April 2020 HSC Section 4 - Plastic and Reconstructive Problems

Annals of Plastic Surgery • Volume 83, Number 2, August 2019

Siotos et al

who received excision and 1 additional medical treatment. However, this difference was not statistically significant (OR, 4.34; REM; 95% CI, 0.38 – 49.20; P = 0.24; I 2 = 79%; Fig. 5). (5)Excision + 1 adjuvant drug versus excision + radiation: Five stud- ies 40,43,47,48,50 were included, resulting in 88 patients among the exci- sion + 1 adjuvant drug group and 74 among the excision + radiation group. Odds for recurrence were significantly higher for the exci- sion + 1 adjuvant drug group (OR, 3.22; REM; 95% CI, 1.35, 7.67; P = 0.008; I 2 = 0%; Fig. 6). (6)Steroids versus IFN in the excision + 1 adjuvant drug group: Two studies 41,53 were included resulting in 91 patients who received ste- roids as adjuvant treatment and 29 who received IFN among patients of the excision + 1 adjuvant drug group. Odds for recurrence were similar among patients of the 2 groups (OR, 0.99; REM; 95% CI, 0.03, 36.31; P = 1.00; I 2 = 92%; Fig. 7). We also investigated the 7 treatments presented in the 14 studies in a network fashion (Fig. 8). The total number of direct comparisons was 12. Out of 14 studies, 8 studies compared 2 treatments; 5 studies compared 3 treatments; 1 study compared 4 treatments (Table 2). Most subjects underwent excision + radiation (313 of 996), and the fewest subjects underwent excision + skin grafting (26 of 996). Odds ratio estimates and the associated 95% CIs in comparison to no excision were calculated (Table 3) and were as follows: excision + pressure ” 0.18 (0.01 – 7.07); excision + 2 adjuvant drugs, 0.47 (0.02 – 12.82); excision + radiation, 0.39 (0.04 – 3.31); excision + skin grafting, 0.58 (0.00 – 76.10); excision + 1 adjuvant drug, 1.76 (0.17 – 21.35), and exci- sion only, 2.17 (0.23 – 23.95). The only statistically significant difference identified was for the lower rates of recurrences for patients who received excision + radiation, in comparison to excision only. SUCRAvalues were calculated based on the cumulative ranking probability curve of each treatment (Figure, Supplemental Digital Content 2, http://links.lww. com/SAP/A365). The values, ordered from the most effective treatment to the least effective, were excision + pressure, 79.2; excision + radiation, 70.7; excision + 2 adjuvant drugs, 62.2; excision + skin grafting, 54.0; no excision, 42.0; excision + 1 adjuvant drug, 23.6; and excision only, 18.4. Gelman diagnostic statistics and trace plots are available as online sup- plemental material (Document, Supplemental Digital Content 3, http:// links.lww.com/SAP/A366). DISCUSSION Keloids and hypertrophic scars are difficult to treat and often re- cur following treatment. Excision is commonly used because it yields the most rapid visible reduction of the keloid. However, as with other treatments, excision is associated with a high rate of recurrence. Although the risk of recurrence is thought to be lower when combined with an ad- juvant therapy, there is no consensus regarding which adjuvant therapy best prevents recurrence. In this study, we sought to compare the effect of various adjuvant modalities in treating keloids. Of the multiple paired comparisons that we performed, exci- sion + radiation proved more effective at reducing the odds of keloid re- currence than excision + 1 adjuvant drug. This inference was drawn from5 studies comparing a total of 162 patients. Of the groups we compared, this was the only one in which we found a significant difference in recur- rence rate. In the comparison group, the most commonly used nonradiation adjuvant modalities were steroid injections. However, according to the paired meta-analysis excision + radiation did not consistently outperform other modalities. The results of our paired meta-analysis agree with the results of the network meta-analysis, according to which excision + radi- ation was proven superior at preventing recurrences in comparison to excision only. In addition, excision + pressure was found to lead to less recurrences than the other treatment modalities; however, this differ- ence was not statistically significant.

No excision 2.17 (0.23 – 23.95) 1.76 (0.17 – 21.35) 0.47 (0.02 – 12.82) 0.39 (0.04 – 3.31) 0.18 (0.01 – 7.07) 0.58 (0.00 – 76.10) Excision + 2 adjuvant drugs 2.13 (0.08 – 46.20) 4.59 (0.31 – 64.78) 3.72 (0.41 – 32.66) 0.82 (0.05 – 9.84) 0.38 (0.01 – 18.75) 1.22 (0.01 – 128.38) Excision + radiation 2.59 (0.30 – 24.31) 5.61 ( 1.27 – 32.04) 4.56 (0.97 – 27.77) 1.22 (0.10 – 20.23) 0.46 (0.02 – 16.58) 1.52 (0.01 – 143.17) excision + pressure 5.59 (0.14 – 200.34) 12.17 (0.41 – 386.06) 9.85 (0.37 – 290.32) 2.64 (0.05 – 149.46) 2.16 (0.06 – 66.42) 3.23 (0.01 – 670.48) Excision + skin grafting 1.72 (0.01 – 294.12) 3.74 (0.04 – 496.71) 3.02 (0.05 – 307.35) 0.82 (0.01 – 134.56) 0.66 (0.01 – 81.29) 0.31 (0.00 – 89.93) Posterior means of OR between column and row (95% Bayesian credible intervals) under the Bayesian REM are reported. Odds ratio of less than 1 favors the treatment in the column, and OR greater than 1 favors the treatment in the row. Significant ORs are in bold. Excision only 0.46 (0.04 – 4.30) 0.81 (0.15 – 4.24) 0.22 (0.02 – 3.25) 0.18 ( 0.03 – 0.79) 0.08 (0.00 – 2.41) 0.27 (0.00 – 22.87) Excision + 1 adjuvant drug 0.57 (0.05 – 5.74) 1.23 (0.24 – 6.50) 0.27 (0.03 – 2.41) 0.22 (0.04 – 1.04) 0.10 (0.00 – 2.71) 0.33 (0.00 – 20.42)

Excision + Skin Grafting

Pressure

Excision +

Radiation

Excision +

Excision + 2

Adjuvant Drugs

Excision + 1

Adjuvant Drug

No Excision Excision Only

TABLE 3. OR Estimates for Keloid Recurrence Obtained From Network Meta-analysis

www.annalsplasticsurgery.com

© 2019 Wolters Kluwer Health, Inc. All rights reserved.

133