Use of New Technologies in Implant-Based Breast Reconstruction

Abstract

Outcomes after mastectomy and prosthetic-based breast reconstruction have improved immensely since the development of the first tissue expander and breast implant in the 1960s. One major factor contributing to our improved outcomes over the past two decades is the increasing availability and improvement of perfusion assessment technology. Instrumental methods now exist which allow surgeons to assess tissue viability intraoperatively, and provide actionable, objective data that augments clinical assessment. In this article, the authors detail two commercially available, state-of-the-art technologies that surgeons may use to assist in mastectomy flap assessment and facilitate the reconstructive process.

Wound healing remains a challenging clinical problem for surgeons. Nowhere is this more apparent than in breast reconstruction. We have moved from the creation of a simple breast mound in the 1960s to the current state-of-the-art reconstructions which closely resemble the natural breast. Whether the reconstruction is implant- or autologous tissue-based, we are judged by our final, well-healed results. Critical to achieving such results is the ability to predict tissue viability and perfusion. Oxygen is an essential elemental requirement in meeting the energy demands of the wound at all phases along the continuum of wound healing. It is vital for glucose metabolism, cell proliferation, microbial defense, angiogenesis, and extracellular matrix deposition and repair. ^{1 2}

Oxygen availability is a function of supply and metabolic demand. Any local or systemic factors depriving the tissue of adequate perfusion with oxygenated blood may affect skin flap healing. Therefore, assessment of these parameters at all time points is crucial for successful healing. Early detection of compromise during the intraoperative and immediate postoperative period is of utmost importance in implant-based breast reconstruction. Assessment during this period allows preservation of native skin through precise resection of devitalized tissue that has not declared itself clinically yet. ³ The ideal assessment techniques should be minimally invasive, rapid, repeatable, and inexpensive, and provide a clinically meaningful end point for decision-making, so that appropriate interventions may be implemented to restore or enhance perfusion and tissue oxygenation.

Perfusion Assessment

Elevating breast skin flaps introduces the risk of vascular compromise and the associated potential for delayed wound healing, tissue necrosis, and increased potential for infection. ^{4 5 6} Clinical judgment by assessing flap color, turgor, temperature, capillary refill, and dermal bleeding still play a critical role in perfusion assessment. Until the 21st century, clinical assessment was the primary method of evaluating the overall tissue quality in the operating room. ^{3 6 7 8 9 10 11 12} Subsequently, fluorescein became the most widely available gold standard method for perfusion assessment, despite its few shortcomings. ^{3 13} Indeed, no device or new technology has the capability of accounting for all systemic and local factors at play when determining the integrity of local tissues. However, useful adjuncts are available to the reconstructive surgeon, which can provide objective data intraoperatively. This, in turn, permits early intervention that can mitigate the risk of postoperative complications, and assist in flap design pre- and intraoperatively. Over the following sections, we will describe and report our experience on two recent and innovative technologies that allow for objective assessment of tissue quality in real-time.

Indocyanine Green Fluorescence Imaging

Investigation of vascular networks and tissue perfusion beyond clinical assessment during the late 20th century required injection of fluorescein, followed by ultraviolet (UV) light evaluation. This was widely used due to its lower cost and good correlation with skin viability. ³ Nonetheless, fluorescein as a technique was prone to certain disadvantages such as adverse drug reactions, anaphylaxis, rapid diffusion into the interstitium yielding imprecise results, single dosing due to prolonged half-life precluding repeat examinations, difficulty interpreting in darker skinned patients, need for UV light source, superficial visualization with the naked eye, and a tendency to underpredict survival. ^{3 6 14 15 16 17} Furthermore, there was no acceptable methods for any quantitative evaluation. ^{11 18}

Indocyanine green (ICG) fluorescence imaging was developed in the 1960s as a means of evaluating perfusion. ICG laser fluorescence was initially developed in Canada to evaluate cardiac perfusion and was subsequently transported into plastic surgery where it finally took hold. The new technology provided real-time, correlative assessment of tissue perfusion both intraoperatively and postoperatively, as well as the feasibility to be used multiple times during the same operative procedure.

Indocyanine green is a water-soluble, iodinated dye with excitation and absorption properties in the near-infrared range, detected to a depth of 2 to 3 mm within the studied tissue. It is easily produced for clinical applications as a sterile lyophilized green powder, which can be dissolved in sterile water and injected intravenously (IV) into the patient. The half-life of ICG is 2 to 4 minutes, with 4% remaining after ∼20 minutes. It remains protein bound intravascularly and is exclusively metabolized by the liver and excreted in the bile. ICG should be avoided in patients with liver disease and uremia, and is contraindicated in patients with allergies to ICG and iodine, since it is administered parenterally. ² These characteristics make ICG suitable for visualization of skin perfusion intraoperatively. In addition, due to its low toxicity profile, sequential administrations can be performed depending on the dye's clearance characteristics. ⁸ Typically, during breast surgery, a dose of 5 to 10 mg is injected as an IV bolus. A camera-associated laser then excites the ICG molecules, thereby causing the loss of an electron which can be detected by the near infrared camera. Visualization usually occurs within 8 to 10 seconds.

The most sophisticated system is the SPY Elite (Novadaq-Stryker) intraoperative perfusion assessment system. This provides a camera, laser, computer hardware, and assessment software—SPY-Q (Novadaq-Stryker) software—for additional viewing options which allow for fluorescence differential analysis. The final result is a series of captured images highlighting differential zones of perfusion. Most ICG-based technologies use both gray scale ( Fig. 1a ) and colorized images ( Fig. 1b ) for assessment. In gray-scale imaging, darker gray to black indicates poor perfusion, while increasingly white areas represent good perfusion. Similarly, on the colored image, scale traversing from blue to red represents poor to good perfusion. Quantification of the percentage fluorescence within an image using the SPY-Q software can also be performed, yielding relative or absolute objective values ( Figs. 1b , 2a , and 3a ). This data can be used intraoperatively to make flap revisions, resect tissue, or delay the procedure. Unfortunately, such quantification is available only with the floor-mounted SPY camera (Novadaq-Stryker) running the SPY-Q software. Overall, the utility of intraoperative ICG assessment has been well demonstrated in many retrospective and prospective studies, showing decreased rates of postoperative necrosis and flap loss when compared with clinical judgment alone. ^{6 8 11 12 14 15 16 18 19 20 21 22 23 24 25} Clinical applications of ICG perfusion analysis in plastic surgery are numerous and include the following: evaluating the viability of skin and nipple sparing mastectomy flaps, pedicled flaps, and free flaps in breast surgery. Unfortunately, the devices can tend to over-read potential necrosis, which may lead to higher resection rates than are actually necessary. ^{6 26} A more recent application of ICG technology has been observed in the realm of sentinel node evaluation and microlymphatic surgery. ICG is readily taken up by the lymphatics and flows through the lymph nodes when injected into the breast skin or between the digits in the upper or lower extremities. The dye can be traced in the subcutaneous tissue to outline lymphatic channels for anastomosis or guide oncologic surgeons to sentinel nodes for biopsy. ^{27 28}

Fig. 1.

( a ) Intraoperative ICG analysis in gray scale revealing perfusion defects adjacent to the incision after skin sparing mastectomy and immediate expander-based breast reconstruction. ( b ) Colorimetric overlay with relative perfusion metrics quantitating the intraoperative perfusion deficit predictive of necrosis. ( c ) Associated breast postoperatively demonstrating full-thickness necrosis in the corresponding regions of poor perfusion.

Fig. 2.

( a ) Nipple sparing mastectomy (NSM) flap intraoperatively evaluated with ICG angiography displaying a well perfused nipple. ( b ) NSM from Fig. 2a assessed with our hyperspectral imaging device displaying decreased nipple tissue oxygen saturation. Minor epidermolysis occurred.

Fig. 3.

( a ) Intraoperative ICG angiography of NSM with perfusion defect laterally predictive of necrosis. ( b ) Intraoperative hyperspectral image of NSM in Fig. 3a revealing adequate tissue oxygen saturation over region of limited perfusion. ( c ) Hyperspectral image of the same NSM at 6 days' follow-up in the office. Tissue oxygen saturation stable relative to intraoperative assessment. Tegaderm bra in place. ( d ) NSM at 2 weeks' follow-up. Minor epidermolysis and bruising laterally in region correlating with lowest absolute perfusion in Fig. 3a . However, full-thickness necrosis did not occur as predicted by ICG angiography.

Indocyanine Green Imaging Devices

Indocyanine green devices currently available include but are not limited to Novadaq-Stryker's SPY and SPYPHI, PDE (Hamamatsu Photonics), Fluoptics (Fluoptics Inc.), and Photonvue (licensed from Fluoptics by Invuity). Our experience has been exclusively with the SPY Elite floor-mounted device. The other four devices above are handheld and do not allow for quantification of percentage fluorescence.

Clinical Outcomes

Our practice performs ICG perfusion analysis on all primary implant- and autologous-based breast reconstructions. ICG data was collected for a 15-month period from 2008 to 2009. One hundred and six patients were evaluated. Procedures included 8 free transverse rectus abdominis muscle (TRAM)/ deep inferior epigastric perforator (DIEP) flaps, 41 pedicled TRAM flaps, 4 latissimus dorsi flaps, 68 tissue expander and implant operations, and 21 breast reductions. Six patients experienced surgical site complications, of which three were associated with TRAM flaps. Intraoperatively, clinical assessment suggested adequate perfusion, which contradicted the SPY data. These areas were not resected or debrided based on clinical grounds. All three of these questionable TRAM flap patients later developed necrosis in the areas of poor perfusion, as predicted by the ICG perfusion imagery. The other three skin flap complications occurred following expander-based breast reconstruction ( Fig. 1a, b ). Again, skin was preserved on clinical grounds, contradictory to ICG perfusion analysis. All three patients developed necrosis requiring excision and re-closure ( Fig. 1c ). One expander was lost.

Indocyanine green-based technology has been invaluable in assessing skin viability in our prepectoral direct-to-implant breast reconstructions. Our most recent study included 50 patients for a total of 73 breasts. As part of our protocol, we assess skin perfusion with the temporary sizer in the prepectoral plane beneath the fresh mastectomy flap. Our decision to proceed depends on the viability of the overlying skin covering the implant size of choice. The surgery is converted to an expander-based reconstruction or closed primarily if inadequate perfusion was revealed intraoperatively with the appropriate sizer in place. In our preliminary series, one breast had a region of tenuous perfusion on ICG analysis requiring intraoperative resection; the patient healed without complication. In addition, six breasts were presented postoperatively with minor delayed wound healing; however, there were no cases of full-thickness skin necrosis. ²

Indocyanine Green Limitations

The limitations associated with ICG perfusion analysis should not be overstated in comparison to its utility; however, the most significant drawback is its exorbitant cost. After 10 years of availability and only 25% market penetration, it is clear that a majority of surgeons do not have access to this technology. The most widely used system, SPY Elite (Stryker-Novadaq), is costly to own despite innovative leasing or placement programs. The high cost of disposables per case is also difficult to justify for many hospital systems. In addition, use of ICG requires intravenous access for dye administration, and it cannot be used in patients with severe iodine allergies. Finally, the system is bulky with limited portability. These limitations hamper its mobility, rendering it impractical for an office setting. The evolving handheld devices are a response to these challenges but cost remains a significant issue as does the mandatory need for IV access.

Moreover, sequential monitoring is restricted by the clearance of the dye. ^{2 7} Random binding of ICG to vascular endothelium and venous congestion can cause pooling of ICG and limit clearance from the bloodstream. This interferes with the ability to visualize first pass contrast, and potentially obscures subsequent intraoperative assessments of perfusion. Numerical outputs used to augment clinical judgment can also be falsely increased by this phenomenon, resulting in false negative results and potentially negative outcomes.

Finally, perfusion is not a surrogate for tissue oxygenation ( Figs. 2 and 3 ). The ICG injected into a patient's bloodstream will course through a patent vessel regardless (to some extent) of how much oxygen is bound to the hemoglobin within it. This is particularly concerning as the wound edge becomes increasingly distant from the perforator supplying a given perforasome. ¹⁷ Each zone of tissue between the perforator and the wound edge is extracting oxygen and nutrients, as blood courses to the periphery of the flap. In the interim, ICG remains protein bound intravascularly and is exclusively metabolized by the liver. Consequently, the fluorescence seen throughout a well-perfused mastectomy flap and wound edge in ICG angiography provides no direct information about oxygenation of the same tissue. This oxygenation versus perfusion conundrum leads us into an overview of our next technology.

Hyperspectral Imaging

Hyperspectral imaging is a near-infrared imaging technology that has been, until recently, largely restricted to the research realm. This method provides measurements reflecting the proportion of oxygenated versus deoxygenated hemoglobin within the tissue being visualized. It measures and analyzes these proportions in such a manner that tissue perfusion is also evaluated. ⁷ Moreover, it samples both the static and pulse modulated components, enabling it to provide a more representative and accurate estimate of the oxygen supply at the microvascular level. ²

The senior author (G.E.J.) has had the opportunity to trial the device marketed by Kent Imaging Inc. The device is a handheld camera which provides an intuitive color image assessment overlay of tissue oxygen saturation ( Figs. 2 and 3 ). The outputs provided by the software reflect the absolute value of tissue oxygenation as a percentage in a clinically meaningful value for the surgeon. Compared with ICG laser fluorescence devices, hyperspectral imaging equipment is substantially cheaper, does not have any disposables, does not require invasive injection, and can be used in sequential monitoring at any time interval

Rapid cycling of the camera allows it to monitor saturation changes occurring in a matter of seconds to minutes, whereas ICG-based assessments require clearance of the dye which may take up to 20 minutes. Due to its convenient portability, we have used this device in the OR, at the bedside, and in the office ( Fig. 3 ). The authors have no financial interest in the device.

Two recent analyses of near-infrared spectroscopy using adherent skin probes have been recently published. In the first study published in 2016, Chen and colleagues reported a sensitivity and specificity of 100% with false positive and false negative rates of zero. ⁷ Kagaya et al published a more recent review in 2018 reporting similar results of 99.1 and 99.9% sensitivity and specificity, respectively. ³⁰ Both papers conclude that near-infrared spectroscopy are more reliable to detect vascular compromise during postoperative flap evaluations, and are able to do so earlier than conventional methods such as physical examination or Doppler monitoring. Such accurate evaluation also led to higher reported salvage rates. The studies presented in these reviews evaluated different commercial platforms, utilizing a variety of sensor modalities on more localized measurements. However, promising data on hyperspectral imaging was published earlier this decade, demonstrating the same quantitative metrics over larger fields of view. ^{31 32} The problem with these adherent devices is the need to glue the probe to dry skin and the high cost of the probes and their parent devices. The most widely used continuous near-infrared spectral monitoring device in the USA is the Vioptix system (Vioptix, Inc.), which we have successfully used to monitor free cutaneous flaps in breast reconstruction. These devices can visualize tissue chromophores to a depth between 10 and 20 mm. ⁷

Hand-held hyperspectral devices, by contrast, rely on more superficial reflectance in the 1- to 3-mm range at most. They provide sensitive tissue oxygen saturation measurements which correlate extremely well with both ICG-based assessments of perfusion as well as clinical evaluation. The device allows for multiple, sequential imaging from a wide variety of angles and the results are almost instantaneous. Our experience with the Kent device has been highly encouraging. Its sensitivity and specificity seem very similar to that of ICG-based devices, without the high hardware costs, need for IV access, or ongoing costs of disposables. The hyperspectral device appears to be accurate and can be used repeatedly within a short time frame to facilitate clinical decision-making. Refinements in the software allow for melanin correction in darker skins and the potential use of a rapid-cycling mode to achieve a video-like capture similar to that used in ICG-based technologies. Anecdotally, the incidence of false readings is gratifyingly low, and its ease of use and significantly lower relative cost makes this a very desirable technology in clinical practice.

Conclusion

The incorporation of ICG angiography and hyperspectral imaging devices into clinical practice make the adoption of direct-to-implant, immediate pre-pectoral, and autologous breast reconstruction safer and more predictable. These technologies provide objective evidence that can assist in intraoperative decision-making, leading to better patient outcome. The use of ICG angiography and hyperspectral imaging should augment, not supersede, the clinical assessment.