Intralesional Corticosteroid Administration in the Treatment of Keloids: A Scoping Review on Injection Methods

Keloids are fibroproliferative disorders resulting from chronic inflammation in the dermis and often cause pain, pruritus, a tight sensation, cosmetic concerns, and occasionally movement restriction [1, 2]. The quality of life is impacted, especially when pain, pruritus, or functional restrictions are experienced [3, 4]. Intralesional corticosteroid administration (ICA) is a first-line treatment in current practice, of which triamcinolone acetonide (TAC) is most commonly used [5]. Corticosteroids induce keloid regression through many different mechanisms, including but not limited to the inhibition of leukocyte and monocyte migration and phagocytosis resulting in reduced inflammation, increased vasoconstriction resulting in reduced oxygen and nutrient delivery to the wound bed, and the inhibition of keratinocytes and fibroblasts resulting in reduced re-epithelialization and new collagen formation [5]. Even though ICA is a first-line keloid treatment, clinical results of this treatment are still highly variable and often suboptimal [6, 7]. Treatment results may strongly be influenced by clinicians’ preferences, such as the type, volume, and concentration of the corticosteroid, the number and interval of treatment sessions, the needle and syringe size, and the manual injection techniques. A search of PROSPERO, CENTRAL, and PubMed identified one systematic review from 2008 that aimed to include (randomized) controlled trials on the most effective concentration and treatment interval of intralesional TAC in keloids. However, the authors did not retrieve any studies that fit their inclusion and exclusion criteria [8]. To our knowledge, there is currently no consensus on different aspects of ICA in keloids among physicians. We hypothesized that a wide variety in this first-line treatment may exist in scientific practice as well. Identifying the (non)uniformity of intralesional corticosteroid treatment protocols is of an exploratory nature that is suited to a scoping review. The aim of this scoping review was to map the details of ICA in keloids described in randomized controlled trials (RCTs), hence presenting the scientific practice of a first-line treatment for keloids in the best available evidence up to date.

This scoping review was developed in accordance with the methodological framework devised by the Joanna Briggs Institute [9]. Prior to starting this study, a predefined charting form was designed with items that we aimed to address. Registration of the review protocol in PROSPERO could not be performed as this is a scoping review. All other items of the Preferred Reporting Items for Systematic reviews and Meta-analyses extension for Scoping Reviews (PRISMA-ScR) checklist were addressed (online suppl. File 1; for all online suppl. material, see www.karger.com/doi/10.1159/000529220).

A systematic search was performed on PubMed, Ovid MEDLINE, Ovid EMBASE, and CENTRAL for articles from inception to January 17, 2022. The search combined terms on (1) keloid and (2) intralesional therapy or injection in title or abstract, and involved relevant MeSH terms and synonyms. Inclusion of terms on generic and chemical names of corticosteroids was not attempted, so that no corticosteroid treatment could be missed. In the context of a scoping review, no explicit outcomes were stated in the search, aiming to capture all relevant papers, irrespective of used outcome measures. The search was limited to studies written in English, French, Dutch, and German. The full electronic search is described in online supplementary 2.

Following the search, all identified citations were imported into EndNote version 20 bibliographic management software. Two reviewers performed the title and abstract screening (Q.Y. and G.K.), and two reviewers (Q.Y. and G.K. or J.L.) independently assessed the full texts on eligibility. Discrepancies were resolved by a fourth reviewer (AW). Reference lists of the included articles were checked for additional studies of interest. Studies were eligible if they included (1) humans of any race, gender, and age, (2) with one or more keloids of any size, in any anatomic location, with any duration and with any etiology, and (3) were RCTs that involved intralesional corticosteroids. Only RCTs were included, because of the high level of evidence inherent to its study design. Articles on both keloids and hypertrophic scars were included only if separated data for keloids were available.

Data charting was performed by QY and GK in duplicate. Data collected on the study population included the used (1) diagnostic criteria for the keloid, (2) number of patients and keloids that received intralesional corticosteroids, (3) location of the keloid, and (4) country of research origin. Data collected on the treatment were details of the used (5) drugs and dosing, including the type, volume, and concentration of the corticosteroid, the number and interval of treatment sessions, and local anesthetics, (6) equipment, including the needle and syringe size, and (7) manual injection techniques including the level of injection, angle of injection, speed of injection, number of injections per session, and endpoint of infiltration. Additionally, study design-related data were collected, which included (8) outcome measures, (9) follow-up period and recurrence, and (10) adverse events. A concise summary was provided for each charted item.

A risk-of-bias assessment was performed by two reviewers (Q.Y. or J.L.) to determine the quality of current RCTs on this first-line treatment of keloids. Version 2 of the Cochrane risk-of-bias tool for randomized trials was used.

The search yielded 1,723 unique articles, of which 38 articles were included for data extraction (shown in Fig. 1 flowchart). Of these included studies, only one study compared the effect of two different doses of TAC directly [10], thirty-four studies investigated different therapies for keloids, with the TAC group being either the treatment or the control arm, and three studies primarily investigated the effect of local anesthetics on injection pain.

The diagnostic criteria were not described by twenty-four (63%) studies. Six studies (16%) separately presented data of both keloids and hypertrophic scars [11‒16], of which only three studies specified the criteria for distinction between them, with the keloid being an entity that extends over the margins of the original injury and the hypertrophic scar one that remains within the confines of the original injury [12, 14, 16]. Six studies on both keloids and hypertrophic scars were excluded, because data of two entities were pooled.

TAC was used in 37 (97.4%) studies. Only one study used a combination of betamethasone disodium phosphate 2 mg/mL and betamethasone dipropionate 5 mg/mL [17]. The dose of TAC per cm² of keloid could only be compared among ten (26%) studies and differed by a factor of 20 [6, 7, 10, 12, 17‒22] (Table 1). Whereas one study reported to inject 0.1 mL of TAC 10 mg/mL (dose: 1 mg TAC) per cm² of keloid [12], other studies reported to inject 0.5 mL of TAC 40 mg/mL (dose: 20 mg TAC) per cm² of keloid [6, 18, 19]. The used concentrations of TAC were 40 mg/mL (39%), 20 mg/mL (18%), 10 mg/mL (24%), and 5 mg/mL (2.7%). In five (13%) studies, the used concentration of TAC was unclear. The injected volume was only reported in twenty (53%) studies. The maximum dose administered per session was mentioned in twelve (32%) studies and varied from 20 to 80 mg (Table 1).

The total number of treatment sessions varied from one [20] to eight [11, 13, 16], with the median being four sessions. Reported treatment intervals differed from weekly [16, 23] to monthly [15, 17, 24‒27]. The most commonly described treatment interval was every 4 weeks (50%), followed by every 3 weeks (29%), weekly (15%), and every 2 weeks (6%). Six studies mentioned that treatment would be stopped if complete lesion flattening or clearance occurred [10, 18, 28‒31] (Table 1).

The effect of local anesthetics on injection pain was investigated by three studies [15, 21, 22] (Table 1). One study demonstrated that 1:1 mixture of lidocaine 1% and TAC did not alleviate needle puncture pain, whereas EMLA cream did [21]. Another study revealed that the mean VAS score of needle puncture and steroid infiltration was significantly lower among patients that received prior cooling with icepack, compared to no pre-treatment (p < 0.001) and lidocaine and prilocaine 25/25 mg/g (EMLA) cream (p < 0.05) [22]. The third study demonstrated that application of −10⁰C using a cooling device to the skin before injection leads to lower pain scores [15]. Among the RCTs that did not primarily focus on injection pain, local anesthetics were used in seven (20%) studies and included lidocaine 1% [11, 13], lidocaine 2% [19], and EMLA cream [32]. Two studies used a lidocaine-TAC mixture as treatment [32, 33]. Oral analgesics were administrated in one study [18].

Reported needle sizes were 26 (brown, 0.45 mm), 27 (medium gray, 0.40 mm), and 30-gauge (yellow, 0.30 mm), which were used in, respectively, one [17], seven [11, 12, 14, 16, 18, 19, 34], and three studies [28, 31, 35]. Twenty-seven (71%) studies did not report the needle size. The syringe size was only specified in four studies, being 1 mL [11, 16, 35, 36]. Seven (18%) studies reported the use of an insulin syringe, without mentioning its size [6, 18, 20, 23, 28, 31, 37] (Table 1).

The level of injection was only specified in eleven studies [11, 13, 16, 17, 20, 23, 31, 33, 35, 38, 39]. Aggarwal et al. [31] reported that the level of injection was at a depth of 3–7 mm, depending upon the size of the lesion. Other terms used for the injection level were “mid-lesion” [12], “body of keloid” [13, 16, 35], “intralesionally” [38], and “into the keloid” [20, 23, 33, 39]. Hietanen et al. [33] specifically mentioned that care was taken not to inject under the keloid mass or too close to the epidermis to avoid local side effects. Khalid et al., Khan et al., and Saha et al. injected the indurated part of a keloid [11, 13, 35]. The angle of injection was described in three studies, being “parallel to the skin” [21], “at 60-degree” [36], and “at 45-degree angle against skin surface” [12]. Nine studies reported administration with multiple injections [11, 13, 17, 18, 23, 31, 33, 35, 38], being separated by 1 cm in six studies [11, 13, 17, 18, 31, 35]. Two studies mentioned the speed of injection, being 0.5–1.0 mL per minute and 0.1 mL per 10–15 s [21, 22]. Blanching was reported in ten (26%) studies as an endpoint indicating adequate TAC distribution [6, 11, 13, 17, 23, 28, 31‒33, 35] (Table 1).

A wide variety of outcome measures was used. The most frequently used outcome measure was keloid size, reported in twelve (34%) studies as either the height, surface area, or volume [19, 20, 23, 27, 29‒32, 35, 38, 40, 41]. Ten (29%) [18, 26, 28, 31, 32, 34, 36, 37, 42, 43] and seven (20%) [6, 7, 12, 25, 33, 44, 45] studies used the Vancouver Scar Scale (VSS) and the Patient and Observer Scar Assessment Scale (POSAS). Remarkably, from 2017 onward, 15 of 22 (68%) studies used either one of these scales. Patient satisfaction was evaluated in eight (23%) studies [17, 19, 22, 27, 28, 30, 41, 42] (Table 2). The treatment results reported by the studies are summarized in online supplementary 3.

The follow-up period in the included studies varied from 1 month [30, 34] to 1 year [29, 35]. Only six (17%) studies had a follow-up of 6 months or longer [7, 19, 23, 29, 35, 41], of which two had a follow-up of 1 year [29] or until recurrence occurred within that year [35]. Eleven (31%) studies did not report the follow-up period. Recurrence was only identified in two studies with a follow-up period of 18 weeks and 1 year [6, 35]. In the latter, the reported recurrence rate was 36% (8 of 22 patients) [35]. In other studies with a minimum follow-up period of 6 months, the recurrence rate was either not specified [7, 19] or zero [23, 29, 41] (Table 2).

Twenty-three (61%) studies reported adverse events. Atrophy (5–75%) and telangiectasia (10–80%) occurred the most and were described in fourteen (37%) studies [13, 17, 18, 25, 28, 30, 31, 33, 37, 40‒42, 44, 45]. The majority (79%) concerned atrophy of the skin. Subcutaneous atrophy was not specified in any study. Four (11%) studies reported that the number of adverse events was zero [14, 19, 22, 27], while in eleven (29%) studies the adverse events were not specified [7, 11, 12, 15, 16, 20, 21, 24, 32, 34, 38] (Table 2).

Twenty-nine (76%) articles were judged as having a high risk of bias. The most common reason for high risk of bias was the measurement of outcome (shown in Fig. 2).

This scoping review mapped the relevant details of ICA in keloids described in RCTs. It functions as a baseline measurement of the scientific practice of a first-line treatment for keloids in the best available evidence to date. Incomplete reporting and substantial heterogeneity were identified for different aspects of this pivotal treatment.

First, the dose of corticosteroid per area of keloid, which is an essential factor that defines treatment effect, could only be extracted from ten (26%) studies. Second, needle (71%) and syringe (89%) sizes were mostly not specified, even though the choice of needle and syringe diameters can influence dose administration. For the syringe, this can be illustrated with Pascal’s law (shown in Fig. 3a). Larger syringes generate lower injection pressure that may cause inadequate drug delivery, while small syringes generate high pressure, which increases the risk of drug delivery into the surrounding healthy tissue. Considering needle sizes, Poiseuille’s law illustrates that 30-gauge needles are insufficient to deliver adequate medication to firmer keloids (shown in Fig. 3b) [46]. Third, the level of injection is a relevant detail that was reported only by the minority (29%) of studies. Notably, a consensus study of 2019 mentioned that the solid central fibrotic mass of the lesion should not be injected, because the drug will not infiltrate the tissue adequately and the induced rising pressure may cause pain. Instead, it was recommended to penetrate the scar from its border with the normal skin and to inject the deepest part of the scar which is softer than the central core, and/or the periphery of the lesion where the inflammation is particularly pronounced [47]. However, it should be noted that deep injection may increase the risk of subcutaneous atrophy. Considering the endpoint of infiltration, blanching is a way to indicate appropriate drug distribution in the dermis, but was only mentioned in ten (26%) studies.

Apart from the details on ICA, the study design was also incompletely described in most studies. Details on the follow-up period (31%), recurrence rate (66%), and adverse events (29%) were frequently lacking among studies. Only four studies have a follow-up of 6 months or longer, and only two studies have a follow-up 12 months or longer. Long-term powered RCTs are necessary to define treatment outcome. Studies on keloid treatments that do not report long-term treatment effects have little clinical value. Finally, quality assessment of studies revealed a considerable risk of bias in reported treatment results.

TAC is the uniformly used corticosteroid in all but one included study. Different types of corticosteroids exist for intralesional administration. Practical reasons such as drug availability and familiarity with the drug may now be the main reasons for the uniformity in the use of TAC. However, its favorable pharmacokinetic properties may be a more important reason of TAC being the preferred intralesional corticosteroid for keloid treatment originally. Studies have reported that TAC has a substantially longer duration of action in the tissue than betamethasone or prednisolone [48‒51].

Apart from the type of corticosteroid, substantial heterogeneity exists in most other aspects of ICA among included studies. First, TAC doses varied from 1 mg per cm² of keloid [12] to 20 mg per cm² of keloid, a dose difference of 20 times [6, 18, 19]. Second, reported maximum doses per session varied from 20 mg to 80 mg. Systemic adverse events may occur at higher doses of TAC and include Cushing’s syndrome. Based on a systematic review of the literature from 1950 to 2012 of reported cases of Cushing’s syndrome following intralesional TAC used for the treatment of scars, Fredman et al. [52] recommended that intralesional dosage of TAC should generally not exceed 40 mg per month in adults. Third, treatment sessions varied from one to eight, and treatment intervals varied from weekly to monthly. Ud-Din et al. [53] demonstrated that treatment response is associated with the number of treatment sessions, with an odds ratio of 3.9 (95% CI: 1.06–14.7, p = 0.041) for patients receiving more than one session. Treatment interval was studied among sixteen keloids by weekly volume measurements after single TAC injection. Volume reduction was found to be most profound within the first 2 weeks post-injection. This was not significant anymore between 3 and 4 weeks, and a reversal trend was identified after 4 weeks (p ≤ 0.05) [54]. In line with this study, the mostly (79%) used treatment interval among RCTs in our study is 3–4 weeks. Furthermore, heterogeneity was identified in needle sizes which varied from 26-gauge (0.45 mm) to 30-gauge (0.30 mm) and the level and angle of injection for which different descriptions were used. Considering local anesthesia, treatment regimens varied from no anesthetics to oral analgesics. Reducing injection pain should be aimed, as pain may affect therapy compliance [55]. Based on the studies that focused on pain relief in ICA, infiltration of the corticosteroid should take place gently, as the speed of drug administration affects pain [21, 22]. Furthermore, local anesthesia using EMLA cream and skin cooling seemed to be effective in attenuating pain [15, 21, 22], while the mixture of an anesthetic with corticosteroid was not [21].

Apart from treatment details, study designs were also very heterogeneous. Outcome measures varied from keloid sizes reported as height, surface area, or volume, to VSS, POSAS, pain and itch scores, patients’ satisfaction, and self-defined efficacy rates. The substantial heterogeneity in used outcome measures makes studies incomparable.

The value of this scoping review encompasses mapping a first-line treatment for keloids and investigating current research conduct [9]. Even though our scoping review included only RCTs, which are studies that provide the best evidence, overall incomplete reporting and high heterogeneity were identified. The current body of literature is therefore not amenable to data synthesis, and evidence-based treatment recommendation cannot be generated. A limitation of our scoping review is that it included keloids of all sizes, locations, duration, and etiology, even though differences may exist within the same entity [56]. However, differentiation in keloid characteristics could not be made in our study, which is inherent to pooling of data in the best available evidence. In future studies, differentiation between different types of keloids should be considered. Also inherent to available data from current RCTs, the definition of a keloid was frequently not specified, leaving the possibility that other entities such as hypertrophic scars could be included erroneously. Future studies should use a uniform definition, which yet needs to be delineated.

In conclusion, this scoping review highlights the incomplete reporting and substantial heterogeneity on ICA in keloids among current RCTs. Standardization of treatment protocol and study design is urgently needed in order to make future keloid studies and treatment results of ICA comparable.

Incomplete reporting and substantial heterogeneity exist on the intralesional corticosteroid treatment in keloids among current randomized controlled trials.

It is not applicable because this study is based exclusively on published literature.

The authors have no conflicts of interest to declare.

The authors did not receive funding.

Q Yin: conceptualization – equal, data curation – lead, formal analysis – lead, investigation – lead, methodology – lead, project administration – lead, and writing – original draft – lead. JMI Louter: conceptualization – equal, investigation – equal, methodology – supporting, and writing – review and editing – supporting. FB Niessen and O Lapid: supervision – equal, and writing – review and editing – supporting. S Gibss and PPM van Zuijlen: supervision – supporting, and writing – review and editing – supporting. G Tasdemir-Kilic: data curation – equal and investigation – equal. A Wolkerstorfer: conceptualization – lead, methodology – lead, supervision – lead, and writing – review and editing – lead.