Abstract
Perioperative antibiotic treatment duration in skin reconstruction with dermal substitutes is not well established. This study compares the incidence of infective complications after two different durations of perioperative antibiotic treatment in patients undergoing surgical reconstruction with skin dermal substitutes (SDS) after excision of skin cancer. Infective complications at the site of SDS were compared in subjects undergoing surgical reconstruction who received either a > 24‐hour (extended protocol) or a ≤ 24‐hour (short protocol) perioperative antibiotic treatment. Of 116 patients undergoing SDS surgical reconstruction, 62 (53.4%) received an extended schedule, and 54 (46.6%) received a short schedule. The two groups were similar for gender, age, comorbidities, American Society of Anesthesiologists score, and type of skin cancer. Overall incidence rate of infection was 20.7% (24/116). No differences in terms of risk of infection were observed between the two groups (OR: 1.04, 95% CI: 0.42‐2.55; P = .937). Patients undergoing SDS reconstruction in the limb/foot had a higher risk of infection in comparison with those undergoing SDS reconstruction in the chest/head (OR: 2.69, 95% CI: 1.06‐6.86; P = .038). The short protocol was demonstrated to be beneficial to patients undergoing surgical reconstruction with SDS. A ≤ 24‐hour perioperative antibiotic schedule did not increase the infection rate, potentially allowing a reduction of antibiotic exposure.
Keywords: dermal matrix, perioperative antibiotic schedule, skin cancer, surgical reconstruction
1. INTRODUCTION
Bioengineered skin dermal substitutes (SDS), such as Integra or Matriderm, represent a novel therapeutic option for the treatment of damaged tissue for which autograft reconstruction is not feasible.1, 2, 3, 4, 5, 6
SDS provide a 3D matrix at the ulcer bed and help with focal vascularisation and fibroblast activity, which in turn leads to the optimal and natural development of skin tissue.7, 8, 9
While clinical management of SDS is well known, it is not the same for perioperative antibiotic schedules in these settings. Indeed, very limited data are available for infectious complications following SDS surgical reconstruction, particularly for patients affected by skin cancer. A very wide range of the incidence (0%‐85%) of infection has been reported, 10 and infections are associated with a potential risk of graft loss. 11 However, there is general agreement on the importance of an effective perioperative antibiotic treatment in plastic surgery.12, 13, 14, 15, 16 In the absence of specific guidelines on the issue, however, different protocols are being implemented, and many of the antibiotics are administered for several days after surgery, with potential consequences because of antibiotic overuse. 17 In a different setting (breast reconstruction), 24 hours of antibiotics were equivalent to extended oral antibiotics treatment, 18 and similar results were reported in other studies.19, 20, 21
2. CLINICAL PROBLEM ADDRESSED
To our knowledge, the optimal length of perioperative antibiotic treatment (extended vs short) in patients undergoing surgical reconstruction with SDS after excision of skin cancer has never been investigated.
The aim of the present study was to compare the incidence of infectious complications following two different durations of perioperative antibiotic treatment (>24 hour, extended protocol; ≤24 hour, short protocol) in patients undergoing surgical reconstruction with SDS after excision of skin cancer.
3. PATIENTS AND METHODS
3.1. Study setting and design
We reviewed all medical records of patients admitted to Policlinico of Modena (Modena, Italy) for surgical skin reconstruction with SDS after excision of skin cancer between September 2014 and June 2017. In these patients, the width of the surgical excision, the presence of ulceration, and the implantation of a medical device make antibiotic treatment necessary, as suggested by current recommendations.16, 22
In March 2016, the intra‐hospital “Committee on the Good Use of Antibiotics” of the University Hospital “Policlinico of Modena” modified the perioperative antibiotic protocol for dermatological procedures in order to reduce the exposure of patients to antimicrobials and, consequently, the antimicrobial resistance. In the new protocol, the antibiotic schedule for SDS implantation has been reduced from 7 days to a single dose before the surgical procedure.
We compared the incidence of infectious complications in patients undergoing surgical reconstruction with SDS before and after the establishment of the new perioperative antibiotic protocol.
Any surgical skin reconstruction with the application of SDS was deemed eligible for the study. Both Integra (Integra Life Sciences, Plainsboro, New Jersey) and Matriderm (MedSkin Solutions, Billerbeck, Germany) dermal substitutes were used in a classic two‐stage operative procedure.
3.2. Procedures
Patients underwent reconstruction with SDS after wide local excision for skin cancer according to appropriate surgical guidelines.23, 24 All wounds were prepared in a standardised way. Cases of full‐thickness scalp resection comprehensive of periosteum with concomitant blurring of the outer cortex could be included. After a Betadine wash, the Integra or Matriderm sheet was applied to the surgical wound and then trimmed to size and secured in place with surgical staples. Before dermal matrix application, each wound (maximum width and length) was measured to estimate the size of the defect. When the size of the defect was larger than a standard‐sized dermal matrix sheet, a further sheet could be used to fill the resulting defect. A non‐adherent gauze dressing was applied on the substitute, and a compression dressing was also applied.
3.3. Antibiotic treatment: extended vs short protocol
Perioperative antibiotics were prescribed in all cases, but patients followed two different regimens in accordance with two study periods. Between September 2014 and February 2016 (period A), the duration of antibiotic treatment was 7 days (extended protocol: cefazolin 2 g intravenous [IV] before the skin incision, followed by one of the following 6‐day regimens: amoxicillin/clavulanate 1 g orally every 8 hours, or ceftriaxone 1 g intramuscularly every 24 hours, or cefixime 400 mg orally every 24 hours). Instead, between March 2016 and June 2017 (period B), the perioperative antibiotics were administered for up to 24 hours (short protocol: a single dose of cefazolin 2 g IV before the skin incision; for patients with an American Society of Anesthesiologists (ASA) score ≥3, perioperative protocol recommended additional doses of 1 g IV cefazolin after 6 hours and every 8 hours for up to 24 hours).
3.4. Postoperative period
During the postoperative period, all patients were followed up in a specific dermatologic outpatient clinic dedicated to wound care management by qualified personnel for dermal matrix management. Patients were regularly medicated twice a week with povidone‐iodine gel at the edge of the wound, a non‐adherent gauze on the matrix, and a compression dressing. In cases of haematoma or seroma formation between the wound bed and the template, fluid collections were promptly evacuated. With signs of neovascularisation and neodermis formation, the silicone top layer of Integra was removed, on average, on day 21 after its application. In case of reconstruction with Matriderm, the same dressing was applied twice a week until a neodermis was formed, on average, 2 to 3 weeks after dermal matrix application. In case of dermal matrix infection or need for surgical revision because of involved surgical margins, a further dermal substitute could be applied after surgical wound revision.
The time span between dermal matrix application and second surgical procedure was variable, between 20 and 30 days, according to wound conditions and patient's status. For most cases, a split‐thickness skin graft, usually taken from a lateral aspect of the thigh, was positioned on the neodermis and secured with surgical staples and standard dressings.
Finally, patients underwent weekly dressing changes in a hospital setting until complete wound healing was achieved.
3.5. Diagnosis of infection
Diagnosis of infection, at the dermal substitute site, was based on local signs of infection (pain, redness, tenderness, warmth, swelling, delayed healing), with or without the presence of systemic signs of infection (ie, fever, hypotension). When local signs of infection were present, a microbiologic swab for culture from the surgical site was performed. Antimicrobial treatment was prescribed based on susceptibility tests and on the hospital antimicrobial resistance rates.
3.6. Statistical Analysis
The incidence of infection was compared between the patients receiving an extended or a short perioperative antibiotic treatment.
All variables of interest were summarised by frequencies and percentage values or by mean values and the relative standard deviation. The two groups were compared by using the non‐parametric Pearson's χ 2 test or Mann‐Whitney test, when appropriate. Univariate logistic regression models were performed to explore potential risk factors for infection; we considered the following risk factors: gender, age, ASA score (<3 or ≥3), perioperative antibiotic treatment (≤24 hours, short protocol or >24 hours, extended protocol), site of SDS reconstruction (head/chest or limb), type of SDS (Integra or Matriderm), presence of chronic renal impairment, presence of diabetes mellitus, presence of hypertension, smoking status, and dimensions of surgical area (<20 cm2; 20‐60 cm2; and >60 cm2).
Analysis was performed using IBM SPSS Statistics 21.0 (IBM, Armonk, New York). Significance threshold was set at P < .05.
4. RESULTS
4.1. Patient populations
Between September 2014 and June 2017, 116 patients underwent a surgical intervention with SDS reconstruction. In particular, 62 patients received a > 24‐hour perioperative antibiotic schedule (extended protocol), while 54 followed the short perioperative antibiotic protocol. The average duration of antibiotic treatment in the two groups of patients was 6.6 days and 0.5 day, respectively.
Demographic characteristics and clinical conditions of the whole sample, stratified by the two groups, are summarised in Table 1. A total of 77 patients (66.4%) were males, and the mean age was 73 years (±15.1). Overall, 78 patients (67.2%) had hypertension, 20 (17.2%) had diabetes mellitus, 16 (13.8%) had chronic renal impairment, 22 (19%) were former or current smokers, and 45 (38.8%) had an ASA score ≥3. The two groups were comparable for all characteristics except for a higher use of Integra in the group receiving an extended antibiotic protocol. A total of 110 patients (94.8%) underwent surgery for skin cancer: 45 patients (38.8%) had squamous cell carcinoma, 34 (29.3%) had basal cell carcinoma, 12 (10.3) had melanoma, and 19 (16.4%) had another type of skin cancer.
TABLE 1.
Characteristics of patients who underwent SDS intervention who received perioperative antibiotic treatment for ≤24 hours (short protocol) and for >24 hours (extended protocol)
| Characteristics | Overall (n = 116), n (%) | Short protocol (n = 54), n (%) | Extended protocol (n = 62), n (%) | P‐value |
|---|---|---|---|---|
| Gender | .102 | |||
| Female | 29 (34) | 14 (26) | 25 (40) | |
| Male | 77 (66) | 40 (74) | 37 (60) | |
| Age (years): | .763 | |||
| Mean (±SD) | 72.6 (15.2) | 72.9 (15.3) | 72.8 (15.2) | |
| ASA score | .244 | |||
| <3 | 71 (61) | 30 (56) | 41 (66) | |
| ≥3 | 45 (39) | 24 (44) | 21 (34) | |
| Underline disease | .494 | |||
| SCC | 45 (39) | 20 (37) | 25 (40) | |
| BCC | 34 (29) | 20 (37) | 14 (23) | |
| Melanoma | 12 (10) | 5 (9) | 7 (11) | |
| Other skin cancer | 19 (17) | 7 (13) | 12 (19) | |
| No cancer | 6 (5) | 2 (4) | 4 (7) | |
| Site of SDS | .072 | |||
| Head/chest | 83 (71) | 43 (77) | 40 (65) | |
| Limbs | 33 (29) | 11 (23) | 22 (35) | |
| Type of SDS | .001 | |||
| Integra | 98 (84) | 39 (72) | 59 (95) | |
| Matriderm | 18 (16) | 15 (28) | 3 (5) | |
| Infection | .937 | |||
| Yes | 24 (21) | 11 (20) | 13 (21) | |
| No | 92 (79) | 43 (80) | 49 (79) | |
| Diabetes mellitus | .518 | |||
| Yes | 20 (17) | 8 (15) | 12 (19) | |
| No | 96 (83) | 46 (85) | 50 (81) | |
| Renal impairment | .434 | |||
| Yes | 16 (14) | 6 (11) | 10 (16) | |
| No | 100 (86) | 48 (89) | 52 (84) | |
| Hypertension | .360 | |||
| Yes | 78 (67) | 34 (63) | 44 (71) | |
| No | 38 (33) | 20 (37) | 18 (29) | |
| Smoking status | .909 | |||
| Former/current smoker | 22 (19) | 10 (19) | 12 (19) | |
| No smoker | 94 (81) | 44 (81) | 50 (81) |
Note: Bold represents statistical significance.
Abbreviations: ASA, American Society of Anesthesiologists; BCC, basal cell carcinoma; SCC, squamous cell carcinoma; SDS, skin dermal substitutes.
Surgical SDS reconstruction involved head/chest site in 83 cases (71%) and limbs in 33 (29%) cases. In particular, patients presented with lesions on the scalp in 44 cases (37.9%), on the face in 28 cases (24.1%), on the chest in 11 cases (9.5%), on the arm or the hand in 9 cases (7.8%), on the leg in 12 cases (10.3%), and on the foot in 12 cases (10.3%) (Table 1).
4.2. SDS outcomes
SDS intervention required a median duration of hospitalisation of 2 days (range: 1‐11 days). The grafting of the SDS has occurred in 87 patients (75.0%) without complications. In 6 (5.1%) patients, surgical SDS reconstruction was complicated by haematoma, and in 24 (20.7) patients, it was complicated by surgical site infection.
4.3. Incidence of infections
The overall incidence of infection at the site of SDS was 20.7% (24 episodes/116 patients).
No differences in the infection rate were observed between the patients who received >24‐hour perioperative antibiotic treatment and those who received ≤24‐hour perioperative antibiotic treatment (21% vs 20%; P = .937).
Univariate logistic regression models were performed to individuate potential risk factors for infection (Table 2). SDS surgical reconstruction at the limbs, in comparison with the chest/head, was associated with a higher infection rate (OR: 2.69, 95% CI: 1.06‐6.86; P = .038).
TABLE 2.
Univariate logistic regression model of the risk factors for infection of patients who underwent surgical SDS reconstruction at the University Hospital of Modena (September 2014 to June 2017)
| Characteristics | OR | 95% CI | P‐value |
|---|---|---|---|
| Gender | |||
| Female | R.C. | ||
| Male | 0.81 | 0.32‐2.05 | .652 |
| Age (continuous) | 1.01 | 0.98‐1.04 | .603 |
| Perioperative antibiotic treatment | |||
| Short protocol | R.C. | ||
| Extended protocol | 1.04 | 0.42‐2.55 | .937 |
| ASA score | |||
| <3 | R.C. | ||
| ≥3 | 0.93 | 0.37‐2.36 | .933 |
| Underline disease | |||
| SCC | R.C. | ||
| BCC | 1.46 | 0.53‐4.04 | .468 |
| Other skin cancer | 0.52 | 0.15‐1.83 | .308 |
| No cancer | Not estimable | ||
| Site od SDS | |||
| Head/chest | R.C. | ||
| Limbs | 2.69 | 1.06‐6.86 | .038 |
| Type of SDS | |||
| Matriderm | R.C. | ||
| Integra | 1.36 | 0.36‐5.16 | .648 |
| Renal impairment | |||
| No | R.C. | ||
| Yes | 2.73 | 0.88‐8.49 | .082 |
| Diabetes mellitus | |||
| No | R.C. | ||
| Yes | 1.35 | 0.44‐4.18 | .602 |
| Hypertension | |||
| No | R.C. | ||
| Yes | 1.23 | 0.46‐3.29 | .674 |
| Smoke | |||
| No smoker | R.C. | ||
| Former/current smoker | 2.11 | 0.75‐5.98 | .158 |
| Surgical area | |||
| <20 cm2 | R.C. | ||
| 20‐60 cm2 | 2.20 | 0.81‐5.99 | .123 |
| ≥60 cm2 | 0.74 | 0.14‐3.99 | .729 |
Abbreviations: ASA, American Society of Anesthesiologists; BCC, basal cell carcinoma; SCC, squamous cell carcinoma; SDS, skin dermal substitutes.
4.4. Isolates
The most common pathogen isolated was Staphylococcus aureus (n = 10; 41.6%), followed by Pseudomonas aeruginosa (n = 7; 29.2%), Proteus mirabilis (n = 2; 8.4%), Enterococcus faecalis (n = 1; 4.2%), Corynebacterium striatum (n = 1; 4.2%), and Bacillus spp. (n = 1; 4.2%) (Figure 1). In two cases (8.4%), the microbiologic swab demonstrated negative results. Of 24 patients who developed infection at the site of SDS, only 3 had an infection because of multidrug‐resistant (MDR) microorganisms. Two cases (0.017%) were infected by methicillin‐resistant S. aureus (one in the extended protocol group and one in the short protocol group) and one case (0.008%) by P. mirabilis resistant to carbapenems (occurred in the extended protocol group). In addition, for the MDR microorganism infection rate, no differences were observed between the patients who received an extended perioperative antibiotic treatment and those who received a short perioperative antibiotic schedule.
FIGURE 1.
Prevalence of microorganisms responsible for local infection in the patients undergoing surgical positioning of skin dermal substitutes at the University Hospital of Modena (September 2014 to June 2017)
5. DISCUSSION
Surgical excision and reconstruction with SDS of large, nodular, ulcerated skin cancer—which is normally super‐infected—is considered a contaminated surgery with device implantation requiring perioperative antibiotic treatment. The administration of perioperative antibiotic also has the aim of reducing the risk of SDS failure, which is primarily caused by the onset of infection. 11
To our knowledge, this is the first study that compared two different protocols of perioperative antibiotic treatment in patients undergoing SDS reconstruction for excision of skin cancer, namely, an extended (7 days) and a short (<24 hours) protocol. Comparing the two patient groups, no differences in the rate of infection (20% vs 21%) were found. This result is of major relevance as it suggests that prolongation of perioperative antibiotics in this type of surgical patients, at an inherently high risk of infection, does not reduce the rate of infection. A short schedule for perioperative antibiotic treatment allowed the reduction of antibiotic exposure by 6 days per patient, with potential benefits in terms of health care expenditure and reduced onset of resistance.
Furthermore, we evaluated factors associated with a higher risk of infection. Remarkably, the only factor associated with a higher risk of infection was reconstruction at the limbs, compared with that of chest/head. Other factors, such as ASA score ≥3 or dimension of the surgical bed, were not associated with a higher risk of infection.
Our study has limitations. First, it is a retrospective, non‐randomised study; moreover, the extended protocol was not standardised, and patients received different antibiotic agents, in most cases amoxicillin/clavulanic acid.
5.1. Innovation
The study suggests a hypothesis that has never been explored before: in patients undergoing surgical SDS reconstruction after excision of skin cancer, an extended perioperative antibiotic schedule (7 days) does not reduce the incidence of surgical site infection compared with a short protocol (1 day) of antibiotic treatment. Randomised controlled trials, with larger populations, could give a more accurate response to the optimal duration of perioperative antibiotic treatment in patients undergoing surgical SDS reconstruction.
CONFLICT OF INTEREST
The authors declare no conflicts of interest.
ACKNOWLEDGEMENTS
Editorial assistance was provided by Aashni Shah (Polistudium SRL, Milan, Italy), and this study was supported by internal funds.
Fiorentini C, Bedini A, Mandel VD, et al. Comparison of two perioperative antibiotic schedules in patients undergoing surgical reconstruction with dermal matrix after excision of skin cancer. Int Wound J. 2020;17:937–943. 10.1111/iwj.13354
REFERENCES
- 1. O'Brien FJ. Biomaterials & scaffolds for tissue engineering. Mater Today. 2011;14(3):88‐95. [Google Scholar]
- 2. Langer R, Vacanti J. Tissue engineering. Science. 1993;260(5110):920. [DOI] [PubMed] [Google Scholar]
- 3. Edlich RF, Winters KL, Woodard CR, Britt LD, Long WB III. Massive soft tissue infections: necrotizing fasciitis and purpura fulminans. J Long Term Eff Med Implants. 2005;15:57‐65. [DOI] [PubMed] [Google Scholar]
- 4. Frame JD, Still J, Lakhel‐LeCoadou A, et al. Use of dermal regeneration template in contracture release procedures: a multicenter evaluation. Plast Reconstr Surg. 2004;113:1330‐1338. [DOI] [PubMed] [Google Scholar]
- 5. Wainwright DJ. Use of an acellular allograft dermal matrix (AlloDerm) in the management of full‐thickness burns. Burns. 1995;21:243‐248. [DOI] [PubMed] [Google Scholar]
- 6. Magnoni C, De Santis G, Fraccalvieri M, et al. Integra in scalp reconstruction after tumor excision: recommendations from a multidisciplinary advisory board. J Craniofac Surg. 2019;30(8):2416‐2420. [DOI] [PubMed] [Google Scholar]
- 7. van der Veen VC, van der Wal MBA, van Leeuwen MCE, Ulrich MMW, Middelkoop E. Biological background of dermal substitutes. Burns J Int Soc Burn Inj. 2010;36(3):305‐321. [DOI] [PubMed] [Google Scholar]
- 8. Böttcher‐Haberzeth S, Biedermann T, Reichmann E. Tissue engineering of skin. Burns J Int Soc Burn Inj. 2010;36(4):450‐460. [DOI] [PubMed] [Google Scholar]
- 9. Andreassi A, Bilenchi R, Biagioli M, D'Aniello C. Classification and pathophysiology of skin grafts. Clin Dermatol. 2005;23(4):332‐337. [DOI] [PubMed] [Google Scholar]
- 10. Widjaja W, Tan J, Maitz PKM. Efficacy of dermal substitute on deep dermal to full thickness burn injury: a systematic review. Anz J Surgery. 2017;87(6):446‐452. [DOI] [PubMed] [Google Scholar]
- 11. Ferrari B, Reggiani C, Francomano M, et al. Clinical factors influencing the outcomes of a cellular dermal matrix for treatment of skin cancers: a retrospective study. Adv Skin Wound Care. 2019; (In Press). [DOI] [PubMed] [Google Scholar]
- 12. Wilensky JS, Rosenthal AH, Bradford CR, Rees RS. The use of a bovine collagen construct for reconstruction of full‐thickness scalp defects in the elderly patient with cutaneous malignancy. Ann Plast Surg. 2005;54:297‐301. [PubMed] [Google Scholar]
- 13. Chalmers RL, Smock E, Geh JLC. Experience of Integra® in cancer reconstructive surgery. J Plast Reconstr Aesthetic Surg. 2010;63:2081‐2090. [DOI] [PubMed] [Google Scholar]
- 14. Muangman P, Deubner H, Honari S, et al. Correlation of clinical outcome of integra application with microbiologic and pathological biopsies. J Trauma. 2006;61:1212‐1217. [DOI] [PubMed] [Google Scholar]
- 15. Khan MA, Ali SN, Farid M, Pancholi M, Rayatt S, Yap LH. Use of dermal regeneration template (Integra) for reconstruction of full‐thickness complex oncologic scalp defects. J Craniofac Surg. 2010;21:905‐909. [DOI] [PubMed] [Google Scholar]
- 16. Ariyan S, Martin J, Lal A, et al. Antibiotic prophylaxis for preventing surgical‐site infection in plastic surgery: an evidence‐based consensus conference statement from the American Association of Plastic Surgeons. Plast Reconstr Surg. 2015;135(6):1723‐1739. 10.1097/PRS.0000000000001265. [DOI] [PubMed] [Google Scholar]
- 17. Bae‐Harboe YS, Liang CA. Perioperative antibiotic use of dermatologic surgeons in 2012. Dermatol Surg. 2013;39(11):1592‐1601. [DOI] [PubMed] [Google Scholar]
- 18. Phillips BT, Fourman MS, Bishawi M, et al. Are prophylactic postoperative antibiotics necessary for immediate breast reconstruction? Results of a prospective randomized clinical trial. J Am Coll Surg. 2016;222(6):1116‐1124. 10.1016/j.jamcollsurg.2016.02.018. [DOI] [PubMed] [Google Scholar]
- 19. Sevin A, Senen D, Sevin K, Erdogan B, Orhan E. Antibiotic use in abdominoplasty: prospective analysis of 207 cases. J Plast Reconstr Aesthet Surg. 2007;60:379‐382. [DOI] [PubMed] [Google Scholar]
- 20. Caniello M, Passerotti GH, Goto EY, Voegels RL, Butugan O. Antibiotics in septoplasty: is it necessary? Braz J Otorhinolaryngol. 2005;71:734‐738. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Man LX, Beswick DM, Johnson JT. Antibiotic prophylaxis in uncontaminated neck dissection. Laryngoscope. 2011;121:1473‐1477. [DOI] [PubMed] [Google Scholar]
- 22. Surgical Antimicrobial Prophylaxis Clinical Guideline. Version 2.0. Adelaide, AU: Government of South Australia; 2017. www.sahealth.sa.gov.au/wps/wcm/connect/6bb523804358edbd883b9ef2cadc00ab/Surgical+Antimicrobial+Prophylaxis_v2.0_23112017.pdf?MOD=AJPERES&CACHEID=ROOTWORKSPACE‐6bb523804358edbd883b9ef2cadc00ab‐mMHP6uy. [Google Scholar]
- 23. Trakatelli M, Morton C, Nagore E, et al. Update of the European guidelines for basal cell carcinoma management. Eur J Dermatol. 2014;24(3):312‐329. 10.1684/ejd.2014.2271. [DOI] [PubMed] [Google Scholar]
- 24. Stratigos A, Garbe C, Lebbe C, et al. Diagnosis and treatment of invasive squamous cell carcinoma of the skin: European consensus‐based interdisciplinary guideline. Eur J Cancer. 2015;51(14):1989‐2007. 10.1016/j.ejca.2015.06.110. [DOI] [PubMed] [Google Scholar]