Histological Validity and Clinical Evidence for Use of Fractional Lasers for Acne Scars

Substrate studied

Though fractional lasers are widely used for facial[25–28] dermatological indications, there are only two studies that studied the histological effects of lasers on the facial skin.[2029]

The other studies where facial skin was studied were either not formally directed towards histological assessment[18] or could not demonstrate a marked clinical improvement[20] even though there was a histological increase in collagen and elastin.

Another issue is that the substrates used in studies are ex vivo models[124–68–18] and their replicability on facial skin is doubtful [Figure 2]. Extrapolating data from studies done for rhytides[25–27] are not useful as the dose and depth required are less as compared to acne scars. Also, in rhytides,[101215] the periocular area is targeted, which is not the site for acne scars. Some studies employed a two-step procedure where the histology was done on ex vivo animal or ex vivo human model,[101215] and the clinical evaluation was done on the in vivo facial skin. This is not scientifically replicable as there are differences in the tissue responses in different substrate models.[29] Some studies have tried to simulate the in vivo characteristics by maintaining the temperature of the ex vivo skin samples at 37°C but in spite of stringent quality controls, it is difficult to replicate a live human tissue.[6]

A commonly used substrate is the abdominoplasty tissue[912–141629] which has been used in both in vivo and ex vivo settings. This tissue has anatomically a thick (almost 3 cm) fibroadipose tissue, which is not representative of the histology of the facial skin.[18] Thus the laser tissue dynamics cannot be extrapolated to the facial skin. The marked differences between the facial skin and abdominal skin in terms of moisture content, sebaceous glands and vascularity affect the histological depth achieved on the facial skin while using settings derived from the extrafacial skin. Instead of dispersing energy in a straight line from the epidermis to the dermis, the laser energy probably travels laterally along blood vessels, diverges around sebaceous glands, or possibly travels down a hair shaft without damaging the surrounding skin.[29] Thus the depth achieved on the facial skin is less as compared to the abdominal skin by a factor of 28%.[29]

The importance of this is that the depth of penetration provided by various laser manufacturers does not account for this and thus the dosimetry suggested is probably not reproducible for use in facial acne scars.

Even if we factor in the accepted variability of depth,[29] there is very little in vivo data on the histological depth of the normal facial skin which are crucial to extrapolate the variable depth generated by various fractional lasers.[1–20] It has been estimated that the facial skin depth (forehead, nose, medial and lateralcheeks, lips and chin) is about 2196 μm, which is composed of the epidermis (105 μm), papillary dermis (105 μm) and reticular dermis (1986 μm).[15] The predicted depth achieved (2–5 mm) with ablative fractional lasers[11131828] on abdominoplasty tissues may not be clinically applicable or relevant on the facial skin. Moreover, any AFR with a depth beyond 2200 μm (ex vivo substrate) will have no additional advantage in real in vivo scenarios, as on the face, it will exceed the total skin depth (2196 μm).

Time of biopsy

The time of biopsy also varies in studies ranging from zero to few hours to 24 h after the laser sittings. A few studies[452026] have also assessed the tissue after 2–3 months when the collagen remodelling takes place which is more useful as it helps to assess the end result achieved. A recent study though has found that the histological changes can be seen up to 6 months to 1 year after the procedure.[30] Thus, ideally a biopsy at 6 or 9 months would give a more relevant picture of the extent of neocollagenisation which has a role in acne scar improvement. The drawback of repeated biopsies though is that the same area cannot be biopsied each time, and thus, the second biopsy can never replicate the exact tissue response. The laser that we use (Dermablate 2940, microspot mode Ascepelion) has a demonstrable effect on collagenisation and a tissue response up to a depth of 1500–2000 μm (data provided by the company) which is evident only after 2–3 months.

Storage

This varies from freezing to storing the tissue in formalin. Paraffin-embedded tissue sections require dehydration of the tissue samples, which causes shrinkage. This is an important point to consider when evaluating the microcolumn lesion depth and width.[1–20]

Sectioning and depth assessment

The microcolumn injury produced has roughly a conical shape. Thus a difference in sectioning angles may change the interpretation of the depth.[111316] To circumvent this, a laborious technique is to take multiple serial sections so that an average tissue depth/diameter can be assessed. Another way – by no means easy – is to perform treatments in triplicate to provide an accurate representation of the tissue injury at a given treatment parameter.[13] The depth/diameter of an MTZ has also variable assessment methods. In a study by Hantash et al.,[514] the maximum dermal ablation width of the laser channels was used whereas Skovbølling et al.[16] used the width at a defined localisation at the interface of the lower one-third and the upper two-thirds of the microcolumn. To complicate matters, another study[6] determined the depth of each MTZ by measuring the distance from the stratum corneum level to the deepest non-viable cell within the confines of any lesion. This lack of homogeneity makes comparison of studies difficult.

It is difficult by vertical sectioning to accurately locate the widest and deepest extent of an MTZ. A slight change in the orientation of the tissue and the sectioning angle can give variable tissue depths. Histologically, MTZ or microcolumns created are usually cylindrical, and conventionally, sectioning is done perpendicular to the surface (vertical sectioning).[23412] This requires that the sample must be oriented to ensure that vertical sections are perpendicular to the skin surface, presuming that the MTZ have the same orientation. Presuming that the MTZ has a uniform diameter, which is not the case, the depth can be accurately measured with a single vertical section anywhere through the MTZ. However, logically the measurement of diameter requires vertical sectioning through the axis of the MTZ which can only be done with serial vertical sectioning where the section with the maximum width is presumed to be that closest to the axis. This process requires multiple sequential slides and most centres rarely employ this procedure which is a very laborious process with a high probability of human error.[11] This is also because most histopathological centres are burdened with routine pathology tests, and assessment by horizontal sectioning of fractional lasers may not be a priority. To circumvent most of these issues, Zelickson et al.[11] suggested replacing the conventional vertical method of sectioning with the horizontal sectioning. Though Zelickson et al.[11] have demonstrated that the horizontal sectioning method is superior, they were ably assisted by the semi-automatic fractional histology interpretation routine (SAFHIR) software, a luxury that most other centres are not endowed with. The use of a software-based analysis of images (J program) was used in another study[18] which makes comparison of data with the other studies difficult.[1–1012–1820]

Stains used

Apart from haematoxylin and eosin (H and E), the various stains employed include nitro blue tetrazolium chloride (NBTC), lactate dehydrogenase (LDH) and Masson's trichrome stains, and the demanding terminal deoxynucleotidyl transferase-mediated X-dUTP nick end labelling (TUNEL) stain which requires the fluorescence excitation microscopy for evaluation. If H and E is used as a marker for thermal damage, it is more likely to result in a relatively smaller width than NBTC-stained lesions. This is because the threshold to detect thermal cell damage is lower for the NBTC stain as compared to H and E staining. Some studies use specialised immunohistochemical stains for heat shock proteins and elastin (anti-HSP70, anti-HSP47 and anti-elastin).[30] Though there are merits of using stains like LDC and TUNEL stains, there is no homogeneity in the stains used in various studies which adds to the difficulty in comparing studies.[2–2028]

Wavelength

It has been observed that the optical window of 1.2 and 1.8 μm is optimal for skin treatments as the penetration depth of the laser is up to the upper dermis. On comparing three wavelengths, 1.32, 1.45 and 1.54 μm, it was found that 1.54 μm was well absorbed by water yet virtually not at all by melanin, allowing deep penetration into the skin with minimal side effects in the pigmented skin. Thus, this was the most commonly used wavelength initially. As ablative fractional lasers evolved, they were found to have a depth that was more than that for the nonablative fractional lasers [Table 1]. What is relevant to note is that on analysing the results of the various histological studies,[1–20] it is obvious that as far as the facial skin is concerned, there are very little difference in depth if the right dose is used, and at least in theory, all the three wavelengths (1540, 2940 and 10600 nm) should be clinically effective in treating acne scars [Figure 3a].[35]

Pulse number/duration

It has been shown that both the DeepFX[14] and Lux 2940[11] achieve a greater depth if there is an increase in the number of pulses. It has been shown that[1225] if the pulse width of the AFR (2940 nm) is increased to 2 ms, the heat impacting on the dermis increases whereas if it is lowered (0.25 ms), a deeper penetration can be achieved. In a study on an ex vivo model with the fractional erbium laser, reducing the pulse duration from 2–5 to 0.25 ms caused a 20–25% deeper penetration of the microbeams.[12] A similar effect has been noted with the fractional carbon dioxide laser[16] where the depth increased by 10-fold when the pulse duration was reduced from 5–10 to 2 ms.

Passes/energy density

The density of the 1550-nm lasers ranges from 100 to 1000 microbeam/probe.[963637] Obviously, higher the density, more the dose (joule/cm²) that can be achieved. By manipulating the density of the microcolumns, a more aggressive treatment may be achieved in a single treatment. This can be achieved by giving multiple passes. However, when retreating or passing over a treated area multiple times, a non-uniform damage is achieved.

Regardless of the mode of delivery, with repeated passes all devices lead to a randomly distributed placement of MTZs. Clustering of MTZs (i.e. placement of individual MTZs into a location that was already damaged) can occur with subsequent passes and lead to alterations in the MTZ size and fill factor. It can be conjectured that the approach of giving multiple passes to achieve a high density and energy is logical and in spite of numerous passes (1–30), the healing can still occur.

A study by Trelles et al.[25] using the Sciton Profile Er:YAG system showed that multiple passes produce photothermally related effects in the dermis. The repeated pulses first ablated the epidermis which created a ‘window’ that helped the dermis with its high water content to heat up adding to the residual thermal damage zone. This effect is useful for acne scars as there is collagen remodelling.

Contrary to this view, another study[9] pointed out that multiple-pass treatments, which are recommended for all fractional laser devices, created an increased broad-based destruction of the DEJ over a larger surface area. This was because the higher energy multiple-passtreatments demonstrated a more superficial treatment than anticipated for their respectivefluences that may be attributed to the potential absorption and concentration of energy or heat at the DEJ.

The balance between wound healing, neo-collagenesis, coagulation and remodelling for optimal skin tightening and rejuvenation with fractional technology warrants further investigations. Bulk damage can be caused with higher density settings at modest microbeam energies or, conversely, with higher microbeam energies at modest densities.[37] In Asian skin, it has been confirmed that it is better to increase the dose/microbeam than increase the density to achieve an equivalent energy.[37]

Histological depth of fractional lasers

The depth of penetration of fractional lasers is probably a crucial parameter for improving acne scars.[1–20] The most important factors predicting the depth are the density and the energy settings. An increased energy level leads to a non-linear increase in the depth of penetration. A study comparing two NAFR[6] lasers (Fraxel SR750 vs SR1500) found that the depth increased from 386±68 μm to 826±108 μm on increasing the dose from 6 to 40 mJ using the Fraxel SR750. The newer Fraxel SR 1500 achieved a deeper penetration (470±56 μm, 1408±53 μm) for proportionately higher energy (6–100 mJ). It is striking that for lower energies (up to 20 mJ), there was no difference or only a marginally significant difference in the lesion width and depth between the two devices. However, for higher energies (over 20 mJ), there was a statistically significant difference in diameter and depth. A study using the fractional CO₂ laser found a proportional increase in the depth and width depending on the density and dose (0.03–0.15 mm; dose 90–100 mJ).[18] This study used density levels 1-3 which predicted the overlap ranging from −10% to +10%. An increase in the density caused an increased depth of ablation while an increase in the dose (energy) affected the width of coagulation. Paradoxically, a study which used the fractional Er:YAG laser (2940 nm)[17] revealed that though the standard predicted depth of ablation ranged between 25 and 1500 mm, the actual achieved depth was much greater than the predicted.

Very few studies have formally compared the laser depth of fractional lasers. Farkas et al.[13] demonstrated that even within the same technology, AFR (ActiveFX, DeepFX, Fraxel repair), the depth of penetration may vary [Figure 3b]. It was initially believed that it is not possible to compare NAFR and AFR as the tissue responses would vary. This is not the case as studies with detailed tissue depth analysis[1114] have shown [Figure 3c] that the tissue response is remarkably similar. All the three laser systems (2940, 1540 and 10,600 nm) produce an MTZ which has an inner and outer diameter, a zone of ablation and a deeper zone of coagulation, and thus, probably have similar effects on the skin. The only difference is that the AFR lasers have less width of damage than the NAFR [Figure 3c].

The existing studies [Table 1] have not formally explored the correlation between depth and energy. For the 1550-nm fractional laser, it has been previously estimated that for every millijoule of increased energy, the depth of coagulation increased roughly by a factor of 10 μm (10 mJ/100–150 μm).[9] Similarly, for the fractional CO₂ (10,600 nm), a formula to estimate the depth has been proposed, where the ablation depth (μm)=12 μm/mJ×energy level (mJ).[16] The fractional erbium:YAG laser has a predicted ablation that starts above a defined threshold (ablation threshold), which is 1.6 J/cm². Energy densities less than 1.6 J/cm² heat up the tissue and may lead to thermal damages. Each subsequent increase in energy (J/cm²) ablates 5 μm of the tissue. Thus, the depth is predicted by the formula depth, D (μm)=5 μm/J/cm² × (F-N ×1.5 J/cm²), where F is the fluence (J/cm²) and N is the number of stacked pulses (Dermablate microspot mode; Ascepelion).

On statistical analyses of the existing data [Table 1], we found that there was a statistical correlation between the dose and depth achieved by the three fractional lasers [Table 2 and Figure 4a–c]. As different fractional lasers have different settings (mJ/microbeam), it is practically useful to statistically analyse the mean depth (μm/mJ) achieved by the lasers used [Table 2]. The data given in Table 2 can give a rough assessment of the probable depth that can be achieved depending on the dose and laser used.

Table 2 Correlation between the energy and the depth of penetration of fractional lasers

Close

Figure 4

(a) An illustration of the correlation of the energy/depth of the Er:glass laser. (b) An illustration of the correlation of the energy/depth of the Er:YAG laser. (c) An illustration of the correlation of the energy/depth of the CO₂ laser

Export to PPT

The two studies[1114] where comparative data are available regarding the depth achieved show variable results. While Figure 3c reveals that the fractional Er:YAG laser is superior to other lasers, Figure 3b shows that the fractional CO₂ laser achieves a depth comparable to the fractional Er:YAG laser. Interestingly, Figure 3c shows that the NAFR (1440 nm) laser achieves a similar depth as the AFR (10600 nm) laser. These variations are probably because of the variable settings and substrates used. To overcome these variations, we statistically analysed the depth of penetration [Table 1] achieved by various fractional lasers, to average out the minor variations in doses. Our analysis revealed that the mean depth of penetration varied [Figure 5a–c], with NAFR laser (Er:glass, 679 μm) achieving a depth that was less than that of the AFR lasers (Er:YAG, 825 μm, and CO₂ 895 μm; P<0.05). On the facial skin though, the difference in the histological depth of the three fractional lasers (1540, 2940 and 10600 nm) would probably be similar for superficial atrophic acne scars [Figure 3a]. For deeper scars, the AFR lasers, at least in theory, would be better as they penetrate deeper than NAFR lasers, but as seen in Figure 3a, the depth achieved is probably not deep enough to ameliorate the ice pick scars [Figure 3a].

Close

Figure 5

Bootstrap plot - central tendency analysis of the depth of penetration (X-axis, μm) of the (a) Er:glass laser (mean 679 μm, median 580 μm, mid-range 884 μm), (b) Er:YAG laser (mean 825 μm, median 525 μm, mid-range 1162.5 μm) and (c) CO₂ laser (mean 895 μm, median 600 μm, mid-range 2520 μm)

Export to PPT

Thus on statistical analysis of the existing data, we can conjecture that there is evidence, both mathematical and substrate based, which can help predict the depth achieved by a particular laser. But the extrapolation of existing data to in vivo facial skin would depend on replicating the exact settings (pulses, duration, passes, energy and density) which is rarely the case. This and the variability of the tissues studied [Figure 2] probably account partially for the disparity in subjective and objective clinical results for acne scars [Table 3].

Table 3 Summary of studies of AFR fractional laser in acne scars