There are important histopathologic distinctions among different scarring processes that may guide prognosis and management. However, distinguishing a hypertrophic scar from a keloid histopathologically is sometimes very difficult.⁴ Keloids and hypertrophic scars both differ from normal skin as they have increased deposition of connective tissue as well as rich vasculature.⁵ In the literature there is conflicting information regarding the vasculature of keloids and hypertrophic scars. Many theories suggest the role of hypoxia in the formation of pathologic scars. Kischer et al.⁶ found an increased number of occluded or partially occluded

microvessels in hypertrophic and keloidal scars in comparison to normal scars and normal dermis because of endothelial cell swelling and increased endothelial cell density. Additionally the degree of vascular occlusion was higher in keloids with fewer and more flattened vessels, especially centrally within the lesion, suggesting less vascular supply in a keloid as compared with a hypertrophic scar. The growth of fibroblasts and increased collagen deposition may result in this occlusion.⁷,⁸ On the contrary, other studies have demonstrated increased vasculature of keloids and hypertrophic scars in comparison with normal scars and normal skin.¹ Amadeu et al.⁴ reported that the volume occupied by vessels in the papillary dermis was 79.7% higher in hypertrophic scars (P < 0.01) and 62.5% higher in keloids (P < 0.05) as compared with normal skin. Similar findings were noted in the reticular dermis, with hypertrophic scars having 62.9% higher vessel volume (P < 0.025) and keloids having 68.5% higher vessel volume than normal skin (P < 0.001). The benefit of pressure dressings and vascular lasers for the treatment of these pathologic scars contradicts our understating of the role of hypoxia in their pathogenesis. At this point, further research needs to be conducted to fully understand this interplay.

Aside from their vasculature, keloids and hypertrophic scars each have a high mesenchymal cell density and inflammatory cell infiltration, with an absence of subepidermal appendages including sebaceous glands and rete ridges.¹ Both lesions have a predominant active fibroblast cell type, but keloids tend to have more quiescent forms.⁹ The dermis of a keloid comprises characteristic large, thick collagen bundles that are brightly eosinophilic and hyalinized with hematoxylin and eosin staining as demonstrated in Figure 5-3. These bundles are closely packed with thin fibrils and form the characteristic “keloidal collagen.” Lee et al.¹⁰ reported distinct keloidal collagen in approximately 55% of all keloids included in their study. They concluded that the histologic absence of keloidal collagen does not rule out the diagnosis of a keloid, but its presence is of high diagnostic value as it is only discovered in keloidal scars. Keloidal collagen is arranged in a nodular fashion in contrast to the parallel architecture of collagen in ordinary scars. In comparison, hypertrophic scars contain distinct smaller nodules present in the dermis (demonstrated in Fig. 5-4), aiding in the histologic distinction from keloids and normal scars, which often lack these nodules. The compaction of the collagen bundles in keloids has no effect on the overlying epidermis.⁷,¹¹ Additionally, hypertrophic scars have an accumulation of myofibroblasts expressing α-smooth muscle actin, whereas keloidal myofibroblasts often lack the expression of α-smooth muscle actin.⁵

In contrast to keloids and hypertrophic scars, the histologic appearance of striae can be very subtle and may be difficult to distinguish from normal skin (Fig. 5-5). As with other scar types, the histologic appearance of striae depends on the stage of evolution. In earlier (younger) striae, there is often dermal edema with perivascular infiltrates. The epidermis can appear normal in earlier stages, but later may flatten and have blunting of the rete ridges (atrophy). Earlier striae may also demonstrate elastolysis and a relative lack of mast cells. There are structural changes seen in collagen bundles with prominent fibroblasts and reduced microfibrils. Earlier striae have been referred to as striae rubra, as they often appear erythematous in color.¹²

Well-established striae may be termed striae alba, owing to their hypopigmented (white) appearance over time. Striae alba demonstrate epidermal atrophy and decreased dermal thickness. Collagen fibers are generally found running parallel to the skin and transverse to the direction of the individual stria. There are variable alterations in the elastic fibers with reduced and fragmented dermal elastin on special stains. There is also loss of skin appendages including hair follicles and adnexal structures, as with all scars.⁹

There are very few studies that examine the histologic findings associated with clinical dyspigmentation in scars. However, Travis et al.¹³ evaluated the optical and histologic properties of wounds and their relationship with dyspigmentation in a porcine model. Hyperpigmented scars, hypopigmented scars, and uninjured tissue were evaluated by fixing and embedding these samples for histologic examination using Azure B stain and primary antibodies to
S100B, HMB45, and α-melanocyte-stimulating hormone (α-MSH) for comparison. They discovered no statistically significant difference in melanocyte number between hyperpigmented and hypopigmented scars and uninjured skin samples. There was, however, a statistically significant difference in the amount of melanin and α-MSH, and immunohistochemical evidence of stimulated melanocytes in hyperpigmented versus hypopigmented scars.

Whereas there has been extensive research conducted on certain scar types, elucidation of the relationship between clinical scar appearance and the corresponding histology may help guide treatment and evaluation of the response. For example, Ozog et al.¹⁴ delineated the histologic changes
in collagen typing between laser-treated and -untreated burn scars. It is well described that normal skin contains a combination of type I and type III collagen; fetal skin contains a higher proportion of type III collagen, and burn scars contain a higher proportion of type I to type III collagen. A course of three ablative fractionated CO₂ laser treatments to burn scars eventuated in a collagen profile approaching that of normal skin, with a posttreatment increase in type III collagen as demonstrated by the Herovici stain (Fig. 5-6). Additionally, Taudorf et al.¹⁵ found statistically significant clinical improvement in various scar types (normal, hypertrophic, and atrophic) after three monthly nonablative fractional laser treatments (P < 0.0001 vs. untreated), with corresponding histology indicative of collagen remodeling. There is a predominance of thickened collagen within a scar as compared with the normal surrounding dermis. The healing process that follows fractionated photothermal injury ultimately leads to the remodeling and reorganization of collagen that begins to approach that of normal skin. A later study completed by Connolly et al.¹⁶ discovered that, counterintuitively, treatment of erythematous burn scars with a fractionated CO₂ laser led to a statistically significant increase in vascular density as determined by anti-CD31 immunostaining, despite a decrease in clinically apparent erythema during the treatment course (Fig. 5-7). These histologic findings as illustrated above have both supported and challenged our understanding of the mechanisms leading to clinical improvement in scars. Perhaps we may use these predicted and unforeseen findings to further propel scientific discovery and improved management.

To date, light microscopy of tissue has been the principal technique to evaluate the histology and microstructure of the skin. In vivo reflectance confocal microscopy (RCM) is a relatively new, noninvasive technology that has recently been utilized for the evaluation of various types of scars.¹⁷,¹⁸,¹⁹,²⁰ RCM provides real-time viewing with similar
resolution to classical histology, without tissue damage (Fig. 5-8). Additionally, it may be used to detect the dynamic microscopic changes of a particular skin lesion over time. RCM may provide promising in vivo evaluation and comparison of many scar types, and may also be used for the pre- and posttreatment analyses of scars, among other dermatologic conditions.

Figure 5-8 provides a clinical comparison of a hypertrophic scar with the associated confocal microscopic imaging. At a depth of 152.4 µm below stratum corneum, collagen fibers and bundles (red arrow) are present inside the lesion since it is raised compared with adjacent normal skin.