Vol. 117 No. 3 March 2014
Can tissue spectrophotometry and laser Doppler flowmetry help to identify patients at risk for wound healing disorders after neck dissection? Nils H. Rohleder, MD, DDS,a Sandra Flensberg, MD,b Florian Bauer, MD, DDS,a Stefan Wagenpfeil, MSc, PhD,c Craig J. Wales, MD, DDS,d Steffen Koerdt, MD,a Klaus D. Wolff, MD, DDS, PhD,a Frank Hölzle, MD, DDS, PhD,e Timm Steiner, MD, DDS,e and Marco R. Kesting, MD, DDS, PhDa Technische Universität München, Munich, Germany; Saarland University, Homburg, Germany; Southern General Hospital, Glasgow, Scotland; RWTH Aachen University, Aachen, Germany
Objective. Microcirculation and oxygen supply in cervical skin were measured with an optical, noninvasive method in patients with or without radiotherapy before neck dissection. The course of wound healing was monitored after the surgical procedure to identify predictive factors for postoperative wound healing disorders. Study Design. Tissue spectrophotometry and laser Doppler flowmetry were used to determine capillary oxygen saturation, hemoglobin concentration, blood flow, and blood velocity at 2-mm and 8-mm depths in the cervical skin of 91 patients before neck dissection in a maxillofacial unit of a university hospital in Munich, Germany. Parameters were evaluated for differences between patients with irradiation (24) and without (67) and patients with wound healing disorders (25) and without (66) (univariate or multivariate statistical analyses). Results. Velocity at 2 mm was lower in irradiated skin (P ¼ .016). Flow at 2 mm was higher in patients with wound healing disorders (P ¼ .018). Conclusions. High flow values could help to identify patients at risk for cervical wound healing disorders. (Oral Surg Oral Med Oral Pathol Oral Radiol 2014;117:302-311)
Since the first description of a neck dissection (ND) by Jawdynski in 1888 and refinements by Crile in the early 20th century, cervical lymphadenectomy has become the predominant surgical treatment option for cervical lymph node metastases.1 Radical ND (RND) procedures, which comprise removal of the sternocleidomastoid muscle with the cranial accessory nerve, internal jugular vein, and carotid artery together with the lymphatic tissue, have increasingly been replaced with less radical selective neck dissection (SND).2 Thus, postoperative morbidity has been reduced.3 However, RND is more effective than SND in controlling particular tumors such as malignant melanoma.4 After surgical resection of tumors such as oral squamous cell carcinoma (OSCC), radiotherapy decreases the risk of local recurrence.5 Adjuvant radiotherapy after Coauthors Rohleder, Flensberg, and Bauer contributed equally to this work as first authors. a Department of Oral and Maxillofacial Surgery, Technische Universität München. b Department of Anaesthesiology, Technische Universität München. c Institute for Medical Biometry, Epidemiology and Medical Informatics, Saarland University. d Department of Oral and Maxillofacial Surgery, Southern General Hospital. e Department of Oral and Maxillofacial Surgery, RWTH Aachen University. Received for publication Jul 31, 2013; returned for revision Sep 22, 2013; accepted for publication Nov 25, 2013. Ó 2014 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2013.11.497
302
OSCC resection usually consists of up to 66 Gy, with attempts to lower overall doses while maintaining outcomes.6,7 Radiotherapy to the head and neck has been found to influence cutaneous gene expression of extracellular matrix factors and bears the risk of acute and long-term side effects such as xerostomia.7,8 In regions such as the abdominal wall, it has been found to increase perioperative complications in terms of negatively affecting the course and results of wound healing.9 In the head and neck, the association of radiotherapy with the development of postoperative wound healing disorders is not completely understood. Patients requiring surgery in previously irradiated tissue (for example, because of local recurrence or late lymph node metastases) are assumed to be at greater risk for developing delayed wound healing; however, sparse literature is available to confirm this idea. Recently, Pellini et al.10 investigated risk factors for wound healing disorders after neck dissection. The authors identified preoperative chemoradiotherapy and the mode of ND as predictive factors
Statement of Clinical Relevance The “blood flow” parameter (measured with laser Doppler flowmetry) is a predictive value for cervical postoperative wound healing disorders, independent of previous radiotherapy or none. This helps to identify patients with a risk for a dehiscence after selective neck dissection.
OOOO Volume 117, Number 3
Fig. 1. Cervical wound healing disorder developed after selective neck dissection in a female patient.
for wound complications (RND and modified RND increased the risk in opposition to SND). No clinical tool is available to determine the individual risk for the development of wound healing disorders in patients requiring ND. Radiation induces vascular endothelial injuries that subsequently lead to tissue ischemia and atrophy.11 Development of fibrosis in irradiated tissue might be attributable to hypoxia, which induces profibrotic cytokines and collagen formation in laboratory experiments and is considered to induce angiogenesis.12,13 Oxygen tension in irradiated skin has been found to be decreased in animal models, but no clear result has been obtained with regard to whether hypoxia is the predominant factor for the induction of angiogenic cytokines.13 The measurement of oxygenation in irradiated skin in humans has been carried out by means of transcutaneous oxygen microelectrodes and has shown a reduction in oxygen tension in patients with previous radiotherapy. However, the results are again inconsistent.13-15 Another method for the detection of hypoxia in irradiated human tissue is the immunohistochemical staining of the hypoxia marker carbonic anhydrase IX.13 However, these methods are not applicable in daily clinical routine. This prospective study investigates, for the first time, the use of noninvasive simultaneous tissue spectrophotometry and laser Doppler flowmetry for the measurement of tissue oxygenation and microperfusion parameters in the cervical skin of patients with or without previous local radiotherapy. Measurements were conducted at the bedside on the day before SND. Postoperatively, patients were monitored for the development of wound healing disorders of the ND incision (Figure 1). Assessment of the variable “wound healing disorder” was dichotomized (no, complete primary wound healing; yes, appearance of any minor or
ORIGINAL ARTICLE Rohleder et al. 303
Fig. 2. Device used for simultaneous tissue spectrophotometry and laser Doppler flowmetry (O2C; Lea Medizintechnik GmbH, Giessen, Germany). Left, Portable device with live monitoring and recording software for parameters oxygen saturation, relative hemoglobin concentration, blood flow, and blood velocity. Right, Detailed photography of the laser sensor, which is attached to the main device. When applied to the region of interest on the skin, the aforementioned parameters are simultaneously measured at both 2 mm and 8 mm tissue depths.
major dehiscence). Data were analyzed to test the following assumptions: “History of cervical radiotherapy alters local tissue oxygenation/perfusion” (null hypothesis I) and “Simultaneous tissue spectrophotometry and laser Doppler flowmetry can predict wound healing disorders after neck dissection” (null hypothesis II). These null hypotheses were tested to achieve the 2 objectives of the study: to investigate the influence of cervical radiotherapy on tissue oxygenation/perfusion and wound healing after neck dissection (objective I) and to identify a predictive marker for the development of postoperative wound healing disorders after neck dissection (objective II). For this reason, the study design included preoperative measurement of the tissue oxygenation/perfusion parameters and also postoperative monitoring of the course of wound healing, and an analysis of the medical history of the patients with respect to cervical radiotherapy. A potential predictive marker for development of a wound healing disorder would then be evaluated by statistical correlation of preand postoperative data.
PATIENTS AND METHODS Patients A total of 91 patients who had undergone surgical procedures for OSCC or other oral malignancies in the Department of Oral and Maxillofacial Surgery, Technische Universität München, Munich, Germany, between January 1, 2012, and December 31, 2012, were included. All patients were treated by resection of the cervical lymph nodes, tumor resection, and reconstruction by using local or free flaps. Clinical data were obtained before surgery (e.g., history of cervical radiotherapy in patients with secondary surgical procedures, alcohol/nicotine abuse) and during the hospital
ORAL AND MAXILLOFACIAL SURGERY 304 Rohleder et al.
OOOO March 2014
Table I. Patient characteristics compared between patients without or with history of cervical radiotherapy
Characteristic Age (y), mean; median; range* Sex, n (%)y Male Female Alcohol abuse, n (%)y Nicotine abuse, n (%)y Postoperative cervical wound healing disorder, n (%)y
All patients, No radiotherapy, Radiotherapy, n (%) n (%) n (%) 91 (100) 67 (73.6) 24 (26.4) 59.9; 60.0; 20-86
60.6; 61; 20-86
57.8; 58.5; 29-86
61 (67.0) 30 (33.0) 41 (45.1)
46 (75.4) 21 (70.0) 32 (78.0)
15 (24.6) 9 (30.0) 9 (22.0)
54 (59.3)
40 (74.1)
14 (25.9)
25 (27.5)
15 (60.0)
10 (40.0)
Table II. Patient characteristics compared between patients without or with wound healing disorder after neck dissection
Characteristic Age (y), mean; median; range* Sex, n (%)y Male Female Alcohol abuse, n (%)y Nicotine abuse, n (%)y History of cervical radiotherapy, n (%)y
All patients, Regular wound Wound healing n (%) healing, n (%) disorder, n (%) 91 (100) 66 (72.5) 25 (27.5) 59.9; 60.0; 20-86
59.4; 60; 20-86 60.9; 60; 43-86
61 (67.0) 30 (33.0) 41 (45.1)
41 (67.2) 25 (83.3) 27 (65.9)
20 (32.8) 5 (16.7) 14 (34.1)
54 (59.3)
36 (66.7)
18 (33.3)
24 (26.4)
14 (58.3)
10 (41.7)
Percentages in column “All patients” refer to total patient collective. Percentages in the subsequent columns refer to the number of cases in each category. No statistically significant difference was evident between the groups, as calculated with Mann-Whitney U test or Fisher exact test, as marked. *Mann-Whitney U test. y Fisher exact test.
Percentages in column “All patients” refer to total patient collective. Percentages in the subsequent columns refer to the number of cases in each category. No statistically significant difference was evident between the groups, as calculated with Mann-Whitney U test or Fisher exact test, as marked. *Mann-Whitney U test. y Fisher exact test.
stay (monitoring for development of cervical wound healing disorders). The medical ethics committee of the Technische Universität München approved the research reported in this article; this research also complies with the ethical rules for human experimentation stated in the Declaration of Helsinki (institutional review board registration No. 263309). All patients gave written informed consent.
measurement of the oxygen saturation and relative hemoglobin concentration parameters is performed with white light. The degree of oxygen saturation of hemoglobin is calculated by measurement of the color change of the blood (venous blood, blue; oxygenated blood, bright red). The relative hemoglobin concentration parameter is determined from absorption measurements within the illuminated tissue volume.
Measurement of parameters indicating oxygen saturation and microperfusion In a noninvasive procedure, the oxygen saturation of hemoglobin, the relative hemoglobin concentration, the blood flow (vessel filling), and the blood flow velocity in microvessels of cervical skin were determined by means of a mobile device (O2C; Lea, Giessen, Germany; Figure 2, left) connected to a glass fiber probe (see Figure 2, right). This system allows the monitoring of the oxygen supply in various layers of tissue. In the present study, the aforementioned parameters were measured at depths of 2 mm and 8 mm. The superficial parameters were recorded by one channel, whereas those of deeper tissues, predominantly the cervical subcutaneous fatty tissue, were recorded by a second channel. The measurement principle of the device makes use of both a laser Doppler technique and tissue spectrophotometry (using white light). Laser light is used to determine the blood flow and the blood flow velocity parameters by measuring the Doppler shift of the erythrocytes caused by their movement within the vessels. Simultaneous
Surgical procedure All SNDs were carried out under standardized conditions by 2 operators from a team of 9 surgeons. Skin incision was carried out 2 cm below the lower margin of the mandible with a scalpel and was extended from the mental region to the sternocleidomastoid muscle, displaying the external jugular vein and the great auricular nerve. In the case of bilateral ND, the incision was extended to the opposite sternocleidomastoid muscle. Subplatysmal flaps were raised, and levels I to III were dissected as a standard for staging ND. For positive nodal disease or a bilateral tumor, levels I to V were removed on the more affected side, and levels I to III were removed on the contralateral side. Closure was carried out after the placement of a Redon drain on the posterior end of the wound (2 in case of bilateral ND) and subcutaneous suturing (including platysma) with single-knot technique (Vicryl 3-0 resorbable suturing material; Ethicon, Norderstedt, Germany). Skin closure was carried out by using continuous suturing (Ethilon 4-0 nylon suture; Ethicon).
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Rohleder et al. 305
Fig. 3. Box plot diagrams indicating results of comparison of parameters measured with simultaneous tissue spectrophotometry and laser Doppler flowmetry in cervical skin of patients who had never previously received radiotherapy (RTx) and patients that were previously irradiated at the neck. Columns indicate parameters oxygen saturation (A), relative hemoglobin amount (B), blood flow (C), and blood velocity (D). Upper row shows parameters as measured at a 2 mm tissue depth; lower row shows parameters as measured at an 8 mm tissue depth. The lines within the boxes indicate the medians; the top edge of the boxes represents the 75th percentile, and the bottom edge represents the 25th percentile. The range is shown as a vertical line (extreme values are not shown). Statistical analysis found a significant difference between the values measured for blood velocity at a 2 mm tissue depth: a median of 17.57 arbitrary units (AU) was measured in patients with no RTx, whereas a median of 14.77 AU was measured in patients with previous cervical RTx (P ¼ .016). None of the remaining parameters was affected by RTx in medical history.
Statistical analysis IBM SPSS version 21.0 (IBM, Armonk, NY, USA) was used. For univariate analysis, differences with respect to age and the measured parameters were analyzed by using the Mann-Whitney U test. Differences with respect to sex, nicotine abuse, and alcohol abuse were analyzed by using the Fisher exact test. For multivariate analysis, a linear regression analysis (with the independent variables age, sex, alcohol abuse, nicotine abuse, wound healing disorder (yes/no), history of radiotherapy (yes/ no), and interaction between wound healing disorder and history of radiotherapy) was performed separately for each of the outcome variables oxygen saturation, relative hemoglobin concentration, blood flow, and blood velocity (each at 2-mm and 8-mm depth) as a multiple approach. Stepwise forward and backward variable
selection approaches were applied to account for potential multicollinearity. Receiver operating characteristic (ROC) analysis was carried out, and ROC curves were created to show the diagnostic value of respective outcome measures. All P values are given as 2-sided and are subject to a local significance level of 5%.
RESULTS For the 91 patients included, patient characteristics and measured oxygenation/microperfusion parameters were compared (1) between the group without any previous cervical radiotherapy and the group with a history of cervical radiotherapy and (2) between the group with regular wound healing and the group with postoperative wound healing disorders after SND.
ORAL AND MAXILLOFACIAL SURGERY 306 Rohleder et al.
OOOO March 2014
Table III. Results of comparison of parameters measured with tissue spectroscopy and laser Doppler flowmetry between patients without or with history of cervical radiotherapy Parameter Oxygen saturation (%) At 2 mm tissue depth At 8 mm tissue depth Relative hemoglobin concentration (AU) At 2 mm tissue depth At 8 mm tissue depth Blood flow (AU) At 2 mm tissue depth At 8 mm tissue depth Blood velocity (AU) At 2 mm tissue depth At 8 mm tissue depth
P*
No RTx (median; range)
RTx (median; range)
51.7; 28.95-69.81 72.49; 19.41-87.53
54.92; 8.51-75.14 72.48; 9.76-564.94
.307 .488
67.08; 43.01-91.05 43.49; 25.80-67.71
64.46; 37.01-81.26 43.77; 28.77-70.31
.234 .505
78.33; 9.79-170.34 173.36; 45.92-365.19
71.49; 13.59-446.58 141.51; 59.65-262.95
.608 .062
17.57; 9.63-27.35 29.23; 12.70-50.02
14.77; 9.09-50.17 24.85; 13.35-76.14
.016y .156
RTx, cervical radiotherapy in medical history; AU, arbitrary units. *Calculated with Mann-Whitney U test. y Statistically significant difference at a significance level of 5%.
Patient characteristics compared between patients without or with history of cervical radiotherapy Sixty-seven patients (73.6%) had no previous radiotherapy, and 24 (26.4%) were previously irradiated in the neck region. No difference was evident between the groups with respect to age, sex, the number of patients with alcohol abuse, or the number of patients with nicotine abuse. No difference was observed with respect to cervical postoperative wound healing disorders (total, 25 [27.5%]; no radiotherapy, 15 [60% of patients with wound healing disorders; 22.4% of patients with no radiotherapy]; vs with radiotherapy, 10 [40% of patients with wound healing disorders; 41.7% of patients with radiotherapy]) (P ¼ .108) (Table I). Patient characteristics compared between patients without or with wound healing disorder after neck dissection Sixty-six patients (72.5%) had regular wound healing, and 25 (27.5%) had a cervical wound healing disorder. No difference was evident between the groups with respect to age, sex, the number of patients with alcohol abuse, or the number of patients with nicotine abuse. No significant difference was observed between the groups with respect to the rate of patients with a history of cervical radiotherapy (total, 24 [26.4%]; regular wound healing, 14 [58.3% of patients with history of cervical radiotherapy; 21.2% of patients with regular wound healing]; vs with wound healing disorder, 10 [41.7% of patients with history of cervical radiotherapy; 40% of patients with wound healing disorder]) (P ¼ .108) (Table II). Comparison of measured parameters between patients without or with history of cervical radiotherapy A significant difference was evident for the blood velocity parameter, which was decreased in patients with
previous radiotherapy at 2 mm tissue depth (without radiotherapy, 17.57 arbitrary units [AU] [range, 9.6327.35]; with history of radiotherapy, 14.77 AU [range, 9.09-50.17]; P ¼ .016) but not at 8 mm tissue depth (Figure 3, D). All other measured parameters (oxygen saturation, relative hemoglobin concentration, and blood flow) were no different between patients without and patients with history of cervical radiotherapy, either at 2 mm or at 8 mm depth (see Figure 3, A-C; Table III). Comparison of measured parameters between patients without or with wound healing disorder after neck dissection A significant difference was evident for the blood flow parameter, which was higher in patients with development of a postoperative cervical wound healing disorder at 2 mm tissue depth (regular wound healing, 70.41 AU [range, 9.79-196.59]; with wound healing disorder, 97.62 AU [range, 26.29-446.58]; P ¼ .018) but not at 8 mm tissue depth (Figure 4, C). All other measured parameters (oxygen saturation, relative hemoglobin concentration, and blood velocity) were no different between patients without and patients with development of a cervical wound healing disorder, either at 2 mm or at 8 mm depth (see Figure 4, A, B, D; Table IV). Results of ROC curve analysis and multivariate analysis For further investigation of the previously described results obtained by means of univariate analyses, measured parameters were investigated for their predictive value with respect to the development of a cervical wound healing disorder by means of an ROC curve analysis. The parameter blood flow at 2 mm tissue depth was identified as a predictive factor for development of a wound healing disorder (area under
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Rohleder et al. 307
Fig. 4. Box plot diagrams indicating results of comparison of parameters preoperatively measured with simultaneous tissue spectrophotometry and laser Doppler flowmetry in patients with regular wound healing after neck dissection and patients who developed a wound healing disorder (WHD) after neck dissection. Columns indicate parameters oxygen saturation (A), relative hemoglobin amount (B), blood flow (C), and blood velocity (D). Upper row shows parameters as measured at a 2 mm tissue depth; lower row shows parameters as measured at an 8 mm tissue depth. The lines within the boxes indicate the medians; the top edge of the boxes represents the 75th percentile, and the bottom edge represents the 25th percentile. The range is shown as a vertical line (extreme values are not shown). Statistical analysis found a significant difference between the groups with respect to the values measured for blood flow at a 2 mm tissue depth: a median of 70.41 arbitrary units (AU) was measured in patients with no WHD, whereas a median of 97.62 AU was measured in patients with postoperative WHD (P ¼ .018). The remaining parameters showed no differences between patients with or without development of WHD.
the curve, 0.661; P ¼ .018) (Figure 5, E). Neither the blood flow parameter at 8 mm tissue depth (see Figure 5, F) nor the other investigated parameters of oxygenation or perfusion exhibited an association with development of a wound healing disorder (see Figure 5, A-D, G, H). Results of ROC analysis are detailed in Table V. Multiple linear regression analyses confirmed the results obtained by univariate investigations.
DISCUSSION Positive and negative biologic effects of ionizing radiation have been discussed since initial observations of skin inflammation and hair loss in the early years of x-ray radiography. The first book about the adverse
biologic effects of radiation and protective measures was written by the dentist William H. Rollins in 1904.16 Today, radiation therapy is known to cause acute side effects (e.g., nausea, erythema, edema), long-term side effects (fibrosis, mucosal dryness resulting in xerostomia/xerophthalmia, osteoradionecrosis of the jaws, risk of secondary malignancies), and cumulative side effects, despite the process of irradiation itself being painless.17-19 These effects vary depending on the treated organ. Although radiotherapy has become an essential therapeutic option against cancer, the development of wound healing disorders, even after surgery occurring long after the radiotherapy, has a negative effect on treatment outcome and patient compliance.20 As
ORAL AND MAXILLOFACIAL SURGERY 308 Rohleder et al.
OOOO March 2014
Table IV. Results of comparison of parameters measured with tissue spectroscopy and laser Doppler flowmetry between patients without or with wound healing disorder after neck dissection Parameter Oxygen saturation (%) At 2 mm tissue depth At 8 mm tissue depth Relative hemoglobin concentration (AU) At 2 mm tissue depth At 8 mm tissue depth Blood flow (AU) At 2 mm tissue depth At 8 mm tissue depth Blood velocity (AU) At 2 mm tissue depth At 8 mm tissue depth
P*
No WHD (median; range)
WHD (median; range)
51.89; 8.51-75.14 73.08; 33.37-87.53
52.21; 23.32-68.06 70.54; 9.76-564.94
.588 .126
64.64; 43.01-91.05 40.34; 25.80-67.71
70.52; 37.01-83.28 44.54; 28.77-70.31
.174 .284
70.41; 9.79-196.59 161.19; 45.92-365.19
97.62; 26.29-446.58 157.74; 84.99-262.95
.018y .255
16.28; 9.09-27.35 27.71; 12.70-50.02
17.35; 10.37-50.17 31.41; 17.73-76.14
.139 .197
WHD, cervical wound healing disorder after neck dissection; AU, arbitrary units. *Calculated with Mann-Whitney U test. y Statistically significant difference at a significance level of 5%.
mentioned earlier, this has been reported predominantly for regions such as the abdominal wall.9 For the head and neck region, the literature is less extensive with respect to the problem of radiogenic wound healing disorders. Sufficient oxygen supply at the cellular level, which is delivered by the microvascular blood circulation, is generally accepted as being a prerequisite for regular wound healing. Schmitt et al.21 observed decreased tissue perfusion after radiotherapy in 4 of 5 head and neck carcinoma patients with functional magnetic resonance imaging. In this study, cervical oxygenation and perfusion after radiotherapy were measured, for the first time, with noninvasive, real-time, simultaneous tissue spectrophotometry and the laser Doppler flowmetry method, allowing for a comparison of the relevant microvascular blood circuit system in various tissue layers. Tissue depths of 2 and 8 mm were chosen to allow for a comparison between the superficial cutaneous and the subcutaneous layer in cervical skin. The results indicate a less significant effect of previous radiotherapy on the parameters oxygen saturation, relative hemoglobin concentration, and blood flow, which showed no difference at depths of 2 or 8 mm between patients with and without irradiation. The blood velocity parameter was significantly decreased at 2 mm depth of cervical skin in patients with a history of local radiation. However, the results indicate a moderate influence on tissue oxygenation and perfusion, and the authors are of the opinion that this indicates a minor effect of cervical radiotherapy on postoperative wound healing. Moreover, the statistical comparison between patients with regular wound healing and patients with postoperative wound healing disorders found no correlation with a history of local radiotherapy (see Table II). By contrast, the blood flow parameter as obtained by laser Doppler flowmetry has been identified as a
predictive value for the development of a cervical wound healing disorder after SND in univariate and multivariate analyses (see Tables IV and V; see Figure 4, C, and Figure 5, E), being significantly higher in patients with a postoperative dehiscence of the ND wound. The parameters oxygen saturation, relative hemoglobin concentration, and blood velocity showed no difference between patients with or without a wound healing disorder. These results are initially surprising, given that a low oxygen saturation and blood supply are assumed to indicate a risk for the development of a dehiscence, with strong perfusion being thought to promote primary wound healing. However, a study of the literature found that observations indicating insufficient tissue oxygenation in chronic wounds have predominantly been obtained from wounds other than in the head and neck region, such as from patients with venous ulcers of the lower extremity, and that oxygenation and perfusion have rarely been investigated in cervical skin wounds.22 What might be the pathophysiologic origins of wound healing disorders after SND in patients with a high blood flow value? We can speculate that the transection of vessels during the ND incision leads to a decrease in local tissue perfusion. Tissue with preoperatively high perfusion might be more susceptible to a sudden deficiency of blood circulation than tissue with a preoperatively already lower blood supply and therefore might be more predisposed to development of a dehiscence. An additional possibility might be an increased tendency for small secondary bleeding within the wound, which might lead to subcutaneous/cutaneous hematomas within the wound and might consecutively induce an increased rate of infection, thereby causing dehiscences. However, the results are surprising, and further research has to be conducted to confirm the findings and to investigate their physiologic causes and implications.
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Rohleder et al. 309
Fig. 5. A to H, Receiver operating characteristic (ROC) curve analyses of parameters with respect to their predictive value regarding the development of a wound healing disorder. ROC curves are drawn by plotting the sensitivity (true-positives out of the positives) on the axis of ordinates and 1 minus the specificity (false-positives out of the negatives) at various threshold settings. An ROC curve near to the diagonal line indicates a coincidental result, given that values on the diagonal line result from an equal rate of true-positive and 1 minus false-positive values. Accordingly, a perfect ROC curve would initially rise in a vertical manner (this would mean 100% true-positive values with 0% false values). Areas under the curves (AUCs) are one measure for the quality of a parameter investigated in an ROC curve. AUCs were calculated, and the parameter of blood flow at 2 mm tissue depth was identified as a predictive value for the development of a postoperative wound healing disorder (P ¼ .018).
No threshold value could be postulated on the basis of the current results, given that the measured values showed a broad range in both groups, but such a value could be helpful for prognosis and a surgical approach. A subsequent study with a larger number of cases
should therefore be conducted to specifically investigate and further characterize the prognostic properties of the blood flow parameter. As a temporary guide, the authors suggest the value of 100 AU as a critical threshold (it is approximately the median value of 97.62 AU
ORAL AND MAXILLOFACIAL SURGERY 310 Rohleder et al.
OOOO March 2014
Table V. Results of receiver operating characteristic curve analysis of parameters measured with tissue spectroscopy and laser Doppler flowmetry between patients without or with wound healing disorder after neck dissection Parameter Oxygen saturation At 2 mm tissue depth At 8 mm tissue depth Relative hemoglobin concentration At 2 mm tissue depth At 8 mm tissue depth Blood flow At 2 mm tissue depth At 8 mm tissue depth Blood velocity At 2 mm tissue depth At 8 mm tissue depth
Area under curve
Mean error
Confidence interval low
Confidence interval high
P*
0.463 0.396
0.066 0.067
0.334 0.264
0.592 0.527
.588 .126
0.593 0.573
0.070 0.068
0.455 0.440
0.730 0.706
.174 .284
0.661 0.578
0.063 0.065
0.536 0.450
0.785 0.705
.018y .255
0.601 0.588
0.067 0.064
0.470 0.463
0.731 0.713
.139 .197
*Calculated with Mann-Whitney U test. y Statistically significant difference at a significance level of 5%.
in the group of patients with wound healing disorders and exceeds the 75th percentile of the values of patients with regular wound healing; see Figure 4, C). In cases with values above this threshold, the patients should be informed about the elevated risk for a dehiscence. Special caution should be considered for closure of the ND incision (e.g., careful bipolar sealing of small bleeding vessels, avoidance of continuous suturing, and advising the patient not to recline the head). This might lower the rate of wound healing disorders after SND and would also offset any expense during the surgical procedure (need for more suturing material when applying interrupted sutures and slightly prolonged surgical procedure time). Postoperative time on the ward would be reduced, and nursing, drugs, and dressing changes would be reduced in cases of primary wound healing. Another aspect to be addressed as a consequence of the results might be the application of hyperbaric oxygen therapy (HBOT). Positive effects of HBOT when used for the treatment of chronic cutaneous wounds in irradiated skin have been suggested.23 HBOT application is partly based on the assumption of a reduced oxygen supply in irradiated skin. This study found that no difference in oxygen saturation or microperfusion is evident in irradiated cervical skin. An investigation of these parameters postoperatively in irradiated patients with the development of a dehiscence would be of interest, as would a comparison of pre- and postoperative values. A potential drawback of this study might be the method of tissue spectrophotometry and laser Doppler flowmetry itself: this combination has been employed in recent head and neck surgery studies,24 but a point to consider when using this method is the influence of tissue conditions on light transmissibility. Although little is known about the significance of this issue with respect to tissue spectrophotometry and laser Doppler flowmetry, irradiated skin might be less transmissible
for the sensor light because of fibrosis, which might lead to incomparable results with respect to nonirradiated skin.25 A positive aspect of this method is its easy applicability: it is portable (and therefore appropriate for bedside use) and needs only a short amount of time to be completed. Patient compliance is good, because the procedure is noninvasive and the application of the sensor is hardly perceptible. As reported earlier, the validity of the laser Doppler flowmetry method has been investigated in more than a thousand publications and has been found to offer accurate, reproducible, and reliable results.24 Its combination with simultaneous measurement of oxygen tension by tissue spectrophotometry prevents the aforementioned disadvantages associated with invasive methods such as transcutaneous microelectrodes.24 It has been stated that “simultaneous noninvasive laser Doppler flowmetry and tissue spectrophotometry gives precise information of capillary perfusion and tissue oxygen saturation at the same time.”24 The authors conclude that radiotherapy alters the microvascular blood velocity in cervical skin at 2 mm tissue depth (confirmation of null hypothesis I) but that radiotherapy seems not to be a risk factor for cervical wound healing disorders after SND. Blood flow has been identified as a predictive value for postoperative wound healing disorders, independent of previous radiotherapy or none (confirmation of null hypothesis II). This might assist the surgeon to identify patients with a risk for a dehiscence after SND and to apply techniques for its prevention. REFERENCES 1. Ferlito A, Johnson JT, Rinaldo A, et al. European surgeons were the first to perform neck dissection. Laryngoscope. 2007;117: 797-802. 2. Robbins KT, Medina JE, Wolfe GT, Levine PA, Sessions RB, Pruet CW. Standardizing neck dissection terminology: official
OOOO Volume 117, Number 3
3. 4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
report of the Academy’s Committee for Head and Neck Surgery and Oncology. Arch Otolaryngol Head Neck Surg. 1991;117:601-605. Muzaffar K. Therapeutic selective neck dissection: a 25-year review. Laryngoscope. 2003;113:1460-1465. O’Brien CJ, Petersen-Schaefer K, Ruark D, Coates AS, Menzie SJ, Harrison RI. Radical, modified, and selective neck dissection for cutaneous malignant melanoma. Head Neck. 1995;17:232-241. Wang SJ, Patel SG, Shah JP, et al. An oral cavity carcinoma nomogram to predict benefit of adjuvant radiotherapy. JAMA Otolaryngol Head Neck Surg. 2013;139:554-559. Goto Y, Kodaira T, Furutani K, et al. Clinical outcome and patterns of recurrence of head and neck squamous cell carcinoma with a limited field of postoperative radiotherapy. Jpn J Clin Oncol. 2013;43:719-725. Tribius S, Sommer J, Prosch C, et al. Xerostomia after radiotherapy. What mattersdmean total dose or dose to each parotid gland? Strahlenther Onkol. 2013;189:216-222. Mueller CK, Thorwarth M, Schultze-Mosgau S. Late changes in cutaneous gene expression patterns after adjuvant treatment of oral squamous cell carcinoma (OSCC) by radiation therapy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2010;109: 694-699. Wang J, Boerma M, Fu Q, Hauer-Jensen M. Radiation responses in skin and connective tissues: effect on wound healing and surgical outcome. Hernia. 2006;10:502-506. Pellini R, Mercante G, Marchese C, et al. Predictive factors for postoperative wound complications after neck dissection. Acta Otorhinolaryngol Ital. 2013;33:16-22. Hopewell JW, Calvo W, Jaenke R, Reinhold HS, Robbins ME, Whitehouse EM. Microvasculature and radiation damage. Recent Results Cancer Res. 1993;130:1-16. Falanga V, Zhou L, Yufit T. Low oxygen tension stimulates collagen synthesis and COL1A1 transcription through the action of TGF-beta1. J Cell Physiol. 2002;191:42-50. Westbury CB, Pearson A, Nerurkar A, et al. Hypoxia can be detected in irradiated normal human tissue: a study using the hypoxic marker pimonidazole hydrochloride. Br J Radiol. 2007;80:934-938. Thorn JJ, Kallehave F, Westergaard P, Hansen EH, Gottrup F. The effect of hyperbaric oxygen on irradiated oral tissues: transmucosal oxygen tension measurements. J Oral Maxillofac Surg. 1997;55:1103-1107. Rudolph R, Tripuraneni P, Koziol JA, McKean-Matthews M, Frutos A. Normal transcutaneous oxygen pressure in skin after radiation therapy for cancer. Cancer. 1994;74:3063-3070.
ORIGINAL ARTICLE Rohleder et al. 311 16. Kathren RL. William H. Rollins (1852-1929): x-ray protection pioneer. J Hist Med Allied Sci. 1964;19:287-294. 17. Lara PC, Russell NS, Smolders IJ, Bartelink H, Begg AC, CocoMartin JM. Radiation-induced differentiation of human skin fibroblasts: relationship with cell survival and collagen production. Int J Radiat Biol. 1996;70:683-692. 18. Li CY, Chen XH, Tao XA, Xia J, Cheng B. The development and inflammatory features of radiotherapy-induced glossitis in rats. Med Oral Patol Oral Cir Bucal. 2011;16:e348-e353. 19. Thorn JJ, Hansen HS, Specht L, Bastholt L. Osteoradionecrosis of the jaws: clinical characteristics and relation to the field of irradiation. J Oral Maxillofac Surg. 2000;58:1088-1093:discussion 1093-1095. 20. Dormand EL, Banwell PE, Goodacre TE. Radiotherapy and wound healing. Int Wound J. 2005;2:112-127. 21. Schmitt P, Kotas M, Tobermann A, Haase A, Flentje M. Quantitative tissue perfusion measurements in head and neck carcinoma patients before and during radiation therapy with a non-invasive MR imaging spin-labeling technique. Radiother Oncol. 2003;67:27-34. 22. Wipke-Tevis DD, Stotts NA, Williams DA, Froelicher ES, Hunt TK. Tissue oxygenation, perfusion, and position in patients with venous leg ulcers. Nurs Res. 2001;50:24-32. 23. Olascoaga A, Vilar-Compte D, Poitevin-Chacon A, ContrerasRuiz J. Wound healing in radiated skin: pathophysiology and treatment options. Int Wound J. 2008;5:246-257. 24. Holzle F, Loeffelbein DJ, Nolte D, Wolff KD. Free flap monitoring using simultaneous non-invasive laser Doppler flowmetry and tissue spectrophotometry. J Craniomaxillofac Surg. 2006;34: 25-33. 25. Rajan V, Varghese B, Van Leeuwen TG, Steenbergen W. Influence of tissue optical properties on laser Doppler perfusion imaging, accounting for photon penetration depth and the laser speckle phenomenon. J Biomed Opt. 2008;13:024001.
Reprint requests: Marco R. Kesting, MD, DDS, PhD Department of Oral and Maxillofacial Surgery Technische UniversitätMünchen Ismaninger Str. 22 81675 Munich Germany [email protected]; [email protected]
Can tissue spectrophotometry and laser Doppler flowmetry help to identify patients at risk for wound healing disorders after neck dissection? Nils H. Rohleder, MD, DDS,a Sandra Flensberg, MD,b Florian Bauer, MD, DDS,a Stefan Wagenpfeil, MSc, PhD,c Craig J. Wales, MD, DDS,d Steffen Koerdt, MD,a Klaus D. Wolff, MD, DDS, PhD,a Frank Hölzle, MD, DDS, PhD,e Timm Steiner, MD, DDS,e and Marco R. Kesting, MD, DDS, PhDa Technische Universität München, Munich, Germany; Saarland University, Homburg, Germany; Southern General Hospital, Glasgow, Scotland; RWTH Aachen University, Aachen, Germany
Objective. Microcirculation and oxygen supply in cervical skin were measured with an optical, noninvasive method in patients with or without radiotherapy before neck dissection. The course of wound healing was monitored after the surgical procedure to identify predictive factors for postoperative wound healing disorders. Study Design. Tissue spectrophotometry and laser Doppler flowmetry were used to determine capillary oxygen saturation, hemoglobin concentration, blood flow, and blood velocity at 2-mm and 8-mm depths in the cervical skin of 91 patients before neck dissection in a maxillofacial unit of a university hospital in Munich, Germany. Parameters were evaluated for differences between patients with irradiation (24) and without (67) and patients with wound healing disorders (25) and without (66) (univariate or multivariate statistical analyses). Results. Velocity at 2 mm was lower in irradiated skin (P ¼ .016). Flow at 2 mm was higher in patients with wound healing disorders (P ¼ .018). Conclusions. High flow values could help to identify patients at risk for cervical wound healing disorders. (Oral Surg Oral Med Oral Pathol Oral Radiol 2014;117:302-311)
Since the first description of a neck dissection (ND) by Jawdynski in 1888 and refinements by Crile in the early 20th century, cervical lymphadenectomy has become the predominant surgical treatment option for cervical lymph node metastases.1 Radical ND (RND) procedures, which comprise removal of the sternocleidomastoid muscle with the cranial accessory nerve, internal jugular vein, and carotid artery together with the lymphatic tissue, have increasingly been replaced with less radical selective neck dissection (SND).2 Thus, postoperative morbidity has been reduced.3 However, RND is more effective than SND in controlling particular tumors such as malignant melanoma.4 After surgical resection of tumors such as oral squamous cell carcinoma (OSCC), radiotherapy decreases the risk of local recurrence.5 Adjuvant radiotherapy after Coauthors Rohleder, Flensberg, and Bauer contributed equally to this work as first authors. a Department of Oral and Maxillofacial Surgery, Technische Universität München. b Department of Anaesthesiology, Technische Universität München. c Institute for Medical Biometry, Epidemiology and Medical Informatics, Saarland University. d Department of Oral and Maxillofacial Surgery, Southern General Hospital. e Department of Oral and Maxillofacial Surgery, RWTH Aachen University. Received for publication Jul 31, 2013; returned for revision Sep 22, 2013; accepted for publication Nov 25, 2013. Ó 2014 Elsevier Inc. All rights reserved. 2212-4403/$ - see front matter http://dx.doi.org/10.1016/j.oooo.2013.11.497
302
OSCC resection usually consists of up to 66 Gy, with attempts to lower overall doses while maintaining outcomes.6,7 Radiotherapy to the head and neck has been found to influence cutaneous gene expression of extracellular matrix factors and bears the risk of acute and long-term side effects such as xerostomia.7,8 In regions such as the abdominal wall, it has been found to increase perioperative complications in terms of negatively affecting the course and results of wound healing.9 In the head and neck, the association of radiotherapy with the development of postoperative wound healing disorders is not completely understood. Patients requiring surgery in previously irradiated tissue (for example, because of local recurrence or late lymph node metastases) are assumed to be at greater risk for developing delayed wound healing; however, sparse literature is available to confirm this idea. Recently, Pellini et al.10 investigated risk factors for wound healing disorders after neck dissection. The authors identified preoperative chemoradiotherapy and the mode of ND as predictive factors
Statement of Clinical Relevance The “blood flow” parameter (measured with laser Doppler flowmetry) is a predictive value for cervical postoperative wound healing disorders, independent of previous radiotherapy or none. This helps to identify patients with a risk for a dehiscence after selective neck dissection.
OOOO Volume 117, Number 3
Fig. 1. Cervical wound healing disorder developed after selective neck dissection in a female patient.
for wound complications (RND and modified RND increased the risk in opposition to SND). No clinical tool is available to determine the individual risk for the development of wound healing disorders in patients requiring ND. Radiation induces vascular endothelial injuries that subsequently lead to tissue ischemia and atrophy.11 Development of fibrosis in irradiated tissue might be attributable to hypoxia, which induces profibrotic cytokines and collagen formation in laboratory experiments and is considered to induce angiogenesis.12,13 Oxygen tension in irradiated skin has been found to be decreased in animal models, but no clear result has been obtained with regard to whether hypoxia is the predominant factor for the induction of angiogenic cytokines.13 The measurement of oxygenation in irradiated skin in humans has been carried out by means of transcutaneous oxygen microelectrodes and has shown a reduction in oxygen tension in patients with previous radiotherapy. However, the results are again inconsistent.13-15 Another method for the detection of hypoxia in irradiated human tissue is the immunohistochemical staining of the hypoxia marker carbonic anhydrase IX.13 However, these methods are not applicable in daily clinical routine. This prospective study investigates, for the first time, the use of noninvasive simultaneous tissue spectrophotometry and laser Doppler flowmetry for the measurement of tissue oxygenation and microperfusion parameters in the cervical skin of patients with or without previous local radiotherapy. Measurements were conducted at the bedside on the day before SND. Postoperatively, patients were monitored for the development of wound healing disorders of the ND incision (Figure 1). Assessment of the variable “wound healing disorder” was dichotomized (no, complete primary wound healing; yes, appearance of any minor or
ORIGINAL ARTICLE Rohleder et al. 303
Fig. 2. Device used for simultaneous tissue spectrophotometry and laser Doppler flowmetry (O2C; Lea Medizintechnik GmbH, Giessen, Germany). Left, Portable device with live monitoring and recording software for parameters oxygen saturation, relative hemoglobin concentration, blood flow, and blood velocity. Right, Detailed photography of the laser sensor, which is attached to the main device. When applied to the region of interest on the skin, the aforementioned parameters are simultaneously measured at both 2 mm and 8 mm tissue depths.
major dehiscence). Data were analyzed to test the following assumptions: “History of cervical radiotherapy alters local tissue oxygenation/perfusion” (null hypothesis I) and “Simultaneous tissue spectrophotometry and laser Doppler flowmetry can predict wound healing disorders after neck dissection” (null hypothesis II). These null hypotheses were tested to achieve the 2 objectives of the study: to investigate the influence of cervical radiotherapy on tissue oxygenation/perfusion and wound healing after neck dissection (objective I) and to identify a predictive marker for the development of postoperative wound healing disorders after neck dissection (objective II). For this reason, the study design included preoperative measurement of the tissue oxygenation/perfusion parameters and also postoperative monitoring of the course of wound healing, and an analysis of the medical history of the patients with respect to cervical radiotherapy. A potential predictive marker for development of a wound healing disorder would then be evaluated by statistical correlation of preand postoperative data.
PATIENTS AND METHODS Patients A total of 91 patients who had undergone surgical procedures for OSCC or other oral malignancies in the Department of Oral and Maxillofacial Surgery, Technische Universität München, Munich, Germany, between January 1, 2012, and December 31, 2012, were included. All patients were treated by resection of the cervical lymph nodes, tumor resection, and reconstruction by using local or free flaps. Clinical data were obtained before surgery (e.g., history of cervical radiotherapy in patients with secondary surgical procedures, alcohol/nicotine abuse) and during the hospital
ORAL AND MAXILLOFACIAL SURGERY 304 Rohleder et al.
OOOO March 2014
Table I. Patient characteristics compared between patients without or with history of cervical radiotherapy
Characteristic Age (y), mean; median; range* Sex, n (%)y Male Female Alcohol abuse, n (%)y Nicotine abuse, n (%)y Postoperative cervical wound healing disorder, n (%)y
All patients, No radiotherapy, Radiotherapy, n (%) n (%) n (%) 91 (100) 67 (73.6) 24 (26.4) 59.9; 60.0; 20-86
60.6; 61; 20-86
57.8; 58.5; 29-86
61 (67.0) 30 (33.0) 41 (45.1)
46 (75.4) 21 (70.0) 32 (78.0)
15 (24.6) 9 (30.0) 9 (22.0)
54 (59.3)
40 (74.1)
14 (25.9)
25 (27.5)
15 (60.0)
10 (40.0)
Table II. Patient characteristics compared between patients without or with wound healing disorder after neck dissection
Characteristic Age (y), mean; median; range* Sex, n (%)y Male Female Alcohol abuse, n (%)y Nicotine abuse, n (%)y History of cervical radiotherapy, n (%)y
All patients, Regular wound Wound healing n (%) healing, n (%) disorder, n (%) 91 (100) 66 (72.5) 25 (27.5) 59.9; 60.0; 20-86
59.4; 60; 20-86 60.9; 60; 43-86
61 (67.0) 30 (33.0) 41 (45.1)
41 (67.2) 25 (83.3) 27 (65.9)
20 (32.8) 5 (16.7) 14 (34.1)
54 (59.3)
36 (66.7)
18 (33.3)
24 (26.4)
14 (58.3)
10 (41.7)
Percentages in column “All patients” refer to total patient collective. Percentages in the subsequent columns refer to the number of cases in each category. No statistically significant difference was evident between the groups, as calculated with Mann-Whitney U test or Fisher exact test, as marked. *Mann-Whitney U test. y Fisher exact test.
Percentages in column “All patients” refer to total patient collective. Percentages in the subsequent columns refer to the number of cases in each category. No statistically significant difference was evident between the groups, as calculated with Mann-Whitney U test or Fisher exact test, as marked. *Mann-Whitney U test. y Fisher exact test.
stay (monitoring for development of cervical wound healing disorders). The medical ethics committee of the Technische Universität München approved the research reported in this article; this research also complies with the ethical rules for human experimentation stated in the Declaration of Helsinki (institutional review board registration No. 263309). All patients gave written informed consent.
measurement of the oxygen saturation and relative hemoglobin concentration parameters is performed with white light. The degree of oxygen saturation of hemoglobin is calculated by measurement of the color change of the blood (venous blood, blue; oxygenated blood, bright red). The relative hemoglobin concentration parameter is determined from absorption measurements within the illuminated tissue volume.
Measurement of parameters indicating oxygen saturation and microperfusion In a noninvasive procedure, the oxygen saturation of hemoglobin, the relative hemoglobin concentration, the blood flow (vessel filling), and the blood flow velocity in microvessels of cervical skin were determined by means of a mobile device (O2C; Lea, Giessen, Germany; Figure 2, left) connected to a glass fiber probe (see Figure 2, right). This system allows the monitoring of the oxygen supply in various layers of tissue. In the present study, the aforementioned parameters were measured at depths of 2 mm and 8 mm. The superficial parameters were recorded by one channel, whereas those of deeper tissues, predominantly the cervical subcutaneous fatty tissue, were recorded by a second channel. The measurement principle of the device makes use of both a laser Doppler technique and tissue spectrophotometry (using white light). Laser light is used to determine the blood flow and the blood flow velocity parameters by measuring the Doppler shift of the erythrocytes caused by their movement within the vessels. Simultaneous
Surgical procedure All SNDs were carried out under standardized conditions by 2 operators from a team of 9 surgeons. Skin incision was carried out 2 cm below the lower margin of the mandible with a scalpel and was extended from the mental region to the sternocleidomastoid muscle, displaying the external jugular vein and the great auricular nerve. In the case of bilateral ND, the incision was extended to the opposite sternocleidomastoid muscle. Subplatysmal flaps were raised, and levels I to III were dissected as a standard for staging ND. For positive nodal disease or a bilateral tumor, levels I to V were removed on the more affected side, and levels I to III were removed on the contralateral side. Closure was carried out after the placement of a Redon drain on the posterior end of the wound (2 in case of bilateral ND) and subcutaneous suturing (including platysma) with single-knot technique (Vicryl 3-0 resorbable suturing material; Ethicon, Norderstedt, Germany). Skin closure was carried out by using continuous suturing (Ethilon 4-0 nylon suture; Ethicon).
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Rohleder et al. 305
Fig. 3. Box plot diagrams indicating results of comparison of parameters measured with simultaneous tissue spectrophotometry and laser Doppler flowmetry in cervical skin of patients who had never previously received radiotherapy (RTx) and patients that were previously irradiated at the neck. Columns indicate parameters oxygen saturation (A), relative hemoglobin amount (B), blood flow (C), and blood velocity (D). Upper row shows parameters as measured at a 2 mm tissue depth; lower row shows parameters as measured at an 8 mm tissue depth. The lines within the boxes indicate the medians; the top edge of the boxes represents the 75th percentile, and the bottom edge represents the 25th percentile. The range is shown as a vertical line (extreme values are not shown). Statistical analysis found a significant difference between the values measured for blood velocity at a 2 mm tissue depth: a median of 17.57 arbitrary units (AU) was measured in patients with no RTx, whereas a median of 14.77 AU was measured in patients with previous cervical RTx (P ¼ .016). None of the remaining parameters was affected by RTx in medical history.
Statistical analysis IBM SPSS version 21.0 (IBM, Armonk, NY, USA) was used. For univariate analysis, differences with respect to age and the measured parameters were analyzed by using the Mann-Whitney U test. Differences with respect to sex, nicotine abuse, and alcohol abuse were analyzed by using the Fisher exact test. For multivariate analysis, a linear regression analysis (with the independent variables age, sex, alcohol abuse, nicotine abuse, wound healing disorder (yes/no), history of radiotherapy (yes/ no), and interaction between wound healing disorder and history of radiotherapy) was performed separately for each of the outcome variables oxygen saturation, relative hemoglobin concentration, blood flow, and blood velocity (each at 2-mm and 8-mm depth) as a multiple approach. Stepwise forward and backward variable
selection approaches were applied to account for potential multicollinearity. Receiver operating characteristic (ROC) analysis was carried out, and ROC curves were created to show the diagnostic value of respective outcome measures. All P values are given as 2-sided and are subject to a local significance level of 5%.
RESULTS For the 91 patients included, patient characteristics and measured oxygenation/microperfusion parameters were compared (1) between the group without any previous cervical radiotherapy and the group with a history of cervical radiotherapy and (2) between the group with regular wound healing and the group with postoperative wound healing disorders after SND.
ORAL AND MAXILLOFACIAL SURGERY 306 Rohleder et al.
OOOO March 2014
Table III. Results of comparison of parameters measured with tissue spectroscopy and laser Doppler flowmetry between patients without or with history of cervical radiotherapy Parameter Oxygen saturation (%) At 2 mm tissue depth At 8 mm tissue depth Relative hemoglobin concentration (AU) At 2 mm tissue depth At 8 mm tissue depth Blood flow (AU) At 2 mm tissue depth At 8 mm tissue depth Blood velocity (AU) At 2 mm tissue depth At 8 mm tissue depth
P*
No RTx (median; range)
RTx (median; range)
51.7; 28.95-69.81 72.49; 19.41-87.53
54.92; 8.51-75.14 72.48; 9.76-564.94
.307 .488
67.08; 43.01-91.05 43.49; 25.80-67.71
64.46; 37.01-81.26 43.77; 28.77-70.31
.234 .505
78.33; 9.79-170.34 173.36; 45.92-365.19
71.49; 13.59-446.58 141.51; 59.65-262.95
.608 .062
17.57; 9.63-27.35 29.23; 12.70-50.02
14.77; 9.09-50.17 24.85; 13.35-76.14
.016y .156
RTx, cervical radiotherapy in medical history; AU, arbitrary units. *Calculated with Mann-Whitney U test. y Statistically significant difference at a significance level of 5%.
Patient characteristics compared between patients without or with history of cervical radiotherapy Sixty-seven patients (73.6%) had no previous radiotherapy, and 24 (26.4%) were previously irradiated in the neck region. No difference was evident between the groups with respect to age, sex, the number of patients with alcohol abuse, or the number of patients with nicotine abuse. No difference was observed with respect to cervical postoperative wound healing disorders (total, 25 [27.5%]; no radiotherapy, 15 [60% of patients with wound healing disorders; 22.4% of patients with no radiotherapy]; vs with radiotherapy, 10 [40% of patients with wound healing disorders; 41.7% of patients with radiotherapy]) (P ¼ .108) (Table I). Patient characteristics compared between patients without or with wound healing disorder after neck dissection Sixty-six patients (72.5%) had regular wound healing, and 25 (27.5%) had a cervical wound healing disorder. No difference was evident between the groups with respect to age, sex, the number of patients with alcohol abuse, or the number of patients with nicotine abuse. No significant difference was observed between the groups with respect to the rate of patients with a history of cervical radiotherapy (total, 24 [26.4%]; regular wound healing, 14 [58.3% of patients with history of cervical radiotherapy; 21.2% of patients with regular wound healing]; vs with wound healing disorder, 10 [41.7% of patients with history of cervical radiotherapy; 40% of patients with wound healing disorder]) (P ¼ .108) (Table II). Comparison of measured parameters between patients without or with history of cervical radiotherapy A significant difference was evident for the blood velocity parameter, which was decreased in patients with
previous radiotherapy at 2 mm tissue depth (without radiotherapy, 17.57 arbitrary units [AU] [range, 9.6327.35]; with history of radiotherapy, 14.77 AU [range, 9.09-50.17]; P ¼ .016) but not at 8 mm tissue depth (Figure 3, D). All other measured parameters (oxygen saturation, relative hemoglobin concentration, and blood flow) were no different between patients without and patients with history of cervical radiotherapy, either at 2 mm or at 8 mm depth (see Figure 3, A-C; Table III). Comparison of measured parameters between patients without or with wound healing disorder after neck dissection A significant difference was evident for the blood flow parameter, which was higher in patients with development of a postoperative cervical wound healing disorder at 2 mm tissue depth (regular wound healing, 70.41 AU [range, 9.79-196.59]; with wound healing disorder, 97.62 AU [range, 26.29-446.58]; P ¼ .018) but not at 8 mm tissue depth (Figure 4, C). All other measured parameters (oxygen saturation, relative hemoglobin concentration, and blood velocity) were no different between patients without and patients with development of a cervical wound healing disorder, either at 2 mm or at 8 mm depth (see Figure 4, A, B, D; Table IV). Results of ROC curve analysis and multivariate analysis For further investigation of the previously described results obtained by means of univariate analyses, measured parameters were investigated for their predictive value with respect to the development of a cervical wound healing disorder by means of an ROC curve analysis. The parameter blood flow at 2 mm tissue depth was identified as a predictive factor for development of a wound healing disorder (area under
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Rohleder et al. 307
Fig. 4. Box plot diagrams indicating results of comparison of parameters preoperatively measured with simultaneous tissue spectrophotometry and laser Doppler flowmetry in patients with regular wound healing after neck dissection and patients who developed a wound healing disorder (WHD) after neck dissection. Columns indicate parameters oxygen saturation (A), relative hemoglobin amount (B), blood flow (C), and blood velocity (D). Upper row shows parameters as measured at a 2 mm tissue depth; lower row shows parameters as measured at an 8 mm tissue depth. The lines within the boxes indicate the medians; the top edge of the boxes represents the 75th percentile, and the bottom edge represents the 25th percentile. The range is shown as a vertical line (extreme values are not shown). Statistical analysis found a significant difference between the groups with respect to the values measured for blood flow at a 2 mm tissue depth: a median of 70.41 arbitrary units (AU) was measured in patients with no WHD, whereas a median of 97.62 AU was measured in patients with postoperative WHD (P ¼ .018). The remaining parameters showed no differences between patients with or without development of WHD.
the curve, 0.661; P ¼ .018) (Figure 5, E). Neither the blood flow parameter at 8 mm tissue depth (see Figure 5, F) nor the other investigated parameters of oxygenation or perfusion exhibited an association with development of a wound healing disorder (see Figure 5, A-D, G, H). Results of ROC analysis are detailed in Table V. Multiple linear regression analyses confirmed the results obtained by univariate investigations.
DISCUSSION Positive and negative biologic effects of ionizing radiation have been discussed since initial observations of skin inflammation and hair loss in the early years of x-ray radiography. The first book about the adverse
biologic effects of radiation and protective measures was written by the dentist William H. Rollins in 1904.16 Today, radiation therapy is known to cause acute side effects (e.g., nausea, erythema, edema), long-term side effects (fibrosis, mucosal dryness resulting in xerostomia/xerophthalmia, osteoradionecrosis of the jaws, risk of secondary malignancies), and cumulative side effects, despite the process of irradiation itself being painless.17-19 These effects vary depending on the treated organ. Although radiotherapy has become an essential therapeutic option against cancer, the development of wound healing disorders, even after surgery occurring long after the radiotherapy, has a negative effect on treatment outcome and patient compliance.20 As
ORAL AND MAXILLOFACIAL SURGERY 308 Rohleder et al.
OOOO March 2014
Table IV. Results of comparison of parameters measured with tissue spectroscopy and laser Doppler flowmetry between patients without or with wound healing disorder after neck dissection Parameter Oxygen saturation (%) At 2 mm tissue depth At 8 mm tissue depth Relative hemoglobin concentration (AU) At 2 mm tissue depth At 8 mm tissue depth Blood flow (AU) At 2 mm tissue depth At 8 mm tissue depth Blood velocity (AU) At 2 mm tissue depth At 8 mm tissue depth
P*
No WHD (median; range)
WHD (median; range)
51.89; 8.51-75.14 73.08; 33.37-87.53
52.21; 23.32-68.06 70.54; 9.76-564.94
.588 .126
64.64; 43.01-91.05 40.34; 25.80-67.71
70.52; 37.01-83.28 44.54; 28.77-70.31
.174 .284
70.41; 9.79-196.59 161.19; 45.92-365.19
97.62; 26.29-446.58 157.74; 84.99-262.95
.018y .255
16.28; 9.09-27.35 27.71; 12.70-50.02
17.35; 10.37-50.17 31.41; 17.73-76.14
.139 .197
WHD, cervical wound healing disorder after neck dissection; AU, arbitrary units. *Calculated with Mann-Whitney U test. y Statistically significant difference at a significance level of 5%.
mentioned earlier, this has been reported predominantly for regions such as the abdominal wall.9 For the head and neck region, the literature is less extensive with respect to the problem of radiogenic wound healing disorders. Sufficient oxygen supply at the cellular level, which is delivered by the microvascular blood circulation, is generally accepted as being a prerequisite for regular wound healing. Schmitt et al.21 observed decreased tissue perfusion after radiotherapy in 4 of 5 head and neck carcinoma patients with functional magnetic resonance imaging. In this study, cervical oxygenation and perfusion after radiotherapy were measured, for the first time, with noninvasive, real-time, simultaneous tissue spectrophotometry and the laser Doppler flowmetry method, allowing for a comparison of the relevant microvascular blood circuit system in various tissue layers. Tissue depths of 2 and 8 mm were chosen to allow for a comparison between the superficial cutaneous and the subcutaneous layer in cervical skin. The results indicate a less significant effect of previous radiotherapy on the parameters oxygen saturation, relative hemoglobin concentration, and blood flow, which showed no difference at depths of 2 or 8 mm between patients with and without irradiation. The blood velocity parameter was significantly decreased at 2 mm depth of cervical skin in patients with a history of local radiation. However, the results indicate a moderate influence on tissue oxygenation and perfusion, and the authors are of the opinion that this indicates a minor effect of cervical radiotherapy on postoperative wound healing. Moreover, the statistical comparison between patients with regular wound healing and patients with postoperative wound healing disorders found no correlation with a history of local radiotherapy (see Table II). By contrast, the blood flow parameter as obtained by laser Doppler flowmetry has been identified as a
predictive value for the development of a cervical wound healing disorder after SND in univariate and multivariate analyses (see Tables IV and V; see Figure 4, C, and Figure 5, E), being significantly higher in patients with a postoperative dehiscence of the ND wound. The parameters oxygen saturation, relative hemoglobin concentration, and blood velocity showed no difference between patients with or without a wound healing disorder. These results are initially surprising, given that a low oxygen saturation and blood supply are assumed to indicate a risk for the development of a dehiscence, with strong perfusion being thought to promote primary wound healing. However, a study of the literature found that observations indicating insufficient tissue oxygenation in chronic wounds have predominantly been obtained from wounds other than in the head and neck region, such as from patients with venous ulcers of the lower extremity, and that oxygenation and perfusion have rarely been investigated in cervical skin wounds.22 What might be the pathophysiologic origins of wound healing disorders after SND in patients with a high blood flow value? We can speculate that the transection of vessels during the ND incision leads to a decrease in local tissue perfusion. Tissue with preoperatively high perfusion might be more susceptible to a sudden deficiency of blood circulation than tissue with a preoperatively already lower blood supply and therefore might be more predisposed to development of a dehiscence. An additional possibility might be an increased tendency for small secondary bleeding within the wound, which might lead to subcutaneous/cutaneous hematomas within the wound and might consecutively induce an increased rate of infection, thereby causing dehiscences. However, the results are surprising, and further research has to be conducted to confirm the findings and to investigate their physiologic causes and implications.
OOOO Volume 117, Number 3
ORIGINAL ARTICLE Rohleder et al. 309
Fig. 5. A to H, Receiver operating characteristic (ROC) curve analyses of parameters with respect to their predictive value regarding the development of a wound healing disorder. ROC curves are drawn by plotting the sensitivity (true-positives out of the positives) on the axis of ordinates and 1 minus the specificity (false-positives out of the negatives) at various threshold settings. An ROC curve near to the diagonal line indicates a coincidental result, given that values on the diagonal line result from an equal rate of true-positive and 1 minus false-positive values. Accordingly, a perfect ROC curve would initially rise in a vertical manner (this would mean 100% true-positive values with 0% false values). Areas under the curves (AUCs) are one measure for the quality of a parameter investigated in an ROC curve. AUCs were calculated, and the parameter of blood flow at 2 mm tissue depth was identified as a predictive value for the development of a postoperative wound healing disorder (P ¼ .018).
No threshold value could be postulated on the basis of the current results, given that the measured values showed a broad range in both groups, but such a value could be helpful for prognosis and a surgical approach. A subsequent study with a larger number of cases
should therefore be conducted to specifically investigate and further characterize the prognostic properties of the blood flow parameter. As a temporary guide, the authors suggest the value of 100 AU as a critical threshold (it is approximately the median value of 97.62 AU
ORAL AND MAXILLOFACIAL SURGERY 310 Rohleder et al.
OOOO March 2014
Table V. Results of receiver operating characteristic curve analysis of parameters measured with tissue spectroscopy and laser Doppler flowmetry between patients without or with wound healing disorder after neck dissection Parameter Oxygen saturation At 2 mm tissue depth At 8 mm tissue depth Relative hemoglobin concentration At 2 mm tissue depth At 8 mm tissue depth Blood flow At 2 mm tissue depth At 8 mm tissue depth Blood velocity At 2 mm tissue depth At 8 mm tissue depth
Area under curve
Mean error
Confidence interval low
Confidence interval high
P*
0.463 0.396
0.066 0.067
0.334 0.264
0.592 0.527
.588 .126
0.593 0.573
0.070 0.068
0.455 0.440
0.730 0.706
.174 .284
0.661 0.578
0.063 0.065
0.536 0.450
0.785 0.705
.018y .255
0.601 0.588
0.067 0.064
0.470 0.463
0.731 0.713
.139 .197
*Calculated with Mann-Whitney U test. y Statistically significant difference at a significance level of 5%.
in the group of patients with wound healing disorders and exceeds the 75th percentile of the values of patients with regular wound healing; see Figure 4, C). In cases with values above this threshold, the patients should be informed about the elevated risk for a dehiscence. Special caution should be considered for closure of the ND incision (e.g., careful bipolar sealing of small bleeding vessels, avoidance of continuous suturing, and advising the patient not to recline the head). This might lower the rate of wound healing disorders after SND and would also offset any expense during the surgical procedure (need for more suturing material when applying interrupted sutures and slightly prolonged surgical procedure time). Postoperative time on the ward would be reduced, and nursing, drugs, and dressing changes would be reduced in cases of primary wound healing. Another aspect to be addressed as a consequence of the results might be the application of hyperbaric oxygen therapy (HBOT). Positive effects of HBOT when used for the treatment of chronic cutaneous wounds in irradiated skin have been suggested.23 HBOT application is partly based on the assumption of a reduced oxygen supply in irradiated skin. This study found that no difference in oxygen saturation or microperfusion is evident in irradiated cervical skin. An investigation of these parameters postoperatively in irradiated patients with the development of a dehiscence would be of interest, as would a comparison of pre- and postoperative values. A potential drawback of this study might be the method of tissue spectrophotometry and laser Doppler flowmetry itself: this combination has been employed in recent head and neck surgery studies,24 but a point to consider when using this method is the influence of tissue conditions on light transmissibility. Although little is known about the significance of this issue with respect to tissue spectrophotometry and laser Doppler flowmetry, irradiated skin might be less transmissible
for the sensor light because of fibrosis, which might lead to incomparable results with respect to nonirradiated skin.25 A positive aspect of this method is its easy applicability: it is portable (and therefore appropriate for bedside use) and needs only a short amount of time to be completed. Patient compliance is good, because the procedure is noninvasive and the application of the sensor is hardly perceptible. As reported earlier, the validity of the laser Doppler flowmetry method has been investigated in more than a thousand publications and has been found to offer accurate, reproducible, and reliable results.24 Its combination with simultaneous measurement of oxygen tension by tissue spectrophotometry prevents the aforementioned disadvantages associated with invasive methods such as transcutaneous microelectrodes.24 It has been stated that “simultaneous noninvasive laser Doppler flowmetry and tissue spectrophotometry gives precise information of capillary perfusion and tissue oxygen saturation at the same time.”24 The authors conclude that radiotherapy alters the microvascular blood velocity in cervical skin at 2 mm tissue depth (confirmation of null hypothesis I) but that radiotherapy seems not to be a risk factor for cervical wound healing disorders after SND. Blood flow has been identified as a predictive value for postoperative wound healing disorders, independent of previous radiotherapy or none (confirmation of null hypothesis II). This might assist the surgeon to identify patients with a risk for a dehiscence after SND and to apply techniques for its prevention. REFERENCES 1. Ferlito A, Johnson JT, Rinaldo A, et al. European surgeons were the first to perform neck dissection. Laryngoscope. 2007;117: 797-802. 2. Robbins KT, Medina JE, Wolfe GT, Levine PA, Sessions RB, Pruet CW. Standardizing neck dissection terminology: official
OOOO Volume 117, Number 3
3. 4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
report of the Academy’s Committee for Head and Neck Surgery and Oncology. Arch Otolaryngol Head Neck Surg. 1991;117:601-605. Muzaffar K. Therapeutic selective neck dissection: a 25-year review. Laryngoscope. 2003;113:1460-1465. O’Brien CJ, Petersen-Schaefer K, Ruark D, Coates AS, Menzie SJ, Harrison RI. Radical, modified, and selective neck dissection for cutaneous malignant melanoma. Head Neck. 1995;17:232-241. Wang SJ, Patel SG, Shah JP, et al. An oral cavity carcinoma nomogram to predict benefit of adjuvant radiotherapy. JAMA Otolaryngol Head Neck Surg. 2013;139:554-559. Goto Y, Kodaira T, Furutani K, et al. Clinical outcome and patterns of recurrence of head and neck squamous cell carcinoma with a limited field of postoperative radiotherapy. Jpn J Clin Oncol. 2013;43:719-725. Tribius S, Sommer J, Prosch C, et al. Xerostomia after radiotherapy. What mattersdmean total dose or dose to each parotid gland? Strahlenther Onkol. 2013;189:216-222. Mueller CK, Thorwarth M, Schultze-Mosgau S. Late changes in cutaneous gene expression patterns after adjuvant treatment of oral squamous cell carcinoma (OSCC) by radiation therapy. Oral Surg Oral Med Oral Pathol Oral Radiol Endod. 2010;109: 694-699. Wang J, Boerma M, Fu Q, Hauer-Jensen M. Radiation responses in skin and connective tissues: effect on wound healing and surgical outcome. Hernia. 2006;10:502-506. Pellini R, Mercante G, Marchese C, et al. Predictive factors for postoperative wound complications after neck dissection. Acta Otorhinolaryngol Ital. 2013;33:16-22. Hopewell JW, Calvo W, Jaenke R, Reinhold HS, Robbins ME, Whitehouse EM. Microvasculature and radiation damage. Recent Results Cancer Res. 1993;130:1-16. Falanga V, Zhou L, Yufit T. Low oxygen tension stimulates collagen synthesis and COL1A1 transcription through the action of TGF-beta1. J Cell Physiol. 2002;191:42-50. Westbury CB, Pearson A, Nerurkar A, et al. Hypoxia can be detected in irradiated normal human tissue: a study using the hypoxic marker pimonidazole hydrochloride. Br J Radiol. 2007;80:934-938. Thorn JJ, Kallehave F, Westergaard P, Hansen EH, Gottrup F. The effect of hyperbaric oxygen on irradiated oral tissues: transmucosal oxygen tension measurements. J Oral Maxillofac Surg. 1997;55:1103-1107. Rudolph R, Tripuraneni P, Koziol JA, McKean-Matthews M, Frutos A. Normal transcutaneous oxygen pressure in skin after radiation therapy for cancer. Cancer. 1994;74:3063-3070.
ORIGINAL ARTICLE Rohleder et al. 311 16. Kathren RL. William H. Rollins (1852-1929): x-ray protection pioneer. J Hist Med Allied Sci. 1964;19:287-294. 17. Lara PC, Russell NS, Smolders IJ, Bartelink H, Begg AC, CocoMartin JM. Radiation-induced differentiation of human skin fibroblasts: relationship with cell survival and collagen production. Int J Radiat Biol. 1996;70:683-692. 18. Li CY, Chen XH, Tao XA, Xia J, Cheng B. The development and inflammatory features of radiotherapy-induced glossitis in rats. Med Oral Patol Oral Cir Bucal. 2011;16:e348-e353. 19. Thorn JJ, Hansen HS, Specht L, Bastholt L. Osteoradionecrosis of the jaws: clinical characteristics and relation to the field of irradiation. J Oral Maxillofac Surg. 2000;58:1088-1093:discussion 1093-1095. 20. Dormand EL, Banwell PE, Goodacre TE. Radiotherapy and wound healing. Int Wound J. 2005;2:112-127. 21. Schmitt P, Kotas M, Tobermann A, Haase A, Flentje M. Quantitative tissue perfusion measurements in head and neck carcinoma patients before and during radiation therapy with a non-invasive MR imaging spin-labeling technique. Radiother Oncol. 2003;67:27-34. 22. Wipke-Tevis DD, Stotts NA, Williams DA, Froelicher ES, Hunt TK. Tissue oxygenation, perfusion, and position in patients with venous leg ulcers. Nurs Res. 2001;50:24-32. 23. Olascoaga A, Vilar-Compte D, Poitevin-Chacon A, ContrerasRuiz J. Wound healing in radiated skin: pathophysiology and treatment options. Int Wound J. 2008;5:246-257. 24. Holzle F, Loeffelbein DJ, Nolte D, Wolff KD. Free flap monitoring using simultaneous non-invasive laser Doppler flowmetry and tissue spectrophotometry. J Craniomaxillofac Surg. 2006;34: 25-33. 25. Rajan V, Varghese B, Van Leeuwen TG, Steenbergen W. Influence of tissue optical properties on laser Doppler perfusion imaging, accounting for photon penetration depth and the laser speckle phenomenon. J Biomed Opt. 2008;13:024001.
Reprint requests: Marco R. Kesting, MD, DDS, PhD Department of Oral and Maxillofacial Surgery Technische UniversitätMünchen Ismaninger Str. 22 81675 Munich Germany [email protected]; [email protected]
