Fax: (852) 2636-4346
Quantitative measurement of post-irradiation neck fibrosis based on the young modulus
Description of a new method and clinical results
Article first published online: 23 JUL 2002
Copyright © 2002 American Cancer Society
Volume 95, Issue 3, pages 656–662, 1 August 2002
How to Cite
Leung, S.-F., Zheng, Y., Choi, C. Y. K., Mak, S. S. S., Chiu, S. K. W., Zee, B. and Mak, A. F. T. (2002), Quantitative measurement of post-irradiation neck fibrosis based on the young modulus. Cancer, 95: 656–662. doi: 10.1002/cncr.10700
- Issue published online: 23 JUL 2002
- Article first published online: 23 JUL 2002
- Manuscript Accepted: 21 FEB 2002
- Manuscript Revised: 21 JAN 2002
- Manuscript Received: 21 SEP 2001
- Hong Kong Polytechnic University, Hong Kong. Grant Number: GT112
- radiation therapy;
Postirradiation fibrosis is one of the most common late effects of radiation therapy for patients with head and neck carcinoma. An objective and quantitative method for its measurement is much desired, but the criteria currently used to score fibrosis are mostly semiquantitative and partially subjective.
The Young Modulus (YM) is a physical parameter that characterizes the deformability of material to stress. The authors measured the YM in soft tissues of the neck, at defined reference points, using an ultrasound probe and computer algorithm that quantified the indentation (deformation) on tissue due to a measured, applied force. One hundred five patients who had received previous radiation therapy to the entire neck were assessed, and the results were compared with the hand palpation scores and with a functional parameter represented by the range of neck rotation, and all results were correlated with symptoms.
The YM was obtained successfully in all patients examined. It had a significant positive correlation with the palpation score and a significant negative correlation with the range of neck rotation. The YM was significantly higher on the side of the neck that received a boost dose of radiation, although the corresponding palpation scores were similar. The results of all three measurement methods were correlated with symptoms.
Postirradiation neck fibrosis can be measured in absolute units based on the YM. The results showed a significant correlation with hand palpation scores, with restriction of neck rotation, and with symptoms. Compared with the palpation method, the YM is more quantitative, objective, focused on small subregions, and better discriminates regions subject to differential radiation dose levels. Its inclusion in the Analytic category of the Late Effects of Normal Tissues-SOMA system should be considered to facilitate comparative studies. Cancer 2002;95:656–62. © 2002 American Cancer Society.
Postirradiation fibrosis of the neck is one of the most common late effects of radiation therapy for patients with malignancies of the head and neck. The most popularly used methods for its objective documentation and scoring are based on hand palpation.1–3 The information obtained is only semiquantitative in nature and can be expressed only in ordinal rather than absolute units. Although fibrosis is categorized as one of the objective parameters in the Late Effects of Normal Tissues (LENT) subjective/objective/management/analytic (SOMA) system,1 its assessment has an inherent element of subjectivity. The Young Modulus (YM) is a physical parameter that characterizes the deformability of physical material. It is defined as stress (applied force per unit area) / strain (resulting deformation) for a material under compression.4 We applied this concept to soft tissue of the neck for quantification of postirradiation fibrosis. The measurement involves the simultaneous measurement of a small applied force and the extent of the resulting indentation. An earlier developmental study had addressed the reliability of the method in terms of intraoperator variation on repeat measurements, interoperator variation, and variation by measurement site. The following issues are addressed in the current study: 1) the feasibility of performing the measurement on a large number of patients in an outpatient setting and 2) the correlation of results with the hand palpation score and functional outcome in terms of restriction in rotation range of the neck, with subsites of the neck given different radiation doses and with symptoms.
MATERIALS AND METHODS
The study was approved by the Clinical Ethics Committee of the authors' affiliated institutions. Written informed consent was obtained from all patients at the time of recruitment.
The Measurement Device
The basic requisite for determining the YM is simultaneous measurement of an applied force and the associated deformation. The details of the measurement device had been described in an earlier publication.5 In summary, it comprises a hand-held probe bearing at its tip an ultrasound transducer measuring 9 mm in diameter, an ultrasound emitter and receiver, a force measurement device (load cell and load driver/amplifier), and a personal computer. When the probe is applied onto the skin surface of the lateral neck (Fig. 1), ultrasound signals are emitted and reflected from the bony surface of the cervical vertebrae, thus allowing measurement of the tissue thickness between the skin and the spine. The alterations in tissue thickness (deformation) as the result of an applied force are recorded simultaneously and are displayed in real time by a custom-developed computer program. The YM is calculated from the load-indentation response curve, using an analytical solution of indentation on an elastic layer, by assuming that the tissue is a near incompressible elastic material with a Poisson ratio of 0.45.6, 7 The influence of technical and operational factors (including the effects of indentation rate, reliability on repeat measurement by the same operator, interoperator variations, and site dependency) in normal volunteers and irradiated patients had been addressed in previous developmental studies by the authors.5
The Measurement Procedure
The individual assumed a sitting position, with the neck neither extended nor flexed and with the eyes looking forward. Two reference points, located 3 cm and 7 cm (Fig. 1) inferior to the lowermost point of the mastoid on each side of the neck, were used for measurement. These points were chosen because they overlie the cervical spine, which provides a soft tissue-bone interface for the reflection of ultrasound signals, and because they can be defined reasonably consistently with reference to an anatomic landmark. Localization was facilitated by use of a small, transparent, acetate grid (not shown). Immediately before the actual measurement, the probe was applied with gentle pressure on the measurement point several times (loading and unloading) to achieve tissue preconditioning and to ensure that a stable ultrasound echo could be obtained. After applying a light force of 0.5 N as a preload, a force (load) of 5 N was applied (or a smaller force was applied, once a 30% indentation of tissue thickness was reached, to avoid discomfort of the individual). Each measurement was conducted 3 times (i.e., 3 cycles of the loading-unloading sequence) within approximately 10 seconds, and the 3 measurement results were averaged to give the YM value of the point.
One hundred and five patients who had been followed for at least 2 years after they completed radiation therapy for nasopharyngeal carcinoma and who did not have evidence of disease at the time of recruitment were recruited from an oncology follow-up clinic over a 12-month period. In the early part of the recruitment period, only patients who were assessed with have neck fibrosis by palpation were recruited, followed in the later part of the period by the recruitment of unselected consecutive patients (n = 69 patients). The reason for the selection of patients with palpable fibrosis in the early recruitment period was to avoid the possibility of skewing the distribution of the study population toward nil or mild fibrosis. The reason for recruiting only patients who had been treated for a single type of malignancy was to exploit the advantage of uniformity of radiation therapy techniques that had been used and the opportunity to assess the effect of a boost dose (parapharyngeal boost) given to the upper part of one side of the neck in some patients8 (Fig. 2). For each patient, the YM was measured for the two reference points, as described above (3 cm and 7 cm below the lowermost point of the mastoid), on each side of the neck. The average of the results for the two measurement sites was used for subsequent analysis. For comparison of the YM scores between the two sides of the neck, the YM scores of the upper and lower measurement sites were analyzed separately, because the upper site was subject to the unilateral boost dose of radiation in some patients, and the lower site was not.
Clinical Palpation Score
The clinical palpation score for each side of the neck was obtained for each patient independently by one or more of three raters (two radiation oncologists and one nurse specialist); 68% of patients were assessed by at least two raters. The scoring criteria were as follows: Grade 0, nil or equivocal presence of palpable fibrosis; Grade 1, unequivocal presence of palpable fibrosis of mild degree; Grade 2, moderately severe fibrosis change; and Grade 3, severe fibrosis. The palpation raters did not communicate about their experience of palpation rating throughout the study period and did not participate in the measurement of the YM.
Measurement of Rotation Range of the Neck
The patient sat upright on a chair with eyes looking straight forward, and a protractor was positioned horizontally above the patient's head with the radial center of the protractor overlying the center of rotation of the head. An extension rod on the protractor helped to align the rotation center and the tip of the nose, which was used as a reference mark. The maximal left lateral and right lateral rotation ranges were measured, with attention to avoiding shoulder movement during the procedure. The range of rotation to the left side was taken to reflect the degree of fibrosis of the right of the neck in the analysis, and vice versa. To allow for potential inaccuracies in measurement, the result was rounded up to the nearest 5°, and two measurements were made for each side by the same observer. The average of the two rounded-up measurements was taken for analysis.
Symptoms Related to Fibrosis
The patients in the latter part of the study period (unselected patients; n = 69 patients) were asked in an interview about the presence of two symptoms: 1) discomfort in the neck (graded nil, mild, moderate, or severe) and 2) the experience of restriction in turning the head during activities of daily living (graded yes or no). The interview was carried out after the palpation and neck rotation measurements by one interviewer. The presence of moderate-to-severe discomfort and the presence of the experience of restriction in turning the head were used as the end points for correlation with the scores of the three different measurement methods.
Radiation Therapy Technique
The radiation therapy details were extracted from the patients' clinical records after the study period. All patients had both sides of the whole length of the neck irradiated. The biological effective dose (BED) Gy39 to the upper-middle neck ranged from 110 Gy3 to 116.9 Gy3, and the unilateral boost dose to the upper neck ranged from 30.3 Gy3 to 33.3 Gy3. The boost port was directed from a posterolateral direction from one side, with the posterior border covering the vertebral body and the lower border typically around the junction of cervical vertebrae 2 and 3 (Fig. 2).
A linear regression method was used to assess the relation between the YM and the palpation score as well as between the YM and the neck rotation score. The slope of the regression line for each rater was determined, and their 95% confidence intervals (95%CI) were calculated. The purpose was to determine whether the three raters were consistent with respect to this relation between the YM and the palpation score. A chi-square test was used to assess the following: differences in YM scores among patient groups with different palpation scores, difference in YM scores between both sides of the neck, and difference in palpation scores between patients with and without symptoms. The Wilcoxon (rank sums) test was used to compare the YM scores of patients with and without symptoms and to compare the neck rotation range of patients with and without symptoms.
YM versus Palpation Score
Figure 3 shows the distribution of YM scores with respect to palpation scores. The YM score differed significantly among patient groups with palpation scores of 3, palpation scores of 2, and palpation scores of 0 or 1 (P < 0.05). This applied to all three palpation raters. Linear regression lines (not shown) for YM versus palpation score were plotted for each of the three palpation raters: All showed a significant positive correlation (Table 1). The overlap of confidence intervals (Table 1) suggests that the relation held irrespective of which rater determined the palpation score.
|Palpation rater||Slope of regression line||P value of slope||95%CI of slope|
|Rater 1||26.9||< 0.01||20.2–33.6|
|Rater 2||29.1||< 0.01||21.0–37.4|
|Rater 3||26.0||< 0.01||19.0–33.1|
YM versus Neck Rotation Range
The relation between the UM and the neck rotation range is summarized in Figure 4. A significant negative correlation was found (P value of slope = 0.0001).
Relation of YM to Radiation Dose
Forty-six percent of patients had received a radiation boost dose to one side of the upper neck. The YM score was significantly higher on the boosted site compared with the contralateral nonboosted site (Table 2), although the palpation score was not significantly different on the two sides in this subset of patients (data not shown). In the other patients who had received the same dose to both sides of the upper neck, the YM scores were similar on both sides (Table 2).
|Patient subset||YM of right side (mean ± SD)||YM of left side (mean ± SD)||P value|
|Equal dose to both sides (n = 57 patients)||64 ± 56 kPa||60 ± 46 kPa||0.27|
|Only right side boosted (n = 29 patients)||81 ± 46 kPa||56 ± 39 kPa||< 0.001|
|Only left side boosted (n = 19 patients)||62 ± 36 kPa||89 ± 53 kPa||< 0.01|
Relation to Symptoms
Thirteen percent of assessed patients had moderate-to-severe discomfort in the neck, and 23% of patients had a sensation of restriction in turning the head. A significantly higher score was found for each of the three measurement methods (YM score, palpation score, and reduction in neck rotation range) in patients with moderate-to-severe discomfort and in patients with a sensation of restricted turning of the head (Tables 3, 4).
|Measurement method||Median (interquartile range)||P value|
|Patients without discomfort||Patients with discomfort|
|Young Modulus (kPa)||49 (30–75)||82 (60–151)||0.02b|
|Palpation Grade 2–3 (%)||28||89||< 0.01c|
|Neck rotation range (degrees)||55 (47–65)||40 (40–50)||0.02b|
|Measurement method||Median (interquartile range)||P value|
|Patients without restricted movement||Patients with restricted movement|
|Young Modulus (kPa)||49 (27–77)||61 (51–126)||0.03b|
|Palpation Grade 2–3 (%)||43||75||< 0.01c|
|Neck rotation range (degrees)||55 (50–65)||40 (32–50)||< 0.01b|
Fibrosis is one of the most common late effects of radiation therapy, but its objective quantification in the context of clinical practice is not easy. This limitation is particularly apparent in contemporary oncology practice, with increasing demand for quantified outcome parameters for the calculation of normal tissue complication probabilities.
The significant correlation of the YM with both the palpation score and a functional outcome indicator of fibrosis, and with the patients' symptoms, suggests that the new parameter can address meaningfully the characteristics that are measured by conventional methods. Such a correlation, per se, does not prove the superiority of one measurement method over the others, and the choice of methods also would depend on their intrinsic characteristics, such as objectivity, reliability, sensitivity, quantitative/qualitative nature, utility (correlation with symptoms and quality of life), and cost. With respect to these considerations, it is evident that the different methods of measurement of fibrosis have different spectrums of strengths and limitations.
The obvious limitations of the hand palpation method include its inherent subjectivity and that it is only semiquantitative rather than quantitative in nature. Although the range of neck rotation can serve as a functional indicator of fibrosis, it provides only an indirect measure of fibrosis; it may be affected by cervical spinal problems, such as spondylosis; and it cannot address fibrosis within subregions of the neck. These limitations are overcome largely by the current method of measuring tissue deformability. Furthermore, the YM is sensitive to differences in radiation dose levels given to different subregions of the neck. The lack of a corresponding difference in the palpation score in these patients may be related in part to the study design, in that the raters gave an average score for each side of the neck rather than separate scores for subregions of the neck, although it is also possible that the palpation method is less sensitive for detecting the effect of a difference in dose of about 30 Gy3. The correlation of the YM score with the patients' symptoms also supports the validity of the new measurement method.
It is noteworthy that there is a degree of spread of the YM value among each palpation grade, and this can be seen as a partial discordance of measurement results between the two measurement methods. In the absence of a gold standard for measuring fibrosis, it should not be assumed that such discordance in measurement results is due to the inaccuracy of the YM, but it should lead to rethinking the clinical definition of fibrosis. It is noted that there are differences in the nature of information acquired by the two measurement methods. The palpation score is actually a form of processed information that, during the acquisition process, may involve the reduction of information (e.g., categorizing a score of 2.5 as either 2.0 or 3.0) and the addition of information (e.g., comparison with the rater's previous experience). The YM value corresponds more to original raw data and, conceptually, is a much more simple and original way of representing fibrosis. In fact, the raters in the current study frequently encountered instances in which assigning a palpation score was difficult. The availability of YM data may facilitate research on factors that influence information processing by the human mind during the palpation process. For example, the discordance in palpation scores from YM measurement may be due to the palpation rater being influenced by visual cues about the slim/obese body build of the individual, or to the rater's interpretation of the amount of fatty tissue at the site. Such subjective interpretations may represent a form of adding value to the stiffness parameter, and research in these aspects may provide better insight into which parameter is the most clinically relevant gold standard for scoring fibrosis.
Regarding the reliability and general applicability of YM measurements, our previous developmental study had shown that the current measurement method for YM had high consistency when measured by different operators on normal participants (95%CI ± 15%) and patients (95%CI ± 13%), Consistency also was high with repeated measurements by the same operator (95%CI ± 7%).5 Such variations are very small compared with the significant increase in average YM values by 200–300% from mild fibrosis to severe fibrosis shown in the current study. Although the measurement points in the current study were confined to the overlying the cervical spine, measurements at more anterior locations can be achieved by tilting the ultrasound probe to obtain a soft tissue-bone interface. With experience, we also found that the availability of a bony surface was not always essential, and ultrasound signals also could be reflected from soft tissue interfaces and even from a skin-air interface on the exit side. The reproducibility of results with alternative measurement devices in other centers is realistic, because the YM is a universally measurable physical characteristic. This is analogous to the weight of an object being independent of different weighing devices. The ability to describe fibrosis in absolute units (kPa) is an important step toward observer-independent comparisons. This is particularly relevant for late radiation morbidity scoring, which may involve different clinicians with different experience over a long follow-up period, and for cross-center comparisons.10 Part of the SOMA grading definitions for soft tissue fibrosis also may pose practical difficulties when applied to the neck. For example, grading based on the percentage of muscle involved by fibrosis1, 3 is much more applicable to the irradiated limb than the irradiated neck. This is understandable, because previous studies on soft tissue fibrosis were based largely on irradiated limb tissue.11–13 This is in contrast to the scarcity of studies on assessment of soft tissue fibrosis of the irradiated neck and underlies the need to explore alternative assessment methods and grading systems tailored to neck fibrosis.
Because the current measurement method is suited to a small target site, site dependency is a concern, with attendant advantages and disadvantages. On the disadvantageous side, the method has a more stringent requirement for measurement site consistency, so that interpatient comparisons and intrapatient comparison over time are meaningful. Our previous study had shown that the YM value changed by up to ± 14% (95%C.I.) at sites 1 cm adjacent to the measuring site.5 Such a requirement for site consistency is not as important for the hand palpation and neck rotation measurement methods, which can be seen as more robust in this regard. Conversely, this site sensitivity can be exploited for fine discrimination of tissue stiffness among different subsites of the neck that may have been subjected to different radiation dose levels, such as in a shrinking-field technique. Conceivably, tattoo marks could be used to facilitate repeat examinations over time. There is potential for the creation of a fibrosis map by multiple site measurement in each individual. Measurements also could be extended to the operated neck. Analysis of such information may provide a more comprehensive picture of fibrosis. Such extensive assessment was not performed in the current cohort of patients because of resource limitations within the study.
Resource requirement also appears to be a relative disadvantage of the current measurement method. Although the measurement can be performed in the outpatient clinic setting, the average time for measurement for one reference point was about 1.5 minutes, which was appreciably longer than it took to perform simple palpation and neck rotation range measurements. Additional time also was required for computation of data after the measurement. We expect that the technique can be learned by an ultrasonography technician with several hours' training. With the current technology, we believe that there is a good possibility that the measurement time can be shortened appreciably and that the apparatus can be made more portable and automated. The potential of the method to be used by nonmedical staff at separate venues, the suitability of data for storage and transmission in electronic media, and the wealth of original information acquired compared with simple palpation are favorable factors in the cost-effectiveness equation. In fact, the current measurement method is the only one of its kind that has been developed and is ready for application in the clinical service setting rather than being restricted to the research setting.
In conclusion, the current study has established a new prototype method to quantify postirradiation neck fibrosis. It allows the definition of fibrosis in absolute physical units rather than arbitrary ordinal units. The results are correlated significantly with the results from conventional measurement methods and with the patient's symptoms. It has the advantages of objectivity, site sensitivity, and correlation with sites subject to different radiation dose levels. Its practicability of measurement in the outpatient clinic setting has been demonstrated, and the data acquired are suitable for electronic storage and retrieval. We advocate incorporation of this parameter in the Analytic category of the LENT SOMA system to facilitate more correlative studies with other SOMA criteria in a broader spectrum of patients postradiation therapy. There are research opportunities in the areas of establishing multiple-site deformability profiles and the use of alternative designs for instruments with enhanced automation.
The authors acknowledge the contributions of colleagues at the Chinese University of Hong Kong: in the Department of Clinical Oncology, Prof. P. J. Johnson, F.R.C.P., for article review, Frankie Mo, B.Sc., for statistical support, and Ricky Chau, M.Sc., for preparation of isodose diagrams; and, in the Department of Orthopedics and Traumatology, Prof. L. K. Hung for development of the study concept.
- 4Biomechanics: mechanical properties of living tissues. New York: Springer-Verlag, 1981..