of wear allows a proportionate increase in fat graft amount without crowding. The expansion effect is also angiogenic as has been documented by MRI.

Experience with the VAC as a means of wound management has proven that microangiogenesis is a direct result of negative mechanical pressure ( ). The extensive VAC data on vascular in-growth and improvement in skin graft take coupled with the MRI findings from Brava expanded breasts support the authors’ thesis that increased microcirculation, combined with the larger interstitial space created by the expansion may both contribute to the potential for increased graft volumes and increased diffusion gradients. This is substantiated by Baker et al, who reported 80–90% long-term survival of grafts with volumes in excess of 250 mL.