OUP user menu

Emerging Technologies for Assessing HER2 Amplification

Frédérique Penault-Llorca MD, PhD, Michael Bilous MD, Mitch Dowsett PhD, Wedad Hanna MD, Robert Yoshiyuki Osamura MD, Josef Rüschoff MD, Marc van de Vijver MD, PhD
DOI: http://dx.doi.org/10.1309/AJCPV2I0HGPMGBSQ 539-548 First published online: 1 October 2009


Patients with human epidermal growth factor receptor-2 (HER2)+ breast cancer are eligible for trastuzumab treatment; therefore, accurate assessment of HER2 status is essential. Until recently, only 2 methods were validated for determining the HER2 status of breast tumors in the routine diagnostic setting: immunohistochemical analysis and fluorescence in situ hybridization (FISH).

Recently, bright-field in situ hybridization techniques such as chromogenic in situ hybridization (CISH) and silver-enhanced in situ hybridization (SISH), which combine features of immunohistochemical analysis and FISH, have been introduced for the determination of HER2 status. These new techniques use a peroxidase enzyme–labeled probe with chromogenic detection, instead of a fluorescent-labeled probe, allowing results to be visualized by standard bright-field microscopy. Thus, the histologic features and HER2 status of a specimen can be evaluated in parallel. Moreover, signals do not decay over time. This review discusses recent publications regarding CISH and SISH testing, including results scoring and concordance between FISH and immunohistochemical analysis.

Key Words:
  • HER2
  • Breast cancer
  • Immunohistochemical analysis
  • Chromogenic in situ hybridization
  • CISH
  • Fluorescence in situ hybridization
  • FISH

The human epidermal growth factor receptor-2 (HER2) protein is a member of the HER family of receptors and has 3 domains: an extracellular ligand-binding site, which receives signals from outside the cell; a short lipophilic transmembrane segment, which transmits signals across the cell membrane; and an intracellular portion with tyrosine-kinase activity, which activates cytoplasmic signal transduction to the nucleus.

HER2 overexpression occurs in approximately 18% to 20% of human breast cancers15 and is associated with a poor prognosis, including aggressive disease and shorter survival.4,5 In addition, HER2 positivity has a strong predictive value and is typically associated with increased response to anthracycline-based chemotherapy and a poor response to hormonal therapy.4,610 Many studies have demonstrated the correlation of HER2 overexpression with amplification of the HER2 gene.1115

Trastuzumab (Herceptin, F. Hoffmann-La Roche, Basel, Switzerland) is a humanized monoclonal antibody that has been specifically developed to target HER2. It has been shown to confer a significant survival benefit in the treatment of women with HER2+ advanced or metastatic breast cancer, as monotherapy16 or in combination with cytotoxic agents.17,18

Recently, the results of 4 clinical trials have demonstrated that the use of trastuzumab in the adjuvant setting is as effective and beneficial as it is in advanced or metastatic disease.1921 Of particular interest is the HERA (HERceptin Adjuvant) trial, which is investigating the use of single-agent trastuzumab over 1 and 2 years in the adjuvant setting compared with observation in patients with HER2+ early breast cancer.20,22 When trastuzumab was administered every 3 weeks for 1 year following adjuvant chemotherapy, disease- and recurrence-free survival were significantly prolonged. Indeed, the risk of disease-free survival events was reduced by 46% and the risk of distant recurrence by 51% in patients receiving trastuzumab.20 After a median follow-up of 2 years, 1 year of adjuvant trastuzumab therapy was shown to have a significant overall survival benefit.22 These important results underline the need for HER2 testing of patients with breast cancer at primary diagnosis.

Accurate determination of HER2 status is key in identifying patients who may benefit from trastuzumab therapy because the likelihood of a response is directly related to the level of HER2 overexpression. Therapeutic benefit from treatment with trastuzumab is seen only in patients with HER2+ disease: patients whose tumors show strong overexpression of HER2, as indicated by an immunohistochemical score of 3+, and/or whose tumors show HER2 gene amplification, as determined by a positive fluorescence in situ hybridization (FISH) result.17,23 Furthermore, the level of HER2 amplification is linked to the level of response to trastuzumab in the neoadjuvant setting.24

Chromogenic in situ hybridization (CISH) and silver-enhanced in situ hybridization (SISH) are new bright-field techniques that have been introduced for the determination of HER2 status.25 This article reviews the recent literature and technical evolutions concerning bright-field in situ hybridization (ISH) testing for HER2 gene amplification, including concordance with immunohistochemical analysis and FISH, and the current recommended scoring systems.

Established HER2 Testing Methods

Various techniques have been used to determine HER2 status in breast cancer samples, including Southern, Northern, and Western blotting; enzyme-linked immunosorbent assay; and polymerase chain reaction. However, only 2 technologies for HER2-status determination are currently validated, Food and Drug Administration–approved,26 and broadly used in the routine diagnostic setting: immunohistochemical analysis, which identifies HER2 protein expression on the cell surface; and FISH, which determines the degree of HER2 gene amplification. Both methods are highly specific and reproducible when performed under standardized and validated conditions.

Immunohistochemical Testing

HER2 expression can be evaluated by a typical immunohistochemical staining pattern of tumor cells and is interpreted by means of a specially devised scoring system. The immunohistochemical test is relatively quick and inexpensive, with a success rate generally higher than ISH, and the results can be viewed using a conventional bright-field microscope. Stained slides do not degrade and can, therefore, be stored for long periods. In addition, immunohistochemical testing allows tumor cell morphologic features to be viewed in parallel. However, immunohistochemical analysis is susceptible to interobserver variability and, as with any assay, requires standardization and validation.

Immunohistochemical analysis is widely used as the primary test for the determination of HER2 status. The most commonly used scoring system regards immunohistochemical 3+ tumors as HER2+ (uniform intense membrane staining of >30% of invasive tumor cells), and patients with such tumors are eligible for treatment with trastuzumab. Tumors with a score of 0 or 1+ (no staining or weak, incomplete membrane staining in any proportion of tumor cells) are negative for HER2 overexpression, and patients with such tumors are not eligible for trastuzumab treatment. A score of 2+ (complete membrane staining that is nonuniform or weak in intensity but with obvious circumferential distribution in ≥10% of cells) is generally considered equivocal, and these samples have to be retested using methods that detect gene amplification (FISH or CISH).26

FISH Testing

FISH is a sensitive and specific method that identifies the number of copies of the HER2 gene, often in conjunction with the number of chromosome 17 centromere (CEP17) copies. This technique uses formalin-fixed, paraffin-embedded tissue and standard 4- to 5-μm sections. It can also be applied to cell blocks or cytologic specimens. Scoring is generally seen as being more objective and quantitative than immunohistochemical scoring, although the reproducibility of results is dependent on the thickness of the tissue sections and interpretation or recognition of the invasive component. DNA is more stable than protein, and preanalytic factors have less impact on test results compared with immunohistochemical analysis. FISH is, however, more time-consuming and expensive than immunohistochemical analysis, requiring the use of a fluorescence microscope and additional training. Furthermore, the fluorescent signals produced by FISH are far less stable than those of immunohistochemical analysis, and decay may occur within a few weeks. Therefore, FISH results must be captured and stored within this time frame. Automated morphometric image analysis of FISH is now possible, and good correlation (r 2 = 0.90) has been shown between HER2/CEP17 ratios calculated from image analysis and those resulting from manual counting.27

Emerging HER2 Testing Methods

Chromogenic In Situ Hybridization

CISH is a modification of the FISH method for the detection of HER2 gene amplification. The assay can be used as a primary test to detect HER2 status or to retest equivocal immunohistochemical results (immunohistochemical score, 2+). There are 2 types of CISH assay: standard CISH, which uses an HER2 probe only; and dual-color CISH (dc-CISH), which uses an HER2 probe and a CEP17 probe. Inclusion of a CEP17 probe in dc-CISH allows the HER2/CEP17 ratio to be calculated, enabling exclusion of chromosome 17 polysomy.

Principle of the CISH Method

The CISH method is based on the ability of labeled nucleic acid probes to hybridize in situ to specific sections of complementary nucleic acid within the sample. As with immunohistochemical analysis and FISH, the CISH assay uses formalin-fixed, paraffin-embedded tissue and sections of 4 to 5 μm thickness. However, the CISH assay detects gene amplification and tissue morphologic features by using a peroxidase enzyme–labeled probe for chromogenic detection instead of a fluorescent dye. Results can, therefore, be evaluated by using a standard bright-field microscope.

Stages of the CISH Assay

The main stages of the CISH assay are as follows: tissue deparaffinization, heat pretreatment and enzyme digestion, denaturation and hybridization of the HER2 probe to the tissue sample, a stringent wash to remove any unbound probe, detection of the HER2 probe, counterstaining, and use of a bright-field microscope to view the samples. The main difficulties encountered are overfixation and underfixation and overdigestion and underdigestion of tissue samples, all of which can lead to erroneous results, including loss of signal.

CISH Scoring

When the CISH method was first introduced, a cutoff point of 6 HER2 gene copies was used to define amplification.25 The supplier of a commercial CISH kit, the SPoT-Light CISH Detection Kit (Zymed Immunoproducts, Invitrogen, Carlsbad, CA), has recently updated this scoring system as follows: no amplification, defined as 5 or fewer copies of the HER2 gene present per nucleus in more than 50% of tumor cells Image 1A ; low amplification, defined as 6 to 10 copies, or small clusters, or a mixture of multiple copies and small clusters of the HER2 gene present per nucleus in more than 50% of cancer cells; and high amplification, defined as more than 10 copies, or large clusters, or a mixture of multiple dots and large clusters of the HER2 gene present per nucleus in more than 50% of cancer cells Image 1B . It should be noted that the American Society of Clinical Oncology/College of American Pathologists (ASCO/CAP) guidelines for assessing HER2 status using a FISH single-probe assay (no CEP17 probe) also recommend a cutoff of 6 or more copies for HER2 positivity.26

With the SPoT-Light CISH Detection Kit, “no amplification” can be further divided into polysomy, defined as 3 to 5 copies of the HER2 gene present per nucleus in more than 50% of tumor cells, and diploid, defined as 1 or 2 copies of the HER2 gene present per nucleus in more than 50% of tumor cells. Borderline samples, or samples with low-level amplification, present the greatest challenge when dealing with FISH and CISH interpretation because interobserver variability can be substantial in these cases.

Dual-Color CISH

dc-CISH is based on the same principle as 2-color FISH, ie, HER2 assessment is reported as an HER2/CEP17 ratio, with a value of 2 or more being indicative of HER2 gene amplification.28 Contrary to standard CISH, 2 separate probes are used: a digoxigenin-labeled HER2 probe (resulting in a green signal), and a biotin-labeled CEP17 probe (resulting in a red signal).28 This allows a HER2/CEP17 ratio to be calculated and excludes chromosome 17 polysomy. Moreover, dc-CISH enables the assessment of HER2 status in conjunction with histopathologic examination and the ease of bright-field microscopy.

Image 1

Chromogenic in situ hybridization analysis showing HER2 gene not amplified (A) and amplified (B).

Silver-Enhanced In Situ Hybridization

A fully automated technique to detect chromogenic signals has recently been introduced by Ventana Medical Systems, Tucson, AZ, using INFORM HER2 SISH technology.29 Automation improves efficiency and consistency, provides high throughput of samples, and decreases the risk of error. Furthermore, automation allows HER2 and CEP17 assays to be performed on contiguous slides.29

SISH is scored similarly to CISH, with no amplification defined as 5 or fewer copies of the HER2 gene present per nucleus in more than 50% of tumor cells Image 2A and high amplification defined as more than 10 copies, or large clusters, or a mixture of multiple dots and large clusters of the HER2 gene present per nucleus in more than 50% of cancer cells Image 2B .

SISH is faster to perform than FISH and requires only a conventional light microscope, making it particularly suited to routine use in pathology laboratories. The development of a new software application (Ventana Medical Systems, Illkirch, France) for use with the Ventana Image Analysis System (VIAS, Ventana, Tucson) facilitates the routine use of SISH in the clinical setting. VIAS produces a digital image of the microscopic field of view and counts HER2 gene copies or CEP17 copies on contiguous slides, as marked by the silver dots of the SISH assay. When the minimum number of selected cells is reached, VIAS calculates the corresponding HER2/CEP17 ratio.

Comparison of Bright-Field ISH With Immunohistochemical Analysis and FISH

Advantages of Bright-Field ISH

Table 1 shows a comparison of immunohistochemical analysis, FISH, and bright-field ISH, listing the advantages and disadvantages of each method. Bright-field ISH offers some of the advantages of immunohistochemical analysis and FISH, with one of the main advantages being the option to evaluate gene copy number and tissue histopathologic features simultaneously. In common with immunohistochemical staining, the bright-field ISH signal does not decay, so slides may be stored for a long period at room temperature, creating a permanent test record. The bright-field ISH method does not require the use of costly equipment, such as a fluorescence microscope, because samples can be viewed using a standard bright-field microscope. In addition, the overall processing costs involved with CISH are less than those involved with FISH.30 Regarding processing, it has been reported that the time required for the CISH and FISH assays is 2 days but that the actual hands-on time is higher for FISH than for CISH.31 One reason is that recognition and localization of the tumor and signal enumeration are much easier with CISH. Furthermore, because the SISH protocol is automated, it can be completed within 6 hours.32

Disadvantages of Bright-Field ISH

Until recently, CISH did not have an intrinsic control for the chromosome 17 copy number and, consequently, no control for chromosome 17 polysomy. (A new, validated dc-CISH kit is now commercially available to address this problem.) Moreover, bright-field ISH techniques are relatively new and have not been broadly established in pathology laboratories. Immunohistochemical analysis is the preferred initial testing method in most pathology laboratories but is susceptible to variations in the testing protocol, antibody selection and dilution, and so on. Thus, careful validation, preferably against FISH as the standard, is required when establishing immunohistochemical analysis as the routine test for HER2 status.

Image 2

Silver-enhanced in situ hybridization analysis showing HER2 gene not amplified (A) and amplified (B).

View this table:
Table 1

Concordance Between Bright-Field ISH and Immunohistochemical Analysis and FISH

Tanner et al25 first reported CISH as an additional method for assessing HER2 gene amplification and compared CISH with immunohistochemical analysis and FISH using archival breast cancer samples. CISH was successful in 157 of 160 attempted samples. Of these, 120 were not amplified by FISH or CISH, 27 were amplified by both methods, and 10 were amplified by FISH but not by CISH, giving a concordance of 94%. The κ coefficient was 0.81 (95% confidence interval, 0.69–0.92), indicating strong correlation between FISH and CISH results. In this study, immunohistochemical analysis was found to be less sensitive but generally in good agreement with FISH and CISH results. The concordance between FISH and CISH was 100% for amplified cases and 92% for nonamplified cases.

Table 2 25,3356 and Table 3 31,3436,4042,44,47,49,5153,5564 show a number of studies assessing CISH as a method for HER2 testing. The concordance between immunohistochemical analysis and CISH,43,65 FISH and CISH,34,57,60,66 and CISH, immunohistochemical analysis, and FISH58,59 has been analyzed. Most of these studies report high rates of concordance (>85%) and, thus, strong agreement of these methods.

Generally, discordance between FISH and CISH results was seen in samples that were judged to have low-level amplification by FISH. This discordance has been attributed to differences in sample materials (eg, using freshly frozen material vs paraffin-embedded tissue),49 discrepancies in the nature of the 2 tests,52,58,66 and differences in scoring strategies.65,66 Whereas the FISH assay results are scored using the HER2/CEP17 ratio (ie, a 2-probe assay) or the HER2 gene copy number (ie, a single-probe assay), the CISH assay in these cases directly assessed the HER2 gene copy number. This may have caused problems when dealing with polysomy or aneusomy of chromosome 17 and led to samples being scored as “low-level amplified” in error. It should be noted that breast carcinomas with low-level HER2 gene amplification are uncommon (~1%–3% of all carcinomas), and interobserver variability is greatest in this category. The recently introduced option of using dc-CISH28 may address this problem in the future.

A study by Bhargava et al31 assessed the new HER2 dc-CISH polymer detection system for clinical testing. The authors were particularly interested in learning whether this new system was effective in tumors with borderline or low-level HER2 amplification. Overall, in defining tumors as amplified or not amplified, concordance between FISH and CISH was 100%. Of 6 borderline cases by FISH, 3 were also borderline by CISH, 2 were low-level amplified, and 1 was high-level amplified. Of 16 low-level amplified cases by FISH, 6 were also low-level amplified by CISH, 1 was borderline, and the remaining 9 were high-level amplified by CISH. When the CISH assay was repeated using the conventional single-probe kit (not adjusting for chromosome 17 polysomy), discordant interpretations were observed in 2.9% of cases. A number of other studies have also used this method.57,60,67 A recent evaluation of 4 cell lines and 40 invasive breast cancer samples by FISH and dc-CISH demonstrated high concordance between these methods (91%; κ coefficient = 0.82), indicating that dc-CISH is a viable alternative to FISH.28 Although the level of concordance in this study28 did not reach 95%, FISH and dc-CISH were 96% concordant in 30 breast cancer samples assayed for HER2 and TOP2A gene amplification.68

View this table:
Table 2

High concordance has also been demonstrated between SISH and FISH in 100 invasive breast carcinoma cases (98% and 94%, respectively)69 and in 292 primary breast carcinomas.32 In addition, 5 pathologists evaluating 99 biopsy samples of invasive breast cancer using immunohistochemical analysis, FISH, and SISH70 observed high concordance (96%) between FISH and SISH. These results comply with the ASCO/CAP requirement of more than 95% concordance of amplified/nonamplified samples.26 Interobserver concordance was high (92.8%–95.2%), demonstrating that SISH is relatively easy to interpret and is equally as reliable as FISH.

View this table:
Table 3

Interlaboratory and Interobserver CISH Concordance

Interlaboratory concordance of CISH has also been studied,31,52,60,65 as has interobserver variability of assessing CISH results.66 κ coefficients are generally used as a measure of the agreement between 2 methods, a value of 1 implying perfect agreement. Bilous et al36 reported good concordance between interlaboratory CISH results. In this study, the κ coefficient reported was 0.67, indicating good agreement between complete sets of results. In common with other studies, some discrepancies were noted for samples defined as having low-level amplification by FISH. van de Vijver et al51 recently reported high interlaboratory CISH concordance from their ring study using 5 international pathology laboratories, with 95% concordance for nonamplified cases and 92% concordance for amplified cases between laboratories. In a study to assess interobserver variability of CISH results, 80 cases of invasive breast carcinoma were evaluated by FISH and CISH.66 CISH results were independently evaluated by 3 pathologists. CISH scoring was highly reproducible among the 3 pathologists, resulting in a concordance of 91%.

CISH Validation

Accurate HER2 status determination by immunohistochemical analysis, FISH, or CISH is essential for the identification of patients eligible for trastuzumab therapy. When methods are standardized, immunohistochemical analysis/CISH and FISH/CISH demonstrate a high level of concordance, typically 90% to 100%.

Factors that can lead to inaccuracies in HER2 testing by all methods include inadequate storage of tissue samples, variation in the fixation protocol, processing of the assays without using controls, and lack of formal training and education. Factors that particularly influence the performance of CISH are variations in heat pretreatment and enzyme digestion.

If CISH is to be implemented as part of routine diagnostic testing, the results achieved should be compared with one of the following: immunohistochemical analysis or FISH results during the setup phase because these are the most commonly used methods; results achieved by other successful protocols; or external controls with known HER2 status, such as cell lines or tissue samples. Protocols without validation should never be used.

Different initiatives can be implemented to ensure quality assurance compliance. CISH testing results obtained in one laboratory can be confirmed by sending the same samples to a reference center or by retesting a certain number in house with a separate method (ie, immunohistochemical analysis or FISH). Comprehensive training and education in practical interpretation should also be undertaken.

CISH and Prognosis

A growing number of studies have examined the prognostic value of using CISH to determine HER2 status.39,47,71 In 173 patients with primary breast cancer, significant correlation was observed between survival analysis results assessed by CISH and immunohistochemical analysis, suggesting that HER2 amplification detected by CISH could be used as a prognostic factor in breast cancer.39 Furthermore, it has been suggested that CISH may provide more prognostic information compared with immunohistochemical analysis.71 In a study of 107 patients with metastatic breast cancer, it was shown that, despite significant correlation between immunohistochemical and CISH results, there was a difference in prognostic values of compared methods during the course of metastatic disease.71 There was a significant difference in the progression-free interval between HER2– and HER2+ cases as determined by CISH but no significant difference in patients whose samples were tested with immunohistochemical analysis. These findings were reflected in an analysis of HER2 amplification detection methods in endometrial carcinomas, indicating that CISH may be a more informative method than immunohistochemical analysis regarding the prognostic value of HER2 status.47

CISH and Response Prediction

Hofmann et al59 recently demonstrated that CISH results were very sensitive in predicting response to trastuzumab therapy, highlighting this method as a good alternative to FISH for the assessment of HER2 gene amplification. The HER2 status of metastatic patients with breast cancer enrolled in a first-line, phase 2, open-label, 3-weekly trastuzumab monotherapy trial (WO16229) was tested by immunohistochemical analysis, FISH, and CISH.42,72 Results from 86 patients were compared, and 71 were found to be HER2+ by all 3 techniques (82.6% concordance). There was 90% concordance between CISH and FISH (κ = 0.66) and 87% concordance between CISH and immunohistochemical analysis (κ = 0.64). Concordance between FISH and immunohistochemical analysis (88%) was almost identical to that observed between CISH and immunohistochemical analysis. Clinical response was divided into 4 response categories: progressive disease, stable disease, partial response, and complete response. Patients with a partial or complete response were classified as responders, whereas patients with progressive or stable disease were classified as nonresponders. Nineteen patients responded to trastuzumab therapy and all had 3+ immunohistochemical scores. Seventeen of these patients were CISH+ and 16 were FISH+, indicating that immunohistochemical analysis appeared to be the most sensitive of the 3 tests in this study, CISH being in second place with 89.5% sensitivity.

CISH and the HER2 Testing Algorithm

An HER2 testing algorithm has been proposed and is recommended in the majority of national and international HER2 testing guidelines.65 Specimens with equivocal 2+ immunohistochemical scores must be reassessed for amplification of the HER2 gene. In the past, this has almost always been done by FISH. More recently, CISH has been proposed as an addition to the HER2 testing algorithm,58,69 and current opinion advocates retesting specimens with equivocal 2+ immunohistochemical scores with FISH or bright-field ISH Figure 1 .


Bright-field ISH has become an accepted method for HER2 testing. Numerous studies have shown that CISH is an effective test for the determination of HER2 status, with generally excellent concordance with immunohistochemical analysis and FISH. High concordance of interlaboratory CISH results has also been demonstrated. In addition, there is growing evidence that SISH is as reliable as CISH and FISH for determining HER2 status. Bright-field ISH assays are easy to use and less expensive than the FISH assay because they require only a standard bright-field microscope. Stained samples do not decay over time, meaning that they can be stored at room temperature. Bright-field ISH is, therefore, a viable alternative to FISH in laboratories in which FISH testing is not, or cannot be, established. An updated HER2-testing algorithm has been proposed to include bright-field ISH for the retesting of specimens with equivocal 2+ immunohistochemical scores.

Figure 1

The recommended human epidermal growth factor receptor 2-testing algorithm. Adapted with permission from Hanna and Kwok.58 FISH, fluorescence in situ hybridization; IHC, immunohistochemical analysis; ISH, in situ hybridization.


We thank Karen Runcie for editorial support, and Katherine Wilson for medical writing support during the preparation of this manuscript.


  • F. Hoffmann-La Roche, Basel, Switzerland, provided funding for the preparation of this manuscript.

  • Dr Penault-Llorca is a member of the Roche pathology advisory board, for which she receives an honorarium, and she has received research grants from Roche and GlaxoSmithKline and free immunohistochemistry or in situ hybridization kits from Abbott, DAKO, Zymed, Ventana, and Cliniscience; additionally, Ventana is one of the immunohistochemistry vendors for anatomic pathology at her institution. Dr Bilous is a member of the advisory boards for Roche in Australia and Europe, for which he receives an honorarium. Dr Dowsett has received honoraria from Roche for participation on advisory boards and giving lectures and has received grant funding for his department. Dr Hanna is a member of the Roche pathology advisory board, for which she receives an honorarium, and Ventana is one of the immunohistochemistry vendors for anatomic pathology at her institution. Dr Osamura is a member of the Roche pathology advisory board, for which he receives an honorarium, and Ventana is one of the immunohistochemistry vendors for anatomic pathology at his institution. Dr Rüschoff is a member of the Roche HERA international pathology advisory board, for which he receives an honorarium, and he has received invitations to give oral presentations for Ventana in 2009. Dr van de Vijver is a member of the Roche pathology advisory board and has given oral presentations for Roche and has received honoraria for these activities.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.
  62. 62.
  63. 63.
  64. 64.
  65. 65.
  66. 66.
  67. 67.
  68. 68.
  69. 69.
  70. 70.
  71. 71.
  72. 72.
View Abstract