Sonoelastographic evaluation with the determination of compressibility ratio for symmetrical prostatic regions in the diagnosis of clinically significant prostate cancer.

AIM: Sonoelastography is a technique that assesses tissue hardness/compressibility. Utility and sensitivity of the method in prostate cancer diagnostics were assessed compared to the current gold standard in prostate cancer diagnostics i.e. systematic biopsy.
MATERIAL AND METHODS: The study involved 84 patients suspected of prostate cancer based on elevated PSA levels or abnormal per rectal examination findings. Sonoelastography was used to evaluate the prostate gland. In the case of regions with hardness two-fold greater than that of symmetric prostate area (strain ratio >2), targeted biopsy was used; which was followed by an ultrasound-guided 8- or 10-core systematic biopsy (regardless of sonoelastography-indicated sites) as a reference point.
RESULTS: The mean age of patients was 69 years. PSA serum levels ranged between 1.02 and 885 ng/dl. The mean prostate volume was 62 ml (19-149 ml). Prostate cancer was found in 39 out of 84 individuals. Statistically significant differences in strain ratios between cancers and benign lesions were shown. Sonoelastography guided biopsy revealed 30 lesions - overall sensitivity 77% (sensitivity of the method - 81%). Sonoelastographic sensitivity increased depending on cancer stage according to the Gleason grading system: 6-60%, 7-75%, 8-83%, 9/10-100%. The estimated sensitivity of systematic biopsy was 92%.
CONCLUSIONS: Sonoelastography shows higher diagnostic sensitivity in prostate cancer diagnostics compared to conventional imaging techniques, i.e. grey-scale TRUS, Doppler ultrasound. It allows to reduce the number of collected tissue cores, and thus limit the incidence of complications as well as the costs involved. Sonoelastography using the determination of compressibility ratio for symmetrical prostatic regions may prove useful in the detection of clinically significant prostate cancer.

Introduction

Prostate cancer is the most frequently diagnosed solid tumor in men, exceeding the number of diagnosed lung and colon cancers. Considering its detectability at 214 per 1000 cases, it is the second most common cause of death among male Western Europeans, and thus presents one of the largest medical issues in Western countries(.

Early-stage prostate cancer is usually asymptomatic and in most cases located in the peripheral prostate. Imaging techniques play a crucial role in cancer detection, localization as well as the assessment of its progression. TRUS (transrectal ultrasonography) guided systematic biopsy is the current standard diagnostic procedure in patients suspected of prostate cancer based on elevated PSA levels (prostate-specific antigen) and/or abnormal DRE (digital rectal examination) findings. Between 8 and 10 specimens are collected from the peripheral part of the prostate gland, from the most posterior and extreme lateral zone.

Since the early 90s of last century, sonoelastography (SE) has been used as an alternative imaging technique for prostate gland evaluation based on the compressibility of prostate tissues. The method, originally described by Ophir(, was for the first time used in the diagnostics of prostate cancer in 2002(. Konig( and Pallwein( research teams reported on the diagnostic sensitivity of this technique, which was estimated at 84% and 80%, respectively. Miyanaga et al.( assessed the sensitivity of sonoelastography in detecting untreated prostate cancer as 93%.

In light of the above-mentioned studies as well as due to an insufficiency of other ultrasound techniques used in prostate cancer diagnostics, such as grey-scale TRUS with 40% sensitivity further improved by 16%( in Doppler methods, sonoelastography (SE) seems a valuable diagnostic tool for detection, localization and progression assessment of prostate cancer. Results of targeted prostate biopsies obtained using the sonoelastography method as well as widely accepted grey-scale TRUS-guided 8-10 core systematic biopsy were presented in the study.

Material and methods

Patients

The study was approved by the appropriate Bioethical Committee. Patients gave their written informed consent to participate in the study. The observation included 84 patients (tab. 1) suspected of prostate cancer based on: elevated PSA levels confirmed in two subsequent tests at least one month apart (following the exclusion of prostate inflammatory conditions) or abnormal per rectal examination findings (e.g. palpable nodular lesions within the prostate gland) or patients who have undergone systemic biopsy with persistent or increasing PSA levels above the norm. Prostate biopsies were performed between May 14th 2011 and February 2nd 2013 in the Department of Urology of the Mazovia Regional Hospital in Siedlce. Histopathology results were obtained using core needle sonoelastography-targeted biopsies and TRUS-guided systematic biopsies.

Tab. 1

Characteristic of patients

Age	58–84 years (mean 69 years)
PSA	1.02–885 ng/dl (mean 15.4 ng/dl)
DRE	35 patients with abnormal digital rectal examination (DRE) out of 84 (41%) patients
Prostate gland volume	19–149 ml (mean 61.8 ml)

Procedure

In the first stage, ultrasonography and sonoelastography assessments were performed with Aplio XG (Toshiba, Japan) using an end-fire transrectal probe, micro-convex, 6 MHz.

Prostate compressibility/hardness was assessed at three levels: around the base of the prostate just below the bladder, at mid-height of the prostate and around the top of the prostate, while additional cross-sections were performed in the cases of gland volume of more than 60 ml or gland largest size of more than 6 cm. Compression performed using mechanical pressure with transrectal probe allowed to obtain images of color encoded deformations, as in accordance with the following scheme: the highest tissue compressibility (soft) – red, the lowest tissue compressibility (hard) – blue, medium tissue compressibility – green. Decompression cycle with optimum sinusoidal function in a rate control chart was selected for analyses. The color scale was optimized in some cases by adjusting its resolution, which allowed to obtain a better image of deformations within the symmetrical prostate areas. Dark blue-encoded regions, i.e. characterized by low compressibility, were classified as areas suspected of neoplastic transformation. This was followed by identifying a region of prostate symmetrical to the suspected area on the grey scale and calculating the strain ratio. Thus the assessment of tissue compressibility was performed based on a comparison between symmetrical prostate regions and relative differences in hardness, and not on the absolute values. In the case of strain ratio ≥2 (range 2–27), the site was classified for core needle biopsy, as shown in fig. 1.

Fig. 1

Patient aged 76 years; PSA – 31.4; histopathological examination of this specimen revealed prostate cancer with Gleason score of 8 (4 + 4). A. Sonoelastography image of prostate gland in the ROI 1 (yellow) shows a suspected area (compressibility ratio of symmetrical areas – strain ratio = 4.7); B. An analogous B-mode image with marked areas used to determine the compressibility ratio and to select sites for targeted biopsy (SE)

Elastography images and the corresponding B-mode images were archived in the form of cine-loops as well as static images to be used on PC during the second stage.

In the second stage, biopsy of each site indicated by SE as “suspicious” was performed with BK Medical Pro Focus Ultrasound (Denmark) using a biplane 5–10 MHz transrectal probe equipped with a probe channel for prostate core needle biopsy, followed by systematic biopsy, during which eight or ten tissue cores were collected regardless of sonoelastography indicated regions. The biopsy was performed using Tru-Cut 18 G needle. The obtained tissue samples provided with a protocol were secured in separate cassettes and sent for histopathological examination.

Once the diagnostic process was completed, patients with indications for surgical treatment were qualified for open or laparoscopic radical prostatectomy and provided with a therapy. In these cases, elastography and B-mode images were additionally correlated with histopathological specimens obtained during surgical treatment, as shown in fig. 2.

Fig. 2

Patient aged 52 years. A. Sonoelastography – biopsy of the indicated area revealed tumor with Gleason score of 8 (4 + 4); ROI 1, strain ratio 4.1; B. B-mode image of the cancerous region; C. Radical laparoscopic prostatectomy specimen; D. Fixed specimen with visible cancerous lesions (arrow)

Results

The study involved 84 patients suspected of prostate cancer, who showed elevated PSA levels ranging between 1.2 ng/dl and 885 ng/dl. In total 42 cases of neoplastic lesions, including 39 adenocarcinomas and 3 cases of prostatic intraepithelial neoplasia (PIN), were diagnosed based on targeted and systematic biopsies.

A total of 782 tissue cores were sampled during systematic biopsies, which allowed to identify 36 adenocarcinomas and one PIN lesion. Targeted biopsies of areas indicated during sonoelastography as low compressibility lesions, i.e. hard, identified 30 adenocarcinomas and 3 PIN-like lesions – a total of 185 tissue cores were sampled for this purpose. Fig. 3 shows one of the areas indicated by sonoelastography as suspected, hard (strain ratio of 9.8), the tissue of which proved cancerous.

Fig. 3

Patient aged 62 years; PSA = 20.09. Histopathological examination of this specimen revealed prostate cancer with Gleason score of 6 (3 + 3) (strain ratio = 9.8)

Targeted biopsy of areas indicated in sonoelastography evaluation failed to identify 9 cases of prostate adenocarcinoma (23%):

Four cases involve low-stage cancers having a Gleason score of 6 (3 + 3); two of these cases were initially diagnosed as atypical small acinar proliferation (ASAP) during histopathological examination and the diagnosis of adenocarcinoma was possible only due to additional histopathological examination.

Two undiagnosed cases resulted from operator error – sides of prostate gland were incorrectly specified.

In one case, sonoelastography failed to detect differences in the hardness of the examined peripheral area of symmetrical prostate regions – 8-core systematic biopsy revealed a neoplastic lesion in one specimen, which had a Gleason score of 8 (4 + 4).

In another patient, no neoplastic lesions were found in two tissue cores sampled during sonoelastography guided targeted biopsy, whereas a systematic biopsy performed in the same patient revealed tumors in four out of ten tissue cores, which had Gleason scores of 8 (4 + 4), 7 (4 + 3), 7 (4 + 3) and 6 (3 + 3).

The last case concerns a patient, whose sonoelastography guided targeted biopsy did not reveal cancer, but systematic biopsy showed neoplastic lesions in two out of ten tissue cores, graded 7 (3 + 4) and 7 (3 + 4) according to Gleason system.

TRUS-guided systematic biopsy failed to reveal three prostate cancer cases out of a total of 39 adenocarcinomas identified based on both used diagnostic methods (7.7%).

In one case, after a negative 10-core systematic biopsy, targeted biopsy of the area indicated during sonoelastography assessment revealed cancer of Gleason score 7 (3 + 4); fig. 4.

Fig. 4

Patient aged 80 years; PSA 11.45 ng/ml; 10- core systematic biopsy failed to reveal tumor, whereas targeted biopsy of SE-indicated area revealed prostate cancer with Gleason score of 7 (3 + 4) (biopsy involved ROI 3; strain ratio = 2.3)

In the second case, after a negative 8-core systematic biopsy, sonoelastography-guided targeted biopsy revealed a tumor of Gleason score 7 (3 + 4).

In the third case, one of the specimens collected during a biopsy of areas indicated during sonoelastography assessment showed a neoplastic lesion of Gleason score 6 (3 + 3), whereas systematic biopsy of 10 tissue cores showed no evidence of tumor.

Furthermore, a correlation between tumor grade in accordance with the Gleason grading system and the diagnostic sensitivity of the described imaging method, as shown in tab. 2, was found during the analysis of histopathological results for the specimens collected during sonoelastography-guided biopsy.

Tab. 2

Correlations between cancer stage according to the Gleason score and the diagnostic sensitivity of sonoelastography

Gleason Score	The number of positive elastography findings/ the number of cancers (sensitivity in percent)
6	6/10 (60%)
7	9/12 (75%)
8	10/12 (83%)
9–10	5/5 (100%)

Statistical analysis

STATISTICA 10 (StatSoft Inc., USA) was used in statistical analysis, which involved three groups of lesions as well as their strain ratio values, i.e.:

strain ratio for lesions with confirmed neoplastic character (SR1);

strain ratio for inflammatory and postoperative lesions with excluded neoplastic transformation based on histopathological examination (SR2);

strain ratio for lesions with negative biopsy findings and showing no other prostate pathologies – inflammation or medical history of postoperative treatment (SR3).

Next, strain ratio values were grouped in accordance with the Gleason grading system as grades 6 to 10, with strain ratio values for Gleason 9 and Gleason 10 included in one group. The Shapiro–Wilk W test w as used to assess the analyzed data for normality distribution. Since the variables were not normally distributed, non-parametric Mann–Whitney test was used for further analysis. The Kruskal–Wallis test was used to analyze the strain ratio values of malignant lesions graded 6–10 in accordance with the Gleason grading system. The α-level w as s et a t 0.05. T he m ean S R w as 7.88 (range 2.3–26.5) for lesions with confirmed neoplastic character; 6.03 (range 2.0–27.0) for inflammatory and postoperative lesions; and 3.57 (range 2.0–17.3) in the case of patients with no other lesions. The results are shown in tab. 3.

Tab. 3

SR (strain ratio) assessment in the study groups of patients

	Mean	Med	Min.	Max.	SD
SR1	7,88	5,35	2,3	26,5	6,05
SR2	6,03	4,25	2,0	27,0	5,48
SR3	3,57	2,9	2,0	17,3	2,65
SR non-neo	4,44	3,4	2,0	27,0	4,04

List of abbreviations: Mean – mean value, Med – median, Min. – minimum value, Max. – maximum value, SD – standard deviation

No statistically significant differences in strain ratio values were found between confirmed neoplastic lesions (SR1) and inflammatory/postoperational lesions (SR2). Statistically significant differences in strain ratio values were found between patients with histopathologically confirmed proliferative lesions (SR1) and patients with negative biopsy findings who showed no other prostate pathologies (SR3); statistically significant differences between group SR2 and SR3 were also distinct. Additionally, proliferative lesions (SR1) were compared with the group of benign lesions, involving SR2 and SR3; the obtained results also showed statistically significant differences (tab. 4, figs. 5–8).

Tab. 4

The Mann–Whitney U Test

Variable	Z	p
SR 01/02	-1,23	0,22
SR 01/03	-4,80	0,000002
SR 02/03	-3,44	0,0006
SR 01/02,03	-3,96	0,00007

List of abbreviations: Z – The Mann–Whitney U Test value for each group; p – p value; SR 01/02 – SR1 strain ratio vs. SR2 strain ratio, SR 01/03 – SR1 strain ratio vs. SR3 strain ratio; SR 01/02,03 – strain ratio comparison: SR1 vs. SR2 and SR3 (as a total group of benign lesions)

Fig. 5

A comparison between a group of lesions with confirmed cancerous character (SR1) and inflammatory/postoperative lesions with no confirmed cancerous character (SR2)

Fig. 8

A comparison: cancerous lesions (SR1) vs. postinflammatory and postoperative lesions (SR2) as well as non-pathological lesions within the prostate gland (SR3)

A comparison between a group of lesions with confirmed cancerous character (SR1) and lesions with no confirmed pathology in histopathological examination (SR3)

A comparison: postinflammatory and postoperative lesions (SR2) vs. lesions with no confirmed pathology (SR3)

No statistically significant differences were found between strain ratio values in groups graded 6–10 in the Gleason system: p = 0.3809.

Conclusions and discussion

Sonoelastography is a valuable diagnostic tool for prostate cancer detection. The possibility to assess prostate tissue compressibility/hardness allows to detect pathological lesions. It is a proven fact that cancerous tissue, due to an increased cellular density and structural impairment, is depicted as a harder region compared to healthy prostate tissue. The diagnostic criteria originally described by Konig( include: lesion hardness, repeatability when changing probe inclination, size of at least 5 mm. These criteria were subsequently modified by Pallwein(, who suggested a three-grade system for prostate lesion assessment in terms of suspected neoplastic transformation (tab. 5).

Tab. 5

Sonoelastographic evaluation system – according to Pallwein

Degree	Description	Cancer patients (%)
1	Uniform dispersion, uniform hardness/compressibility	2,3–11,9
2	Non-homogeneous increase in hardness/compressibility, variable red and blue areas, each colored dot with a diameter of <5 mm, no repeatability when changing probe inclination	26,4–28,8
3	Focal increase in hardness/compressibility – homogenous, asymmetric focal lesions >5 mm in size, repeatable after a change in probe inclination	68–82,4

In a study by Aboumarzouk(, involving a retrospective metaanalysis of 16 studies on the use of sonoelastography in prostate cancer diagnostics, which included a total of 2278 patients, a sensitivity of 71–82% was obtained for this method.

In the present study, where all sites subjected to a targeted biopsy met grade 3 criterion according to Pollwein, the overall sensitivity of biopsy targeted at sonoelastography-indicated regions was 77% and even 81%, following the exclusion of operator error cases. This sensitivity corresponds to the one described above in the “BJU” study.

Operator errors resulted from the necessity to use two devices for the assessment, i.e. Aplio XG (Toshiba, Japan) for sonoelastography and BK Medical Pro Focus (Denmark) for biopsies. Such errors may be readily eliminated and are unlikely to occur provided that the overall transrectal evaluation, including biopsy, is performed using one device for the same examination.

Furthermore, it should be noted that the sensitivity of a 10-core systematic biopsy in the conducted study was, according to the determined scheme, 92% (assuming that cases identified using both methods: systematic biopsy and sonoelastography-guided targeted biopsy account for 100%).

The currently used golden standard for the diagnosis of prostate cancer is a 10-core ultrasound-guided biopsy, which is associated with the risk of a number of complications; and the most common of these include: hematospermia (blood in the ejaculate) – 37.4%; bleeding from the bladder – 14.5%; fever – 0.8%; urosepsis – 0.3%; rectal bleeding – 2.2%; urinary retention – 0.2%; prostatitis – 1%; epididymitis – 0.7%. Some of these require hospitalization, expensive treatment and significantly impair the quality of life of diagnosed patients. In the light of the above, the number of tissue cores sampled during diagnostic procedures is also significant. In our study, a total of 782 tissue cores were sampled during the traditional systematic biopsy, whereas a total of 185 tissue cores were collected in the same number of patients for the diagnostics using biopsy targeted at sonoelastography-indicated sites. Thus, the number of invasive diagnostic procedures (core biopsy) using sonoelastography decreased to less than ¼ compared to the currently used golden standard, which undoubtedly allows to reduce complications and costs of treatment as well as to improve the quality of life of diagnosed patients.

In the study group of patients, it was shown that the sonoelastography-guided targeted biopsy, following the exclusion of cases resulting from operator error, failed to identify eight cases of cancer confirmed by the 10-core systematic biopsy. Tab. 2 shows the influence of the degree of cancer malignancy graded in accordance with the Gleason system on the elastographic sensitivity. It may be concluded from the table that the cancer stage may affect the sensitivity of sonoelastography in prostate cancer detection.

End-fire transrectal probe was used in our study for sonoelastography evaluation of the prostate gland. Prostate gland compression was generated by an operator and involved applying rhythmic probe pressure on the examined object. Compression records in the form of a sinusoid were displayed on a screen. Such a technique allows the compression to be generated mainly in the probe axis, i.e. on the central part of the examined prostate gland, while a lower compression force is applied on the lateral areas, resulting in a colored sonoelastography scan. In order to overcome this technical problem, color scale was used only for the initial evaluation in our study, while classification for biopsy was based on a comparison of symmetrical prostate areas as well as the differences in their compressibility. Thus, the assessment was based on symmetrical prostate areas (compressed with the same force) rather than the absolute hardness of the suspected areas, and involved identifying the relative differences in hardness. Areas of more than two-fold hardness compared to symmetrical prostate regions were qualified as suspected and subjected to a targeted biopsy. Other factors significantly affecting prostate diagnostics using sonoelastography include medical history of prostate conditions, previous diagnostic examinations or surgical procedures of the prostate gland.

It seems that a medical history of prostatitis and diagnostic biopsy as well as treatment due to prostatic adenoma, e.g. TURP, irreversibly alter gland compressibility and may impair elastographic evaluation. This is confirmed by statistical analysis, which showed significant differences between strain ratios of cancers and benign lesions, whereas no statistically significant differences regarding strain ratio values were found between cancer patients and patients with previous prostate surgeries.

Even before the application of sonoelastography in ultrasonographic diagnostics it had been noticed in B-mode imaging that in the case of large volume tumors involving the whole organ, the ultrasonographic image may appear as normal – this is a so-called superscan phenomenon(. A similar situation occurs in sonoelastographic imaging – in the case of large volume tumor involving the whole organ, the differences in compressibility are minor or none. In three patients with advanced prostate cancer involving the whole gland [positive systematic biopsies and sonoelastography-guided biopsies; Gleason grade 8 (4+4)] the hardness ratio of symmetrical prostate areas ranged between 2.0 and 2.7, with the mean of 7.88 for neoplastic lesions. It seems, in such cases, that elastography assessing prostate hardness in absolute values (kPa) – shear wave elastography – would allow for a correct diagnosis. It should be noted, on the other hand, that these cases are clinically evident (PSA, DRE), which allows to avoid errors in targeted biopsy considering the hardness of symmetrical areas.

Summary

The study assessed the usefulness of sonoelastography in prostate cancer diagnostics. The results of sonoelastography guided targeted biopsy were compared with the results obtained in grey-scale TRUS-guided 8- or 10-core systematic biopsy, which is currently a standard procedure in prostate cancer diagnostics. The overall sensitivity of the method was 77%. It seems that the use of sonoelastography in selecting areas suspected of being cancerous allows to significantly reduce the number of collected tissue cores, and thus limit both the incidence of complications and the costs involved as well as to improve the quality of life of diagnosed patients. The study suggests that sonoelastography evaluation may be affected by the degree of cancer progression, although these findings require confirmation based on a larger number of cases.

Elastographic evaluation meets some limitations associated with the technique of examination, which is dependent on the operator as well as on the used equipment, especially in the case of applying mechanical compression on the prostate gland. The results may be further affected by previous prostate inflammatory conditions as well as diagnostic and surgical procedures.

Undoubtedly, further studies are necessary to verify the role of elastography in the diagnostics of prostate cancer as well as to determine not only the potential and usefulness of this method but also its limitations.