Changes in Skin Vascularity in a Murine Model for Post-Mastectomy Radiation

Abstract

BACKGROUND

Post-mastectomy radiation causes persistent injury to the breast microvasculature and the prevailing assumption is that longer delays prior to breast reconstruction allow for recovery of blood supply. This study utilizes a murine model to examine the effects of radiation on skin vascularity to help determine when radiation-induced effects on the microvasculature begin to stabilize.

STUDY DESIGN

Isogenic Lewis rats were divided into two groups: XRT (radiation therapy) (n=24) and Control (n=24). XRT rats received a breast cancer therapy human dose-equivalent of radiation to the groin, whereas Control rats received no radiation. Animals were sacrificed at 4, 8, 12, and 16 weeks after completion of radiation. The vasculature was injected with Microfil and groin skin was harvested for radiomorphometric analysis by micro-computed tomography. ANOVA with post-hoc Tukey tests was utilized to determine significance between groups.

RESULTS

Augmentation in vascularity was observed in the XRT group at 4 weeks following radiation compared to the Control group (p= 0.045). Vessel number was decreased at 12 weeks (p= 0.002) and at 16 weeks (p= 0.001) in the XRT rats compared to Control rats. Vessel separation in the XRT group was higher than in the Control group at 12 weeks (p= 0.009) and 16 weeks (p= 0.001). There was no change in vessel number and separation between weeks 12 and 16.

CONCLUSIONS

A period of augmented skin vascularity is seen after radiation injury followed by decreased vascularity which demonstrates stabilization at approximately 12 weeks in this murine model. This model can be used to further study breast flap vascularity and the optimization of the timing of delayed breast reconstruction.

INTRODUCTION

Breast cancer is the most common malignancy and second most common cause of cancer death for women in the United States(1) Post-mastectomy radiation therapy is an essential component of multimodal treatment of breast cancer in appropriately selected patients and it has been shown to improve disease-free as well as overall survival rates and decreases local recurrence (2–7). Although the benefits of post-mastectomy radiation therapy have been well established, the deleterious effects to the radiated field present challenges when considering breast reconstruction.

Radiation-induced soft tissue injury is associated with higher rates of surgical morbidity, including poor wound healing, infection and chronic pain (8–14). Based on well-recognized clinical observations of the pernicious effects of radiation on both prosthetic and autologous breast reconstruction, some have recommended a delay period before performing breast reconstruction (15). Although it has been suggested that waiting longer periods of time from completion of radiation could be beneficial, there is little available data to guide reconstructive surgeons as to how long one should wait.

Murine models for radiation injury have been developed and used to study multiple pertinent clinical questions outside of breast reconstruction (16–18). Thanik et al (16) reported progressively worsening blood flow to ventral abdominal skin after radiation by laser Doppler imaging over a six week observation period. Similarly, Deshpande et al (18) found a significant degradation of mandibular vessel quality and size after radiation with use of micro-computed tomography. However, limited work has been done to specifically evaluate changes in the soft tissue microvasculature following radiation in relation to subsequent breast reconstruction.

This study utilizes a murine model to examine the effects of radiation on skin vascularity in order to help determine when radiation-induced effects on the microvasculature begin to stabilize. Though some clinicians have suggested delays in microsurgical breast reconstruction for up to a year after radiation to allow time for recovery (19), we posit that such extended delays are not warranted and deprive women of a more prompt reconstruction. We hypothesize that the damage to the skin vasculature from a human equivalent dose of a breast cancer irradiation protocol will stabilize at a point less than 4 months. Furthermore we posit that a truncated time point of vascular stabilization may then serve to guide the optimal timing of delayed breast reconstruction.

MATERIALS AND METHODS

Adult male isogenic Lewis rats (mean weight 350g) were obtained through our institution’s University Lab Animal Medicine (ULAM) department in compliance with their sub-division of the University Committee on the Use and Care of Animals (UCUCA). Rats were acclimated a minimum of 7 days in a light and temperature controlled housing facility prior exposure to radiation. They were also fed a standard hard chow and water without restriction during the acclimation period. The animals were then randomly assigned to Radiation (n= 24) or Control (n= 24) groups (Figure 1). Animals in the radiation group were exposed to radiation after the acclimation period while animals in the control groups received no radiotherapy after acclimation.

Figure 1.

Schematic showing the experimental groups. Animals in the XRT group received radiation while animals in the Control groups not radiated.

Radiation Treatment

Prior to each administration of radiation, rats were anesthetized with inhaled isoflurane in a closed chamber. Post-mastectomy radiation therapy typically consists of 50–60 Gy administered as 2Gy fractions over 5 to 6 weeks (20). Through collaboration with the Department of Radiation Oncology at our institution’s Comprehensive Cancer Center, we calculated a post-mastectomy human dose-equivalent radiation regimen of 5 fractions of 5.6Gy/fraction (total 28Gy) using previously described methodology (21, 22). Fractions were administered once daily over five consecutive days at a dose rate of 147.7cG/minute using a Philips RT250 orthovoltage unit (250 kV X-rays, 15 mA; Kimtron Medical, Oxford, CT). Radiation was delivered in a radiation chamber with the rat covered with a protective lead shield with only the lower abdominal area exposed for XRT. Dosimetry was carried out using an ionization chamber connected to an electrometer system which is directly traceable to a National Institute of Standards and Technology calibration.

Perfusion and Quantitative Vascular Analysis

At the end of each time point (4, 8, 12, and 16 weeks), rats in the XRT and Control groups were sacrificed and Microfil perfusion was performed. Animals were anesthetized with inhalational isoflurane (4%). Subsequently, a thoracotomy and left ventricular catheterization was performed. Following vessel fixation with heparinized normal saline and buffered formalin solution, the vasculature was injected with Microfil MV122 (Flow Tech; Carver, MA) as previously described (18). After allowing the vascular resin to cure for 12 hours at −20°C, a 3 × 3 cm region of skin from the left groin was harvested for micro-computed tomography (CT) angiography.

The microvasculature was quantified by radiomorphometric analysis of micro-CT images. Using MicroView ABA 2.2 software (GE Healthcare, Milwaukee, WI), scans were reconstructed and reoriented in a 3-dimensional x, y, and z plane. The 3×3 cm biopsy from the left groin was selected as the region of interest (ROI) and the scans were cropped for quantitative analysis. Analysis of vascularity in the ROI was accomplished by setting a global grayscale threshold of 350 to differentiate vessels from surrounding tissue. MicroView uses algorithms to assign relative densities to the voxels based on the contrast content within the vessels. MicroView reported four metrics, all of which control for the size of the ROI: vessel volume fraction, vessel number, vessel thickness, vessel separation. Vessel volume fraction (VVF) represents the fraction of specimen occupied by the volume of vessels within the ROI tissue. Vessel number (VN) represents the mean number of vessels encountered by stochastic 1-mm lines in the ROI. Vessel thickness (VT) represents the intraluminal diameter of the vessel and is reported in mm. Vessel separation (VS) is the mean distance between the mid axes of two vessels within the ROI and is reported in mm.

Statistical Analysis

Statistical analysis was performed using SPSS Statistics software V. 20 (IBM; Armonk, NY). The data were compared using one-way analysis of variance (ANOVA). Two tailed t-tests were also performed in order to compare the XRT and Control groups at each specific time point. Post-hoc analysis was performed by either Tukey or Games-Howell method, depending on the homogeneity of variances. Significance was defined as p < 0.05.

RESULTS

All of the XRT animals tolerated the radiation administration without significant complications and all animals in both groups survived to their designated time points. However, not all of the vascular perfusions were successful, decreasing the number of samples available for evaluation in some of the groups with the final number of samples per group presented in Table 1.

Table 1.

Number of animals in the XRT and Control groups with successful vascular perfusion and micro-CT evaluation.

Study Time Point	XRT (n)	Control (n)
4 week	5	5
8 week	5	5
12 week	5	5
16 week	6	4

Evaluation of the representative maximal intensity projections (MIP) in the study groups revealed an increase in vascular density within the region of interest when non-radiated controls were compared to radiated skin at 4 weeks (Figure 2). The vascular density in radiated skin at 12 weeks was noticeably decreased in comparison to controls. Radiomorphometric analysis of the mean vessel separation in skin samples was consistent with the trends observed from the MIPs (Figure 3). A significant decrease in vessel separation (VS) was observed in radiated skin at 4 weeks compared to controls (0.628 ± 0.060 mm vs 0.806 ± 0.155mm, p= 0.044). There was no difference in the vessel separation between XRT and Control groups at 8 weeks (0.767 ± .089 mm vs 0.868 ± 0.290 mm, p= 0.478). Vessel separation increased in radiated skin at 12 weeks (1.177 ± 0.262 mm vs 0.751 ± 0.081 mm, p= 0.009) and 16 weeks (1.233 ± .075 mm vs 0.805 ± 0.069 mm, p= 0.001) relative to controls. The greatest increase in VS was from 8 weeks to 12 weeks with a 53% increase (p= 0.007). There was no difference in the vasculature of radiated skin at 12 weeks and 16 weeks (p= 0.999).

Figure 2.

Representative Maximal Intensity Projections (MIP) of non-radiated skin (left), radiated skin at 4 weeks (middle) and 12 weeks (right).

Figure 3.

Vessel Seperation (VS) stereologic metrics showing a gradual increase in the mean vessel separation over time in radiated animals, with stabilization after 12 weeks. ** indicate p<0.05 comparing mean values of VS in XRT animals between observation time points. * indicate p<0.05 between XRT and Control at each time point.

An increase in the mean vessel number (VN) was observed in radiated skin specimens relative to controls at 4 weeks (1.5573 ± 0.172 mm⁻¹ vs 1.247 ± 0.236 mm⁻¹, p= 0.045) (Figure 4). No difference in VN was observed between XRT and Control groups at 8 weeks (1.294 ± 0.166 mm⁻¹ vs 1.2375 ± 0.422 mm⁻¹, p= 0.785). The mean VN decreased in radiated skin at 12 weeks (0.8686 ± 0.164 mm⁻¹ vs1.3190 ± 0.150 mm⁻¹, p= 0.002) and 16 weeks (0.7989 ± 0.045 mm⁻¹ vs 1.2056 ± 0.086 mm⁻¹, p= 0.001) in comparison to non-radiated controls. Similar to vessel separation, the greatest decrease in vessel number occurred between 8 and 12 weeks, with a 49% decrease (p=.043). There was no change in the VN between 12 week and 16 week time points in the XRT group (p=0.969).

Figure 4.

Vessel Number (VN) stereologic metrics showing an increase in the mean vessel number followed by a gradual decrease and stabilization after 12 weeks in XRT animals. ** indicate p<0.05 comparing mean values of VN in XRT animals between observation time points. * indicate p<0.05 between XRT and Control at each time point.

In evaluating vessel volume fraction (VVF), radiated specimens at 4 weeks had a significantly greater mean VVF than similar specimens at 8 weeks (p=0.003), 12 weeks (p=0.000) and 16 weeks (p=0.007). No significant differences were found in comparisons between radiated specimens at 8, 12 and 16 weeks. Similarly, in assessing vessel thickness (VT), no differences were found between radiated specimens and controls, and between study time points. Radiomorphometric analysis of skin showed no statistical difference between the control groups at all of the time points for all vascular metrics.

DISCUSSION

The principle of delaying breast reconstruction after radiation therapy allows for acute radiation injury to subside. It is well established that radiation therapy confers significant morbidity in the setting of both implant and autologous breast reconstruction (23, 24). As a result of a paucity of data demonstrating an optimal time to perform delayed breast reconstruction, plastic surgeons have recommended various arbitrary periods of delay in their practices, ranging in general from 6 months to 2 years. However, because radiation-induced injury does not have a clearly defined end point, waiting for extended periods of time may not provide a clinical benefit to patients. For example, some vascular sclerotic changes seen in coronary artery disease are thought to in fact, worsen over time after radiation therapy (25). Because no studies to date have attempted to determine an optimal time to perform delayed breast reconstruction, development of a murine model that depicts microvascular changes within the skin after radiation therapy is an important first step in determining when radiation-induced injury begins to stabilize.

In this investigation, radiated skin exhibited a decrease in vessel separation and an increase in vessel number and vessel volume fraction relative to non-radiated skin at four weeks after radiation. These changes reflect an initial increase in the vascular density early on after radiation (Figure 2), suggesting a hyper-vascular response. A recent study on the effect of low-dose radiation on ischemic cutaneous flaps, demonstrated an upregulation of angiogenic chemokines, an increase in systemic progenitor cell mobilization (26). Skin perfusion and vascularity were found to increase on laser Doppler and whole mount evaluation respectively. This described pathway may explain the early vascular changes we observed, however this intial trend has not been reported with higher doses of radiation. One explanation may have to do with the radiation regimen employed in this study. As opposed to a single large dose of radiation described in many animal models, smaller fractionated doses of radiation were delivered to mimic the fractionated radiation delivery utilized in clinical practice. The individual fractionated doses delivered in our study (5.6 Gy) were very similar to the single 5 Gy radiation dose delivered in above study by Thanik et al. (26).

The hyper-vascular changes trend downward and at eight weeks after radiation, the skin vascularity in the XRT group is similar to controls. This downward trend in the skin vascularity continues to 12 weeks after radiation exposure at which time the radiated skin demonstrates an increase in vessel separation and a decrease in vessel number when compared to controls. Though not statistically significant, an overall downward trend in the VVF from 8 to 12 weeks is appreciated and might reach significance with improved power. Taken together, these three metrics suggest a hypo-vascularity in the skin relative to baseline, consistent with findings from studies on radiated murine mandibles (27). This decrease in vascularity is likely driven by apoptosis of hematopoietic cells within the skin exposed to radiation (28). Preservation of vascularity studied with similar techniques in radiated murine mandibles has been shown to result in improved fracture healing (29). On further observation at 16 weeks, the stereologic metrics for vessel number and vessel separation are unchanged from the findings at 12 weeks (figures 3 and 4). This suggests that the pathologic effect on the skin vasculature is stabilizing.

The microvascular changes observed in this study parallel some findings appreciated on clinical examination of skin following radiation. The initial hypervascular response likely coincides with acute clinical findings of hyperemia observed early on after PMRT. Hyperemia ultimately subsides over weeks to months and less obvious changes in the skin and soft tissue persist. The findings from this study demonstrate that there is a point in time following radiation therapy when stabilization of microvascular injury occurs. Understanding that the skin vascularity will likely never return to its pre-radiation state, we postulate that reconstruction at any time after this critical time point of stabilization would be associated with minimal additional morbidity. A retrospective review of a single center clinical experience with delayed autologous reconstruction by Momoh et al., found that there were no significant differences in success rates and complications when reconstructions were performed before or after 6 months following PMRT (30); it should be noted that very few reconstructions were performed within 1 to 3 months of radiation given the acute skin injury typically encountered during that time period. Similar results were also found when complication rates were examined in reconstructions performed at a time before or after 12 months following PMRT. These findings led to a conclusion that it is likely safe to proceed with reconstruction as early as 3 to 6 months after radiation.

This study was limited by the fact that only radiation-induced changes to the skin microvasculature were examined. In clinical practice, the effects of post-mastectomy radiation to the surrounding chest wall muscles and subcutaneous fat are also relevant to the timing of delayed breast reconstruction. The specific time period for observed changes in animals would differ from what is observed in radiated patients and as such cannot be directly applied to clinical cases. A study on the effect of radiation on skin vascularity in breast cancer patients is currently underway. Furthermore, in regards to delayed free flap breast reconstruction, damage to radiated recipient vessels such as the internal mammary vessels undoubtedly plays a major role in the execution and outcomes of microvascular anastomoses. By establishing a murine model for post-mastectomy radiation, these other important factors can now be investigated as well. Another limitation is that the study period might not have been long enough to capture additional recovery or worsening of the microvasculature beyond the observed stabilization. Though the observation period in this study is much longer than others (16–18) investigating the effects of radiation on tissue, additional studies with longer observation periods will be needed to confirm that no further vascular deterioration or improvement occurs after stabilization.

CONCLUSIONS

A murine model for post-mastectomy radiation is described to examine the microvascular effects of radiation within the skin. An initial period of hyper-vascularity was observed at 4 weeks and this may be consistent with an acute inflammatory reaction resulting from the radiation. Over time, animals that received radiation demonstrated decreased skin vascularity which went on to stabilize at approximately 12 weeks. This model can now be used to further study optimization of the timing of delayed breast reconstruction and investigate potential interventions directed at improving mastectomy skin flap vascularity prior to translation to human subjects.

Figure 5.

Vessel Volume Fraction (VVF) stereologic metrics showing an initial increase in the mean VVF and subsequent decrease at 8, 12 and 16 weeks after radiation. ** indicate p<0.05 comparing mean values of VVF between observation time points. * indicate p<0.05 between XRT and Control at each time point.