Abstract
Objective
The aim of our study was to determine the factors influencing the evolution of the total volume and bone volume of free flaps commonly used in head and neck surgery, with a 30‐month prospective study, to establish volume change predictions and thus propose a degree of overcorrection to be expected before reconstruction.
Study Design
We prospectively included all consecutive free flap.
Setting
Our maxillofacial surgery department between August 2021 and January 2024.
Methods
We collected information on preoperative, per‐operative, and postoperative factors, on patients, surgical techniques, and adjuvant treatments. We measured on each postoperative imaging the overall flap volume and bone volume if applicable. Multivariate mixed models were then used to select clinical parameters associated with volume loss.
Results
We included 166 flaps, performed on 155 patients. The mean age was 60.1 ± 15.1 years. A total of 634 imagings were segmented (487 computed tomography [CT] scans, 77%; 147 magnetic resonance imagings [MRIs], 23%). The use of the superior thyroid or lingual veins for venous anastomosis, such as the use of small couplers, resulted in negative volume changes. Predicted bone volumes decreased by 23% at 30 months for deep circumflex iliac artery (DCIA) free flaps, 19% for fibula free flap (FFF), and 38% for scapular system free flap (SFF).
Conclusion
These findings allow us to envisage a volume overcorrection of around 60% for fasciocutaneous or osteocutaneous flaps, and 75% for muscle or osteomuscular flaps. The choice of vein and microsurgical technique seems to have more impact on the evolution of free flap volume than patient characteristics or adjuvant treatments.
Keywords: CT scans, flap contour, flap volume, free flap, head and neck, MRI, reconstruction
Reconstruction using microvascularized free transfers or free flaps in head and neck surgery has become a gold standard. 1 The removal of large malignant or benign tumors, reconstruction after severe trauma involving complex tissue loss, or the treatment of radiotherapy complications such as osteoradionecrosis all require not only the filling of anatomical defects, but above all the restoration of functions such as speech or mastication. 2 A flap that is either too large or too small will lead to poor functional results. 3 , 4 , 5
Several small retrospective studies have attempted to determine the factors influencing changes in flap volume over time, most often by univariate analysis of a small number of factors.
The results of these studies remain controversial, particularly as regards the effect of adjuvant radiotherapy on transferred tissue. 1 , 6 , 7 , 8 In addition, levels of overcorrection of free flaps vary widely from study to study, ranging from 10% to 40% in the literature. 1 , 6
Through a prospective study with standardized data collection and reproducible measurements on computed tomography (CT) scans and magnetic resonance imaging (MRI), we aimed to shed light on the factors to be taken into account before reconstruction, to predict the ideal flap volume, and to improve patients' quality of life.
Thus, the aim of our study was to determine the factors influencing the evolution of the total volume and bone volume, if any, of flaps commonly used in head and neck surgery, with a 30‐month prospective survey of patients operated on with free flaps in our department, such as fibula free flaps (FFFs), scapular system free flaps (SFFs), lateral arm flaps (LAFs), radial forearm free flaps (RFFFs), or deep circumflex iliac artery (DCIA) flaps. In this way, we can establish volume change predictions for total volume and bone volume of the flaps, and thus propose a degree of overcorrection to be expected before reconstruction.
Methods
Patient Data Collection
We prospectively included all consecutive free flaps performed in our maxillofacial surgery department between August 2021 and January 2024. Exclusion criteria were partial or total flap necrosis, secondary remodeling surgery, skin paddle removal, or secondary bone flap osteotomies. The study was validated by our local ethics committee (CER‐2011‐022) in accordance with the Declaration of Helsinki.
In a standardized way, we collected the following:
-
1.
Preoperative and demographic data. Age, gender, cardiovascular SCORE (Systemic COronary Risk Evaluation), 9 thrombotic risk factors (history of phlebitis or pulmonary embolism, thrombophilia, polycythemia, and cirrhosis), active or past alcohol consumption, active smoking, body mass index (BMI), history of cervicofacial radiotherapy, indication, site of reconstruction, and atherosclerosis/stenosis of lower limb arteries on CT scan in case of fibula flap.
-
2.
Per‐operative data. Type of free flap, presence of a skin paddle, cervical lymph node dissection, per‐operative flap volume, choice of donor and recipient vessels, size of donor artery, type of venous anastomosis (sutures or coupler device), and, if applicable, size of coupler used, duration of arterial anastomosis, duration of venous anastomosis, flap ischemia duration, per‐operative revision of anastomoses, and operating time. The diameter of the donor artery was measured using a microsurgical meter.
-
3.
Postoperative data. Return to the theater for revision of anastomoses, cervical hematoma, infection, and mortality during the hospital stay. The completion of adjuvant radiotherapy and the dose in Grays used, as well as the realization of chemotherapy, were reported.
Imaging Data Collection
We measured on each postoperative imaging between August 2021 and May 2024 the overall flap volume and bone volume in case of bone flap, using the open‐source, free 3D Slicer software version 5.6.1 after importing the data in a DICOM format. The time required for postoperative imaging was not standardized, and its realization was left to the discretion of the patient's referring surgeon, while following international recommendations for reference imaging to be performed 3 months after the end of treatment. The total volume was obtained by contouring in the three axial, coronal, and sagittal planes, on injected CT or T1 FATSAT MRI slices with gadolinium injection. From the slices, the 3D contouring was automatically generated by the software, then after a smoothing step, the volume was measured. Bone volume was obtained by thresholding on injected or noninjected CT scans. Patients for whom only one scan was available during follow‐up were excluded, as well as patients who had undergone a second operation that may modify the flap volume (remodeling, removal of necrosis, defatting, etc.). We compared the performance of segmentations on MRI and on injected CT scans, when both imaging exams were performed on the same day in the patient's follow‐up. The Intraclass Correlation Coefficient (ICC) 10 was measured between the two types of imaging to determine whether it could be incorporated indiscriminately into future models.
Volume Modeling Over Time
For postoperative total volume and bone volume as a function of time after surgery, four hierarchical or mixed models were tested: the simple linear model, polynomial models of degree 2 and 3, and a log‐linear model. These four models were compared using the Akaike Information Criterion (AIC). 11 All these models were hierarchical, to take into account the nonindependence of the data due to repeated measurements for the same patient over time. The model that significantly decreased the AIC the most, and therefore increased likelihood the most, was selected for further analysis.
Once the most suitable model had been determined, each variable was tested independently in univariate analyses, with one parameter of interaction with time. Explanatory variables significantly modifying the slope of the model, that is, having a slope statistically different from 0 using Student's t tests (P‐value < .05), were included in the final multivariate model.
The statistical analyses were performed on R 4.2.2 using the nlme 12 and ggplot 13 packages. All statistical tests were two‐sided, and P‐values ≤ .05 were considered to be statistically significant.
Results
Patient Data Description
We prospectively included all consecutive 307 free flaps, performed on 274 patients in our maxillofacial surgery department between August 2021 and January 2024. On these flaps, after applying the exclusion criteria, we finally carried out our analyses on 166 flaps, performed on 155 patients. Mean age was 60.1 ± 15.1 years, and 45% were women.
The distribution of flaps performed was as follows: 65 RFFF (39%), 51 FFF (31%), 33 SFF (20%), 12 LAF (7%), 3 DCIA (2%), 1 soleus (1%), and 1 antero lateral thigh flap (ALT) (1%); thus, 48% of flaps were fasciocutaneous, 5% muscular, 16% osteomuscular, and 31% osteocutaneous.
In total, 74 free flaps (45%) were exposed to adjuvant radiotherapy and 42 (25%) to chemotherapy. Finally, 10% of patients died during follow‐up. These results are detailed in Table 1.
Table 1.
Patient Data Description
| N (%)/mean ± SE | |
|---|---|
| N (flaps) | 166 |
| N (patients) | 155 |
| Age, y | |
| Mean ± SD | 60.1 ± 15.1 |
| Median | 61.4 |
| Min | 19.5 |
| Max | 92.4 |
| Gender | |
| Female | 76/168 (45%) |
| Male | 92/168 (55%) |
| BMI | |
| Mean ± SD | 24.5 ± 5.1 |
| Median | 23.7 |
| Min | 14.1 |
| Max | 43.8 |
| Cardiovascular SCORE risk | |
| Low risk | 67/164 (41%) |
| Moderate risk | 48/164 (29%) |
| High risk | 22/164 (13%) |
| Very high risk | 27/164 (16%) |
| Venous thrombotic risk | 10/166 (6%) |
| Angio‐CT of the lower limbs | |
| Normal | 11/36 (31%) |
| Proximal atherosclerosis/stenosis | 15/36 (42%) |
| Distal atherosclerosis/stenosis | 10/36 (28%) |
| Toxics | |
| Active smoker | 51/142 (36%) |
| Chronic drinker | 37/129 (21%) |
| Indication | |
| Malignant tumors | 113/166 (68%) |
| Osteoradionecrosis | 24/166 (14%) |
| Malignant tumor sequellas | 17/166 (10%) |
| Benign tumors | 10/166 (6%) |
| Traumatology | 2/166 (1%) |
| Reconstructed structure | |
| Mandible | 59/165 (36%) |
| Maxilla | 34/165 (21%) |
| Tongue | 26/165 (16%) |
| Buccal floor | 16/165 (10%) |
| Maxillo mandibular commissure | 11/165 (7%) |
| Cheek mucosa | 8/165 (5%) |
| Cervical | 2/165 (1%) |
| Lip | 2/165 (1%) |
| Orbital region | 2/165 (1%) |
| Scalp | 2/165 (1%) |
| Soft palate | 2/165 (1%) |
| Cervicofacial radiotherapy history | 53/166 (32%) |
| Flap | |
| RFFF | 65/166 (39%) |
| FFF | 51/166 (31%) |
| SFF | 33/166 (20%) |
| Scapular tip | 20/166 (12%) |
| LD | 8/166 (5%) |
| Scapular crest | 4/166 (2%) |
| TDAP | 1/166 (1%) |
| LAF | 12/166 (7%) |
| With periosteal lateral brachial flap | 3/166 (2%) |
| DCIA | 3/166 (2%) |
| Soleus | 1/166 (1%) |
| ALT | 1/166 (1%) |
| Flap type | |
| Fasciocutaneous | 79/166 (48%) |
| Muscular | 9/166 (5%) |
| Osteomuscular | 27/166 (16%) |
| Osteocutaneous | 51/166 (31%) |
| Skin paddle | 143/166 (86%) |
| Lymph node dissection | 94/166 (57%) |
| Donnor artery | |
| Facial | 86/152 (57%) |
| Superior thyroid | 27/152 (18%) |
| Lingual | 20/152 (13%) |
| Transverse cervical | 9/152 (6%) |
| External carotid | 6/152 (4%) |
| Superficial temporal | 2/152 (1%) |
| Previous pedicle | 1/152 (1%) |
| Internal mammary | 1/152 (1%) |
| Donnor artery diameter, mm | |
| Mean ± SD | 2.3 ± 1.0 |
| Median | 2.0 |
| Min | 1.0 |
| Max | 8.0 |
| Recipient vein | |
| External jugular | 39/140 (28%) |
| LFT | 36/140 (26%) |
| Common facial | 28/140 (20%) |
| Superior thyroid | 14/140 (10%) |
| Lingual | 7/140 (5%) |
| Internal jugular | 6/140 (4%) |
| Transverse cervical | 5/140 (4%) |
| Anterior jugular | 2/140 (1%) |
| Retromandibular vein | 2/140 (1%) |
| Superficial temporal | 1/140 (1%) |
| Vein anastomosis | |
| Coupler | 124/129 (96%) |
| Sutures | 5/129 (4%) |
| Type of vein anastomosis | |
| Terminoterminal | 124/129 (96%) |
| Terminolateral | 5/129 (4%) |
| Coupler size | |
| Mean ± SD | 3.3 ± 0.6 |
| Median | 3.5 |
| Min | 1.0 |
| Max | 4.0 |
| Anastomosis duration, min | |
| Artery | 28.6 ± 15.6 |
| Vein | 14.0 ± 10.9 |
| Ischemia duration | |
| Mean ± SD | 79.7 ± 33.8 |
| Median | 71.5 |
| Min | 23.0 |
| Max | 225.0 |
| Per‐op. anastomosis revision | 11/166 (7%) |
| Artery | 9/11 (82%) |
| Vein | 2/11 (18%) |
| Post‐op. anastomosis revision | 13/166 (8%) |
| Artery | 2/13 (15%) |
| Vein | 5/13 (38%) |
| Postoperative infection drainage | 16/166 (10%) |
| Cervical hematoma drainage | 5/166 (3%) |
| Adjuvant radiotherapy | 74/166 (45%) |
| Mean dose, Gy | 58.2 ± 11.2 |
| Adjuvant chemotherapy | 42/166 (25%) |
Abbreviations: ALT, antero lateral thigh flap; BMI, body mass index; CT, computed tomography; DCIA, deep circumflex iliac artery; FFF, fibula free flap; LAF, lateral arm flap; LD, latissimus dorsi; LFT, linguo‐facial trunk; per‐op., per‐operative; post‐op., postoperative; RFFF, radial forearm free flap; SCORE, Systematic COronary Risk Evaluation; SD, standard deviation; SFF, scapular system free flap; TDAP, thoraco‐dorsal artery perforator.
Imaging Data Description
A total of 634 imaging exams were segmented (487 CT scans, 77%; 147 MRIs, 23%). Patients had an average of 3.8 ± 1.8 imaging exams in their follow‐up, with a maximum of 10. Mean follow‐up was 10.8 ± 7.0 months, with a maximum of 31.1 months (Supplemental Table S1, available online). The average total volumes were, in descending order: SFF (112.5 ± 86.7 cm3), DCIA (92.1 ± 42.2 cm3), FFF (72.6 ± 35.7 cm3), LAF (22.0 ± 12.1 cm3), and RFFF (21.1 ± 11.0 cm3). Mean bone volumes were, in descending order: DCIA (18.8 ± 2.6 cm3), FFF (11.0 ± 3.9 cm3), and SFF (8.6 ± 3.5 cm3). The ICC measured between MRI and CT scan was 0.975 (0.942‐0.989), meaning an excellent reliability. 14 Figure 1 shows an example of a 3D reconstruction of an osteocutaneous flap segmentation.
Figure 1.
Example of total (A, B) and bone (C, D) volume segmentation at 1 month (A, C) and 2.5 years (B, D) postreconstruction by fibula free flap.
Total Volume Modeling Over Time
The model that significantly decreased AIC without overfitting was the mixed log‐linear model. The following variables were included in the multivariate model because they significantly modified the slope over time in univariate analyses: age, cardiovascular risk, active smoker, history of radiotherapy, type of flap, presence or absence of a skin paddle, donor artery and its diameter, and recipient vein. Adjuvant radiotherapy did not significantly modify flap volume over time (slope: −0.093 ± 0.079 cm3 per month, P = .240) and was therefore not included in the multivariate model.
In multivariate analyses, age did not significantly modify volume over time, after accounting for the potential confounding factors (±8.415 ± 5.360 cm3 per month, P = .118 and −4.633 ± 2.863 cm3 per month, P = .107 for patients <50 years and >74 years), nor did cardiovascular risk, active smoking, or history of radiotherapy. Changes in total volume over time depended on the type of flap, after eliminating potential confounding factors: taking fasciocutaneous flaps as reference, the difference in model slope was −11.68 ± 2.454 cm3 per month (P < .001) for osteocutaneous flaps, and −47.55 ± 6.286 cm3 per month (P < .001) for muscular flaps. The presence of a skin paddle resulted in a significant volume difference of −19.15 ± 4.927 cm3 per month (P < .001). Use of the transverse cervical artery also resulted in a volume difference of −17.12 ± 6.863 cm3 per month (P = .014). Similarly, use of the superior thyroid or lingual veins resulted in volume differences of −9.248 ± 4.525 cm3 per month (P = .043) and −10.72 ± 4.190 cm3 per month (P = .011), respectively. Finally, at all other comparable factors, each 1‐mm increase in coupler diameter resulted in an increase in the slope of the total volume model of 9.970 ± 2.785 cm3 per month (P < .001) (Table 2, Figure 2).
Table 2.
Results of the Log‐Linear Mixed Model of Total Volume Versus Time, in Univariate and Multivariate Analyses a
| Univariate | Multivariate | |||
|---|---|---|---|---|
| Slope | P‐value | Slope | P‐value | |
| Age | ||||
| <50 | 3.517 ± 1.683 | .037 | 8.415 ± 5.360 | .118 |
| >74 | −4.435 ± 2.079 | .036 | −4.633 ± 2.862 | .107 |
| Gender | ||||
| Female | 0.469 ± 1.450 | .746 | ||
| BMI | ||||
| <18.5 | −4.338 ± 2.368 | .068 | ||
| >25 | −2.723 ± 1.528 | .077 | ||
| Cardiovascular risk | ||||
| Moderate | −3.575 ± 1.875 | .057 | 1.258 ± 3.993 | .753 |
| High | −8.948 ± 2.209 | <.001 | −1.660 ± 4.119 | .688 |
| Very high | −5.089 ± 2.029 | .013 | 3.723 ± 4.246 | .382 |
| Thrombotic risk | −2.576 ± 3.595 | .474 | ||
| Angioscan of the lower limbs | ||||
| Proximal atherosclerosis | 3.989 ± 2.246 | .079 | ||
| Distal atherosclerosis | 1.855 ± 2.525 | .465 | ||
| Toxics | ||||
| Alcohol | 1.909 ± 1.772 | .282 | ||
| Active smoker | 5.460 ± 1.672 | .001 | −2.165 ± 2.072 | .297 |
| Indication (reference: others) | ||||
| Malignant | −2.587 ± 2.155 | .231 | ||
| ORN/sequella | 1.060 ± 2.390 | .658 | ||
| Cervicofacial radiotherapy history | −4.422 ± 1.460 | .003 | −0.168 ± 4.832 | .972 |
| Flap (reference: RFFF) | ||||
| Lateral brachial | 1.234 ± 2.480 | .619 | ‐ | ‐ |
| Fibula | −8.935 ± 1.627 | <.001 | ‐ | ‐ |
| Scapulodorsal | −21.68 ± 1.966 | <.001 | ‐ | ‐ |
| Flap type (reference: fasciocutaneous) | ||||
| Muscular | −36.57 ± 2.556 | <.001 | −47.55 ± 6.286 | <.001 |
| Osteocutaneous | −9.393 ± 1.343 | <.001 | −11.68 ± 2.454 | <.001 |
| Osteomuscular | −17.38 ± 2.032 | <.001 | ||
| Skin paddle | 9.907 ± 1.937 | <.001 | −19.15 ± 4.927 | <.001 |
| Neck dissection | 3.467 ± 1.437 | .016 | 7.088 ± 3.634 | .053 |
| Donnor artery (reference: facial) | ||||
| External carotid | 2.504 ± 3.172 | .431 | 2.343 ± 4.636 | .614 |
| Transverse cervical | −3.301 ± 4.940 | .505 | −17.12 ± 6.863 | .014 |
| Lingual | −1.783 ± 2.011 | .377 | 4.348 ± 5.193 | .404 |
| Superior thyroid | 2.679 ± 2.500 | .285 | 2.786 ± 3.480 | .424 |
| Superficial temporal | −33.01 ± 5.208 | <.001 | ‐ | ‐ |
| Donnor artery diameter | −4.247 ± 0.502 | <.001 | −2.354 ± 2.774 | .397 |
| Receiver vein (reference: facial) | ||||
| Internal jugular | 4.280 ± 4.167 | .305 | −8.700 ± 20.06 | .665 |
| External jugular | 13.99 ± 2.298 | <.001 | −3.022 ± 3.824 | .430 |
| Anterior jugular | 17.27 ± 5.203 | .001 | −0.172 ± 6.068 | .977 |
| Transverse cervical | 12.14 ± 4.158 | .004 | ‐ | ‐ |
| Superior thyroid | 15.81 ± 4.227 | <.001 | −9.248 ± 4.525 | .043 |
| LFT | 11.88 ± 2.219 | <.001 | −5.444 ± 3.606 | .133 |
| Lingual | 20.12 ± 3.711 | <.001 | −10.72 ± 4.190 | .011 |
| Coupler size | −6.747 ± 1.277 | <.001 | 9.970 ± 2.785 | <.001 |
| Anastomosis duration | ||||
| Artery | 0.050 ± 0.053 | .342 | ||
| Vein | −0.107 ± 0.073 | .142 | ||
| Ischemia | 0.013 ± 0.023 | .576 | ||
| Operating time | −0.006 ± 0.010 | .529 | ||
| Anastomosis revision | ||||
| Per‐operative/postoperative | 0.184 ± 2.076 | .930 | ||
| Postoperative infection | −3.783 ± 2.177 | .083 | ||
| Cervical hematoma | −0.086 ± 5.510 | .987 | ||
| Adjuvant radiotherapy | 0.642 ± 1.440 | .656 | ||
| Dose radiotherapy | −0.093 ± 0.079 | .240 | ||
| Adjuvant chemotherapy | −0.457 ± 1.562 | .770 | ||
Abbreviations: BMI, body mass index; LFT, linguo‐facial trunk; ORN, osteoradionecrosis; RFFF, radial forearm free flap.
The P‐values in bold correspond to a significant result.
Figure 2.
Predictions of the total volume evolution model over time, for different variables: (A) flap type, (B) whether or not adjuvant radiotherapy was performed, and (C) coupler size. RT, radiotherapy.
Bone Volume Modeling Over Time
The model that significantly decreased AIC without overfitting was the mixed linear model. The following variables were included in the multivariate model because they significantly modified the slope over time in univariate analyses: gender, BMI, age, distal atherosclerosis, history of radiotherapy, and an anastomosis revision (Table 3). There were no differences in bone volume evolution according to the type of flap.
Table 3.
Results of the Linear Mixed Model of Bone Volume Versus Time, in Univariate and Multivariate Analyses a
| Univariate | Multivariate | |||
|---|---|---|---|---|
| Slope | P‐value | Slope | P‐value | |
| Age | ||||
| <50 | −0.001 ± 0.034 | .967 | ||
| >74 | −0.021 ± 0.054 | .704 | ||
| Gender | ||||
| Female | 0.070 ± 0.029 | .019 | −0.051 ± 0.188 | .785 |
| BMI | ||||
| <18.5 | −0.114 ± 0.054 | .035 | −0.022 ± 0.094 | .820 |
| >25 | 0.026 ± 0.034 | .442 | 0.025 ± 0.038 | .506 |
| Cardiovascular risk | ||||
| Moderate | −0.007 ± 0.040 | .861 | ||
| High | 0.040 ± 0.092 | .661 | ||
| Very high | 0.015 ± 0.036 | .678 | ||
| Thrombotic risk | −0.067 ± 0.061 | .277 | ||
| Angioscan of the lower limbs | ||||
| Proximal atherosclerosis | −0.049 ± 0.039 | .213 | −0.092 ± 0.062 | .142 |
| Distal atherosclerosis | −0.128 ± 0.045 | .005 | −0.082 ± 0.054 | .133 |
| Toxics | ||||
| Alcohol | 0.048 ± 0.028 | .097 | ||
| Active smoker | 0.004 ± 0.032 | .893 | ||
| Indication (reference: others) | ||||
| Malignant | 0.058 ± 0.041 | .160 | ||
| ORN/sequella | 0.050 ± 0.050 | .322 | ||
| Cervicofacial radiotherapy history | 0.102 ± 0.030 | .001 | −0.075 ± 0.035 | .037 |
| Flap (reference: fibula) | ||||
| DCIA | −0.072 ± 0.074 | .330 | ||
| Scapulodorsal | −0.048 ± 0.041 | .252 | ||
| Number of fragments | 0.013 ± 0.018 | .480 | ||
| Localization (reference: mandible) | ||||
| Maxilla | −0.065 ± 0.039 | .099 | ||
| Skin paddle | 0.089 ± 0.049 | .074 | ||
| Neck dissection | −0.031 ± 0.029 | .297 | ||
| Donnor artery (reference: facial) | ||||
| External carotid | 0.088 ± 0.049 | .074 | ||
| Transverse cervical | 0.026 ± 0.096 | .789 | ||
| Lingual | 0.070 ± 0.040 | .081 | ||
| Superior thyroid | 0.013 ± 0.070 | .852 | ||
| Superficial temporal | 0.138 ± 0.286 | .630 | ||
| Donnor artery diameter | 0.010 ± 0.015 | .493 | ||
| Receiver vein (reference: facial) | ||||
| Internal jugular | 0.020 ± 0.078 | .799 | ||
| External jugular | 0.057 ± 0.071 | .422 | ||
| Superior thyroid | 0.104 ± 0.245 | .670 | ||
| LFT | 0.106 ± 0.083 | .207 | ||
| Lingual | −0.089 ± 0.111 | .428 | ||
| Coupler size | 0.027 ± 0.037 | .455 | ||
| Anastomosis duration | ||||
| Artery | 0.001 ± 0.001 | .329 | ||
| Vein | −0.002 ± 0.001 | .056 | ||
| Ischemia | −0.000 ± 0.000 | .672 | ||
| Operating time | 0.000 ± 0.000 | .351 | ||
| Anastomosis revision | ||||
| Per‐operative/postoperative | −0.106 ± 0.043 | .014 | −0.079 ± 0.048 | .103 |
| Postoperative infection | 0.038 ± 0.086 | .660 | ||
| Cervical hematoma | 0.213 ± 0.211 | .316 | ||
| Adjuvant radiotherapy | 0.013 ± 0.031 | .682 | ||
| Dose radiotherapy | −0.002 ± 0.001 | .102 | ||
| Adjuvant chemotherapy | 0.058 ± 0.029 | .047 | ||
Abbreviations: BMI, body mass index; DCIA, deep circumflex iliac artery; LFT, linguo‐facial trunk; ORN, osteoradionecrosis.
The P‐values in bold correspond to a significant result.
In the multivariate model, the only variable remaining significant was a history of radiotherapy, modifying bone volume by −0.075 ± 0.035 cm3 per month (P = .037) (Supplemental Figure S1, available online). The other variables, notably gender, BMI, distal atherosclerosis on the lower limbs, a maxillary reconstruction, or revision of anastomoses, were nonsignificant after accounting for potential confounding factors. There was, however, a tendency towards reduced bone volume in cases of lower limb atherosclerosis (−0.082 ± 0.054 cm3, P = .133) or anastomosis revision (−0.079 ± 0.048, P = .103).
Predictions Over Time
Finally, we have plotted the different mean predictions of total and bone volume per flap over time, from 0 to 30 months' follow‐up after surgery (Supplemental Table S2, available online). In descending order, the greatest losses in total volume were for SFF (−77%), FFF (−66%), RFFF (−65%), and LAF (−57%). The volume of muscle flaps decreased by 82% after 30 months, compared with 62% for fasciocutaneous flaps.
Predicted bone volumes decrease by 23% at 30 months for DCIA, 19% for FFF, and 38% for SFF, keeping in mind that these results are not significantly different between flaps according to our model.
Discussion
This is the largest prospective volume follow‐up study of free flaps in head and neck reconstruction. Existing studies are mainly retrospective, mainly concern ALT flaps, and compare initial volumes and volumes after a given time using univariate tests. 6 , 7 , 8 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 We included all of the patients' imagings, regardless of the postoperative delay, which was left to the discretion of the referring surgeon. Razavi et al 6 used linear models to study the follow‐up of RFFF and ALT at 6 months after radiation therapy, and found a significant effect of age and radiotherapy on volume. Wang et al 23 studied the evolution of volume at 2 years on 70 patients and found a significant negative effect of BMI reduction and radiotherapy. In 2024, Pfister et al 1 published a meta‐analysis of 14 retrospective studies on the evolution of flap volume, including 119 patients, mainly operated on by RFFF and ALT flaps. The authors found a significant negative effect of radiotherapy on volume. It should be noted, however, that the results of these studies are very disparate. After reviewing the literature, we felt it was essential to study the clinical, surgical, and postoperative parameters influencing the evolution of free flap volume in a prospective study, with standardized measurements and a suitable statistical method.
A particularity of our study was the inclusion of all imaging regardless of the delay in realization. Thus, different stages of volume change were included in our model, such as early postoperative edema or the effect of radiotherapy on the flap. With this type of model, we could define a reliable prediction of volume loss, with the intraoperative volume of the harvested flap as the value at t = 0. Volume evolution in our study at 30 months, with the predictions of our model, is −62% for fasciocutaneous flaps, −67% for osteocutaneous flaps, −74% for osteomuscular flaps, and −82% for muscle flaps. We believe that the follow‐up time was sufficient, as Figure 2 shows a significant flattening of the volume curve from 12 months onwards, with little change thereafter.
These differences were significant in our multivariate model, suggesting greater volume loss in the muscle portions of free flaps. These decreases were much greater than those found in the meta‐analysis of Pfister et al, 1 with a volume reduction of −9.4% at 6 months. However, initial imaging was carried out at very different times depending on the study, from 1 day post‐op 25 to 3 months post‐op. 29 It therefore seems difficult from this study to predict a degree of flap overcorrection, which the authors probably underestimated at 25% to 35%, as the volume at the time of flap harvesting cannot be reflected by postoperative imaging at 3 months. We therefore recommend an overcorrection of around 60% for fasciocutaneous or osteocutaneous flaps and 75% for muscle or osteomuscular flaps. However, it is important to note that it is difficult to modify the thickness of a free fasciocutaneous flap and that we can mainly vary the dimensions of the paddle, that is, width and length. It would therefore be interesting to study the evolution of the 2D dimensions of the skin pallets over time to get a concrete idea of the overcorrection required. Also, we chose not to separate patients with pure muscular flaps or musculocutaneous flaps. Thus, a population with a higher mean BMI would have resulted in a smaller difference in flap volume loss between fasciocutaneous and muscular flaps.
We did not find any effect of adjuvant radiotherapy on the evolution of total flap volume, contrary to several teams. 1 , 6 , 7 , 8 , 21 , 22 , 23 , 27 This factor was not even significant in univariate analyses in our model (slope: ±0.642 ± 1.440 cm3 per month, P = .656). Pfister et al 1 found in their meta‐analysis of 14 studies a significant negative effect of radiotherapy on flap volume, with a very important effect in the study by Haymerle et al 8 (−47.8 [−66.7; −28.8] on mean volume, in a population of 13 patients; 10 of the studies found no effect. This effect of radiotherapy is therefore much debated in the literature, and our large‐scale prospective study did not suggest that it is significant.
Jeong et al 18 argue that a sufficient blood supply would prevent partial fat or skin partial necrosis and maintain flap volume. We found a significant effect of the choice of recipient vein, with a loss of −9.248 ± 4.525 cm3 per month (P = .043) in the case of anastomosis with a small‐diameter vessel such as the superior thyroid vein, after accounting for potential confounding factors; in the same direction, a 1‐mm increase in coupler size results in a 9.970 ± 2.785 cm3 per month (P < .001) increase in total flap volume. These two results suggest the importance of venous anastomoses to avoid chronic tissue suffering and guarantee good maintenance of flap volume over time. Lin et al 31 explained that the choice of recipient vein has a major influence on the success of a free flap, as venous anastomosis is more sensitive to extrinsic compression phenomena. We therefore recommend choosing a good‐caliber vein for the venous anastomosis to maintain volume over time. A missing element from our analysis, but still of major importance, is the impact of these changes in volume on patients' quality of life and on functions such as speech and mastication. It is therefore possible to take advantage of a reduction in the volume of the skin paddle of a fibula flap to improve dental prosthetic rehabilitation. Similarly, a lingual reconstruction flap that remains voluminous can hinder good mobility, while a reconstruction that is too thin can alter the transfer of food to the pharynx during propulsion of the base of the tongue.
There are few studies on the evolution of bone volume in free flaps. 32 , 33 Taxis et al 33 found more significant bone resorption for FFF compared with SFF on 211 CT scans, without finding any influence from other factors, such as age, gender, smoking, diabetes, and adjuvant radiotherapy. Wilkman et al 32 found on 38 patients a bone loss of 14% for SFF, 3% for DCIA, and 1% for FFF after 2 years, with no correlation with factors such as age, gender, or adjuvant radiotherapy. On 260 CT scans of 77 patients, we found a significant negative effect of previous radiotherapy on bone volume evolution (slope: −0.075 ± 0.035 cm3 per month, P = .037) in a multivariate model. There were no other significant effects, including age, gender, cardiovascular risk factors, or adjuvant radiotherapy. The type of bone flap (FFF, SFF, or DCIA) did not significantly modify the evolution of bone volume. In our study, predicted bone volumes at 30 months were −23% for DCIA, −19% for FFF, and −38% for SFF. However, it would be interesting in a future study to investigate the evolution of bone density over time as a function of flap type. Lee and Thiele 34 explained in their study that a history of cervicofacial radiotherapy impairs the healing of free flaps by modifying the regulation of the extracellular matrix by transforming growth factor β1 (TGF β1), reducing the expression of matrix metalloprotease‐1, and increasing the expression of matrix metalloproteinase‐1 inhibitor. Animal studies have shown that irradiated tissue promotes microthrombi, tissue necrosis, and vascular density in free flaps. 35 , 36 Thus, we can imagine that a history of radiotherapy impairs bone consolidation and reduces bone volume after microvascularized transfer.
Conclusion
With this large‐scale prospective cohort of 634 CT scans or MRIs corresponding to 166 patients, we were able to identify factors influencing the evolution of total and bone volume, in particular microsurgical venous anastomotic parameters. Similarly, we ruled out factors with a discussed effect in the literature, such as age or adjuvant radiotherapy. These findings allow us to propose a volume overcorrection of around 60% for fasciocutaneous or osteocutaneous flaps and 75% for muscle or osteomuscular flaps.
Author Contributions
Quentin Hennocq, design, conduct, analysis, presentation of the research; Jean‐Baptiste Caruhel, conduct, presentation of the research; Mourad Benassarou, conduct, presentation of the research; Jebrane Bouaoud, conduct, presentation of the research; André Chaine, conduct, presentation of the research; Angélique Girod, conduct, presentation of the research; Nicolas Graillon, conduct, presentation of the research; Sylvie Testelin, conduct, presentation of the research; Mélika Amor‐Sahli, conduct, presentation of the research; Jean‐Philippe Foy, design, conduct, analysis, presentation of the research; Chloé Bertolus, design, conduct, presentation of the research.
Disclosures
Competing interests
The authors declare that they have no conflicts of interest.
Funding source
None.
Supporting information
Supplementary Figure S1. Predictions of the bone volume evolution model over time, for different variables: (A) flap type and (B) radiotherapy history. DCIA, deep circumflex iliac artery; FFF, fibula free flap; SFF, scapular system free flap.
Supporting Information.
Acknowledgments
We would like to thank Constance Fenoll and Reine Guibert for their help with data collection.
References
- 1. Pfister P, Speck NE, Gahl B, et al. Impact of adjuvant radiochemotherapy on free flap volume in head and neck reconstruction: a systematic review and meta‐analysis. J Plast Reconstr Aesthet Surg. 2024;91:24‐34. 10.1016/j.bjps.2024.02.026 [DOI] [PubMed] [Google Scholar]
- 2. Hsiao HT, Leu YS, Liu CJ, Tung KY, Lin CC. Radial forearm versus anterolateral thigh flap reconstruction after hemiglossectomy: functional assessment of swallowing and speech. J Reconstr Microsurg. 2008;24(2):085‐088. 10.1055/s-2008-1076097 [DOI] [PubMed] [Google Scholar]
- 3. Manrique O, Leland H, Langevin CJ, et al. Optimizing outcomes following total and subtotal tongue reconstruction: a systematic review of the contemporary literature. J Reconstr Microsurg. 2017;33(2):103‐111. 10.1055/s-0036-1593772 [DOI] [PubMed] [Google Scholar]
- 4. Matsui Y, Ohno K, Yamashita Y, Takahashi K. Factors influencing postoperative speech function of tongue cancer patients following reconstruction with fasciocutaneous/myocutaneous flaps—a multicenter study. Int J Oral Maxillofac Surg. 2007;36(7):601‐609. 10.1016/j.ijom.2007.01.014 [DOI] [PubMed] [Google Scholar]
- 5. Uwiera T, Seikaly H, Rieger J, Chau J, Harris JR. Functional outcomes after hemiglossectomy and reconstruction with a bilobed radial forearm free flap. J Otolaryngol. 2004;33(6):356‐359. 10.2310/7070.2004.00356 [DOI] [PubMed] [Google Scholar]
- 6. Razavi CR, Hostetter J, Shukla A, et al. Predictors of free flap volume loss in nonosseous reconstruction of head and neck oncologic defects. Ear Nose Throat J. 2022;101(1):48‐53. 10.1177/0145561320938903 [DOI] [PubMed] [Google Scholar]
- 7. Tarsitano A, Vietti MV, Cipriani R, Marchetti C. Functional results of microvascular reconstruction after hemiglossectomy: free anterolateral thigh flap versus free forearm flap. Acta Otorhinolaryngol Ital. 2013;33(6):374‐379. [PMC free article] [PubMed] [Google Scholar]
- 8. Haymerle G, Enzenhofer E, Lechner W, et al. The effect of adjuvant radiotherapy on radial forearm free flap volume after soft palate reconstruction in 13 patients. Clin Otolaryngol. 2018;43(2):742‐745. 10.1111/coa.13042 [DOI] [PubMed] [Google Scholar]
- 9. Conroy R. Estimation of ten‐year risk of fatal cardiovascular disease in Europe: the SCORE project. Eur Heart J. 2003;24(11):987‐1003. 10.1016/s0195-668x(03)00114-3 [DOI] [PubMed] [Google Scholar]
- 10. Bartko JJ. The Intraclass Correlation Coefficient as a measure of reliability. Psychol Rep. 1966;19(1):3‐11. 10.2466/pr0.1966.19.1.3 [DOI] [PubMed] [Google Scholar]
- 11. Akaike H. A new look at the statistical model identification. IEEE Trans Autom Control. 1974;19(6):716‐723. 10.1109/TAC.1974.1100705 [DOI] [Google Scholar]
- 12. Pinheiro JC, Bates DJ, DebRoy S, Sakar D. The Nlme Package: Linear and Nonlinear Mixed Effects Models, R Version 3. Vol 6. Springer; 2012. [Google Scholar]
- 13. Wickham H. Ggplot2. Springer; 2016. 10.1007/978-3-319-24277-4 [DOI] [Google Scholar]
- 14. Koo TK, Li MY. A guideline of selecting and reporting Intraclass Correlation Coefficients for reliability research. J Chiropr Med. 2016;15(2):155‐163. 10.1016/j.jcm.2016.02.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Bittermann G, Thönissen P, Poxleitner P, Zimmerer R, Vach K, Metzger MC. Microvascular transplants in head and neck reconstruction: 3D evaluation of volume loss. J Craniomaxillofac Surg. 2015;43(8):1319‐1324. 10.1016/j.jcms.2015.06.013 [DOI] [PubMed] [Google Scholar]
- 16. Kimura M, Takada K, Ishibashi K, et al. Quantitative analysis of chronological changes in the volume of flaps used for reconstruction of oral cavity defects. J Oral Sci. 2017;59(4):499‐503. 10.2334/josnusd.16-0683 [DOI] [PubMed] [Google Scholar]
- 17. Yun IS, Lee DW, Lee WJ, Lew DH, Choi EC, Rah DK. Correlation of neotongue volume changes with functional outcomes after long‐term follow‐up of total glossectomy. J Craniofac Surg. 2010;21(1):111‐116. 10.1097/SCS.0b013e3181c46692 [DOI] [PubMed] [Google Scholar]
- 18. Jeong HH, Jeong WS, Choi JW, et al. 3D computer simulation analysis of the flap volume change in total tongue reconstruction flaps. J Craniomaxillofac Surg. 2018;46(5):844‐850. 10.1016/j.jcms.2018.02.013 [DOI] [PubMed] [Google Scholar]
- 19. Cho KJ, Joo YH, Sun DI, Kim MS. Perioperative clinical factors affecting volume changes of reconstructed flaps in head and neck cancer patients: free versus regional flaps. Eur Arch Otrhinolaryngol. 2011;268(7):1061‐1065. 10.1007/s00405-010-1450-5 [DOI] [PubMed] [Google Scholar]
- 20. Kamizono K, Yoshida S, Yasumatsu R, Kadota H. Volumetric changes of transferred free anterolateral thigh flaps in head and neck lesions. Auris Nasus Larynx. 2021;48(4):751‐757. 10.1016/j.anl.2020.12.006 [DOI] [PubMed] [Google Scholar]
- 21. Kagaya Y, Arikawa M, Sekiyama T, Higashino T, Akazawa S. Chronological flap volume and distribution changes after reconstruction of total maxillectomy defect using a rectus abdominis myocutaneous flap. J Plast Reconstr Aesthet Surg. 2021;74(12):3341‐3352. 10.1016/j.bjps.2021.05.021 [DOI] [PubMed] [Google Scholar]
- 22. Joo YH, Hwang SH, Sun DI, Park JO, Cho KJ, Kim MS. Assessment of volume changes of radial forearm free flaps in head and neck cancer: long‐term results. Oral Oncol. 2011;47(1):72‐75. 10.1016/j.oraloncology.2010.11.001 [DOI] [PubMed] [Google Scholar]
- 23. Wang SJ, Zhang WB, Yu Y, Wang T, Yang HY, Peng X. Factors affecting volume change of anterolateral thigh flap in head and neck defect reconstruction. J Oral Maxillofac Surg. 2020;78(11):2090‐2098. 10.1016/j.joms.2020.06.017 [DOI] [PubMed] [Google Scholar]
- 24. Yamazaki M, Suzuki T, Hiraga C, et al. Effect of postoperative radiotherapy for free flap volume changing after tongue reconstruction. Oral Radiol. 2021;37(3):518‐523. 10.1007/s11282-020-00489-0 [DOI] [PubMed] [Google Scholar]
- 25. Yokoi S, Nishio N, Fujimoto Y, et al. Feasibility of virtual surgical simulation in the head and neck region for soft tissue reconstruction using free flap: a comparison of preoperative and postoperative volume measurement. Int J Oral Maxillofac Surg. 2021;50(3):316‐322. 10.1016/j.ijom.2020.07.025 [DOI] [PubMed] [Google Scholar]
- 26. Fujioka M, Masuda K, Imamura Y. Fatty tissue atrophy of free flap used for head and neck reconstruction. Microsurgery. 2011;31(1):32‐35. 10.1002/micr.20811 [DOI] [PubMed] [Google Scholar]
- 27. Yamaguchi K, Kimata Y, Onoda S, Mizukawa N, Onoda T. Quantitative analysis of free flap volume changes in head and neck reconstruction. Head Neck. 2012;34(10):1403‐1407. 10.1002/hed.21944 [DOI] [PubMed] [Google Scholar]
- 28. Woo SH, Kim YC, Jeong WS, Choi JW. A 3‐Dimensional analysis of flap volume change in hemi‐tongue reconstruction. Ann Plast Surg. 2022;89(6):e45‐e50. 10.1097/SAP.0000000000003306 [DOI] [PubMed] [Google Scholar]
- 29. Kim MS, Oh KH, Cho JG, et al. Assessment of chronological volume changes in radial forearm free flaps for tongue cancer. ORL. 2020;82:40‐46. 10.1159/000504196 [DOI] [PubMed] [Google Scholar]
- 30. Sakamoto Y, Takahara T, Ota Y, et al. MRI analysis of chronological changes in free‐flap volume in head and neck reconstruction by volumetry. Tokai J Exp Clin Med. 2014;39(1):44‐50. [PubMed] [Google Scholar]
- 31. Lin Y, He J, Zhang X, Wang H. Intraoperative factors associated with free flap failure in the head and neck region: a four‐year retrospective study of 216 patients and review of the literature. Int J Oral Maxillofac Surg. 2019;48(4):447‐451. 10.1016/j.ijom.2018.08.009 [DOI] [PubMed] [Google Scholar]
- 32. Wilkman T, Apajalahti S, Wilkman E, Törnwall J, Lassus P. A comparison of bone resorption over time: an analysis of the free scapular, iliac crest, and fibular microvascular flaps in mandibular reconstruction. J Oral Maxillofac Surg. 2017;75(3):616‐621. 10.1016/j.joms.2016.09.009 [DOI] [PubMed] [Google Scholar]
- 33. Taxis J, Nobis CP, Grau E, Kesting M, Moest T. Retrospective three‐dimensional analysis of bone resorption volumes of free microvascular scapular and fibular grafts. Br J Oral Maxillofac Surg. 2022;60(10):1417‐1423. 10.1016/j.bjoms.2022.08.002 [DOI] [PubMed] [Google Scholar]
- 34. Lee S, Thiele C. Factors associated with free flap complications after head and neck reconstruction and the molecular basis of fibrotic tissue rearrangement in preirradiated soft tissue. J Oral Maxillofac Surg. 2010;68(9):2169‐2178. 10.1016/j.joms.2009.08.026 [DOI] [PubMed] [Google Scholar]
- 35. Angelos PC, McCarn KE, Winn SR, et al. Development of an irradiated rodent model to study flap revascularization. Arch Facial Plast Surg. 2010;12(2):119‐122. 10.1001/archfacial.2010.16 [DOI] [PubMed] [Google Scholar]
- 36. Guelinckx PJ, Boeckx WD, Fossion E, Gruwez JA. Scanning electron microscopy of irradiated recipient blood vessels in head and neck free flaps. Plast Reconstr Surg. 1984;74(2):217‐226. 10.1097/00006534-198408000-00008 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary Figure S1. Predictions of the bone volume evolution model over time, for different variables: (A) flap type and (B) radiotherapy history. DCIA, deep circumflex iliac artery; FFF, fibula free flap; SFF, scapular system free flap.
Supporting Information.