Pathology (February 2014) 46(2), pp. 113–120
SOFT TISSUE PATHOLOGY
Grading of soft tissue sarcomas: from histological to molecular assessment AGNES NEUVILLE1,2, FRE´DE´RIC CHIBON1,3
AND
JEAN-MICHEL COINDRE1,2
1Department of Biopathology, Institut Bergonie´ and INSERM U916, 2Laboratory of Pathology, University Victor Se´galen Bordeaux, and 3Translational Research, Institut Bergonie´, Bordeaux, France
Summary Several histological grading systems for soft tissue sarcomas have been described since the early 1980s. Their main objective is to select patients for adjuvant chemotherapy. Two histological grading systems are used in daily practice, the National Cancer Institute (NCI) and the French Federation of Cancer Centers Sarcoma Group (FNCLCC) systems. They have been devised by combining histological parameters: number of mitoses per high-power field, the presence of necrosis, cellular and nuclear morphology and the degree of cellularity for the NCI grading; and tumour differentiation, mitotic index and extent of necrosis for the French system. Histological grading is far more appropriate to assess the risk of metastasis. However, several limitations prevent its use: grade cannot be applied to all histological types, its reproducibility is not perfect, a three-grade system generates an intermediate grade with undetermined prognosis, and finally the core needle biopsy, now widely used for the diagnosis of soft tissue sarcoma, is not the best sample to assess the grade. The development of molecular grading in addition to histological grading probably represents the next step. Molecular signatures based on quantitative evaluation of chromosomal complexity such as CINSARC (complexity index in sarcomas) appear as a strong independent predictive factor for metastasis in several types of sarcoma, and even in several other types of cancer. When they can be instituted in daily practice on formalin fixed, paraffin embedded material, molecular signatures will not only provide information on risk of metastasis, but also better understanding of cancer development, response or resistance to evaluated drugs, and potential targets for future treatments. Key words: Grading, molecular signature, sarcoma. Received 7 October, revised 5 November, accepted 11 November 2013
INTRODUCTION Soft tissue sarcomas (STS) represent a heterogeneous group of rare malignant tumours with a wide spectrum in terms of histological type and prognosis.1 Prognosis of STS is dominated by local recurrence and distant metastasis. STS recur locally in less than 10% when located in the limbs and trunk wall but metastasise in 30–50% of cases. Quality of surgical margins is the most important factor for predicting local recurrence, whereas histological grade particularly indicates the probability of distant metastasis and overall survival.2–5 For more than 20 years, the main use of grading has been the selection of patients for chemotherapy. However, the utility of Print ISSN 0031-3025/Online ISSN 1465-3931 DOI: 10.1097/PAT.0000000000000048
#
adjuvant chemotherapy in STS is still debated and the approach of sarcoma therapy has recently undergone shifts from nonspecific cytotoxic agents towards utilisation of molecularly targeted treatments. Moreover, some histotypes have shown different sensitivity to specific cytotoxic drugs. Nowadays, the value of grading is also limited by the universal use of core needle biopsies. In fact, histological grade can be considered as a morphological translation of molecular events that determine tumour aggressiveness; therefore, it should be regarded as a transient practical method that should prompt research in order to establish a definitive system based on molecular parameters. For more than 10 years, biologists and pathologists of the French Sarcoma Group have been working on molecular analyses of adult sarcomas with complex genomics in order to determine which genes and pathways are related to chromosomal complexity and whether or not a link between chromosomal complexity and prognosis exists. This work has led to the description of the CINSARC (Complexity INdex in SARComas) expression profile signature, offering much better prognostic value than histological grading.6
HISTOLOGICAL GRADING Concept of grading The concept of histological grade was introduced by Broders in 1920 and must be clearly distinguished from staging and nomograms. Whereas grading is based on the intrinsic quality of the primary tumour only, staging also takes into account tumour extent. Nomograms assess multiple clinical and histological parameters to calculate the probability of recurrence for a given patient. Histological grade is based on the histological qualities of the primary tumour and is expected to predict tumour aggressiveness. In 1977, Russell et al.7 proposed the first coherent clinicopathological classification separating patients into four stages with a different prognosis and introduced a histological grading applicable to all adult STS. In their system, grading was the most important prognostic factor. However, the grading system they used was rather subjectively based on several histological parameters with no clear definition. To obtain the most efficient system, histological parameters should be clearly defined and selected by multivariate analysis so that only the necessary parameters summarising all prognostic histological information are used. According to the studies that have followed this approach, the best parameters are tumour histotype and subtype and/or differentiation, tumour necrosis, mitotic index8–10 and, as stated in a few, vascular invasion.11 In the early 1980s, several
2014 Royal College of Pathologists of Australasia
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
114
NEUVILLE et al.
grading systems were reported and a critical review of those most important was performed by Oliveira and Nascimento.12 The two systems most commonly used are the National Cancer Institute (NCI) grading described by Costa et al.9 and the French Federation of Cancer Centers Sarcoma Group (FNCLCC) grading described by Trojani et al.8 Description of grading systems The NCI system, published in 1984,9 was based on an analysis of 163 patients and assessment of six histological parameters (histological type, mitoses, necrosis, pleomorphism, cellularity and intercellular matrix). Costa et al. found by multivariate analysis that gross and microscopic tumour necrosis had the greatest impact. The final form of this system uses a group of predefined grade 1 and grade 3 sarcomas and, in tumours not automatically categorised, the amount of necrosis is used to distinguish grade 2 from grade 3 tumours, with a cut-off point of 15% for the extent of necrosis. Shortly after, the French Sarcoma Group proposed a system based on the analysis of 155 patients with assessment of seven histological parameters (tumour differentiation, cellularity, nuclear atypia, presence of malignant giant cells, mitotic count, extent of tumour necrosis and presence of vascular emboli) and multivariate analyses that identified three independent prognostic factors: tumour differentiation, mitotic index and extent of necrosis.8 These parameters are scored 1–3 for differentiation and mitotic index and 0–2 for necrosis. A three-grade system is obtained by summing the scores obtained for each of these three parameters (Table 1). Grade 1 is defined as a total of 2 or 3; grade 2 as a total of 4 or 5; and grade 3 as a total of 6–8. To enhance the reproducibility of the system, the parameters are defined as exactly as possible. The rules for using this Table 1
Definition of parameters in the French grading system*
Pathology (2014), 46(2), February
system and its limitations have been previously described in detail.13 Differentiation is the most controversial parameter and is in fact a mixture of histological type and subtype and/or true differentiation. Table 2 shows the differentiation scores attributed to common sarcoma types. A score of 1 is currently assigned to sarcomas closely resembling normal adult tissue to such a degree as to be confused with benign tumours, such as a well-differentiated leiomyosarcoma. A score of 3 is given to embryonal and poorly-differentiated sarcomas, sarcomas of doubtful histological type, synovial sarcoma, primitive neuroectodermal tumour, osteosarcoma, pleomorphic rhabdomyosarcoma, and pleomorphic liposarcoma. Others of certain histological type such as myxoid liposarcoma are scored 2. Mitotic count is obtained in 10 successive high power fields (HPF) in most mitotic areas. This count is taken to establish the score: 1, 0–9 mitoses; 2, 10–19 mitoses; 3, more than 19 mitoses per 10 HPF. Strict rules should be followed for counting mitoses: early and adequate fixation of tumour, correct macroscopic sampling (one section per 1 or 2 cm of the tumour diameter), good quality sections, and adequate selection of the most mitotic areas at medium power and where mitoses should be counted. A calibrated HPF surface should be performed according to the first study (HPF ¼ 0.1734 mm2).8 Moreover, pathologists should take their time, as counting mitoses hurriedly might dwindle reproducibility. If the mitotic count is close to the cut-off score (i.e., 7–9 or 17–19 mitoses per 10 HPF), recounting is strongly advised. Any necrotic or hypocellular areas as well as zones of ulceration should be avoided. Tumour necrosis should be evaluated at histological and macroscopic levels. However, macroscopic evaluation of necrosis is poorly reproducible and should always be checked under the microscope. Necrosis related to previous surgery or to ulceration must not be taken into account, and nor should haemorrhage or hyalinisation. The score is 0 when there is no necrosis, 1 when the necrotic area is less than 50% and 2 when it is more than 50% of the tumour surface.
Definition of parameters Tumour differentiation (see Table 2) Score 1: Sarcomas closely resembling normal adult mesenchymal tissue and potentially difficult to distinguish from the counterpart benign tumour (e.g., well-differentiated liposarcoma, well-differentiated leiomyosarcoma) Score 2: Sarcomas for which histological typing is certain (e.g., myxoid liposarcoma, myxofibrosarcoma) Score 3: Embryonal and undifferentiated sarcomas, synovial sarcomas, sarcomas of doubtful type
Table 2 Tumour differentiation score according to histological type in the French grading system* Histological type
Grade 1: Total score 2, 3 Grade 2: Total score 4, 5 Grade 3: Total score 6, 7, 8
Well-differentiated liposarcoma Well-differentiated leiomyosarcoma Malignant neurofibroma Myxoid liposarcoma Conventional leiomyosarcoma Conventional MPNST Myxofibrosarcoma Myxoid chondrosarcoma Pleomorphic liposarcoma Dedifferentiated liposarcoma Rhabdomyosarcoma Poorly-differentiated/pleomorphic leiomyosarcoma Poorly-differentiated MPNST Malignant Triton tumour Synovial sarcoma Extraskeletal osteosarcoma Extraskeletal PNET Mesenchymal chondrosarcoma Epithelioid sarcoma Malignant rhabdoid tumour Undifferentiated (spindle cell and pleomorphic) sarcoma
*
*
Mitotic count (established on 10 HPF){ Score 1: 0–9 mitoses per 10 HPF Score 2: 10–19 mitoses per 10 HPF Score 3: More than 19 mitoses per 10 HPF Tumour necrosis Score 0: No necrosis Score 1: Less than 50% of tumour necrosis Score 2: 50% or more than 50% of tumour necrosis Histological grade
Modified from Trojani et al.8 A HPF measures 0.1734 mm2. HPF, high-power field. {
Differentiation score 1 1 1 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3
Modified from Guillou et al.5 MPNST, malignant peripheral nerve sheath tumour; PNET, primitive neuroectodermal tumour.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
GRADING IN SARCOMAS
In accordance with the College of American Pathologists14 and the American Joint Committee on Cancer (AJCC) recommendations,15 this system is preferred over the NCI system, at least in adults. When comparing the performances of these two systems in the same population of patients with STS, the French system performed better.5 It allocated fewer patients to the noninformative grade 2 category and appeared more discriminating and efficient in the selection of patients with tumours of high malignant potential that could benefit from adjuvant chemotherapy. Discrepancies between the two grading systems were nevertheless observed in 34.6% of the cases. A study by soft tissue pathologists from 30 countries showed that the French grading was more widely used than the other systems:16 37.3% for the French system, 24% for the NCI system, 12% for the Broders system and 1.3% for the Markhede system, with 25.3% unspecified. However, the majority of responders (74%) were from Europe where the French system predominates. The main weakness of the French system lies in the attribution of a differentiation score, particularly for tumours with no normal tissue counterpart such as undifferentiated pleomorphic sarcoma.
115
adjuvant chemotherapy: patients with a grade 1 STS should certainly not receive adjuvant chemotherapy since it will be inefficient. On the other hand, a patient with a grade 3 sarcoma is a good candidate for receiving adjuvant chemotherapy due to the high probability of distant metastases. Moreover, it has been shown that adjuvant chemotherapy may be more efficient for patients with a grade 3 sarcoma than for patients with a grade 2 sarcoma. However, the utility of adjuvant chemotherapy in STS is still debated. Limitations of histological grading systems Grading systems have been subject to numerous criticisms.12,18,19 Some object to the fact that all grading systems have been developed and tested in the overall sarcoma group and not specifically in every histological category. However, in a study on 1240 localised STS,17 we showed that, with the exception of malignant peripheral nerve sheath tumour (MPNST) and rhabdomyosarcoma, histological grade remains an important predictor of metastasis in the main histological types of STS (90% of all STS). It is now widely accepted that no current grading system performs well for every type of STS. However, given the number of histological types of STS, it is unrealistic and pointless to develop a grading system for every specific histological type of STS. For the time being, the French or the NCI systems perform correctly for the most common sarcoma types and represent an acceptable alternative. Yet, we should keep in mind that grading is less informative than histological type in dedifferentiated and round cell liposarcomas, alveolar soft part sarcoma, clear cell sarcoma, epithelioid sarcoma, and primitive neuroectodermal tumour (PNET), and that it should not be used on tumours of ‘intermediate malignancy’ such as atypical fibroxanthoma and dermatofibrosarcoma protuberans. Among poorly differentiated sarcomas, it is also important to differentiate leiomyosarcomas from undifferentiated sarcomas since the former have a worse prognosis than the latter.20,21 Moreover, since some subtypes show a differential sensitivity to specific drugs, it is now more and more important to obtain the exact histotype.
Value of grading It is now recognised that histological grade is the most important prognostic factor for adult STS. As the best predictor of metastasis development and tumour mortality, histological grade is a key parameter of the currently used 2010 TNM clinicopathological staging system.15 A grading system allows us to indicate the probability of distant metastases and overall survival. In every study with multivariate analyses testing histological grade and clinical factors in STS, grade is the most important prognostic factor. In a study of 1240 patients with a localised STS, the 5 year metastasis-free survival rate was 91% for grade 1, 71% for grade 2 and 43% for grade 3 (Fig. 1).17 Since grading permits prediction of metastasis, and because the response rate to palliative chemotherapy is better in patients with high grade sarcoma, we use grading to select patients for
SOFT TISSUE SARCOMAS: French sarcoma group (n = 1240) M E T A S T A S I S F R E E S U R V I V A L
1
Grade 1 (n = 157)
0.9 0.8
Grade 2 (n = 511)
0.7 0.6 0.5
Grade 3 (n = 572) 0.4 0.3 0.2
p < 10–29
0.1 1
1
2
3
4
5
6
7
8
9
10
11
12
13
Years Fig. 1 Metastasis-free survival curves for French histological grade in a cohort of 1240 adult patients with a localised soft tissue sarcoma.17
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
116
NEUVILLE et al.
Another weakness of any grading system lies in its reproducibility. When the reproducibility of the French grading system was tested by 15 pathologists on 25 cases, the crude proportion in agreement was 75% for tumour grade, 73% for mitotic index, 74% for differentiation, and 81% for tumour necrosis.22 A reproducibility of only 75% is difficult to accept by clinicians and patients. Reproducibility of diagnosis of histological type was significantly lower with an agreement rate of 61%. However, this reproducibility study was conducted in the early 1980s when immunohistochemistry was not very efficient and molecular analysis was not available. Nowadays, at least among pathologists specialised in soft tissue tumours, the reproducibility of histotyping can be considered as excellent. A further major limitation of almost every grading system is the existence of an intermediate grade which corresponds to undetermined prognosis. This category represents 46% and 58% of adult patients with a STS assessed with the French and the NCI grading systems, respectively,5 meaning that histological grading is useful for only 50% of STS patients. Conventional grading systems have been developed to be used on representative material, i.e., on whole untreated tumours or on surgical biopsies. Core needle biopsies are now widely used for the diagnosis of primary STS and for establishing treatment strategies. Most studies show that core needle biopsies are highly accurate in the diagnosis of malignancies (95%)23 and that sarcoma typing and grading can also be performed with an acceptable degree of accuracy (70–80%).24 But determination of histological grade can essentially only be done with accuracy for high grade (G3) neoplasms. Determination of grade in the intermediate (G2) and low grade (G1) categories is more problematic. It should also be emphasised that core needle biopsies should not be evaluated without a good knowledge of clinical and radiological data. Imaging procedures often provide useful information about the nature of the lump, especially regarding tumour size, borders, and amount of necrosis.
MOLECULAR GRADING Molecular markers hold great promise for refining our ability to establish early prognosis and to predict response to therapy. Identification of molecular prognostic markers in sarcoma started in the early 2000s, with the study reported by Wu¨rl et al.25 The authors showed that expression of the two genes TERT (telomerase reverse transcriptase) and survivin (also named BIRC5) was associated with poor outcome in a series of 89 sarcomas of different subtypes. This series was composed of sarcomas with complex genetics such as leiomyosarcoma, pleomorphic rhabdomyosarcoma and liposarcoma, and in these tumours co-expression of both genes was a significant prognostic factor. Survivin belongs to the family of genes that inhibit apoptosis and is also involved in chromosome segregation. TERT is involved in the immortal status of cells by maintaining telomere size. Even if this two-gene signature has never been validated in an independent group, it has biological meaning that links metastatic outcome with genes involved in chromosome structure and segregation. With the development of microarray technology, approaches to identify molecular signatures have changed and several studies have reported molecular profiling analyses in sarcomas,26 but only a few have been validated on independent groups of patients. Lee and colleagues27 reported on one of the first expression profiling studies leading to a prognostic
Pathology (2014), 46(2), February
signature in sarcomas. Comparing the expression profiles from both metastatic and non-metastatic leiomyosarcoma, they identified 335 genes differentially expressed between primary tumours (20 cases) and metastases (7 cases). Although this signature has significant prognostic value, the study suffered from two main limitations: the small size of the series and the lack of validation of the results, compromising its clinical use. Francis and colleagues28 analysed 177 sarcomas and produced a meaningful expression signature of different histotypes, demonstrating that histological classification fits well with expression profiles of tumours. They also identified 220 genes associated with metastasis events in a subgroup of 89 pleomorphic sarcomas. Both studies developed supervised approaches with the metastatic event as an end-point and selected genes that, on the whole, have no biological links between them. In fact, such an approach has proven to be inefficient for identifying genes or pathways implicated in the metastatic potential acquisition of metastases. Recently, the French Sarcoma Group published the identification of a 67-gene expression signature called CINSARC, satisfying all the criteria to be a clinically applicable prognostic molecular marker.6 With the aim of establishing a possible link between chromosomal complexity and prognosis, this group studied a large series of sarcomas with complex genomics, (i.e., leimyosarcoma, undifferentiated pleomorphic sarcoma, myxofibrosarcoma, pleomorphic liposarcoma, pleomorphic rhabdomyosarcoma and dedifferentiated liposarcoma) by two combined techniques, array-CGH and expression profile. The first step was to assess chromosomal complexity by array-CGH and three types of recurrent profiles were identified: tumours with a simple amplicon profile based on a coamplification corresponding to dedifferentiated liposarcomas, tumours with a few alterations (less than 30) involving full chromosome arms or entire chromosomal gain or loss (‘arm profile’), and tumours with a high level of chromosomal complexity with more than 30 alterations (‘rearranged profile’). Although the metastasisfree survival was not significantly different between the latter two groups of patients (arm and rearranged profiles), there was a positive correlation between the number of genomic alterations and histological grade (79% of grade 1 or 2 sarcomas have less than 30 chromosomal alterations and 74% of grade 3 sarcomas have more than 30 chromosomal alterations, p ¼ 0.001). Furthermore, since histological grade is a marker of tumour aggressiveness, it has been postulated that the latter is related to genomic complexity. A CIN (chromosomal instability) signature established by Carter et al.29 was also tested. This CIN signature has been set up by a computational method for assessing chromosome instability from expression profiles. Based on the expression profile of 70 genes, it has been shown to predict survival in several types of cancers, but not in the studied series of sarcomas. Data from CGH, FNCLCC grade and Carter’s signature were finally integrated together with, in a first step, three t-tests to compare the expression profiles of tumours according to three characteristics: array-CGH profile (35 imbalances), histological grade (grade 3 versus grade 2) and the 70-gene Carter CIN signature. This approach resulted in the selection of 86 genes for array-CGH, 73 genes for grade and 39 genes for the Carter signature. In a second step, a gene ontology analysis was performed in order to identify the underlying pathways. Only genes involved in the most significantly overrepresented pathways were selected: 37 genes for array-CGH, 18 genes for grade and 22 genes for Carter signature. The final gene set called CINSARC contains
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
GRADING IN SARCOMAS
A – Cohort 1
A – Cohort 2 1.0
1.0
n = 42
n = 83
0.8
0.8
0.6
Probability
Probability
117
n = 100
0.4
n = 85
0.4
n = 127 p = 5 × 10–4
0.2
n = 183 p = 1 × 10–7
0.2
0.6
0.0
0.0 0
12
24
36
48
60
72
84
96
108
0
120
12
24
36
48
60
72
84
96
108
120
Time in months
Time in months
B – Cohort 1
B – Cohort 2
1.0
1.0
0.8
0.8
Grade 1 + 2
0.4
Probability
Probability
Grade 1 + 2 0.6
Grade 3 n = 171 p = 0.43
0.2
12
24
0.4
Grade 3 n = 117 p = 4 × 10–3
0.2
0.0 0
0.6
36
48
60
72
84
96
108
0.0
120
0
Time in months
12
24
36
48
60
72
84
96
108
120
Time in months
Fig. 2 Metastasis-free survival curves according to (A) CINSARC molecular grading and (B) French histological grade in two independent groups of adult patients with a soft tissue sarcoma (cohort 1, training group; cohort 2, validation group).
67 genes. These genes belong to chromosome management and mitosis control pathways and are related to genome complexity. Some of these genes have already been reported in prognostic signatures in breast carcinomas, where they are usually classified as proliferation genes, but they are also clearly involved in genome complexity and, in fact, they are related more to chromosome instability than to proliferation. This CINSARC signature was set up on a training group of 183 non-translocated adult soft tissue sarcomas and validated on an independent cohort of 127 tumours (Fig. 2). For all these patients, clinical, histological, treatment and follow-up data were available in a shared tumour bank (https://conticabase. sarcomabcb.org/). Both array-CGH and expression profiling were performed on untreated primary tumours. CINSARC grade is an independent predictive factor for metastasis (in
both training and validation cohorts) for leimyosarcoma and undifferentiated pleomorphic sarcoma (Fig. 3) as well as among tumours of the same histological grade (Fig. 4). CINSARC grade was also of prognostic value in breast carcinoma, lymphoma and GIST when applied to series of such tumours with available expression profile and follow-up data. A subsequent study conducted on frozen tissue from 67 primary untreated GISTs30 showed that CINSARC is a strong predictor of prognosis. The gene with expression most strongly associated with recurrence was AURKA, but the AURKA gene was not amplified. Instead, deletions of p16 (CDKN2A) and/or retinoblastoma (RB1) genes were identified and were likely causal events leading to increased AURKA and CINSARC gene expression, to chromosome rearrangement, and ultimately to aggressiveness. Different studies have shown that 9p21 deletion
B – Undifferentiated sarcomas
A – Leiomyosarcomas 1.0
1.0
n = 32 n = 16
0.8
Probability
Probability
0.8
0.6
Cohort 1 n = 52 p = 2.7 × 10–4
0.4
0.2
0.6
0.4
n = 36 0.0 0
12
24
36
48
60
72
Time in months
84
96
108
120
n = 39
Cohort 1 n = 71 p = 4.69 × 10–4
0.2
0.0 0
12
24
36
48
60
72
84
96
108
120
Time in months
Fig. 3 Metastasis-free survival curves according to CINSARC molecular grading in (A) leiomyosarcomas and (B) undifferentiated sarcomas.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
118
NEUVILLE et al.
Pathology (2014), 46(2), February
A – French grade 1–2
B – French grade 3
1.0
1.0
CINSARC 1
CINSARC 1 0.8
Probability
Probability
0.8
0.6
CINSARC 2
0.4
n = 47 p = 0.009
0.2
0.0
0.6
CINSARC 2
0.4
n = 102 p = 7.6 × 10–6
0.2
0.0 0
12
24
36
48
60
72
84
96
108
Time in months
120
0
12
24
36
48
60
72
84
96
108
120
Time in months
Fig. 4 Metastasis-free survival curves according to CINSARC molecular grading in the French histological grade 1–2 patients and grade 3 patients.
is associated with metastatic outcome in patients with GIST. Nevertheless, to date it is still controversial to attribute this prognostic value specifically to one of the three genes contained in the minimal deleted chromosomal area, i.e. CDKN2A, CDKN2B and MTAP.31 Genomic (and not expression) markers can be considered as the best form of molecular criteria because they presumably drive outcomes. Additionally, their identification in daily practice is relatively easy with CGH-array, FISH or sequencing technologies, which are widely used in pathology. Since frozen tissue is rarely available in GISTs, a genomic index (GI) integrating the number and type of DNA copy number alterations that can be performed on formalin fixed, paraffin embedded tissue was established through array-CGH. This GI is a prognostic factor in GISTs as strong as CINSARC and could be used for identifying poor prognosis patients in the group classified as intermediate risk by the AFIP classification. CINSARC and GI were also tested on frozen samples from 100 primary untreated synovial sarcomas (SS).32 Both CINSARC and GI have independent and validated prognostic values and metastasis development in SS is strongly associated with chromosome complexity. The study from Lagarde et al. offers a biological explanation for the roughly opposite outcomes of paediatric and adult SS patients. Results show that metastatic outcomes are strongly associated with chromosomal complexity in both age strata and that this instability is very frequent in adult but not in paediatric SS. Given that the initial genetic driver event is the same in both groups [the t(X;18) translocation], this suggests that an independent, still unknown, mechanism is permissive to chromosome instability in adult cases and resistant in paediatric cases. No genomic alteration or significantly differentially expressed gene indicative of such mechanisms was identified. Even if this discrepancy could be due to somatic events not detectable using array-CGH, one wonders why it occurs in adults and not in paediatric SS. We hypothesised that the chromosomal complexity could be more likely related to the possible patient predilection, age, or genome ageing. GI is potentially the best overall prognostic biomarker and the most clinically relevant, considering that array-CGH on formalin fixed samples is already used in pathology. The CINSARC genes are involved in the same pathways (BIRC5 belongs to CINSARC signature) that were first reported by Wu¨rl et al.,25 i.e., chromosome integrity, segregation and mitosis. We could speculate from this observation that these mechanisms are among the most important to predict outcome in sarcomas and that the molecular mechanisms leading to distant metastases might be related to the capacity of tumour cells to induce chromosome instability and complexity. It could also suggest that the more rearranged a
genome, the higher the probability of obtaining a gene expression profile permitting cells to complete the process of allowing dissemination and metastasis development. Finally, CINSARC signature has been retrospectively validated in the three types of sarcomas, according to genetics, i.e., complex (leiomyosarcomas, undifferentiated pleomorphic sarcomas, etc), activating mutations (GISTs) and translocation related (SSX). The next step is prospective validation of CINSARC in a clinical trial.
HISTOLOGICAL OR MOLECULAR GRADING? THE WAY AHEAD Histological grading is known as the most important prognostic factor of adult STS considered as a single entity. It is a cheap, quick and easy method for identifying sarcomas with high metastatic potential, and should be used in daily practice for most soft tissue sarcomas. However, its limitations should be kept in mind: no current grading system performs well for every type of sarcoma, and grading is less informative than histotype for some of them. Furthermore, its moderate reproducibility and the existence of an intermediate grade representing almost half of cases and corresponding to undetermined prognosis, should not be forgotten. The current universal use of core needle biopsies is also a limitation for grading, which should certainly be adapted to this type of material and complemented by other parameters such as imaging for a better assessment of tumour necrosis and proliferation index evaluated by Ki-67 or MIB-1 rather than the mitotic index.33 Molecular grading based on quantitative evaluation of chromosomal complexity, such as the CINSARC signature or the GI, has shown better prognostic value as compared to histological grading in several categories of sarcomas: those with complex genomics such as leiomyosarcoma and undifferentiated pleomorphic sarcoma, sarcomas with a reciprocal translocation such as synovial sarcoma, and GIST. However, molecular grading should now be validated in prospective independent series and on other sarcoma types. Molecular grading should also be refined by looking for other categories of genes involved in the metastatic process and by defining driver genes responsible for chromosomal instability. This molecular grading should also be adapted to daily practice. Expression of a set of genes is already in current use in breast carcinoma. MammaPrint, based on the Amsterdam 70-gene breast cancer signature,34 has been cleared by the US Food and Drug Administration (FDA) for use in the USA for lymph node negative breast carcinoma patients under 61 years with tumours of less than 5 cm. However, for more universal use, these molecular signatures need to be adapted to formalin
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
GRADING IN SARCOMAS
fixed, paraffin embedded tissue from core needle biopsies. The potential lack of robustness and reproducibility of the chips was long cited as a limitation of their clinical application, but we now know this not to be the case thanks to the efforts of the international MAQC (MicroArray Quality Control) consortium. The MAQC project involved 137 participants from 51 academic institutions and industrial partners to systematically address the technical reproducibility of microarray measurements within and between laboratories, as well as across different microarray platforms. The results derived from this collaborative effort show microarray measurements to be highly reproducible within and across different microarray platforms. It also demonstrates that microarray technologies are sufficiently reliable to be used for clinical and regulatory purposes.35,36 Nevertheless, the excellent work by the MAQC consortium was performed using high-quality RNA extracted from fresh or frozen samples in optimal conditions. The situation with FFPE samples is quite different in practice since RNA quality is significantly altered and variable across laboratories. Further, and despite the considerable impact it might have on the quality of data obtained during gene expression profiling, the entire pre-analytic phase was not evaluated by this consortium. Nevertheless, it may be possible for the technical obstacles to be overcome, at least partially, by the use of NGS techniques where RNA quality has a smaller impact on result reliability.37 Array-CGH, which could be used on formalin fixed, paraffin embedded tissue could also be a suitable technique for assessing chromosomal complexity. Tumours showing a lower chromosomal complexity than sarcomas with complex genomics, such as GIST and translocation-related sarcomas, are good candidates for such a technique. New techniques using high-throughput sequencing can be used on paraffin blocks and will certainly be very fruitful for the detection of specific genomic abnormalities38 and for molecular grading in the near future. Molecular grading certainly holds great promise for several clinical applications. A first potential use is a better selection of patients for adjuvant treatment, and the best candidate for this is GIST. Adjuvant imatinib is now recommended for localised GIST of more than 3 cm, or for localised high risk and intermediate risk GIST. Seemingly, quite a high proportion of patients with an intermediate risk or a GIST larger than 3 cm will receive imatinib with no benefit, but with possible secondary effects and at a high cost. Pertinent selection of only those patients with a risk of recurrence would be an important improvement in the management of these patients. Children with synovial sarcoma almost systematically receive neoadjuvant and/or adjuvant chemotherapy although most of them have a good prognosis. Molecular grading would be of great help for selecting patients with a high risk of metastasis and those for whom adjuvant chemotherapy could be useful. For GISTs and synovial sarcomas, the GI evaluated by array-CGH from formalin fixed, paraffin embedded tissue can be used. For the most frequent and aggressive sarcomas, i.e., leiomyosarcoma and undifferentiated pleomorphic sarcoma, the GI is not pertinent and an expression profile signature such as CINSARC should be used. Therefore, it is necessary to adapt this signature to formalin fixed, paraffin embedded tissue obtained from core needle biopsies. The value of CINSARC for predicting the response to treatment is unknown and should be evaluated. Nevertheless,
119
a recent study39 has shown its potential to predict value in breast carcinomas. Such molecular signatures, and array-CGH and expression profiling more generally, should be more systematically applied to primary untreated tumours for patients entering clinical trials. This would not only provide a better understanding of genomic complexity, activated signalling pathways and response or resistance to the evaluated drugs, but could help identify potential targets for future treatments. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Address for correspondence: Dr J-M. Coindre, Department of Biopathology, Institut Bergonie´, 229 cours de l’Argonne, 33076 Bordeaux Ce´dex, France. E-mail: [email protected]
References 1. Fletcher CDM, Bridge JA, Hogendoorn CW, Mertens F, eds. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon: IARC, 2013. 2. Coindre JM. Grading of soft tissue sarcomas. Review and update. Arch Pathol Lab Med 2006; 130: 1448–53. 3. Coindre JM, Terrier P, Bui NB, et al. Prognostic factors in adult patients with locally controlled soft tissue sarcoma: a study of 546 patients from the French Federation of Cancer Centers Sarcoma Group. J Clin Oncol 1996; 14: 869–77. 4. Pisters PWT, Leung DHY, Woodruff J, et al. Analysis of prognostic factors in 1041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol 1996; 14: 1679–89. 5. Guillou L, Coindre JM, Bonichon F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers Sarcoma Group grading systems in a population of 410 adult patients with soft tissue sarcoma. J Clin Oncol 1997; 15: 350–62. 6. Chibon F, Lagarde P, Salas S, et al. Validated prediction of clinical outcome in sarcomas and multiple types of cancer on the basis of a gene expression signature related to genome complexity. Nat Med 2010; 16: 781–7. 7. Russell WO, Cohen J, Enzinger F, et al. Clinical and pathological staging system for soft tissue sarcomas. Cancer 1977; 40: 1562–70. 8. Trojani M, Contesso G, Coindre JM, et al. Soft tissue sarcomas of adults: study of pathological prognostic variables and definition of histopathological grading system. Int J Cancer 1984; 33: 37–42. 9. Costa J, Wesley RA, Glatstein E, et al. The grading of soft tissue sarcomas: results of a clinicopathological correlation in a series of 163 cases. Cancer 1984; 53: 530–41. 10. van Unnik JA, Coindre JM, Contesso G, et al. Grading of soft tissue sarcomas: experience of the EORTC soft tissue and bone sarcoma group. Eur J Cancer 1993; 29A: 2089–93. 11. Gustafson P, Akerman M, Alvegard TA, et al. Prognostic information in soft tissue sarcoma using tumour size, vascular invasion and microscopic tumour necrosis: the SIN-system. Eur J Cancer 2003; 39: 1568–76. 12. Oliveira AM, Nascimento AG. Grading in soft tissue tumors: principles and problems. Skeletal Radiol 2001; 30: 543–59. 13. Guillou L, Coindre JM. How should we grade soft tissue sarcomas and what are the limitations? Pathol Case Review 1998; 3: 1–6. 14. Rubin BP, Cooper K, Fletcher CDM, et al. Protocol for the examination of specimens from patients with tumors of soft tissue. October 2013. http:// www.cap.org/apps/docs/committees/cancer/cancer_protocols/2013/Soft Tissue_13protocol_3120.pdf 15. Edge SB, Byrd DR, Compton CC, et al. AJCC Cancer Staging Manual. 7th ed. New York: Springer-Verlag, 2010. 16. Golouh R, Bracko M. What is current practice in soft tissue sarcoma grading? Radiol Oncol 2001; 35: 47–52. 17. Coindre JM, Terrier P, Guillou L, et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer 2001; 91: 1914–26. 18. Brown FM, Fletcher CD. Problems in grading soft tissue sarcomas. Am J Clin Pathol 2000; 114: S82–9. 19. Deyrup AT, Weiss SW. Grading of soft tissue sarcomas: the challenge of providing precise information in an imprecise world. Histopathology 2006; 48: 42–50. 20. Fletcher CD, Gustafson P, Rydholm A, et al. Clinicopathologic re-evaluation of 100 malignant fibrous histiocytomas: prognostic relevance of subclassification. J Clin Oncol 2001; 19: 3045–50.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
120
NEUVILLE et al.
21. Deyrup AT, Haydon RC, Huo D, et al. Myoid differentiation and prognosis in adult pleomorphic sarcomas of the extremity: an analysis of 92 cases. Cancer 2003; 98: 805–13. 22. Coindre JM, Trojani M, Contesso G, et al. Reproducibility of a histopathologic grading system for adult soft tissue sarcoma. Cancer 1986; 58: 306–9. 23. Welker JA, Henshaw RM, Jelinek J, et al. The percutaneous needle biopsy is safe and recommended in the diagnosis of musculoskeletal masses. Outcome analysis of 155 patients in a sarcoma referral center. Cancer 2000; 89: 2677–86. 24. Hoeber I, Spillane AJ, Fisher C, et al. Accuracy of biopsy techniques for limb and limb girdle soft tissue tumors. Ann Surg Oncol 2001; 8: 80–7. 25. Wu¨rl P, Kappler M, Meye A, et al. Co-expression of survivin and TERT and risk of tumour-related death in patients with soft tissue sarcoma. Lancet 2002; 16: 943–5. 26. Nielsen TO, West RB. Translating gene expression into clinical care: sarcomas as a paradigm. J Clin Oncol 2010; 28: 1796–805. 27. Lee YF, John M, Falconer A, et al. A gene expression signature associated with metastatic outcome in human leiomyosarcomas. Cancer Res 2004; 15: 7201–4. 28. Francis P, Namlos HM, Mu¨ller C, et al. Diagnostic and prognostic gene expression signatures in 177 soft tissue sarcomas: hypoxia-induced transcription profile signifies metastatic potential. BMC Genomics 2007; 8: 73. 29. Carter SL, Eklund AC, Kohane IS, et al. A signature of chromosomal intability inferred from gene expression profiles predicts clinical outcome in multiple human cancers. Nat Med 2006; 38: 1043–8. 30. Lagarde P, Pe´rot G, Kauffmann A, et al. Mitotic checkpoints and chromosome instability are strong predictors of clinical outcome in gastrointestinal stromal tumors. Clin Cancer Res 2012; 18: 826–38.
Pathology (2014), 46(2), February
31. Astolfi A, Nannini M, Pantaleo MA, et al. A molecular portrait of gastrointestinal stromal tumors: an integrative analysis of gene expression profiling and high-resolution genomic copy number. Lab Invest 2010; 90: 1285–94. 32. Lagarde P, Pzybyl J, Brulard C, et al. Chromosme instability accounts for reverse metastatic outcomes of pediatric and adult synovial sarcomas. J Clin Oncol 2013; 31: 608–15. 33. Hasegawa T, Yamamoto S, Yokoyama R, et al. Prognostic significance of grading and staging systems using MIB-1 score in adult patients with soft tissue sarcoma of the extremities and trunk. Cancer 2002; 95: 843–51. 34. van de Vijver MJ, He YD, van’t Veer LJ, et al. A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 2002; 347: 1999– 2009. 35. Shi L, Reid LH, Jones WD, et al. The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. Nat Biotechnol 2006; 24: 1151–61. 36. Shi L, Campbell G, Jones WD, et al. The MicroArray Quality Control (MAQC)-II study of common practices for the development and validation of microarray-based predictive models. Nat Biotechnol 2010; 28: 827– 38. 37. Beck AH, Weng Z, Witten DM, et al. 30 -end sequencing for expresion quantification (3SEQ) from archival tumor samples. PLoS One 2010; 5: e8768. 38. Sweeney RT, Zhang B, Zhu SX, et al. Desktop transcriptome sequencing from archival tissue to identify clinically relevant translocations. Am J Surg Pathol 2013; 37: 796–803. 39. Bertucci F, Finetti P, Sabatier R, et al. The CINSARC signature. Prognostic and predictive of response to chemotherapy? Cell Cycle 2010; 9: 1–3.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
SOFT TISSUE PATHOLOGY
Grading of soft tissue sarcomas: from histological to molecular assessment AGNES NEUVILLE1,2, FRE´DE´RIC CHIBON1,3
AND
JEAN-MICHEL COINDRE1,2
1Department of Biopathology, Institut Bergonie´ and INSERM U916, 2Laboratory of Pathology, University Victor Se´galen Bordeaux, and 3Translational Research, Institut Bergonie´, Bordeaux, France
Summary Several histological grading systems for soft tissue sarcomas have been described since the early 1980s. Their main objective is to select patients for adjuvant chemotherapy. Two histological grading systems are used in daily practice, the National Cancer Institute (NCI) and the French Federation of Cancer Centers Sarcoma Group (FNCLCC) systems. They have been devised by combining histological parameters: number of mitoses per high-power field, the presence of necrosis, cellular and nuclear morphology and the degree of cellularity for the NCI grading; and tumour differentiation, mitotic index and extent of necrosis for the French system. Histological grading is far more appropriate to assess the risk of metastasis. However, several limitations prevent its use: grade cannot be applied to all histological types, its reproducibility is not perfect, a three-grade system generates an intermediate grade with undetermined prognosis, and finally the core needle biopsy, now widely used for the diagnosis of soft tissue sarcoma, is not the best sample to assess the grade. The development of molecular grading in addition to histological grading probably represents the next step. Molecular signatures based on quantitative evaluation of chromosomal complexity such as CINSARC (complexity index in sarcomas) appear as a strong independent predictive factor for metastasis in several types of sarcoma, and even in several other types of cancer. When they can be instituted in daily practice on formalin fixed, paraffin embedded material, molecular signatures will not only provide information on risk of metastasis, but also better understanding of cancer development, response or resistance to evaluated drugs, and potential targets for future treatments. Key words: Grading, molecular signature, sarcoma. Received 7 October, revised 5 November, accepted 11 November 2013
INTRODUCTION Soft tissue sarcomas (STS) represent a heterogeneous group of rare malignant tumours with a wide spectrum in terms of histological type and prognosis.1 Prognosis of STS is dominated by local recurrence and distant metastasis. STS recur locally in less than 10% when located in the limbs and trunk wall but metastasise in 30–50% of cases. Quality of surgical margins is the most important factor for predicting local recurrence, whereas histological grade particularly indicates the probability of distant metastasis and overall survival.2–5 For more than 20 years, the main use of grading has been the selection of patients for chemotherapy. However, the utility of Print ISSN 0031-3025/Online ISSN 1465-3931 DOI: 10.1097/PAT.0000000000000048
#
adjuvant chemotherapy in STS is still debated and the approach of sarcoma therapy has recently undergone shifts from nonspecific cytotoxic agents towards utilisation of molecularly targeted treatments. Moreover, some histotypes have shown different sensitivity to specific cytotoxic drugs. Nowadays, the value of grading is also limited by the universal use of core needle biopsies. In fact, histological grade can be considered as a morphological translation of molecular events that determine tumour aggressiveness; therefore, it should be regarded as a transient practical method that should prompt research in order to establish a definitive system based on molecular parameters. For more than 10 years, biologists and pathologists of the French Sarcoma Group have been working on molecular analyses of adult sarcomas with complex genomics in order to determine which genes and pathways are related to chromosomal complexity and whether or not a link between chromosomal complexity and prognosis exists. This work has led to the description of the CINSARC (Complexity INdex in SARComas) expression profile signature, offering much better prognostic value than histological grading.6
HISTOLOGICAL GRADING Concept of grading The concept of histological grade was introduced by Broders in 1920 and must be clearly distinguished from staging and nomograms. Whereas grading is based on the intrinsic quality of the primary tumour only, staging also takes into account tumour extent. Nomograms assess multiple clinical and histological parameters to calculate the probability of recurrence for a given patient. Histological grade is based on the histological qualities of the primary tumour and is expected to predict tumour aggressiveness. In 1977, Russell et al.7 proposed the first coherent clinicopathological classification separating patients into four stages with a different prognosis and introduced a histological grading applicable to all adult STS. In their system, grading was the most important prognostic factor. However, the grading system they used was rather subjectively based on several histological parameters with no clear definition. To obtain the most efficient system, histological parameters should be clearly defined and selected by multivariate analysis so that only the necessary parameters summarising all prognostic histological information are used. According to the studies that have followed this approach, the best parameters are tumour histotype and subtype and/or differentiation, tumour necrosis, mitotic index8–10 and, as stated in a few, vascular invasion.11 In the early 1980s, several
2014 Royal College of Pathologists of Australasia
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
114
NEUVILLE et al.
grading systems were reported and a critical review of those most important was performed by Oliveira and Nascimento.12 The two systems most commonly used are the National Cancer Institute (NCI) grading described by Costa et al.9 and the French Federation of Cancer Centers Sarcoma Group (FNCLCC) grading described by Trojani et al.8 Description of grading systems The NCI system, published in 1984,9 was based on an analysis of 163 patients and assessment of six histological parameters (histological type, mitoses, necrosis, pleomorphism, cellularity and intercellular matrix). Costa et al. found by multivariate analysis that gross and microscopic tumour necrosis had the greatest impact. The final form of this system uses a group of predefined grade 1 and grade 3 sarcomas and, in tumours not automatically categorised, the amount of necrosis is used to distinguish grade 2 from grade 3 tumours, with a cut-off point of 15% for the extent of necrosis. Shortly after, the French Sarcoma Group proposed a system based on the analysis of 155 patients with assessment of seven histological parameters (tumour differentiation, cellularity, nuclear atypia, presence of malignant giant cells, mitotic count, extent of tumour necrosis and presence of vascular emboli) and multivariate analyses that identified three independent prognostic factors: tumour differentiation, mitotic index and extent of necrosis.8 These parameters are scored 1–3 for differentiation and mitotic index and 0–2 for necrosis. A three-grade system is obtained by summing the scores obtained for each of these three parameters (Table 1). Grade 1 is defined as a total of 2 or 3; grade 2 as a total of 4 or 5; and grade 3 as a total of 6–8. To enhance the reproducibility of the system, the parameters are defined as exactly as possible. The rules for using this Table 1
Definition of parameters in the French grading system*
Pathology (2014), 46(2), February
system and its limitations have been previously described in detail.13 Differentiation is the most controversial parameter and is in fact a mixture of histological type and subtype and/or true differentiation. Table 2 shows the differentiation scores attributed to common sarcoma types. A score of 1 is currently assigned to sarcomas closely resembling normal adult tissue to such a degree as to be confused with benign tumours, such as a well-differentiated leiomyosarcoma. A score of 3 is given to embryonal and poorly-differentiated sarcomas, sarcomas of doubtful histological type, synovial sarcoma, primitive neuroectodermal tumour, osteosarcoma, pleomorphic rhabdomyosarcoma, and pleomorphic liposarcoma. Others of certain histological type such as myxoid liposarcoma are scored 2. Mitotic count is obtained in 10 successive high power fields (HPF) in most mitotic areas. This count is taken to establish the score: 1, 0–9 mitoses; 2, 10–19 mitoses; 3, more than 19 mitoses per 10 HPF. Strict rules should be followed for counting mitoses: early and adequate fixation of tumour, correct macroscopic sampling (one section per 1 or 2 cm of the tumour diameter), good quality sections, and adequate selection of the most mitotic areas at medium power and where mitoses should be counted. A calibrated HPF surface should be performed according to the first study (HPF ¼ 0.1734 mm2).8 Moreover, pathologists should take their time, as counting mitoses hurriedly might dwindle reproducibility. If the mitotic count is close to the cut-off score (i.e., 7–9 or 17–19 mitoses per 10 HPF), recounting is strongly advised. Any necrotic or hypocellular areas as well as zones of ulceration should be avoided. Tumour necrosis should be evaluated at histological and macroscopic levels. However, macroscopic evaluation of necrosis is poorly reproducible and should always be checked under the microscope. Necrosis related to previous surgery or to ulceration must not be taken into account, and nor should haemorrhage or hyalinisation. The score is 0 when there is no necrosis, 1 when the necrotic area is less than 50% and 2 when it is more than 50% of the tumour surface.
Definition of parameters Tumour differentiation (see Table 2) Score 1: Sarcomas closely resembling normal adult mesenchymal tissue and potentially difficult to distinguish from the counterpart benign tumour (e.g., well-differentiated liposarcoma, well-differentiated leiomyosarcoma) Score 2: Sarcomas for which histological typing is certain (e.g., myxoid liposarcoma, myxofibrosarcoma) Score 3: Embryonal and undifferentiated sarcomas, synovial sarcomas, sarcomas of doubtful type
Table 2 Tumour differentiation score according to histological type in the French grading system* Histological type
Grade 1: Total score 2, 3 Grade 2: Total score 4, 5 Grade 3: Total score 6, 7, 8
Well-differentiated liposarcoma Well-differentiated leiomyosarcoma Malignant neurofibroma Myxoid liposarcoma Conventional leiomyosarcoma Conventional MPNST Myxofibrosarcoma Myxoid chondrosarcoma Pleomorphic liposarcoma Dedifferentiated liposarcoma Rhabdomyosarcoma Poorly-differentiated/pleomorphic leiomyosarcoma Poorly-differentiated MPNST Malignant Triton tumour Synovial sarcoma Extraskeletal osteosarcoma Extraskeletal PNET Mesenchymal chondrosarcoma Epithelioid sarcoma Malignant rhabdoid tumour Undifferentiated (spindle cell and pleomorphic) sarcoma
*
*
Mitotic count (established on 10 HPF){ Score 1: 0–9 mitoses per 10 HPF Score 2: 10–19 mitoses per 10 HPF Score 3: More than 19 mitoses per 10 HPF Tumour necrosis Score 0: No necrosis Score 1: Less than 50% of tumour necrosis Score 2: 50% or more than 50% of tumour necrosis Histological grade
Modified from Trojani et al.8 A HPF measures 0.1734 mm2. HPF, high-power field. {
Differentiation score 1 1 1 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3
Modified from Guillou et al.5 MPNST, malignant peripheral nerve sheath tumour; PNET, primitive neuroectodermal tumour.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
GRADING IN SARCOMAS
In accordance with the College of American Pathologists14 and the American Joint Committee on Cancer (AJCC) recommendations,15 this system is preferred over the NCI system, at least in adults. When comparing the performances of these two systems in the same population of patients with STS, the French system performed better.5 It allocated fewer patients to the noninformative grade 2 category and appeared more discriminating and efficient in the selection of patients with tumours of high malignant potential that could benefit from adjuvant chemotherapy. Discrepancies between the two grading systems were nevertheless observed in 34.6% of the cases. A study by soft tissue pathologists from 30 countries showed that the French grading was more widely used than the other systems:16 37.3% for the French system, 24% for the NCI system, 12% for the Broders system and 1.3% for the Markhede system, with 25.3% unspecified. However, the majority of responders (74%) were from Europe where the French system predominates. The main weakness of the French system lies in the attribution of a differentiation score, particularly for tumours with no normal tissue counterpart such as undifferentiated pleomorphic sarcoma.
115
adjuvant chemotherapy: patients with a grade 1 STS should certainly not receive adjuvant chemotherapy since it will be inefficient. On the other hand, a patient with a grade 3 sarcoma is a good candidate for receiving adjuvant chemotherapy due to the high probability of distant metastases. Moreover, it has been shown that adjuvant chemotherapy may be more efficient for patients with a grade 3 sarcoma than for patients with a grade 2 sarcoma. However, the utility of adjuvant chemotherapy in STS is still debated. Limitations of histological grading systems Grading systems have been subject to numerous criticisms.12,18,19 Some object to the fact that all grading systems have been developed and tested in the overall sarcoma group and not specifically in every histological category. However, in a study on 1240 localised STS,17 we showed that, with the exception of malignant peripheral nerve sheath tumour (MPNST) and rhabdomyosarcoma, histological grade remains an important predictor of metastasis in the main histological types of STS (90% of all STS). It is now widely accepted that no current grading system performs well for every type of STS. However, given the number of histological types of STS, it is unrealistic and pointless to develop a grading system for every specific histological type of STS. For the time being, the French or the NCI systems perform correctly for the most common sarcoma types and represent an acceptable alternative. Yet, we should keep in mind that grading is less informative than histological type in dedifferentiated and round cell liposarcomas, alveolar soft part sarcoma, clear cell sarcoma, epithelioid sarcoma, and primitive neuroectodermal tumour (PNET), and that it should not be used on tumours of ‘intermediate malignancy’ such as atypical fibroxanthoma and dermatofibrosarcoma protuberans. Among poorly differentiated sarcomas, it is also important to differentiate leiomyosarcomas from undifferentiated sarcomas since the former have a worse prognosis than the latter.20,21 Moreover, since some subtypes show a differential sensitivity to specific drugs, it is now more and more important to obtain the exact histotype.
Value of grading It is now recognised that histological grade is the most important prognostic factor for adult STS. As the best predictor of metastasis development and tumour mortality, histological grade is a key parameter of the currently used 2010 TNM clinicopathological staging system.15 A grading system allows us to indicate the probability of distant metastases and overall survival. In every study with multivariate analyses testing histological grade and clinical factors in STS, grade is the most important prognostic factor. In a study of 1240 patients with a localised STS, the 5 year metastasis-free survival rate was 91% for grade 1, 71% for grade 2 and 43% for grade 3 (Fig. 1).17 Since grading permits prediction of metastasis, and because the response rate to palliative chemotherapy is better in patients with high grade sarcoma, we use grading to select patients for
SOFT TISSUE SARCOMAS: French sarcoma group (n = 1240) M E T A S T A S I S F R E E S U R V I V A L
1
Grade 1 (n = 157)
0.9 0.8
Grade 2 (n = 511)
0.7 0.6 0.5
Grade 3 (n = 572) 0.4 0.3 0.2
p < 10–29
0.1 1
1
2
3
4
5
6
7
8
9
10
11
12
13
Years Fig. 1 Metastasis-free survival curves for French histological grade in a cohort of 1240 adult patients with a localised soft tissue sarcoma.17
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
116
NEUVILLE et al.
Another weakness of any grading system lies in its reproducibility. When the reproducibility of the French grading system was tested by 15 pathologists on 25 cases, the crude proportion in agreement was 75% for tumour grade, 73% for mitotic index, 74% for differentiation, and 81% for tumour necrosis.22 A reproducibility of only 75% is difficult to accept by clinicians and patients. Reproducibility of diagnosis of histological type was significantly lower with an agreement rate of 61%. However, this reproducibility study was conducted in the early 1980s when immunohistochemistry was not very efficient and molecular analysis was not available. Nowadays, at least among pathologists specialised in soft tissue tumours, the reproducibility of histotyping can be considered as excellent. A further major limitation of almost every grading system is the existence of an intermediate grade which corresponds to undetermined prognosis. This category represents 46% and 58% of adult patients with a STS assessed with the French and the NCI grading systems, respectively,5 meaning that histological grading is useful for only 50% of STS patients. Conventional grading systems have been developed to be used on representative material, i.e., on whole untreated tumours or on surgical biopsies. Core needle biopsies are now widely used for the diagnosis of primary STS and for establishing treatment strategies. Most studies show that core needle biopsies are highly accurate in the diagnosis of malignancies (95%)23 and that sarcoma typing and grading can also be performed with an acceptable degree of accuracy (70–80%).24 But determination of histological grade can essentially only be done with accuracy for high grade (G3) neoplasms. Determination of grade in the intermediate (G2) and low grade (G1) categories is more problematic. It should also be emphasised that core needle biopsies should not be evaluated without a good knowledge of clinical and radiological data. Imaging procedures often provide useful information about the nature of the lump, especially regarding tumour size, borders, and amount of necrosis.
MOLECULAR GRADING Molecular markers hold great promise for refining our ability to establish early prognosis and to predict response to therapy. Identification of molecular prognostic markers in sarcoma started in the early 2000s, with the study reported by Wu¨rl et al.25 The authors showed that expression of the two genes TERT (telomerase reverse transcriptase) and survivin (also named BIRC5) was associated with poor outcome in a series of 89 sarcomas of different subtypes. This series was composed of sarcomas with complex genetics such as leiomyosarcoma, pleomorphic rhabdomyosarcoma and liposarcoma, and in these tumours co-expression of both genes was a significant prognostic factor. Survivin belongs to the family of genes that inhibit apoptosis and is also involved in chromosome segregation. TERT is involved in the immortal status of cells by maintaining telomere size. Even if this two-gene signature has never been validated in an independent group, it has biological meaning that links metastatic outcome with genes involved in chromosome structure and segregation. With the development of microarray technology, approaches to identify molecular signatures have changed and several studies have reported molecular profiling analyses in sarcomas,26 but only a few have been validated on independent groups of patients. Lee and colleagues27 reported on one of the first expression profiling studies leading to a prognostic
Pathology (2014), 46(2), February
signature in sarcomas. Comparing the expression profiles from both metastatic and non-metastatic leiomyosarcoma, they identified 335 genes differentially expressed between primary tumours (20 cases) and metastases (7 cases). Although this signature has significant prognostic value, the study suffered from two main limitations: the small size of the series and the lack of validation of the results, compromising its clinical use. Francis and colleagues28 analysed 177 sarcomas and produced a meaningful expression signature of different histotypes, demonstrating that histological classification fits well with expression profiles of tumours. They also identified 220 genes associated with metastasis events in a subgroup of 89 pleomorphic sarcomas. Both studies developed supervised approaches with the metastatic event as an end-point and selected genes that, on the whole, have no biological links between them. In fact, such an approach has proven to be inefficient for identifying genes or pathways implicated in the metastatic potential acquisition of metastases. Recently, the French Sarcoma Group published the identification of a 67-gene expression signature called CINSARC, satisfying all the criteria to be a clinically applicable prognostic molecular marker.6 With the aim of establishing a possible link between chromosomal complexity and prognosis, this group studied a large series of sarcomas with complex genomics, (i.e., leimyosarcoma, undifferentiated pleomorphic sarcoma, myxofibrosarcoma, pleomorphic liposarcoma, pleomorphic rhabdomyosarcoma and dedifferentiated liposarcoma) by two combined techniques, array-CGH and expression profile. The first step was to assess chromosomal complexity by array-CGH and three types of recurrent profiles were identified: tumours with a simple amplicon profile based on a coamplification corresponding to dedifferentiated liposarcomas, tumours with a few alterations (less than 30) involving full chromosome arms or entire chromosomal gain or loss (‘arm profile’), and tumours with a high level of chromosomal complexity with more than 30 alterations (‘rearranged profile’). Although the metastasisfree survival was not significantly different between the latter two groups of patients (arm and rearranged profiles), there was a positive correlation between the number of genomic alterations and histological grade (79% of grade 1 or 2 sarcomas have less than 30 chromosomal alterations and 74% of grade 3 sarcomas have more than 30 chromosomal alterations, p ¼ 0.001). Furthermore, since histological grade is a marker of tumour aggressiveness, it has been postulated that the latter is related to genomic complexity. A CIN (chromosomal instability) signature established by Carter et al.29 was also tested. This CIN signature has been set up by a computational method for assessing chromosome instability from expression profiles. Based on the expression profile of 70 genes, it has been shown to predict survival in several types of cancers, but not in the studied series of sarcomas. Data from CGH, FNCLCC grade and Carter’s signature were finally integrated together with, in a first step, three t-tests to compare the expression profiles of tumours according to three characteristics: array-CGH profile (35 imbalances), histological grade (grade 3 versus grade 2) and the 70-gene Carter CIN signature. This approach resulted in the selection of 86 genes for array-CGH, 73 genes for grade and 39 genes for the Carter signature. In a second step, a gene ontology analysis was performed in order to identify the underlying pathways. Only genes involved in the most significantly overrepresented pathways were selected: 37 genes for array-CGH, 18 genes for grade and 22 genes for Carter signature. The final gene set called CINSARC contains
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
GRADING IN SARCOMAS
A – Cohort 1
A – Cohort 2 1.0
1.0
n = 42
n = 83
0.8
0.8
0.6
Probability
Probability
117
n = 100
0.4
n = 85
0.4
n = 127 p = 5 × 10–4
0.2
n = 183 p = 1 × 10–7
0.2
0.6
0.0
0.0 0
12
24
36
48
60
72
84
96
108
0
120
12
24
36
48
60
72
84
96
108
120
Time in months
Time in months
B – Cohort 1
B – Cohort 2
1.0
1.0
0.8
0.8
Grade 1 + 2
0.4
Probability
Probability
Grade 1 + 2 0.6
Grade 3 n = 171 p = 0.43
0.2
12
24
0.4
Grade 3 n = 117 p = 4 × 10–3
0.2
0.0 0
0.6
36
48
60
72
84
96
108
0.0
120
0
Time in months
12
24
36
48
60
72
84
96
108
120
Time in months
Fig. 2 Metastasis-free survival curves according to (A) CINSARC molecular grading and (B) French histological grade in two independent groups of adult patients with a soft tissue sarcoma (cohort 1, training group; cohort 2, validation group).
67 genes. These genes belong to chromosome management and mitosis control pathways and are related to genome complexity. Some of these genes have already been reported in prognostic signatures in breast carcinomas, where they are usually classified as proliferation genes, but they are also clearly involved in genome complexity and, in fact, they are related more to chromosome instability than to proliferation. This CINSARC signature was set up on a training group of 183 non-translocated adult soft tissue sarcomas and validated on an independent cohort of 127 tumours (Fig. 2). For all these patients, clinical, histological, treatment and follow-up data were available in a shared tumour bank (https://conticabase. sarcomabcb.org/). Both array-CGH and expression profiling were performed on untreated primary tumours. CINSARC grade is an independent predictive factor for metastasis (in
both training and validation cohorts) for leimyosarcoma and undifferentiated pleomorphic sarcoma (Fig. 3) as well as among tumours of the same histological grade (Fig. 4). CINSARC grade was also of prognostic value in breast carcinoma, lymphoma and GIST when applied to series of such tumours with available expression profile and follow-up data. A subsequent study conducted on frozen tissue from 67 primary untreated GISTs30 showed that CINSARC is a strong predictor of prognosis. The gene with expression most strongly associated with recurrence was AURKA, but the AURKA gene was not amplified. Instead, deletions of p16 (CDKN2A) and/or retinoblastoma (RB1) genes were identified and were likely causal events leading to increased AURKA and CINSARC gene expression, to chromosome rearrangement, and ultimately to aggressiveness. Different studies have shown that 9p21 deletion
B – Undifferentiated sarcomas
A – Leiomyosarcomas 1.0
1.0
n = 32 n = 16
0.8
Probability
Probability
0.8
0.6
Cohort 1 n = 52 p = 2.7 × 10–4
0.4
0.2
0.6
0.4
n = 36 0.0 0
12
24
36
48
60
72
Time in months
84
96
108
120
n = 39
Cohort 1 n = 71 p = 4.69 × 10–4
0.2
0.0 0
12
24
36
48
60
72
84
96
108
120
Time in months
Fig. 3 Metastasis-free survival curves according to CINSARC molecular grading in (A) leiomyosarcomas and (B) undifferentiated sarcomas.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
118
NEUVILLE et al.
Pathology (2014), 46(2), February
A – French grade 1–2
B – French grade 3
1.0
1.0
CINSARC 1
CINSARC 1 0.8
Probability
Probability
0.8
0.6
CINSARC 2
0.4
n = 47 p = 0.009
0.2
0.0
0.6
CINSARC 2
0.4
n = 102 p = 7.6 × 10–6
0.2
0.0 0
12
24
36
48
60
72
84
96
108
Time in months
120
0
12
24
36
48
60
72
84
96
108
120
Time in months
Fig. 4 Metastasis-free survival curves according to CINSARC molecular grading in the French histological grade 1–2 patients and grade 3 patients.
is associated with metastatic outcome in patients with GIST. Nevertheless, to date it is still controversial to attribute this prognostic value specifically to one of the three genes contained in the minimal deleted chromosomal area, i.e. CDKN2A, CDKN2B and MTAP.31 Genomic (and not expression) markers can be considered as the best form of molecular criteria because they presumably drive outcomes. Additionally, their identification in daily practice is relatively easy with CGH-array, FISH or sequencing technologies, which are widely used in pathology. Since frozen tissue is rarely available in GISTs, a genomic index (GI) integrating the number and type of DNA copy number alterations that can be performed on formalin fixed, paraffin embedded tissue was established through array-CGH. This GI is a prognostic factor in GISTs as strong as CINSARC and could be used for identifying poor prognosis patients in the group classified as intermediate risk by the AFIP classification. CINSARC and GI were also tested on frozen samples from 100 primary untreated synovial sarcomas (SS).32 Both CINSARC and GI have independent and validated prognostic values and metastasis development in SS is strongly associated with chromosome complexity. The study from Lagarde et al. offers a biological explanation for the roughly opposite outcomes of paediatric and adult SS patients. Results show that metastatic outcomes are strongly associated with chromosomal complexity in both age strata and that this instability is very frequent in adult but not in paediatric SS. Given that the initial genetic driver event is the same in both groups [the t(X;18) translocation], this suggests that an independent, still unknown, mechanism is permissive to chromosome instability in adult cases and resistant in paediatric cases. No genomic alteration or significantly differentially expressed gene indicative of such mechanisms was identified. Even if this discrepancy could be due to somatic events not detectable using array-CGH, one wonders why it occurs in adults and not in paediatric SS. We hypothesised that the chromosomal complexity could be more likely related to the possible patient predilection, age, or genome ageing. GI is potentially the best overall prognostic biomarker and the most clinically relevant, considering that array-CGH on formalin fixed samples is already used in pathology. The CINSARC genes are involved in the same pathways (BIRC5 belongs to CINSARC signature) that were first reported by Wu¨rl et al.,25 i.e., chromosome integrity, segregation and mitosis. We could speculate from this observation that these mechanisms are among the most important to predict outcome in sarcomas and that the molecular mechanisms leading to distant metastases might be related to the capacity of tumour cells to induce chromosome instability and complexity. It could also suggest that the more rearranged a
genome, the higher the probability of obtaining a gene expression profile permitting cells to complete the process of allowing dissemination and metastasis development. Finally, CINSARC signature has been retrospectively validated in the three types of sarcomas, according to genetics, i.e., complex (leiomyosarcomas, undifferentiated pleomorphic sarcomas, etc), activating mutations (GISTs) and translocation related (SSX). The next step is prospective validation of CINSARC in a clinical trial.
HISTOLOGICAL OR MOLECULAR GRADING? THE WAY AHEAD Histological grading is known as the most important prognostic factor of adult STS considered as a single entity. It is a cheap, quick and easy method for identifying sarcomas with high metastatic potential, and should be used in daily practice for most soft tissue sarcomas. However, its limitations should be kept in mind: no current grading system performs well for every type of sarcoma, and grading is less informative than histotype for some of them. Furthermore, its moderate reproducibility and the existence of an intermediate grade representing almost half of cases and corresponding to undetermined prognosis, should not be forgotten. The current universal use of core needle biopsies is also a limitation for grading, which should certainly be adapted to this type of material and complemented by other parameters such as imaging for a better assessment of tumour necrosis and proliferation index evaluated by Ki-67 or MIB-1 rather than the mitotic index.33 Molecular grading based on quantitative evaluation of chromosomal complexity, such as the CINSARC signature or the GI, has shown better prognostic value as compared to histological grading in several categories of sarcomas: those with complex genomics such as leiomyosarcoma and undifferentiated pleomorphic sarcoma, sarcomas with a reciprocal translocation such as synovial sarcoma, and GIST. However, molecular grading should now be validated in prospective independent series and on other sarcoma types. Molecular grading should also be refined by looking for other categories of genes involved in the metastatic process and by defining driver genes responsible for chromosomal instability. This molecular grading should also be adapted to daily practice. Expression of a set of genes is already in current use in breast carcinoma. MammaPrint, based on the Amsterdam 70-gene breast cancer signature,34 has been cleared by the US Food and Drug Administration (FDA) for use in the USA for lymph node negative breast carcinoma patients under 61 years with tumours of less than 5 cm. However, for more universal use, these molecular signatures need to be adapted to formalin
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
GRADING IN SARCOMAS
fixed, paraffin embedded tissue from core needle biopsies. The potential lack of robustness and reproducibility of the chips was long cited as a limitation of their clinical application, but we now know this not to be the case thanks to the efforts of the international MAQC (MicroArray Quality Control) consortium. The MAQC project involved 137 participants from 51 academic institutions and industrial partners to systematically address the technical reproducibility of microarray measurements within and between laboratories, as well as across different microarray platforms. The results derived from this collaborative effort show microarray measurements to be highly reproducible within and across different microarray platforms. It also demonstrates that microarray technologies are sufficiently reliable to be used for clinical and regulatory purposes.35,36 Nevertheless, the excellent work by the MAQC consortium was performed using high-quality RNA extracted from fresh or frozen samples in optimal conditions. The situation with FFPE samples is quite different in practice since RNA quality is significantly altered and variable across laboratories. Further, and despite the considerable impact it might have on the quality of data obtained during gene expression profiling, the entire pre-analytic phase was not evaluated by this consortium. Nevertheless, it may be possible for the technical obstacles to be overcome, at least partially, by the use of NGS techniques where RNA quality has a smaller impact on result reliability.37 Array-CGH, which could be used on formalin fixed, paraffin embedded tissue could also be a suitable technique for assessing chromosomal complexity. Tumours showing a lower chromosomal complexity than sarcomas with complex genomics, such as GIST and translocation-related sarcomas, are good candidates for such a technique. New techniques using high-throughput sequencing can be used on paraffin blocks and will certainly be very fruitful for the detection of specific genomic abnormalities38 and for molecular grading in the near future. Molecular grading certainly holds great promise for several clinical applications. A first potential use is a better selection of patients for adjuvant treatment, and the best candidate for this is GIST. Adjuvant imatinib is now recommended for localised GIST of more than 3 cm, or for localised high risk and intermediate risk GIST. Seemingly, quite a high proportion of patients with an intermediate risk or a GIST larger than 3 cm will receive imatinib with no benefit, but with possible secondary effects and at a high cost. Pertinent selection of only those patients with a risk of recurrence would be an important improvement in the management of these patients. Children with synovial sarcoma almost systematically receive neoadjuvant and/or adjuvant chemotherapy although most of them have a good prognosis. Molecular grading would be of great help for selecting patients with a high risk of metastasis and those for whom adjuvant chemotherapy could be useful. For GISTs and synovial sarcomas, the GI evaluated by array-CGH from formalin fixed, paraffin embedded tissue can be used. For the most frequent and aggressive sarcomas, i.e., leiomyosarcoma and undifferentiated pleomorphic sarcoma, the GI is not pertinent and an expression profile signature such as CINSARC should be used. Therefore, it is necessary to adapt this signature to formalin fixed, paraffin embedded tissue obtained from core needle biopsies. The value of CINSARC for predicting the response to treatment is unknown and should be evaluated. Nevertheless,
119
a recent study39 has shown its potential to predict value in breast carcinomas. Such molecular signatures, and array-CGH and expression profiling more generally, should be more systematically applied to primary untreated tumours for patients entering clinical trials. This would not only provide a better understanding of genomic complexity, activated signalling pathways and response or resistance to the evaluated drugs, but could help identify potential targets for future treatments. Conflicts of interest and sources of funding: The authors state that there are no conflicts of interest to disclose. Address for correspondence: Dr J-M. Coindre, Department of Biopathology, Institut Bergonie´, 229 cours de l’Argonne, 33076 Bordeaux Ce´dex, France. E-mail: [email protected]
References 1. Fletcher CDM, Bridge JA, Hogendoorn CW, Mertens F, eds. Pathology and Genetics of Tumours of Soft Tissue and Bone. Lyon: IARC, 2013. 2. Coindre JM. Grading of soft tissue sarcomas. Review and update. Arch Pathol Lab Med 2006; 130: 1448–53. 3. Coindre JM, Terrier P, Bui NB, et al. Prognostic factors in adult patients with locally controlled soft tissue sarcoma: a study of 546 patients from the French Federation of Cancer Centers Sarcoma Group. J Clin Oncol 1996; 14: 869–77. 4. Pisters PWT, Leung DHY, Woodruff J, et al. Analysis of prognostic factors in 1041 patients with localized soft tissue sarcomas of the extremities. J Clin Oncol 1996; 14: 1679–89. 5. Guillou L, Coindre JM, Bonichon F, et al. Comparative study of the National Cancer Institute and French Federation of Cancer Centers Sarcoma Group grading systems in a population of 410 adult patients with soft tissue sarcoma. J Clin Oncol 1997; 15: 350–62. 6. Chibon F, Lagarde P, Salas S, et al. Validated prediction of clinical outcome in sarcomas and multiple types of cancer on the basis of a gene expression signature related to genome complexity. Nat Med 2010; 16: 781–7. 7. Russell WO, Cohen J, Enzinger F, et al. Clinical and pathological staging system for soft tissue sarcomas. Cancer 1977; 40: 1562–70. 8. Trojani M, Contesso G, Coindre JM, et al. Soft tissue sarcomas of adults: study of pathological prognostic variables and definition of histopathological grading system. Int J Cancer 1984; 33: 37–42. 9. Costa J, Wesley RA, Glatstein E, et al. The grading of soft tissue sarcomas: results of a clinicopathological correlation in a series of 163 cases. Cancer 1984; 53: 530–41. 10. van Unnik JA, Coindre JM, Contesso G, et al. Grading of soft tissue sarcomas: experience of the EORTC soft tissue and bone sarcoma group. Eur J Cancer 1993; 29A: 2089–93. 11. Gustafson P, Akerman M, Alvegard TA, et al. Prognostic information in soft tissue sarcoma using tumour size, vascular invasion and microscopic tumour necrosis: the SIN-system. Eur J Cancer 2003; 39: 1568–76. 12. Oliveira AM, Nascimento AG. Grading in soft tissue tumors: principles and problems. Skeletal Radiol 2001; 30: 543–59. 13. Guillou L, Coindre JM. How should we grade soft tissue sarcomas and what are the limitations? Pathol Case Review 1998; 3: 1–6. 14. Rubin BP, Cooper K, Fletcher CDM, et al. Protocol for the examination of specimens from patients with tumors of soft tissue. October 2013. http:// www.cap.org/apps/docs/committees/cancer/cancer_protocols/2013/Soft Tissue_13protocol_3120.pdf 15. Edge SB, Byrd DR, Compton CC, et al. AJCC Cancer Staging Manual. 7th ed. New York: Springer-Verlag, 2010. 16. Golouh R, Bracko M. What is current practice in soft tissue sarcoma grading? Radiol Oncol 2001; 35: 47–52. 17. Coindre JM, Terrier P, Guillou L, et al. Predictive value of grade for metastasis development in the main histologic types of adult soft tissue sarcomas: a study of 1240 patients from the French Federation of Cancer Centers Sarcoma Group. Cancer 2001; 91: 1914–26. 18. Brown FM, Fletcher CD. Problems in grading soft tissue sarcomas. Am J Clin Pathol 2000; 114: S82–9. 19. Deyrup AT, Weiss SW. Grading of soft tissue sarcomas: the challenge of providing precise information in an imprecise world. Histopathology 2006; 48: 42–50. 20. Fletcher CD, Gustafson P, Rydholm A, et al. Clinicopathologic re-evaluation of 100 malignant fibrous histiocytomas: prognostic relevance of subclassification. J Clin Oncol 2001; 19: 3045–50.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
120
NEUVILLE et al.
21. Deyrup AT, Haydon RC, Huo D, et al. Myoid differentiation and prognosis in adult pleomorphic sarcomas of the extremity: an analysis of 92 cases. Cancer 2003; 98: 805–13. 22. Coindre JM, Trojani M, Contesso G, et al. Reproducibility of a histopathologic grading system for adult soft tissue sarcoma. Cancer 1986; 58: 306–9. 23. Welker JA, Henshaw RM, Jelinek J, et al. The percutaneous needle biopsy is safe and recommended in the diagnosis of musculoskeletal masses. Outcome analysis of 155 patients in a sarcoma referral center. Cancer 2000; 89: 2677–86. 24. Hoeber I, Spillane AJ, Fisher C, et al. Accuracy of biopsy techniques for limb and limb girdle soft tissue tumors. Ann Surg Oncol 2001; 8: 80–7. 25. Wu¨rl P, Kappler M, Meye A, et al. Co-expression of survivin and TERT and risk of tumour-related death in patients with soft tissue sarcoma. Lancet 2002; 16: 943–5. 26. Nielsen TO, West RB. Translating gene expression into clinical care: sarcomas as a paradigm. J Clin Oncol 2010; 28: 1796–805. 27. Lee YF, John M, Falconer A, et al. A gene expression signature associated with metastatic outcome in human leiomyosarcomas. Cancer Res 2004; 15: 7201–4. 28. Francis P, Namlos HM, Mu¨ller C, et al. Diagnostic and prognostic gene expression signatures in 177 soft tissue sarcomas: hypoxia-induced transcription profile signifies metastatic potential. BMC Genomics 2007; 8: 73. 29. Carter SL, Eklund AC, Kohane IS, et al. A signature of chromosomal intability inferred from gene expression profiles predicts clinical outcome in multiple human cancers. Nat Med 2006; 38: 1043–8. 30. Lagarde P, Pe´rot G, Kauffmann A, et al. Mitotic checkpoints and chromosome instability are strong predictors of clinical outcome in gastrointestinal stromal tumors. Clin Cancer Res 2012; 18: 826–38.
Pathology (2014), 46(2), February
31. Astolfi A, Nannini M, Pantaleo MA, et al. A molecular portrait of gastrointestinal stromal tumors: an integrative analysis of gene expression profiling and high-resolution genomic copy number. Lab Invest 2010; 90: 1285–94. 32. Lagarde P, Pzybyl J, Brulard C, et al. Chromosme instability accounts for reverse metastatic outcomes of pediatric and adult synovial sarcomas. J Clin Oncol 2013; 31: 608–15. 33. Hasegawa T, Yamamoto S, Yokoyama R, et al. Prognostic significance of grading and staging systems using MIB-1 score in adult patients with soft tissue sarcoma of the extremities and trunk. Cancer 2002; 95: 843–51. 34. van de Vijver MJ, He YD, van’t Veer LJ, et al. A gene-expression signature as a predictor of survival in breast cancer. N Engl J Med 2002; 347: 1999– 2009. 35. Shi L, Reid LH, Jones WD, et al. The MicroArray Quality Control (MAQC) project shows inter- and intraplatform reproducibility of gene expression measurements. Nat Biotechnol 2006; 24: 1151–61. 36. Shi L, Campbell G, Jones WD, et al. The MicroArray Quality Control (MAQC)-II study of common practices for the development and validation of microarray-based predictive models. Nat Biotechnol 2010; 28: 827– 38. 37. Beck AH, Weng Z, Witten DM, et al. 30 -end sequencing for expresion quantification (3SEQ) from archival tumor samples. PLoS One 2010; 5: e8768. 38. Sweeney RT, Zhang B, Zhu SX, et al. Desktop transcriptome sequencing from archival tissue to identify clinically relevant translocations. Am J Surg Pathol 2013; 37: 796–803. 39. Bertucci F, Finetti P, Sabatier R, et al. The CINSARC signature. Prognostic and predictive of response to chemotherapy? Cell Cycle 2010; 9: 1–3.
Copyright © Royal College of pathologists of Australasia. Unauthorized reproduction of this article is prohibited.
