Introduction

Lung cancer is the leading cause of death from cancer for both men and women in Canada [1]. Screening for lung cancer and safe surgical resection is the most promising strategy to improve lung cancer survival [2,3]. The preferred method of lung cancer resection is by Video Assisted Thoracic Surgery (VATS); however, VATS can make it difficult for the surgeon to accurately identify small and deep subpleural nodules. When this occurs, the surgeon must make a larger thoracotomy to manually palpate and locate these subpleural nodules. Avoiding the need to intraoperatively transition to an open thoracotomy is preferable for improved patient outcomes [4].

We have developed a Tactile-Sensing System (“finger on a stick” lung palpator) that is wireless and provides visual feedback to identify deep subpleural lung nodules [5–7]. Our current lab phantom tumour model consists of silicon and agar-based spheres that have excellent acoustic properties for detection by ultrasound but lack the appropriate elastic properties needed for the tactile imaging experiments. Better ex vivo and in vivo phantom models of lung tumours are needed [5,6,8].

The standard surgical practice of palpation is based on the qualitative assessment of the stiffness of tissue. The Young’s Modulus (YM) can be used to define the stiffness of tissue and is expressed as the relationship between stress and strain of a material. It is calculated from the slope of the stress–strain curve and can be achieved by compressing a sample of tissue by a controlled amount while measuring the resulting forces. This value has been discussed in the context of breast cancer and colorectal cancer, but not for human lung tumours or in healthy lung tissue [9,10]. While this information is relevant on its own, it would also allow for the development of more accurate tumour models that can help translate tactile imaging technology to the operating room to detect subpleural lung nodules.

The purpose of this study was to assess the YM quantitative values of lung nodules in select patients with suspected lung cancer. We hypothesized that there would be a difference between the YM of lung tumours and normal lung parenchyma.  We believed that there may be a correlation between the YM and different histological tissues in human lung (adenocarcinoma, squamous cell carcinoma, normal lung tissue), with an expected range of YM for each tumour type.

Patients and Methods

We examined the YM of a cohort of patients presenting with clinical and radiographic evidence of lung nodules that underwent surgical resection at Victoria Hospital, London Health Sciences Centre, London, Ontario, Canada.  Participants were excluded if they were younger than 18 years of age or were unable to provide informed consent. Patients with suspected metastatic disease, and previous neoadjuvant therapy were included. All patients underwent either a VATS or thoracotomy resection of the tumour by a board-certified thoracic surgeon. Immediately following surgical resection, a 5–10-millimeter section of tumour and a similar-sized sample of healthy adjacent lung tissue were removed from the resected specimen.

To measure the YM of tissue, our team constructed a portable indenting device (Figure 1). The device measured the YM using two different modalities per test. The first method applied a unform force over a small surface area on lung tissue samples such that the tissue was compressed by 15% of its thickness. The device measured the force applied and displacement of the tissue 200 times during the procedure. We defined the results of this process as the “linear YM”. The device then performed an indentation test using the second technique. The indenter compressed the tissue to its maximum strain level and allowed the tissue to return to its original thickness at a rate of 10 Hz. The force applied and displacement of the tissue were measured 200 times. This was repeated for 20 cycles to yield a total of 4000 data points. We defined the results of this process as the “cyclic YM”. Each iteration of the test yielded one linear and cyclic YM result. These measurements were repeated three times for each lung nodule and adjacent lung sample.

Figure 1: Labelled image of the tissue indentation device used to measure the Young’s Modulus.

The displacement and force data collected were converted to microns and micro-Newtons respectively and saved on a microSD card. The YM was calculated according to the equation below:

where E is the YM, n is the number of data points, R is the radius of the indenter tip, F is the applied force, and W is the displacement of the indenter. The coefficient term corrects for the size of the indenter tip [11]. The tissue displacement and force measurement accuracies were within 30–50 microns and 0.2–0.4 micronewtons, respectively.

The size of the tumour, Positron Emission Tomography (PET) Maximum Specific Uptake Value (SUV Max), histological diagnosis, location, and neoadjuvant chemotherapy/radiation, as well as the patient’s age and sex were recorded and stored in our REDCap database to maintain patient privacy and confidentiality. Data analysis was performed using Microsoft Excel Version 16.69.1 and XLSTAT Premium Version 24.3.1. The average cyclic and linear YM was calculated based on the results from each test. Pearson Product Moment correlation and Spearman’s Rank correlation tests were performed to assess the associations between the clinical and pathological variables and measured YM. The results were considered significant if R > 0.70. The frequency distribution of the YM of each tissue histology was assessed for normality using the Shapiro-Wilk test. For nonparametric data, the Mann-Whitney U-test was used to compare the differences in YM for each histology. A threshold p value of 0.05 was used to assess for statistical significance.

Results

The normality of each distribution was assessed by using the Shapiro-Wilk test at a significance level of a=0.05. The null hypothesis of this test stated that the sample was taken from a normal distribution. As displayed in Table 1, the computed p-values were lower than the significance level. Therefore, we did not accept the null hypothesis and the Mann-Whitney U-test was used for our analyses. The linear and cyclic YM results follow a similar distribution and have negligible differences in measurement results for tumour or adjacent lung tissue. Thus, we reported the linear YM results for our analysis. Tumour and adjacent lung YM measurements were compared in different subsets based on the histology of tumour. The complete results of our analysis are displayed in Table 2. On aggregate, the median linear YM for lung tumours (12.73 kPa [2.68–199.10]; p < 0.001) was statistically greater than adjacent lung tissue (6.12 kPa [1.65– 13.19]; p < 0.001) as seen in Table 3 and Figure 2. The median linear YM for adenocarcinomas (12.51 kPa [2.68–199.10]; p < 0.001) was statistically greater than adjacent lung tissue (6.24 kPa [1.65–13.19]; p < 0.001) as displayed in Table 3 and Figure 3. The median linear YM for squamous cell carcinoma (12.88 kPa [5.03–73.96]; p < 0.001) was statistically greater than adjacent lung tissue (6.13 kPa [4.13–9.58]; p < 0.001) described in Table 3 and Figure 4. The various metastatic tumours included colorectal carcinoma, breast carcinoma, adenoid cystic carcinoma, renal cell carcinoma, and melanoma. All metastases were grouped together due to the small sample size of each individual pathological subset. The median linear YM for lung metastases (12.57 kPa [7.40–73.96]; p < 0.001) was statistically greater than for adjacent lung tissue (5.14 kPa [3.28–10.19]; p < 0.001) as seen in Table 3 and Figure 5. For adjacent lung tissue, there were no statistically significant differences in the median YM of healthy lung tissue and fibrotic lung tissue, emphysematous changes, or other various benign pathologies. The median YM values were stratified based on the grade of lung tumours and were not statistically significantly different. For both pathological T-stage and N-stage, there were no statistically significant differences between the median YM values of lung tumours. When considering tumour location, there was no statistically significant difference between median YM values for lung tumours.

The Pearson Product Moment Correlation test was performed between the linear YM and the following variables: patient age; patient sex; the number of cigarette packs (20 cigarettes/pack) smoked per day multiplied by the number of years smoked (smoking pack-years); PET SUV Max; size of lesion on Computed Tomography (CT); and size of lesion from histology. We found no significant correlation between the average linear YM values and these variables since they all had an R < 0.70.

Figure 2: Standard box plots of the median and interquartile range of the Young’s Modulus of all lung tumours compared to adjacent lung tissue.

Figure 3: Standard box plots of the median and interquartile range of the Young’s Modulus of adenocarcinomas compared to adjacent lung tissue.

Figure 4: Standard box plots of the median and interquartile range of the Young’s Modulus of squamous cell carcinomas compared to adjacent lung tissue.

Figure 5: Standard box plots of the median and interquartile range of the Young’s Modulus of various metastases compared to adjacent lung tissue.

Table 1: Median linear vs. cyclic Young’s Modulus for tumour and adjacent ling tissue and Shapiro Wilk test results for normality.

Table 2: Comparison of median linear Young’s Modulus values based on tumour and adjacent normal lung histology.

Table 3: Summary of median Young’s Modulus tumour measurements that were statistically significant compared to adjacent normal lung.

Comment

This study presents the first measurements of the YM of lung tumours and adjacent lung tissue for various pathologies. The measurements were obtained by indentation testing of recent surgically excised lung tumours and adjacent lung tissue. The YM of biological tissue can be estimated through various indirect methods, such as ultrasound, shear wave elastography, or harmonic motion elastography. These indirect methods use imaging techniques to noninvasively mimic palpation below the skin’s surface to determine the YM of tissues. However, the YM values obtained from these non-invasive techniques are only estimates that are qualitative in nature. Our indentation device, based on the work of Egorov et al., has been specifically designed to measure the YM of small tissue samples and produces repeatable quantitative measurement [11].