Lesion Site Is the Key Prognostic Factor for Keloid Patients Receiving Surgery With Adjuvant Radiotherapy : Annals of Plastic Surgery

Secondary Logo

Journal Logo

Burn Surgery and Research

Lesion Site Is the Key Prognostic Factor for Keloid Patients Receiving Surgery With Adjuvant Radiotherapy

Chen, Frank MDa,b; Kuo, Yur-Ren MD, PhDb,c; Huang, Chih-Jen MD, PhDa; Tang, Jen-Yang MD, PhDa; Chiang, Chen-Han MSa; Huang, Ming-Yii MD, PhDa,b

Author Information
doi: 10.1097/SAP.0000000000003315
  • Open


Keloid is a benign excessive tissue progressing from aberrant tissue-remodeling with disordered fibroblast proliferation. It usually forms a tumor-like lesion extending beyond the margins of the initial insult1 and is mainly diagnosed clinically with high accuracy by its clinical presentation.2 The Japan Scar Workshop Scar Scale and Vancouver Scar Scale are often used for the diagnosis of keloid.3 Patients can experience symptoms such as pain, pruritus, anatomic dysfunction, significant cosmetic disfigurement, and psychosocial issues, with impaired quality of life,1,4 so seeking treatment for symptomatic keloids becomes a necessity.

Risk factors for keloid scars include family history, genetic predispositions, pregnancy, hypertension, age as adolescence, and specific ethnic background.5 Asian populations (Orientals) might have higher keloid incidence compared with European populations (Occidentals) (0.15% in Taiwan compared with 0.09% in England).3,6 Anatomic wound site is also a risk factor of keloid. Keloid appears to have higher incidence over high-tension areas such as the chest, scapula, and suprapubic areas,7–9 and often grows along skin tension lines.10

The mechanism of keloid formation is not yet well understood,11 although the pathway of keloid formation could begin from any kind of skin injury from such causes as surgery, heat, chemical burns, or inflammation4; then, immune cells such as macrophages, mast cells, neutrophils, eosinophils, and T cells collaborate with growth factors such as vascular endothelial or platelet-derived ones that correspondingly induce cell proliferation, angiogenesis, re-epithelization, and stimulate fibrosis. Later, cell migration and proliferation of keratinocytes and myofibroblasts are promoted that increase collagen synthesis and deposition, then extracellular matrix (ECM) accumulates below the dermis, ultimately forming a scar.12 A previous study has shown that keloid-derived fibroblast has higher sensitivity for transforming growth factor β1, insulin-like growth factor 1, than do normal skin fibroblasts,13 which might explain why keloid grows faster than normal skin tissue.

Treatments for keloid include intralesional steroids and chemotherapy, compression therapy, cryotherapy, laser therapy, surgical intervention, and radiotherapy.1,4,13 Intralesional steroid is often used, and usually requires a series of injections, resulting in 50% to 100% response rates and 9% to 50% recurrence rates; however, treatment can continue for several months and can be painful.1,13 Compression therapy and silicone gel patch have less evidential effectiveness, with long course and poor compliance,1,13 whereas cryotherapy and laser therapy have also been used, but adverse skin effects are concerning.13 Chemotherapy modalities such as 5-fluorouracil, bleomycin, and mitomycin C have been used but reveal limited evidence,4,13 whereas surgical intervention is a curative method, but causes new skin damage with 45% to 100% recurrence rate.1,4,13,14 Studies have shown adjuvant radiotherapy after surgical intervention has the lowest recurrence rate, lowering the recurrence rate from 45% to 100% with surgery alone to 15% to 25% when paired with adjuvant radiotherapy.4

To further understand the factors that might affect the local control rate of keloid, keloid patients receiving surgical excision with adjuvant radiotherapy in Kaohsiung Medical University Hospital were reviewed.


The radiation oncology department database was used to include keloid patients receiving surgical excision and postoperative radiotherapy, those having completed a radiotherapy course, received first radiotherapy within 24 hours after surgery, and received radiation dose with 15 Gy in the study. Patients having been followed up for less than 6 months and/or aged less than 18 years were excluded. The primary end point was recurrence rate, and recurrence was defined clinically as palpable gross tumor over the original treatment site, with or without subjective symptoms such as pain or pruritus, with secondary end point set as local recurrence-free interval (LRFI) defined as duration from the last day of radiotherapy to disease recurrence. Patients were evaluated by following up at the outpatient department or by telephone if they were unwilling to visit the department due to the COVID-19 pandemic. Recurrence was determined by same plastic surgeon and radiation oncologist.

Tumor excisions for these patients were performed mainly by plastic surgeons with or without flap reconstruction. Adjuvant radiotherapy was performed with electron beam of 6 or 9 million electron volts by Linear accelerator with bolus or photon beam of 50 kV by Xoft Axxent Electronic Brachytherapy (eBx) System (San Jose, CA). Treatment fields encompassed the keloid surgical suture line with an expansion of the surrounding margin of 1 to 1.5 cm. Radiation dose was 15 Gy in 3 fractions. After treatment, patients were followed up at the plastic surgery outpatient department for wound care.

Two end points were analyzed with multiple variants such as sex, tumor location, tumor size, and treatment modality. In univariate analysis, χ2 test was used for recurrence rate and Kaplan-Meier method for LRFI. In multivariate analysis, logistic regression was used for recurrence rate and Cox regression method for LRFI. Values of P < 0.05 indicated statistical significance. All statistical analyses were performed by SPSS software for Windows (SPSS version 19; IBM Corp). The study was approved by the Institutional Review Board of Kaohsiung Medical University Hospital (KMUHIRB-SV[I]-20210071).


From May 2017 to July 2020, 32 patients with 40 keloid lesions were included from our radiation oncology database (Table 1). The mean age for these patients was 37.6 years (range, 18–64 years), and the median follow-up time was 15.3 months (range, 6.57–43.57 months). The overall recurrence rate for these 40 lesions was 52.5%, and the median LRFI was 9.7 months. Because of the small number for lesions from different body locations, we grouped these lesions into head and ear (n = 8), chest (n = 13), shoulder and upper extremities (n = 11), and abdomen and back (n = 8).

TABLE 1 - Patient Characteristics
Characteristic n (%)
Age, mean ± SD, y 37.6 ± 12.6
Median (range) follow-up, mo 15.3 (6.57–43.57)
Sex, lesions
 Male 15 (40)
 Female 25 (60)
Lesion site, lesions
 Head 2 (5)
 Ear 6 (15)
 Shoulder 4 (10)
 Chest 13 (32.5)
 Back 4 (10)
 Abdomen 4 (10)
 Upper extremities 7 (17.5)
Size, lesions
 Over 20 cm2 7 (17.5)
 Under 20 cm2 33 (82.5)
Radiotherapy modality, lesions
 Electron beam 30 (75)
 50 kV photon 10 (25)
 Yes 21 (52.5)
 No 19 (47.5)
SD, standard deviation.

Recurrence rates for males and females, respectively, were 46.7% and 56% (P = 0.567); for head and ear, chest, shoulder and upper extremities, and abdomen and back were 12.5%, 61.5%, 63.6%, and 62.5% (P = 0.093); for lesions over 20 cm2 and below 20 cm2 were 62.5% and 50% (P = 0.527); and for megavoltage electron beam and kilovoltage photon beam were 56.7% and 40% (P = 0.361) (Table 2). Local recurrence-free intervals (months) for male and female were 25.4 and 14.9 (P = 0.623); for head and ear, chest, shoulder and upper extremities, and abdomen and back were 37.3, 11.8, 13.7, and 13.9 (P = 0.181); for lesions over 20 cm2 and below 20 cm2 were 13.9 and 22.4 (P = 0.642); and for megavoltage electron beam and kilovoltage photon beam were 20.3 and 16.2 (P = 0.886) (Table 3).

TABLE 2 - Univariate Analysis for Recurrent Rate
Characteristic No Recurrence, n (%) Recurrence, n (%) P
Sex, lesions 0.567
 Male 8 (53.3) 7 (46.7)
 Female 11 (44) 14 (56)
Lesion Site, lesions 0.093
 Head and ear 7 (87.5) 1 (12.5)
 Chest 5 (38.5) 8 (61.5)
 Shoulder and upper extremities 4 (36.4) 7 (63.6)
 Back and abdomen 3 (37.5) 5 (62.5)
Lesion site, lesions 0.011
 Head and ear 7 (87.5) 1 (12.5)
 Others 12 (37.5) 20 (62.5)
Size, lesions 0.527
 Over 20 cm2 3 (37.5) 5 (62.5)
 Under 20 cm2 16 (50) 16 (50)
Radiotherapy modality, lesions 0.361
 Electron beam 13 (43.3) 17 (56.7)
 50 kV photon 6 (60) 4 (40)

TABLE 3 - Univariate Analysis for Local Recurrence-Free Interval
Characteristic Average LRFI, mo P
Sex, lesions 0.623
 Male 25.4
 Female 14.9
Lesion site, lesions 0.181
 Head and ear 37.3
 Chest 11.8
 Shoulder and upper extremities 13.7
 Back and abdomen 13.9
Lesion site, lesions 0.028
 Head and ear 37.3
 Others 13.7
Size, lesions 0.642
 Over 20 cm2 13.9
 Under 20 cm2 22.4
Radiotherapy modality, lesions 0.886
 Electron beam 20.3
 50 kV photon 16.2

Univariate analysis showed no significant difference for recurrence rate and LRFI, but there was a trend in lesion site where the head and ear had lower recurrence rate than other locations. According to this result, lesions were further classified into 2 groups with head and ear and others, which showed lower recurrence rate (P = 0.011) and longer LRFI (P = 0.028) with lesions over the head and ear than other locations (Table 2, Table 3, Fig. 1).

Kaplan-Meier figure for LRFI of lesion site (P = 0.028). Blue line: lesions over the head and ear. Green line: lesions over other body sites. Vertical axis: recurrence rate. Horizontal axis: LRFI (months).

Multivariate analysis showed that keloid location was a significant prognostic factor in these 4 variants for recurrence rate (P = 0.042) when lesions were classified into 2 groups according to univariate analysis results (Table 4), but not for LRFI (Table 5).

TABLE 4 - Multivariate Analysis for Recurrence Rate
Characteristic 95% Confidence Interval P
Sex 0.139–3.058 0.588
Lesion site (2 groups*) 0.010–0.923 0.042
Lesion size 0.200–5.764 0.934
Radiotherapy modality 0.345–11.548 0.440
*Two groups: head and ear, others.

TABLE 5 - Multivariate Analysis for Local Recurrence-Free Interval
Characteristic 95% Confidence Interval P
Sex 0.357–2.475 0.899
Lesion site (2 groups*) 0.017–1.042 0.055
Lesion size 0.373–2.966 0.923
Radiotherapy modality 0.211–2.456 0.600
*Two groups: head and ear, others.


Surgical excision with adjuvant radiotherapy has been used for treating keloid for decades. Radiotherapy mainly targets proliferating fibroblasts thereby inducing cell apoptosis, senescence, and cell cycle arrest,15 which decreases the amount of keloid fibroblast within normal limits16 and lowers the recurrence rate of keloid. Mankowski et al17 reviewed 72 articles, and they found that adjuvant radiotherapy offered an additional 15% control rate compared with surgery alone, whereas other studies also showed better local control with surgery and adjuvant radiotherapy for keloid patients.4 According to Japanese treatment guidelines for keloid by Ogawa et al,18 recurrence rate has no dose-dependent relationship when biologically equivalent dose (with α/β ratio of 10) over 30 Gy is used, so most facilities use adjuvant radiotherapy for keloid with 15 Gy in 3 fractions.

This study analyzed 40 keloid lesions treated by surgical excision and adjuvant radiotherapy with variants of sex, tumor location, tumor size, and treatment modality. For sex, some studies have shown that males have higher recurrence rate19,20 and less improvement after treatment,3 although no difference for recurrence rate (P = 0.567) and LRFI (P = 0.623) between males and females was found in this study. For tumor size, lesions were separated into larger and smaller ones with the cut point of 20 cm2 according to the Japan Scar Workshop Scar Scale,3 with larger tumors having higher recurrence rate without significant difference (P = 0.527). For treatment modality, a previous study showed the same recurrence rate for electron and superficial x-ray,15 and our study also showed no significant difference for recurrence rate (P = 0.361) and LRFI (P = 0.886) between these 2 treatment modalities, which is similar to previous data.17 Our study showed that superficial x-ray had a trend of lower recurrence rate, but this might be due to selection bias for more head and ear lesions in the superficial x-ray group. Mankowski et al17 found that brachytherapy had the lowest recurrence rate (15%) compared with electron and x-ray radiotherapy (23%), but brachytherapy is performed interstitially mostly, which is more invasive for patients, and medical staff are more easily exposed to radiation. For lesion site, a previous study showed higher recurrence rate at the anterior chest wall, scapular and suprapubic regions, and lower recurrence rate over the ear lobe and other sites.21 Our study showed no recurrence difference over the head and ear, chest, shoulder and upper extremities, and abdomen and back (P = 0.093), although head and ear lesions had a trend of lower recurrence rate than the other 3 locations, so lesions were further classified into 2 categories: head and ear lesions, and others. Subsequently, it was found that head and ear lesions had significantly lower recurrence rate than lesions over other locations by both univariant (P = 0.011) and multivariant analyses (P = 0.042).

There is no well-established mechanism of keloid formation so far4,11; as a consequence, this study assumes some hypotheses to explain why tumor location might play a role in keloid tumor recurrence. First, tension is known to be a risk factor for keloid formation.7,8,10,22 Ogawa et al21 presented the recurrence rate of 38.3% in high-tension areas with adjuvant electron beam radiotherapy and 5.7% to 17.3% in other areas, which also provides clinical evidence that tension might play a role in diversity of recurrence. Second, different thickness of reticular dermis over different locations might affect keloid recurrence, because reticular dermis is where keloid-genic inflammatory cellular activities take place,21 and different thicknesses of reticular dermis might contain different amounts of keloid fibroblasts. Third, different properties of ECM or inflammation cells might have some impact on keloid recurrence. According to a study by Butzelaar et al,23 mechanical properties of skin vary with anatomic location and depend largely on composition of ECM. These differences are also found in vascularization and resident immune cell populations, and site-specific variations in ECM properties as well as macrophage numbers exist. They also found that predilection sites for keloid formation contain larger amounts of collagen compared with nonpredilection sites.23

All the aforementioned reasons might play a role in the mechanism of keloid formation. Finally, different activity of fibroblasts from different body sites might affect keloid recurrence. Previous studies have shown that keloid fibroblast is more sensitive to inflammatory factors than normal skin fibroblast.13 To find whether different activity of fibroblasts from different body sites exists, we compared cell survival of keloid fibroblasts from our patients with normal fibroblast by the MTT method after irradiation to culture medium, which showed higher survival rate for normal fibroblast. Keloid fibroblasts were found to be more radiosensitive than normal fibroblasts, indicating that difference of radiosensitivity exists between fibroblasts. The authors are now collecting keloid fibroblasts from different body sites in an attempt to determine whether fibroblasts from different body sites play a role in keloid recurrence while also trying to find the specific proper radiation dose for keloid over different body sites.

There are some advantages in our study. Enrolled patients all received the same radiation dose, having the first radiotherapy within 24 hours after surgery as per the suggestion of previous studies.4,7 Local recurrence-free interval was also analyzed as our end point, which provided more information than previous studies. Unfortunately, the limitations in our study include a small sample of only 40 lesions with relatively short follow-up time, as well as limited record of adverse effects able to be accessed. Patients showed higher recurrence rates than other studies, especially for lesions over sites other than the head and ear (Table 6), and we assume this might be due to other studies excluding patients with high risk factors such as multiple lesions or family history, and so on. Other than this, some patients did not receive regular follow-up for proper postoperative treatment and follow-up care because keloid is a benign disease, especially while the hospital was under the pandemic of COVID-19, which might possibly have increased the risk of recurrence. Last, because this is a retrospective study, some data are missing. Bias might also exist, especially when determining recurrence and LRFI.

TABLE 6 - Literature Review of Keloid Patients Receiving Surgery and Adjuvant Radiotherapy
Year, Author Radiation Dose (Gray) Results
2006, Slemp and Kirschner 16 15–20 Control rate 65%–99%
2009, Ogawa et al 22 Not mentioned Response rate 67%–98%
2015, Lee and Park 7 12–18 Nonrecurrence rate 81.1%
2015, Shen et al 20 18 Nonrecurrence rate 90.4%
2018, Renz et al 1 12–20 Nonrecurrence rate 94.4%
2019, Petrou et al 24 15 Nonrecurrence rate 94%
2019, Ogawa et al 21 15 for ear Nonrecurrence rate 85%
15 for neck Nonrecurrence rate 83%
15 for chest wall Nonrecurrence rate 61%
15 for scapula Nonrecurrence rate 62%
10–15 for ear Nonrecurrence rate 89%
20 for chest wall Nonrecurrence rate 86%
20 for scapula Nonrecurrence rate 89%
2019, Hsueh et al 3 20 for high tension area Nonrecurrence rate 68%
12 for other area

Recently, studies have found lower recurrence rate with radiation total dose over 20 Gy or biologically equivalent dose (with α/β ratio of 2) over 60 Gy,1 especially for lesions over the anterior chest, scapula, and suprapubic areas.21 Higher doses can provide better local control, but might carry high risk of acute and delayed adverse effects, including secondary malignancy.16,21 We are now modifying our adjuvant radiation dose to keloid lesions over different body sites in order to determine site-specific dosages for keloid patients by clinical and basic research.

To conclude, in the retrospective review, it was found that lesion sites might be a prognostic factor for keloid recurrence. Keloid tumors over the head and ear area achieved good local control rate (87.5%) with adjuvant radiotherapy with 15 Gy. Adjuvant radiation dose escalation for high-recurrence risk areas (other than head and ear) might be required.


The authors thank Hsiu-Ying Ku, PhD, for the assistance with statistical consultation.


1. Renz P, Hasan S, Gresswell S, et al. Dose effect in adjuvant radiation therapy for the treatment of resected keloids. Int J Radiat Oncol Biol Phys. 2018;102:149–154.
2. Ogawa R, Akaishi S, Hyakusoku H. Differential and exclusive diagnosis of diseases that resemble keloids and hypertrophic scars. Ann Plast Surg. 2009;62:660–664.
3. Hsueh WT, Hung KS, Chen YC, et al. Adjuvant radiotherapy after keloid excision: preliminary experience in Taiwan. Ann Plast Surg. 2019;82(1S Suppl 1):S39–S44.
4. Perez CA, Brady LW, Wazer DE, et al. Perez and Brady's Principles and Practice of Radiation Oncology. 7th ed. Philadelphia, PA: Wolters Kluwer; 2019.
5. Maemoto H, Iraha S, Arashiro K, et al. Risk factors of recurrence after postoperative electron beam radiation therapy for keloid: comparison of long-term local control rate. Rep Pract Oncol Radiother. 2020;25:606–611.
6. Glass DA 2nd. Current understanding of the genetic causes of keloid formation. J Investig Dermatol Symp Proc. 2017;18:S50–S53.
7. Lee SY, Park J. Postoperative electron beam radiotherapy for keloids: treatment outcome and factors associated with occurrence and recurrence. Ann Dermatol. 2015;27:53–58.
8. Akaishi S, Akimoto M, Ogawa R, et al. The relationship between keloid growth pattern and stretching tension: visual analysis using the finite element method. Ann Plast Surg. 2008;60:445–451.
9. Ogawa R, Mitsuhashi K, Hyakusoku H, et al. Postoperative electron-beam irradiation therapy for keloids and hypertrophic scars: retrospective study of 147 cases followed for more than 18 months. Plast Reconstr Surg. 2003;111:547–553; discussion 554–545.
10. Ogawa R. Keloid and hypertrophic scars are the result of chronic inflammation in the reticular dermis. Int J Mol Sci. 2017;18.
11. Nangole FW, Agak GW. Keloid pathophysiology: fibroblast or inflammatory disorders? JPRAS Open. 2019;22:44–54.
12. Tripathi S, Soni K, Agrawal P, et al. Hypertrophic scars and keloids: a review and current treatment modalities. Biomedical Dermatology. 2020;4.
13. Xu J, Yang E, Yu N-Z, et al. The radiation therapy in keloids treatment: a comprehensive review of pathomechanism, damage mechanisms and cellular response. Plastic and Aesthetic Research. 2017;4.
14. Lemperle G, Schierle J, Kitoga KE, et al. Keloids: which types can be excised without risk of recurrence? A new clinical classification. Plast Reconstr Surg Glob Open. 2020;8:e2582.
15. Cheraghi N, Cognetta A Jr., Goldberg D. Radiation therapy for the adjunctive treatment of surgically excised keloids: a review. J Clin Aesthet Dermatol. 2017;10:12–15.
16. Slemp AE, Kirschner RE. Keloids and scars a review of keloids and scars, their pathogenesis, risk factors, and management. Curr Opin Pediatr. 2006;18:396–402.
17. Mankowski P, Kanevsky J, Tomlinson J, et al. Optimizing radiotherapy for keloids: a meta-analysis systematic review comparing recurrence rates between different radiation modalities. Ann Plast Surg. 2017;78:403–411.
18. Ogawa R, Akita S, Akaishi S, et al. Diagnosis and treatment of keloids and hypertrophic scars—Japan Scar Workshop Consensus Document 2018. Burns Trauma. 2019;7:39.
19. Bijlard E, Verduijn GM, Harmeling JX, et al. Optimal high-dose-rate brachytherapy fractionation scheme after keloid excision: a retrospective multicenter comparison of recurrence rates and complications. Int J Radiat Oncol Biol Phys. 2018;100:679–686.
20. Shen J, Lian X, Sun Y, et al. Hypofractionated electron-beam radiation therapy for keloids: retrospective study of 568 cases with 834 lesions. J Radiat Res. 2015;56:811–817.
21. Ogawa R, Tosa M, Dohi T, et al. Surgical excision and postoperative radiotherapy for keloids. Scars Burn Heal. 2019;5:2059513119891113.
22. Ogawa R, Yoshitatsu S, Yoshida K, et al. Is radiation therapy for keloids acceptable? The risk of radiation-induced carcinogenesis. Plast Reconstr Surg. 2009;124:1196–1201.
23. Butzelaar L, Niessen FB, Talhout W, et al. Different properties of skin of different body sites: the root of keloid formation? Wound Repair Regen. 2017;25:758–766.
24. Petrou IG, Jugun K, Ruegg EM, et al. Keloid treatment: what about adjuvant radiotherapy? Clin Cosmet Investig Dermatol. 2019;12:295–301.

keloid; lesion site; recurrence; radiotherapy; surgery

Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc.