Introduction

Nasal defects occur following excision of tumors, namely Basal cell carcinoma, Squamous cell carcinoma, melanomas, sarcoma, lymphomas, sweat gland carcinoma and other benign/inflammatory growth, trichoepithelioma, arteriovenous malformations, rhinophyoma; trauma; insect, animal or human bite; post burns (thermal, electrical, chemical); skin necrosis following radiotherapy, sepsis; cosmetic removal of various skin lesions, nevi; congenital craniofacial deformities.

A wide variety of flaps have been devised for nasal defect reconstruction depending upon the nasal subunit(s) involved, extent and size and the layers involved. However, the paramedian forehead flap and the nasolabial flap have been explored far more than other reconstruction modalities, hence they may be considered the workhorse for nasal defect reconstruction. The proponents of either flap had enumerated various pearls and pitfalls of their flaps. Some authors have also compared the two flaps in their studies [1,2,3,4,5,6,7,8,9]. But more often their outcomes and conclusions were not coherent to one another.

Our primary objective was to estimate the frequency of each type of nasal subunit being reconstructed with either the paramedian forehead flap or the nasolabial flap, and also the incidences of various post-operative complications in each group. Our secondary objective was to determine the relative risk of various complications among both the groups of flap reconstruction. The complications included partial or total flap necrosis, alar notching, hematoma or bleeding, following reconstruction with either these flaps.

Materials and Method

Search Strategy

This systematic review was registered with the international prospective register of systematic reviews (PROSPERO), adhering to the standards of the preferred reporting items for systematic reviews and meta-analysis (PRISMA) guidelines. An electronic database search on PubMed, Google Scholar and the Cochrane Library was conducted on December, 2020 using a combination of both Medical Subject Heading (MeSH) terms and plain text related to nasal reconstruction. Studies limited to humans and published in English language were searched since inception. The syntax used for search strategy was as follows:

PubMed: ("paramedian forehead flap") OR ("nasolabial flap") OR ("melolabial flap") OR ("nasal reconstruction").

The Cochrane Library: "paramedian forehead flap" in Title Abstract Keyword OR "nasolabial flap" in Title Abstract Keyword OR "melolabial flap" in Title Abstract Keyword OR "nasal reconstruction" in Title Abstract Keyword.

Google Scholar: "paramedian forehead flap" OR "nasolabial flap" OR "melolabial flap" OR "nasal reconstruction" (excluding patents and citations).

The manuscripts were reviewed manually by two independent authors (SSC, NS) to identify appropriate studies. Duplicate studies were removed. References of appropriate articles were also screened to identify additional related studies. Predetermined inclusion and exclusion criteria (Table 1) were applied for eligibility for inclusion. In case of any discrepancy, a consensus was formed by mutual discussion with other reviewers.

Table 1 Inclusion and exclusion criteria
Full size table

Data Extraction

Two independent reviewers (SA, AKM) extracted the data independently from the included studies in a standardised Data Extraction sheet for various parametersusing Microsoft® Excel®. In case of any discrepancy, a consensus was formed by mutual discussion with other reviewers. The data extracted includes demographic details (Title, authors, country of origin, year of publication, type of study, level of evidence), population details (Total number of patients, number of males/females, number of flaps done and followed-up, patients’ age, comorbidities and addiction), perioperative details (Flaps performed, with any modifications, defect size, subunit reconstructed, follow-up duration), outcomes (Functional outcomes measured by methods like questionnaire, NAFEQ scores, quality of speech, total nasal function; Aesthetic outcomes assessed by questionnaire, visual analog scale, pigmentation, contour deformity, Derriford Appearance Scale 24, height, colour etc.),and complications (necrosis, dehiscence, pin cushioning, hematoma, infection, scar contracture, congestion, notching, revision surgery, donor morbidities).

Statistical Analysis

Two review authors (ADG, SSC) analysed data. The weighted mean of each outcome was calculated based on sample sizes of each included study using the following method: (1) multiply the mean outcome of each study by the study sample size, (2) sum the products to get the total value, (3) sum the sample sizes to get the total weight, and (4) divide the total value by the total weight to provide a weighted mean for each outcome. The meta-analysis was performed using the Microsoft® Excel® 2016, with MetaXL version 5.2, add-in software (developed by EpiGear International Pty Ltd). The summary effect was ascertained using Odds Ratio (OR) which was calculated using the Inverse Variance Heterogenity model. Heterogeneity was ascertained using the I squared statistic. Small study effects like publication bias were evaluated using the Doi plot and Luis Furuya-Kanamori (LFK) index. High heterogeneity in the summary effect was further explored using sensitivity analysis. A p value of less than 0.05 was considered as significant.

Evidence Certainty

The certainty of evidence for the systematic review was assessed by two independent reviewers (ADG, NS) using the GRADEpro GDT: GRADEpro Guideline Development Tool [Software]. McMaster University and Evidence Prime, 2021. Available from gradepro.org. In case of any discrepancy, a consensus was formed by mutual discussion with other reviewers.

Results

Summary of Study and Patient Demography

The electronic database search produced 16,698 results. After title and abstract review 124 citations were identified (Fig. 1), that were considered for full text review. Thirty-eight articles [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38] met our inclusion/exclusion criteria. The other 86 studies were either case reports; or less than 10 patients were followed-up; or complications/outcomes were not recorded independently for the reconstructed flaps; or variations of the forehead/nasolabial flaps have been used. Twenty-seven studies were retrospective, while one study was prospective (Table 2). Out of total 2652 patients in these studies, 2036 patients were followed-up, in which 1443 underwent reconstruction with paramedian forehead flap. In the forehead flap group reported male and female patients were 425 and 399 respectively, while in nasolabial flap group this was 150 and 208 respectively. The weighted mean age of patients in the forehead and nasolabial group is 66.63 and 64.91 yearsrespectively. The weighted mean follow-up in the groups is 10.7 and 21.9 months respectively.

Fig. 1
figure 1

Flowchart of literature search

Full size image
Table 2 Demographic details of the studies included
Full size table

Summary of Flap Demography

Twenty-six studies have reported the relative number of nasal subunits that were reconstructed with either forehead or nasolabial flaps (Supplementary Table 1). Forehead flap was most commonly used for reconstruction of alar defects, immediately followed by sidewall, tip and dorsum (Table 3). Nasolabial flap was predominantly used for alar reconstruction. Our Meta-analysis shows a statistically significant difference in the incidence of alar reconstruction by forehead versus nasolabial flap [pooled odds ratio (OR): 0.3; 95% CI 0.01, 0.92; p = 0.72; I2 = 0%, n = 6 studies] (Fig. 2). To analyse the impact of individual studies on the pooled estimate a sensitivity analysis was done. On excluding the study of Vasalikis et al. [6], it is found that the pooled OR is not significant statistically, anymore (95% CI 0.14, 2.43) (Table 4). The meta-analysis done on incidence of dorsum and columella reconstruction by these flaps shows that the pooled odds ratios are not significant statistically [pooled OR for dorsum reconstruction: 5.97; 95% CI 0.99, 35.95; p = 0.61; I2 = 0%; n = 3 studies and pooled OR for columellar reconstruction: 1.84; 95% CI 0.35, 9.67; p = 0.4; I2 = 0%; n = 4 studies] (Fig. 2). Although, the study of Vasalikis et al. [6] and of Yoon et al. [7] appears outlier for dorsum and columella respectively, the sensitivity analysis performed shows that even on omitting the studies the pooled OR obtained is not significantly different from the overall pooled estimates [pooledOR: 3.02; 95% CI 0.32, 28.4; p = 0.96; I2 = 0%; n = 3 studies; and pooled OR: 4.77; 95% CI 0.65, 35.15; p = 0.37; I2 = 0%; n = 4 studies respectively] (Table 4).

Table 3 Comparative analysis of subunit reconstructed and complication in the two groups
Full size table
Fig. 2.
figure 2

Forest plots of nasal subunit reconstruction in different studies

Full size image
Table 4 Sensitivity analysis for heterogeneity of subunit reconstruction in different studies
Full size table

Summary of Complications and Outcomes

The incidence of complications following nasal reconstruction with nasolabial or forehead flap was reported by 36 studies (Supplementary table 2). The frequency of pin cushioning effect, thick/bulky scar,pigmentation, and subsequent steroid injection seemed more common in forehead flap reconstruction than nasolabial flap, while in the latter nasal obstruction seemed more frequently present (Table 3). All other complications were comparable in both groups. Meta-analysis was performed over data from 9 observational studies [1,2,3,4,5,6,7,8,9] which have reported the relative incidences of complications of both flaps, or the relative number of the nasal subunits reconstructed. Our meta-analysis shows that there is no significant difference in the risk of partial flap necrosis in the forehead flap vs nasolabial flap [pooled OR 1.28; 95% CI 0.46–3.53; p = 0.7; I2 = 0%, n = 6 studies] (Fig. 3). On visual inspection, the study by Uzun et al. [9] appears to be an outlier with OR 9.89 (95% CI 0.17, 560.85). However, on excluding this study the pooled OR is 1.117 (95% CI 0.392, 3.183) which is not significantly different from the overall pooled estimate (p = 0.739) (Table 5). The difference in the risk of complete flap necrosis in-between these flaps is also not significant statistically [pooled OR 1.28; 95% CI 0.19–8.39; p = 0.67; I2 = 0%, n = 4 studies] (Fig. 3). Again, the study by Uzun et al. [9] seems an outlier with OR 9.89 (95% CI 0.17, 560.85). However, on excluding this study the pooled OR is 0.721 (95% CI 0.086, 6.070) which is not significantly different from the overall pooled estimate (p = 0.862) (Table 5). There is no significant difference in the risk of alar notching also [pooled OR 1.81; 95% CI 0.55–5.91; p = 0.27; I2 = 0%, n = 4 studies] (Fig. 3). Here the study by Han et al. [8] appears an outlier with OR = 10.00 (95% CI 0.58, 171.2). On excluding this study, the pooled OR is not significantly different from the overall pooled estimate [OR: 1.453; 95% CI 0.460, 4.59; p = 0.309; I2 = 14.85; n = 4 studies] (Table 5). There is no significant difference in the risk of hematoma/bleeding in the forehead flap vs nasolabial flap [OR 2.04; 95% CI 0.34–12.81; p = 0.79; I2 = 0%, n = 4 studies] (Fig. 3). Although the study by Uzun et al. [9] seems an outlier with OR 9.89 (95% CI 0.17, 560.85), on excluding this study the pooled OR is not significantly different [OR: 1.38; 95% CI 0.19, 10.18; p = 0.85; I2 = 0.0; n = 4 studies] (Table 5). Five studies have compared the paramedian forehead and nasolabial flaps used for nasal reconstruction in terms of functional and aesthetic outcomes of donor and recipient sites [1, 3,4,5, 8]. We refrained from performing meta-analysis of other complications and the outcomes due to paucity of uniform level of information.

Fig. 3.
figure 3

Forest plot of complications (outcome) in different studies

Full size image
Table 5. Sensitivity analysis of complications (outcome) in different studies.
Full size table

Publication Bias

The Doi plot with LFK index of the subunit reconstructions (dorsum, alae, columella) (Fig. 4) and the complications (partial or complete flap necrosis, alar notching, hematoma or bleeding) (Fig. 5) shows no or minor asymmetry. This suggests that there is no publication bias and small study effects, which may be affecting the results of meta-analysis done.

Fig. 4.
figure 4

Doi plot of nasal subunit reconstruction in different studies

Full size image
Fig. 5.
figure 5

Doi plot of complications (outcome) in different studies

Full size image

Evidence Certainty

The certainty of evidence assessed for the various outcomes as per GRADE (Grading of Recommendations, Assessment, Development and Evaluations) criteria are of lowcertainty category. The summary of findings table shows a mild increased risk difference of complications with paramedian forehead flap than the nasolabial flap (Table 6).

Table 6. Certainty of evidence using GRADEpro GDT
Full size table

Discussion

In our review, based on 38 studies, we found that ala is the most commonly reconstructed subunit. Seven studies were dedicated for alar reconstruction [1, 3, 4, 8, 18, 28, 36], four of which [1, 3, 4, 8] have compared the nasolabial flap and paramedian forehead flap for alar reconstruction in their studies. Most of the nasolabial flaps (melolabial is an anatomically more precise description) [13] have been used for alar reconstruction in comparison to other subunits, which is not the case with forehead flap (Table 3). Our meta-analysis shows that there is increased propensity of alar reconstruction by nasolabial versus forehead flap, which is statistically significant. Vasalikis et al. [6] suggested larger alar reconstruction to be done using nasolabial flap, while tip and dorsal defects > 1.5 cm using forehead flap. However, our meta-analysis shows that there is no significant difference in the incidence of dorsal or columellar reconstructions by either flap. The I2 values in Forest plot shows that there is no heterogeneity in these studies (Fig. 2).

We were not able to find any relation of the nasal defect size with either flap type used for reconstruction, due to heterogeneity in the information of the studies. In the algorithm suggested by Uzun et al. [9], nasolabial flaps are preferred for lateral defects in middle or lower third, while forehead flap reconstruction for distal 2/3rd or combined defects. Based on their study of 17 nasal defects’ reconstruction, Han et al. [8] suggested that for a full thickness defect, that is > 2 cm, forehead flap is preferred. While for defects < 2 cm nasolabial flap reconstruction is preferred. In 2001, Drisco and Baker [4] suggested forehead flap reconstruction for nasal alar defects > 1.5 cm, young patient with inconspicuous melolabial fold and cheek involvement, while in other alar defects interpolated cheek flap is preferred. In their follow-up of 83 nasal reconstructions, Vasilakis et al. [6] reported statistically significant difference in the size of the defects (p < 0.001), in their greater diameter, in-between the groups. However, all these studies were retrospective so they would be prone to selection bias.

Our meta-analysis shows that there is no statistically significant difference in the risk of complications (partial or total flap necrosis, alar notching, hematoma/bleeding) in-between the groups. Overall, there is no heterogeneity in most of these studies (Fig. 3). However, in our summary of findings table it seems that there is mild increased risk of alar notching in case of paramedian forehead flap (Table 6). Paddack et al. [2], with a follow-up of 25 nasolabial and 82 paramedian forehead flaps, found that diabetes, hypertension, coronary arterial disease, COPD and smoking were not statistically significant factors in flap failure, although in the latter flap failure shows an increasing trend.

In 1999, Arden et al. [1] in a follow-up of 20 forehead flaps and 18 melolabial flaps for alar reconstruction reported that objective scar measurements (rim thickness difference, donor scar length and width), subjective rating of textural quality, and post-operative alar notching favoured melolabial reconstructions. Patient’s questionnaire results demonstrated a statistically significant (p = 0.026) difference in donor site rating, favouring melolabial group responses. While in 2019, Genova et al. [3] reported that for alar reconstruction forehead flaps gave statistically better aesthetic and functional results (p = 0.03 for both variable) than nasolabial flaps, according to patient satisfaction survey, after a mean follow-up period of 2.3 years. Forehead flaps also had superior alar contour, telangiectasia/erythema, post-operative scar, alar and nostril symmetry from basilar view, compared to nasolabial flaps, according to Surgeons’ questionnaire (comprising of all 5 above variables, each scored as 1–3). Sherris et al. [5] recorded better aesthetic results after forehead flap reconstruction than nasolabial flap, 5.5 versus 4.4 respectively (change on a 1–10 Visual Analog Scale from defect to after reconstruction) for columellar defects. However, due to small group size, authors refrained from any statistical analysis. Drisco and Baker [4] found on a scale of 1–3 that forehead flap reconstruction seemed better in terms of breathing self-assessment, patient result, and observer result of scar and overall. We were not able to perform any meta-analysis on the overall aesthetic outcome from the data of these studies, as different scales for measurement were used.

Despite our sincere efforts there were certain limitations in this study. Statistical significance of the difference in cosmetic and functional outcomes of the paramedian forehead and nasolabial flaps could not be calculated because the included studies have used different parameters and scales to assess them. Meta-analysis of many outcome complications (infection, dehiscence, congestion, pin cushioning, nasal obstruction etc.) and few subunit reconstructions (nasal tip and sidewall) could not be done due to inadequate data. The included studies have different follow-up periods. The variations in nasolabial flap, namely interpolated, islanded flaps, or in forehead flaps, namely 2- or 3-staged flaps, included in our review might be a limitation in our results. Most of the studies in the review are retrospective and some are prospective which could have affected the results. Lastly, confounding factors such as age, gender, co-morbidities of patients, dimension and extent of the flap or the defect etc could not be taken into account due to absence of individual data.

Conclusion

We conclude that for the alar subunit, reconstruction using the nasolabial (melolabial) flap is significantly preferable than the paramedian forehead flap. This does not hold true for other subunits. There is an inclination to reconstruct larger defects by forehead flap. The difference in the risk of complications in the post operative period in either group is statistically insignificant, though there is a mild increased risk in the forehead flap group. The aesthetic outcomes among the groups compared by the studies are contradictory to each other and needs further exploration.