Available online at www.sciencedirect.com
British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
Can we predict which patients are likely to develop severe complications following reconstruction for osteoradionecrosis? Andrew J. Lyons a,∗ , Iain Nixon a , Dionyssia Papadopoulou a , Siobhan Crichton b a b
Head and Neck Unit, Guy’s and St. Thomas’ Hospital NHS Foundation Trust, London, United Kingdom Division of Health and Social Care Research, King’s College, London, United Kingdom
Accepted 29 August 2013 Available online 28 October 2013
Abstract High morbidity has been reported for free vascularised reconstruction for osteoradionecrosis (ORN) and there are no apparent risk factors. A single nucleotide polymorphism in the transforming growth factor beta 1 gene (TGF-1) has been implicated in the cause of ORN and may also predict these complications. We studied a series of 30 consecutive patients who had had reconstruction for severe ORN with free tissue transfer in relation to their outcomes and complications for a number of risk factors including TFG-1 genotype, age, sex, comorbidities, site and stage of tumour, type of initial operation, and dose of radiotherapy or chemoradiotherapy. Two patients died and 2 flaps failed. Using the Clavien–Dindo classification, 16 patients developed grade III complications and 6 grade II. Median (IQR) duration of inpatient stay was 19.5 (12–25) days and the median (IQR) duration of outpatient treatment was 6 (4–11) weeks. No specific risk factors for postoperative complications were identified. In view of the severity and unpredictability of the complications, careful preoperative counselling of patients is essential. © 2013 Published by Elsevier Ltd. on behalf of The British Association of Oral and Maxillofacial Surgeons. Keywords: Osteoradionecrosis; Reconstruction; Complications
Introduction Osteoradionecrosis (ORN) affects between 2% and 22% of patients after radiotherapy to the jaws.1 Small lesions may be treated conservatively2 or not treated at all if asymptomatic, but when serious symptoms such as pathological fracture or fistula (Figs. 1 and 2) occur, free tissue transfer is often the only option. Several authors have alluded to the high rate of morbidity, which can be up to 55%,3–13 when undertaking reconstruction for advanced ORN of the jaws.9 95% survival of flaps is considered the standard acceptable outcome in reconstructions with free vascularised flaps and no previous ∗ Corresponding author at: Head and Neck Unit, Guy’s and St. Thomas’ Hospital NHS Foundation Trust, Great Maze Pond, London SE1 9RT, United Kingdom. Tel.: +44 0207 1884344; fax: +44 02071882180. E-mail address: [email protected] (A.J. Lyons).
radiotherapy, but a rate of 85% has been reported in a series of 53 patients who had reconstruction for ORN.11 Although often successful, free tissue transfer to the necrotic area is likely to be problematic in terms of healing because of radiation damage to all tissues in the field. One of the first problems is to find suitable vessels to give reliability to the flap and avoid arterial or venous failure. This can often be achieved if vessels of the contralateral side of the neck are used for the anastomosis, but this area can also be damaged by a combination of radiotherapy and operation, and it may preclude some types of free tissue transfer because the pedicle is too short, as may be the case with a scapular flap. Accordingly, several procedures have been described to provide reliable arterial supply and venous drainage. They include use of the transverse cervical vessels, transposition of the internal mammary vessels or cephalic vein, or both,
0266-4356/$ – see front matter © 2013 Published by Elsevier Ltd. on behalf of The British Association of Oral and Maxillofacial Surgeons.
http://dx.doi.org/10.1016/j.bjoms.2013.08.017
708
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
those not previously exposed to HBO.6–7 It is therefore difficult to predict which patients will have complications after reconstruction for ORN. One of the most recent theories on the mechanism for damage to the bone and other exposed tissues after radiation is the fibroatrophic process.13 Excessive fibroblastic activity modulated by transforming growth factor beta 1 (TFG-1) is the main factor, and an analysis of patients with a particular single nucleotide polymorphism T instead of C at −509 in the promoter region of the TFG-1 gene showed that it could predict ORN.14 As this genotype seems to have contributed to poor healing as in cases of ORN, it seems logical that the T allele might cause further healing problems after reconstructive surgery to the tissues affected.
Fig. 1. Exposed intraoral pathological fracture.
and incorporation of the loops between a major artery and vein for anastomosis.12 In many cases, as a consequence of inadequate vascular supply or the general physical condition of the patient, or both, surgeons abandon free tissue transfer in favour of a simpler pedicled flap. However, this may also cause problems. The vascularity of free tissue transfer flaps is usually superior to that of pedicled flaps, and a pedicled flap may further predispose the patient to problems with healing. Adequate bony reconstruction is seldom available with pedicled flaps and the lack of a mandible may debilitate the patient in terms of speech, swallowing, mastication, and aesthetics. Although these patients are at risk of wound healing complications that may be fatal, many have an uncomplicated operation and recover at least as well as those who have similar operations but do not have ORN. Although recognised coexisting conditions such as poor nutrition, diabetes, and alcohol misuse may increase the risk of problems with healing, one study found that smoking reduced the risk of ORN recurring.11 Another study found that the longer the onset of severe ORN from the date of radiotherapy, the greater the likelihood of complications.4 However, if treatment with hyperbaric oxygen (HBO) fails then patients with ORN have a higher risk of complications after free tissue transfer than
Patients and methods A total of 30 consecutive patients who had contracted ORN after radiotherapy or chemoradiotherapy were included in the study. At the time of operation all patients had progressive, active, and symptomatic ORN, and were classified as Epstein grade III,15 with the exception of 2 who had previously had reconstruction for ORN between 2007 and 2012 with composite microvascular free tissue transfer. There were 20 fibular flaps, 6 composite radial forearm flaps, one iliac crest flap, and 3 scapular flaps. Where possible the arterial and venous anastomoses were taken from the contralateral side of the neck, as there was likely to be less radiation damage. In 4 patients the neck vessels were unsuitable, and after mobilisation and transposition by a thoracic surgeon, the internal mammary vessels were used instead. The transverse cervical vessels and a transposed cephalic vein were used in one patient, and in 3 the cervical veins were inadequate so the cephalic vein was transposed. In each case the postoperative duration of hospital stay, the severity of complications graded according to the Clavien–Dindo classification,16 the duration of treatment in outpatients, and the need for further operation, were compared with the following variables: sex, smoking, alcohol consumption, chemotherapy or not, dose of radiation to the primary site and to the neck, the interval between radiotherapy and treatment, whether the primary site and neck were operated on, the original tumour site, coexisting conditions, and genotype. Statistical analysis
Fig. 2. Fistula with cervical communication and extensive bony necrosis.
Frequencies and percentages were used to summarise categorical characteristics of patients and treatment, while continuous data were summarised as median (IQR). The severity of complications was compared across groups using Fisher’s exact test or the Kruskall–Wallis test, as appropriate. Duration of stay and outpatient treatment were summarised as median (IQR) and compared across groups using the Mann–Whitney or Kruskall–Wallis test. Associations
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
between duration of stay and continuous characteristics were examined using Spearman’s correlation. All tests were 2sided and probabilities of less than 0.05 were considered significant. Analyses were conducted using STATA MP 12 (StataCorp LP, TX, USA).
709
Results Two patients died postoperatively, one at 4 and one at 18 days. The first had a cerebrovascular accident intraoperatively on the side of the neck where a repeated attempt had been made
Table 1 Associations between patients, treatment, and Clavien–Dindo grade for severity of complications. Data are number unless otherwise stated. No. of patients
Clavien–Dindo grade 0–I
II
p-Value III+
All
30
8
6
16
Genotype CC CT TT
6 14 10
3 2 3
1 4 1
2 8 6
0.52
Sex Male Female
22 8
6 2
3 3
13 3
0.40
Comorbidities None 1+
22 8
5 3
4 2
13 3
0.55
Insulin-dependent diabetes mellitus No Yes
27 3
7 1
6 0
14 2
0.10
Smoking status Never smoked Ex-smoker Current smoker
9 9 12
1 2 5
2 1 3
6 6 4
0.47
Alcohol consumption None Moderate Excessive
11 13 6
4 3 1
2 2 2
5 8 3
0.86
Operation at primary site No Yes
10 20
3 5
3 3
4 12
0.60
Operation in neck No Yes
10 20
2 6
4 2
4 12
0.22
Chemotherapy No Yes
17 13
6 2
3 3
8 8
0.61
Median (IQR) radiation dose (Gy) Primary site Neck Median (IQR) time from reconstruction to treatment (days)
30 30 30
Site Floor of mouth Posterior tongue SCC of the palate Tongue Tonsil Alveolus Buccal Maxilla Oropharynx Parotid Retromolars SCC, squamous cell carcinoma.
8 1 1 4 1 2 1 1 1 4 6
62.5 (55–63.8) 62.5 (55–65) 46 (28.5–75.5) 4 0 1 1 0 0 1 0 0 1 0
64.5 (52–65) 64.5 (52–65) 19.5 (18–40) 2 0 0 0 1 0 0 0 1 2 0
65 (56–65) 65 (58.5–65) 28.5 (22–69) 2 1 0 3 0 2 0 1 0 1 6
0.63 0.70 0.26 – – – – – – – – – – –
710
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
Fig. 3. Wound breakdown following successful free tissue transfer.
to use the stump of the external carotid for arterial supply. After this a vessel was found on the contralateral side of the neck but it was not used initially as it was a scapular flap with a short pedicle. At the time of death the flap was well-perfused. The second patient developed infection and breakdown of soft tissue after 14 days. Ultimately it caused secondary haemorrhage at the anastomosis with the internal mammary artery, which necessitated urgent transfer to the operating room. Unfortunately the patient removed his endotracheal tube 8 h after returning from the operating theatre, and respiratory arrest with fatal hypoxia ensued before reintubation. Two flaps failed because of a lack of arterial perfusion and could not be salvaged. One was a latissimus dorsi with scapular bone, the other a fibular flap that had been anastomosed to the transverse cervical vessels and cephalic vein. Reconstruction with a pectoralis major flap was initially successful, but after 4 weeks it became necrotic after it dehisced from each of its margins in turn. Further microvascular reconstruction with access by an alternative vessel was the only viable option. No other flaps required salvage. A total of 16 patients developed Clavien–Dindo grade III complications (Table 1). Most were serious wound dehiscences that required protracted non-operative inpatient and outpatient care (median IQR 19.5 days); others required a formal operation (Fig. 3). One nasolabial flap and one supraclavicular flap developed minor breakdown, and in 5 pectoralis major myocutaneous flaps in the neck, the skin broke down. There were 2 radial forearm fasciocutaneous flaps, one for a large dehiscence in the oral cavity with neck dehiscence, and one with a large dehiscence in the oral cavity alone. Although 4 patients had grade I complications they were not analysed, as they required no medical or surgical intervention. Grade II complications were analysed (n = 6) but they required pharmacological intervention only (see Table 2). Despite apparent successful free tissue transfer, the 2 fibular flaps that were used to reconstruct previous reconstructions slowly necrosed with non-union of the bony fragments. Duration of stay and the period of outpatient treatment varied widely. The median duration of hospital stay was 20 days (range 7–42). Most patients did not require outpatient treatment for dressings but one was treated for 37 weeks.
There was no apparent significant correlation between any of the adverse outcomes audited when compared with all the variables examined including the genotype of patients (Table 3). A total of 14 patients required an extended duration of wound dressing in outpatients, and some required this to the donor site as well as the head and neck (Fig. 3). Subsequently 5 patients required removal of their titanium reconstruction plates; in 3 this resulted in a disrupted union of the bony fragments with apparent bony necrosis. Although it caused problems with function and appearance, and one patient was offered further free tissue transfer, ultimately the bone healed with no more signs of infection. In these 3 patients the plates had been exposed intraorally or extraorally, or both, because the wound had broken down during their stay in hospital (Fig. 3). Analysis of the TGF-1 genotype at 509C-T indicated that 7/22 patients were homozygous for the CC allele, 6 were heterozygous (CT) and 9 were homozygous for the TT allele. This was not correlated with the severity of complications, or the duration of inpatient stay or outpatient treatment (see Fig. 4).
Discussion Data from this series treated by a single surgeon suggest that patients having reconstruction for ORN should be warned about the likelihood of a slow recovery and the potential for serious complications. No single factor could predict the severity of complications including the TGF-1 genotype, but as most of the patients were being treated for ORN, relatively few did not have the T variant allele at −509 in the TFG-1 gene, which makes conclusions based on the statistical analysis unreliable. In our series some of the worst complications developed in patients without this allele, which suggests that increased production of TGF-1, because of its activation after operation, is not likely to account for the reduced capacity of patients with ORN to heal. According to the fibroatrophic theory, it is excess fibrosis after radiotherapy that causes atrophy of the tissues. In a surgically produced wound excess fibrous tissue may not be as problematic in the short term. However, of the 3 patients who had necrosis and non-healing of the transferred tissue, 2 were homozygous and one was heterozygous for the T allele at −509 of the TGF 1 gene. These numbers are small but it is possible that the tendency towards excess fibrosis has deleterious consequences for reconstruction in the longer term. In this series 2 flaps failed and 2 patients died. One developed a large dehiscence in the neck 17 days postoperatively despite the flap being well-perfused. After returning from the operating theatre he had removed his endotracheal tube and it could not be replaced before he became anoxic and died. The second did not recover consciousness after operation and had developed a dense right hemiplegia caused by a thrombosis or embolus in the left circle of Willis, which had almost
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
711
Table 2 Associations between patients and treatment, and duration of hospital stay and outpatient treatment. Data are median (IQR). Duration of stay
p-Value
Duration of outpatient treatment
p-Value
All
19.5 (12–25)
Genotype CC CT TT
16 (12–33) 20 (14–22) 13.5 (10–27)
0.90
5.5 (4–9) 6 (4–15) 6 (3–9)
0.93
Sex Male Female
19.5 (12–22) 22 (10.5–35)
0.50
7 (4–13) 5 (3.5–7.5)
0.42
Comorbidities None 1+
19 (11–22) 20 (12.5–28.5)
0.61
5 (4–13) 6.5 (4–8)
0.77
Insulin-dependent diabetes mellitus No Yes
19 (11–25) 20 (12–30)
0.68
6 (4–11) 8 (2–37)
0.77
Smoking status Never smoked Ex-smoker Current smoker
20 (19–27) 20 (10–21) 14.5 (12–30.5)
0.59
4 (2–8) 7 (5–13) 7 (4–15)
0.40
Alcohol consumption None Moderate Excessive
12 (10–20) 20 (19–25) 18 (9–39)
0.42
4.5 (3–9) 6 (4–8) 13 (6–20)
0.30
Operation at primary site No Yes
20 (10–22) 19 (12–26)
0.79
5.5 (2–9) 7 (4–15)
0.34
Operation in neck No Yes
20 (10–22) 19 (12–26)
0.83
5.5 (2–13) 7 (4–11)
0.53
Chemotherapy No Yes
14 (10–25) 20 (19–22)
0.13
8 (3–15) 5.5 (4–8)
0.63
Radiation dose Primary site a Neck a Reconstruction to treatment time a
0.134 0.201 −0.232
0.48 0.29 0.22
a
6 (4–11)
−0.242 −0.145 0.046
0.21 0.45 0.81
Results estimate correlation between measures and duration of stay or duration of outpatient treatment.
Table 3 Complications compared against genotype. Genotype
p-Value
CC (n = 6)
CT (n = 14)
TT (n = 10)
Median (IQR) age (years) Median (IQR) time to operation (months)
57.5 (56–63) 52.5 (40–83)
64 (53–67) 27 (20–55)
62.5 (52–70) 31 (22–70)
0.71 0.64
Grade (No. of patients) 0–I II III+ Grade, median (IQR) Median (IQR) duration of inpatient stay (days) Median (IQR) duration of outpatient treatment (weeks)
3 1 2 1(0–3) 16 (12–33) 5.5 (4–9)
2 4 8 3(2–3) 20 (14–22) 6 (4–15)
3 1 6 3(0–3) 13.5 (10–27) 6 (3–9)
0.52
Alcohol consumption (no. of patients) None Moderate Excessive
3 2 1
3 8 3
5 3 2
0.57
Overall median (IQR) duration of outpatient treatment = 6 (4–11) weeks.
0.388 0.90 0.93
712
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
1 death
SCAPULAR 3
2 successful 1 failure
DCIA 1
Grade III 0RN 28 CASES
Composite RFF 6
1 NL
Successful
All succesful Plate removed 1 death 4 PM, 1 SC, 1 tongue 17 successful 2 RFF
FIBULAR 18
2 plates removed
1 failure
FAILED RECONSTRUCTION 2
FIBULAR 2
2 successful
1 PM (necrosed)
1 plate removed 2 collapsed bone
Fig. 4. Summary of results (DCIA, deep circumflex iliac artery flap; NL, nasolabial pedicled flap; PM, pectoralis major pedicled flap; RFF, radial forearm free flap; SC, supraclavicular pedicled flap).
certainly, and at least partly been caused by manipulation of the internal carotid before anastomosis. Although stenosis of the carotid has been reported after radiotherapy to the neck,17 and there was atheroma with no serious narrowing on his preoperative computed tomography (CT) angiogram, we thought that an end-to-side arterial anastomosis could be achieved. Consequently we have abandoned anastomosis to the common carotid, and we now access alternative vessels or use a pedicled flap. A high number of patients required a second procedure during their admission for definitive reconstruction (n = 16). Although 4 of the 7 pedicled flaps dehisced further, they all successfully stopped orocervical communications. However, this study shows that these patients may need extended outpatient treatment. No patients had had HBO treatment at any stage. Despite conflicting evidence for its benefit, a recent Cochrane Review has indicated that it may be of marginal benefit in these cases.18 No particular method of reconstruction seems to cause fewer or more complications. However, as a consequence of reviewing the results of this series, we always insert a free tissue transfer into the neck and oral cavity when reconstructing for ORN, either subcutaneously or exposed as a skin paddle. Since doing this, fewer wounds have broken down seriously.
Conflict of interest None declared. References 1. Støre G, Boysen M. Mandibular osteoradionecrosis: clinical behaviour and diagnostic aspects. Clin Otolaryngol Allied Sci 2000;25:378–84. 2. Delanian S, Depondt J, Lefaix JL. Major healing of refractory mandible osteoradionecrosis after treatment combining pentoxifylline and tocopherol: a phase II trial. Head Neck 2005;27:114–23. 3. Baumann DP, Yu P, Hanasono MM, Skoracki RJ. Free flap reconstruction of osteoradionecrosis of the mandible: a 10-year review and defect classification. Head Neck 2011;33:800–7. 4. Alam DS, Nuara M, Christian J. Analysis of outcomes of vascularized flap reconstruction in patients with advanced mandibular osteoradionecrosis. Otolaryngol Head Neck Surg 2009;141:196–201. 5. Chang DW, Oh HK, Robb GL, Miller MJ. Management of advanced mandibular osteoradionecrosis with free flap reconstruction. Head Neck 2001;23:830–5. 6. Hirsch DL, Bell RB, Dierks EJ, Potter JK, Potter BE. Analysis of microvascular free flaps for reconstruction of advanced mandibular osteoradionecrosis: a retrospective cohort study. J Oral Maxillofac Surg 2008;66:2545–56. 7. Gal TJ, Yueh B, Futran ND. Influence of prior hyperbaric oxygen therapy in complications following microvascular reconstruction for advanced osteoradionecrosis. Arch Otolaryngol Head Neck Surg 2003;129:72–6.
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713 8. Suh JD, Kim BP, Abemayor E, et al. Reirradiation after salvage surgery and microvascular free flap reconstruction for recurrent head and neck carcinoma. Otolaryngol Head Neck Surg 2008;139: 781–6. 9. Iseli TA, Yelverton JC, Iseli CE, Carroll WR, Magnuson JS, Rosenthal EL. Functional outcomes following secondary free flap reconstruction of the head and neck. Laryngoscope 2009;119: 856–60. 10. Suh JD, Blackwell KE, Sercarz JA, et al. Disease relapse after segmental resection and free flap reconstruction for mandibular osteoradionecrosis. Otolaryngol Head Neck Surg 2010;142:586–91. 11. Cannady SB, Dean N, Kroeker A, Albert TA, Rosenthal EL, Wax MK. Free flap reconstruction for osteoradionecrosis of the jaws – outcomes and predictive factors for success. Head Neck 2011;33: 424–8. 12. Jacobson AS, Eloy JA, Park E, Roman B, Genden EM. Vessel-depleted neck: techniques for achieving microvascular reconstruction. Head Neck 2008;30:201–7.
713
13. Delanian S, Lefaix JL. The radiation-induced fibroatrophic process: therapeutic perspective via the antioxidant pathway. Radiother Oncol 2004;73:119–31. 14. Lyons AJ, West CM, Risk J, et al. Osteoradionecrosis in head-and-neck cancer has a distinct genotype-dependent cause. Int J Radiat Oncol Biol Phys 2012;82:1479–84. 15. Epstein JB, Wong FL, Stevenson-Moore P. Osteoradionecrosis: clinical experience and a proposal for classification. J Oral Maxillofac Surg 1987;45:104–10. 16. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–13. 17. Greco A, Gallo A, De Virgilio A, et al. Carotid stenosis after adjuvant cervical radiotherapy in patients with head and neck cancers: a prospective controlled study. Clin Otolaryngol 2012;37:376–81. 18. Bennett MH, Feldmeier J, Hampson N, Smee R, Milross C. Hyperbaric oxygen therapy for late radiation tissue injury. Cochrane Database Syst Rev 2012;5:CD005005.
British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
Can we predict which patients are likely to develop severe complications following reconstruction for osteoradionecrosis? Andrew J. Lyons a,∗ , Iain Nixon a , Dionyssia Papadopoulou a , Siobhan Crichton b a b
Head and Neck Unit, Guy’s and St. Thomas’ Hospital NHS Foundation Trust, London, United Kingdom Division of Health and Social Care Research, King’s College, London, United Kingdom
Accepted 29 August 2013 Available online 28 October 2013
Abstract High morbidity has been reported for free vascularised reconstruction for osteoradionecrosis (ORN) and there are no apparent risk factors. A single nucleotide polymorphism in the transforming growth factor beta 1 gene (TGF-1) has been implicated in the cause of ORN and may also predict these complications. We studied a series of 30 consecutive patients who had had reconstruction for severe ORN with free tissue transfer in relation to their outcomes and complications for a number of risk factors including TFG-1 genotype, age, sex, comorbidities, site and stage of tumour, type of initial operation, and dose of radiotherapy or chemoradiotherapy. Two patients died and 2 flaps failed. Using the Clavien–Dindo classification, 16 patients developed grade III complications and 6 grade II. Median (IQR) duration of inpatient stay was 19.5 (12–25) days and the median (IQR) duration of outpatient treatment was 6 (4–11) weeks. No specific risk factors for postoperative complications were identified. In view of the severity and unpredictability of the complications, careful preoperative counselling of patients is essential. © 2013 Published by Elsevier Ltd. on behalf of The British Association of Oral and Maxillofacial Surgeons. Keywords: Osteoradionecrosis; Reconstruction; Complications
Introduction Osteoradionecrosis (ORN) affects between 2% and 22% of patients after radiotherapy to the jaws.1 Small lesions may be treated conservatively2 or not treated at all if asymptomatic, but when serious symptoms such as pathological fracture or fistula (Figs. 1 and 2) occur, free tissue transfer is often the only option. Several authors have alluded to the high rate of morbidity, which can be up to 55%,3–13 when undertaking reconstruction for advanced ORN of the jaws.9 95% survival of flaps is considered the standard acceptable outcome in reconstructions with free vascularised flaps and no previous ∗ Corresponding author at: Head and Neck Unit, Guy’s and St. Thomas’ Hospital NHS Foundation Trust, Great Maze Pond, London SE1 9RT, United Kingdom. Tel.: +44 0207 1884344; fax: +44 02071882180. E-mail address: [email protected] (A.J. Lyons).
radiotherapy, but a rate of 85% has been reported in a series of 53 patients who had reconstruction for ORN.11 Although often successful, free tissue transfer to the necrotic area is likely to be problematic in terms of healing because of radiation damage to all tissues in the field. One of the first problems is to find suitable vessels to give reliability to the flap and avoid arterial or venous failure. This can often be achieved if vessels of the contralateral side of the neck are used for the anastomosis, but this area can also be damaged by a combination of radiotherapy and operation, and it may preclude some types of free tissue transfer because the pedicle is too short, as may be the case with a scapular flap. Accordingly, several procedures have been described to provide reliable arterial supply and venous drainage. They include use of the transverse cervical vessels, transposition of the internal mammary vessels or cephalic vein, or both,
0266-4356/$ – see front matter © 2013 Published by Elsevier Ltd. on behalf of The British Association of Oral and Maxillofacial Surgeons.
http://dx.doi.org/10.1016/j.bjoms.2013.08.017
708
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
those not previously exposed to HBO.6–7 It is therefore difficult to predict which patients will have complications after reconstruction for ORN. One of the most recent theories on the mechanism for damage to the bone and other exposed tissues after radiation is the fibroatrophic process.13 Excessive fibroblastic activity modulated by transforming growth factor beta 1 (TFG-1) is the main factor, and an analysis of patients with a particular single nucleotide polymorphism T instead of C at −509 in the promoter region of the TFG-1 gene showed that it could predict ORN.14 As this genotype seems to have contributed to poor healing as in cases of ORN, it seems logical that the T allele might cause further healing problems after reconstructive surgery to the tissues affected.
Fig. 1. Exposed intraoral pathological fracture.
and incorporation of the loops between a major artery and vein for anastomosis.12 In many cases, as a consequence of inadequate vascular supply or the general physical condition of the patient, or both, surgeons abandon free tissue transfer in favour of a simpler pedicled flap. However, this may also cause problems. The vascularity of free tissue transfer flaps is usually superior to that of pedicled flaps, and a pedicled flap may further predispose the patient to problems with healing. Adequate bony reconstruction is seldom available with pedicled flaps and the lack of a mandible may debilitate the patient in terms of speech, swallowing, mastication, and aesthetics. Although these patients are at risk of wound healing complications that may be fatal, many have an uncomplicated operation and recover at least as well as those who have similar operations but do not have ORN. Although recognised coexisting conditions such as poor nutrition, diabetes, and alcohol misuse may increase the risk of problems with healing, one study found that smoking reduced the risk of ORN recurring.11 Another study found that the longer the onset of severe ORN from the date of radiotherapy, the greater the likelihood of complications.4 However, if treatment with hyperbaric oxygen (HBO) fails then patients with ORN have a higher risk of complications after free tissue transfer than
Patients and methods A total of 30 consecutive patients who had contracted ORN after radiotherapy or chemoradiotherapy were included in the study. At the time of operation all patients had progressive, active, and symptomatic ORN, and were classified as Epstein grade III,15 with the exception of 2 who had previously had reconstruction for ORN between 2007 and 2012 with composite microvascular free tissue transfer. There were 20 fibular flaps, 6 composite radial forearm flaps, one iliac crest flap, and 3 scapular flaps. Where possible the arterial and venous anastomoses were taken from the contralateral side of the neck, as there was likely to be less radiation damage. In 4 patients the neck vessels were unsuitable, and after mobilisation and transposition by a thoracic surgeon, the internal mammary vessels were used instead. The transverse cervical vessels and a transposed cephalic vein were used in one patient, and in 3 the cervical veins were inadequate so the cephalic vein was transposed. In each case the postoperative duration of hospital stay, the severity of complications graded according to the Clavien–Dindo classification,16 the duration of treatment in outpatients, and the need for further operation, were compared with the following variables: sex, smoking, alcohol consumption, chemotherapy or not, dose of radiation to the primary site and to the neck, the interval between radiotherapy and treatment, whether the primary site and neck were operated on, the original tumour site, coexisting conditions, and genotype. Statistical analysis
Fig. 2. Fistula with cervical communication and extensive bony necrosis.
Frequencies and percentages were used to summarise categorical characteristics of patients and treatment, while continuous data were summarised as median (IQR). The severity of complications was compared across groups using Fisher’s exact test or the Kruskall–Wallis test, as appropriate. Duration of stay and outpatient treatment were summarised as median (IQR) and compared across groups using the Mann–Whitney or Kruskall–Wallis test. Associations
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
between duration of stay and continuous characteristics were examined using Spearman’s correlation. All tests were 2sided and probabilities of less than 0.05 were considered significant. Analyses were conducted using STATA MP 12 (StataCorp LP, TX, USA).
709
Results Two patients died postoperatively, one at 4 and one at 18 days. The first had a cerebrovascular accident intraoperatively on the side of the neck where a repeated attempt had been made
Table 1 Associations between patients, treatment, and Clavien–Dindo grade for severity of complications. Data are number unless otherwise stated. No. of patients
Clavien–Dindo grade 0–I
II
p-Value III+
All
30
8
6
16
Genotype CC CT TT
6 14 10
3 2 3
1 4 1
2 8 6
0.52
Sex Male Female
22 8
6 2
3 3
13 3
0.40
Comorbidities None 1+
22 8
5 3
4 2
13 3
0.55
Insulin-dependent diabetes mellitus No Yes
27 3
7 1
6 0
14 2
0.10
Smoking status Never smoked Ex-smoker Current smoker
9 9 12
1 2 5
2 1 3
6 6 4
0.47
Alcohol consumption None Moderate Excessive
11 13 6
4 3 1
2 2 2
5 8 3
0.86
Operation at primary site No Yes
10 20
3 5
3 3
4 12
0.60
Operation in neck No Yes
10 20
2 6
4 2
4 12
0.22
Chemotherapy No Yes
17 13
6 2
3 3
8 8
0.61
Median (IQR) radiation dose (Gy) Primary site Neck Median (IQR) time from reconstruction to treatment (days)
30 30 30
Site Floor of mouth Posterior tongue SCC of the palate Tongue Tonsil Alveolus Buccal Maxilla Oropharynx Parotid Retromolars SCC, squamous cell carcinoma.
8 1 1 4 1 2 1 1 1 4 6
62.5 (55–63.8) 62.5 (55–65) 46 (28.5–75.5) 4 0 1 1 0 0 1 0 0 1 0
64.5 (52–65) 64.5 (52–65) 19.5 (18–40) 2 0 0 0 1 0 0 0 1 2 0
65 (56–65) 65 (58.5–65) 28.5 (22–69) 2 1 0 3 0 2 0 1 0 1 6
0.63 0.70 0.26 – – – – – – – – – – –
710
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
Fig. 3. Wound breakdown following successful free tissue transfer.
to use the stump of the external carotid for arterial supply. After this a vessel was found on the contralateral side of the neck but it was not used initially as it was a scapular flap with a short pedicle. At the time of death the flap was well-perfused. The second patient developed infection and breakdown of soft tissue after 14 days. Ultimately it caused secondary haemorrhage at the anastomosis with the internal mammary artery, which necessitated urgent transfer to the operating room. Unfortunately the patient removed his endotracheal tube 8 h after returning from the operating theatre, and respiratory arrest with fatal hypoxia ensued before reintubation. Two flaps failed because of a lack of arterial perfusion and could not be salvaged. One was a latissimus dorsi with scapular bone, the other a fibular flap that had been anastomosed to the transverse cervical vessels and cephalic vein. Reconstruction with a pectoralis major flap was initially successful, but after 4 weeks it became necrotic after it dehisced from each of its margins in turn. Further microvascular reconstruction with access by an alternative vessel was the only viable option. No other flaps required salvage. A total of 16 patients developed Clavien–Dindo grade III complications (Table 1). Most were serious wound dehiscences that required protracted non-operative inpatient and outpatient care (median IQR 19.5 days); others required a formal operation (Fig. 3). One nasolabial flap and one supraclavicular flap developed minor breakdown, and in 5 pectoralis major myocutaneous flaps in the neck, the skin broke down. There were 2 radial forearm fasciocutaneous flaps, one for a large dehiscence in the oral cavity with neck dehiscence, and one with a large dehiscence in the oral cavity alone. Although 4 patients had grade I complications they were not analysed, as they required no medical or surgical intervention. Grade II complications were analysed (n = 6) but they required pharmacological intervention only (see Table 2). Despite apparent successful free tissue transfer, the 2 fibular flaps that were used to reconstruct previous reconstructions slowly necrosed with non-union of the bony fragments. Duration of stay and the period of outpatient treatment varied widely. The median duration of hospital stay was 20 days (range 7–42). Most patients did not require outpatient treatment for dressings but one was treated for 37 weeks.
There was no apparent significant correlation between any of the adverse outcomes audited when compared with all the variables examined including the genotype of patients (Table 3). A total of 14 patients required an extended duration of wound dressing in outpatients, and some required this to the donor site as well as the head and neck (Fig. 3). Subsequently 5 patients required removal of their titanium reconstruction plates; in 3 this resulted in a disrupted union of the bony fragments with apparent bony necrosis. Although it caused problems with function and appearance, and one patient was offered further free tissue transfer, ultimately the bone healed with no more signs of infection. In these 3 patients the plates had been exposed intraorally or extraorally, or both, because the wound had broken down during their stay in hospital (Fig. 3). Analysis of the TGF-1 genotype at 509C-T indicated that 7/22 patients were homozygous for the CC allele, 6 were heterozygous (CT) and 9 were homozygous for the TT allele. This was not correlated with the severity of complications, or the duration of inpatient stay or outpatient treatment (see Fig. 4).
Discussion Data from this series treated by a single surgeon suggest that patients having reconstruction for ORN should be warned about the likelihood of a slow recovery and the potential for serious complications. No single factor could predict the severity of complications including the TGF-1 genotype, but as most of the patients were being treated for ORN, relatively few did not have the T variant allele at −509 in the TFG-1 gene, which makes conclusions based on the statistical analysis unreliable. In our series some of the worst complications developed in patients without this allele, which suggests that increased production of TGF-1, because of its activation after operation, is not likely to account for the reduced capacity of patients with ORN to heal. According to the fibroatrophic theory, it is excess fibrosis after radiotherapy that causes atrophy of the tissues. In a surgically produced wound excess fibrous tissue may not be as problematic in the short term. However, of the 3 patients who had necrosis and non-healing of the transferred tissue, 2 were homozygous and one was heterozygous for the T allele at −509 of the TGF 1 gene. These numbers are small but it is possible that the tendency towards excess fibrosis has deleterious consequences for reconstruction in the longer term. In this series 2 flaps failed and 2 patients died. One developed a large dehiscence in the neck 17 days postoperatively despite the flap being well-perfused. After returning from the operating theatre he had removed his endotracheal tube and it could not be replaced before he became anoxic and died. The second did not recover consciousness after operation and had developed a dense right hemiplegia caused by a thrombosis or embolus in the left circle of Willis, which had almost
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
711
Table 2 Associations between patients and treatment, and duration of hospital stay and outpatient treatment. Data are median (IQR). Duration of stay
p-Value
Duration of outpatient treatment
p-Value
All
19.5 (12–25)
Genotype CC CT TT
16 (12–33) 20 (14–22) 13.5 (10–27)
0.90
5.5 (4–9) 6 (4–15) 6 (3–9)
0.93
Sex Male Female
19.5 (12–22) 22 (10.5–35)
0.50
7 (4–13) 5 (3.5–7.5)
0.42
Comorbidities None 1+
19 (11–22) 20 (12.5–28.5)
0.61
5 (4–13) 6.5 (4–8)
0.77
Insulin-dependent diabetes mellitus No Yes
19 (11–25) 20 (12–30)
0.68
6 (4–11) 8 (2–37)
0.77
Smoking status Never smoked Ex-smoker Current smoker
20 (19–27) 20 (10–21) 14.5 (12–30.5)
0.59
4 (2–8) 7 (5–13) 7 (4–15)
0.40
Alcohol consumption None Moderate Excessive
12 (10–20) 20 (19–25) 18 (9–39)
0.42
4.5 (3–9) 6 (4–8) 13 (6–20)
0.30
Operation at primary site No Yes
20 (10–22) 19 (12–26)
0.79
5.5 (2–9) 7 (4–15)
0.34
Operation in neck No Yes
20 (10–22) 19 (12–26)
0.83
5.5 (2–13) 7 (4–11)
0.53
Chemotherapy No Yes
14 (10–25) 20 (19–22)
0.13
8 (3–15) 5.5 (4–8)
0.63
Radiation dose Primary site a Neck a Reconstruction to treatment time a
0.134 0.201 −0.232
0.48 0.29 0.22
a
6 (4–11)
−0.242 −0.145 0.046
0.21 0.45 0.81
Results estimate correlation between measures and duration of stay or duration of outpatient treatment.
Table 3 Complications compared against genotype. Genotype
p-Value
CC (n = 6)
CT (n = 14)
TT (n = 10)
Median (IQR) age (years) Median (IQR) time to operation (months)
57.5 (56–63) 52.5 (40–83)
64 (53–67) 27 (20–55)
62.5 (52–70) 31 (22–70)
0.71 0.64
Grade (No. of patients) 0–I II III+ Grade, median (IQR) Median (IQR) duration of inpatient stay (days) Median (IQR) duration of outpatient treatment (weeks)
3 1 2 1(0–3) 16 (12–33) 5.5 (4–9)
2 4 8 3(2–3) 20 (14–22) 6 (4–15)
3 1 6 3(0–3) 13.5 (10–27) 6 (3–9)
0.52
Alcohol consumption (no. of patients) None Moderate Excessive
3 2 1
3 8 3
5 3 2
0.57
Overall median (IQR) duration of outpatient treatment = 6 (4–11) weeks.
0.388 0.90 0.93
712
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713
1 death
SCAPULAR 3
2 successful 1 failure
DCIA 1
Grade III 0RN 28 CASES
Composite RFF 6
1 NL
Successful
All succesful Plate removed 1 death 4 PM, 1 SC, 1 tongue 17 successful 2 RFF
FIBULAR 18
2 plates removed
1 failure
FAILED RECONSTRUCTION 2
FIBULAR 2
2 successful
1 PM (necrosed)
1 plate removed 2 collapsed bone
Fig. 4. Summary of results (DCIA, deep circumflex iliac artery flap; NL, nasolabial pedicled flap; PM, pectoralis major pedicled flap; RFF, radial forearm free flap; SC, supraclavicular pedicled flap).
certainly, and at least partly been caused by manipulation of the internal carotid before anastomosis. Although stenosis of the carotid has been reported after radiotherapy to the neck,17 and there was atheroma with no serious narrowing on his preoperative computed tomography (CT) angiogram, we thought that an end-to-side arterial anastomosis could be achieved. Consequently we have abandoned anastomosis to the common carotid, and we now access alternative vessels or use a pedicled flap. A high number of patients required a second procedure during their admission for definitive reconstruction (n = 16). Although 4 of the 7 pedicled flaps dehisced further, they all successfully stopped orocervical communications. However, this study shows that these patients may need extended outpatient treatment. No patients had had HBO treatment at any stage. Despite conflicting evidence for its benefit, a recent Cochrane Review has indicated that it may be of marginal benefit in these cases.18 No particular method of reconstruction seems to cause fewer or more complications. However, as a consequence of reviewing the results of this series, we always insert a free tissue transfer into the neck and oral cavity when reconstructing for ORN, either subcutaneously or exposed as a skin paddle. Since doing this, fewer wounds have broken down seriously.
Conflict of interest None declared. References 1. Støre G, Boysen M. Mandibular osteoradionecrosis: clinical behaviour and diagnostic aspects. Clin Otolaryngol Allied Sci 2000;25:378–84. 2. Delanian S, Depondt J, Lefaix JL. Major healing of refractory mandible osteoradionecrosis after treatment combining pentoxifylline and tocopherol: a phase II trial. Head Neck 2005;27:114–23. 3. Baumann DP, Yu P, Hanasono MM, Skoracki RJ. Free flap reconstruction of osteoradionecrosis of the mandible: a 10-year review and defect classification. Head Neck 2011;33:800–7. 4. Alam DS, Nuara M, Christian J. Analysis of outcomes of vascularized flap reconstruction in patients with advanced mandibular osteoradionecrosis. Otolaryngol Head Neck Surg 2009;141:196–201. 5. Chang DW, Oh HK, Robb GL, Miller MJ. Management of advanced mandibular osteoradionecrosis with free flap reconstruction. Head Neck 2001;23:830–5. 6. Hirsch DL, Bell RB, Dierks EJ, Potter JK, Potter BE. Analysis of microvascular free flaps for reconstruction of advanced mandibular osteoradionecrosis: a retrospective cohort study. J Oral Maxillofac Surg 2008;66:2545–56. 7. Gal TJ, Yueh B, Futran ND. Influence of prior hyperbaric oxygen therapy in complications following microvascular reconstruction for advanced osteoradionecrosis. Arch Otolaryngol Head Neck Surg 2003;129:72–6.
A.J. Lyons et al. / British Journal of Oral and Maxillofacial Surgery 51 (2013) 707–713 8. Suh JD, Kim BP, Abemayor E, et al. Reirradiation after salvage surgery and microvascular free flap reconstruction for recurrent head and neck carcinoma. Otolaryngol Head Neck Surg 2008;139: 781–6. 9. Iseli TA, Yelverton JC, Iseli CE, Carroll WR, Magnuson JS, Rosenthal EL. Functional outcomes following secondary free flap reconstruction of the head and neck. Laryngoscope 2009;119: 856–60. 10. Suh JD, Blackwell KE, Sercarz JA, et al. Disease relapse after segmental resection and free flap reconstruction for mandibular osteoradionecrosis. Otolaryngol Head Neck Surg 2010;142:586–91. 11. Cannady SB, Dean N, Kroeker A, Albert TA, Rosenthal EL, Wax MK. Free flap reconstruction for osteoradionecrosis of the jaws – outcomes and predictive factors for success. Head Neck 2011;33: 424–8. 12. Jacobson AS, Eloy JA, Park E, Roman B, Genden EM. Vessel-depleted neck: techniques for achieving microvascular reconstruction. Head Neck 2008;30:201–7.
713
13. Delanian S, Lefaix JL. The radiation-induced fibroatrophic process: therapeutic perspective via the antioxidant pathway. Radiother Oncol 2004;73:119–31. 14. Lyons AJ, West CM, Risk J, et al. Osteoradionecrosis in head-and-neck cancer has a distinct genotype-dependent cause. Int J Radiat Oncol Biol Phys 2012;82:1479–84. 15. Epstein JB, Wong FL, Stevenson-Moore P. Osteoradionecrosis: clinical experience and a proposal for classification. J Oral Maxillofac Surg 1987;45:104–10. 16. Dindo D, Demartines N, Clavien PA. Classification of surgical complications: a new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg 2004;240:205–13. 17. Greco A, Gallo A, De Virgilio A, et al. Carotid stenosis after adjuvant cervical radiotherapy in patients with head and neck cancers: a prospective controlled study. Clin Otolaryngol 2012;37:376–81. 18. Bennett MH, Feldmeier J, Hampson N, Smee R, Milross C. Hyperbaric oxygen therapy for late radiation tissue injury. Cochrane Database Syst Rev 2012;5:CD005005.
