Immunohistochemical assessment of cyclin D1 and p53 is associated with survival in childhood malignant peripheral nerve sheath tumor

Abstract

BACKGROUND:

Malignant peripheral nerve sheath tumor (MPNST) is rare, aggressive soft tissue sarcoma which may affect children.

OBJECTIVE:

We aimed to assess prognostic significance of immunohistochemical (IHC) markers, osteopontin, fibronectin, survivin, cyclin D1 and p53, in pediatric MPNST.

METHODS:

A total of 26 pediatric MPNST patients were enrolled in the current study with a median follow-up of 51 months. IHC staining using commercially available monoclonal antibodies were employed to detect analyzed antigens on tissue microarrays. Eventually, all markers were subclassified to high (H) and low (L) expression categories in all analyzed tumors.

RESULTS:

High IHC expressions of survivin, cyclin D1, osteopontin, fibronectin, and p53 were detected in 18 (69.2%), 13 (50%), 16 (61.5%), 16 (61.5%), and 13 (50%) tumors, respectively. A significant correlation was demonstrated between cyclin D1 and osteopontin ( $p=$ 0.004). Both markers were associated with neurofibromatosis type 1 (NF1) status ( $p=$ 0.041 and $p=$ 0.037, respectively). H-fibronectin was more prevalent in deeply located tumors ( $p=$ 0.046). None of the markers was associated with IRS stage, age at diagnosis, and tumor size. Univariate analysis identified IRS stage, regional lymph node metastases, NF1, and cyclin D1 as variables associated with overall survival (OS), whereas tumor depth, osteopontin, and cyclin D1 – for relapse-free survival (RFS). Subsequent multivariate analysis identified cyclin D1 and p53 as independent variables predicting RFS, whereas cyclin D1 and regional lymph nodes status were independent predictors for OS.

This is the first study to demonstrate the independent prognostic roles of IHC tumor expressions of cyclin D1 and p53 in predicting of survival in children with MPNST. Interestingly, well established clinical prognostic factors (e.g. IRS stage, NF1 status, tumor size and location) did not maintain their significance in multivariate analysis. Osteopontin might also become useful in clinical practice, however larger multicenter studies should investigate its significance.

Keywords: Survivin, p53, cyclin D1, osteopontin, fibronectin, MPNST prognosis

1. Introduction

Malignant peripheral nerve sheath tumors (MPNSTs) are highly aggressive soft tissue sarcomas, arising from the neural crest [1]. In the general population, the incidence is very low (0.001%), whereas among patients with neurofibromatosis type 1 (NF1), the lifetime risk is significantly higher (up to 15%) [2]. In the latter group, MPNSTs most often originate from malignant transformation of plexiform neurofibromas, whereas sporadic cases develop de novo [3]. The majority of cases of MPNST develop in adults, however, up to 20% affect pediatric patients [4]. The common sites of MPNSTs are trunk, extremities, and head and neck area. MPNSTs are known to be rapidly growing masses, which invade surrounding tissues and may involve regional lymph nodes and disseminate to distant sites. They tend to metastasize to lungs, bones, pleura, retroperitoneal space and diaphragm [5].

In children with MPNST, it is essential to undertake wide excision of the tumor before its growth, invasion and dissemination. However, a significant percentage of patients with MPNST are diagnosed in advanced stages, with large, deep-seated lesions, located in unfavorable sites, with lymph nodes and/or distant organ metastases. In such cases, complete primary surgical resection with negative margins is hampered [6]. To improve outcomes, neo-adjuvant preoperative chemotherapy (CTx), and/or radiotherapy (RTx), are tried in children with MPNST to enable tumor shrinkage. However, the exact role of CTx and RTx in MPNST has not been defined yet. Moreover, there are no strict guidelines for risk stratification in patients with MPNST. Therefore, clinicopathological, immunohistochemical and molecular predictors of survival and response to therapy are highly anticipated [7].

In our previous paper, we demonstrated that cyclin D1, p53, fibronectin, survivin, and osteopontin, determined by immunohistochemistry (IHC) in the tumor sample at diagnosis, are promising predictive factors for response to neo-adjuvant CTx in pediatric MPNST [8]. Both mutant p53 and cyclin D1 deregulate cell proliferation and are proven prognostic factors in various malignancies [9, 10]. Also, high tumor expression of survivin has been found to be associated with inferior overall survival (OS) among MPNST patients, including pediatric ones [4]. Expression of extracellular matrix components (fibronectin and osteopontin) is positively related to lymph nodes involvement and tumor progression in some cancers [11, 12].

Most studies evaluating prognostic factors in patients with MPNST enroll adults. To shed a light on prognostic markers in children and adolescents, in the current study we aimed to evaluate the influence of pre-treatment tumor expression of cyclin D1, p53, fibronectin, survivin, and osteopontin and the presence of various clinical factors on the survival in pediatric MPNST.

2. Material and methods

2.1 Patients

Twenty-six pediatric patients with MPNST, treated in Polish centers of oncology and registered in the Polish Pediatric STS Registry between March 1992 and November 2013, were included in this study. Primary histopathological diagnoses were made by pathologists from two independent institutions. The staging was assessed according to both Tumor-Nodes-Metastasis (TNM) and the Intergroup Rhabdomyosarcoma Study (IRS) for pre- and post-surgical stratifications, respectively. In all patients, complete data on the possible clinical prognostic factors, such as patients’ age, primary tumor size and site, disease IRS stage, lymph nodes involvement, the presence of distant metastases, completeness of primary and delayed surgical resections, response to neo-adjuvant CTx and presence of NF1, were available. All patients received the multi-modality treatment according to CWS (Cooperative Weichteilsarkom Study Group) protocols, including surgery, CTx and RTx, as reported in our previous study [18].

2.2 Immunohistochemistry

Formalin-fixed and paraffin embedded (FFPE) tiss-

Table 1
The clinical data and expression pattern of immunohistochemical markers

Case	Sex	NF1	Primary tumor location	TNM	Survi	Cyclin	p53	OPN	Fibro	First line therapy			Relapse free	Outcome –
	(age)		max size in mm	IRS	vin	D1			nectin	CTx	RTx	SL	survival	overall survival
				stages								(completeness)	(months)	(months)

1	F (17y)	No	Left lung and pleura (unfavorable, deep) 250	T2N1M0 IRS II	H	H	H	H	H	Yes	No	No	1	Died (11)
2	F (21y)	Yes	Right lower limb (unfavorable, superficial) 100	T2N1M0 IRS III	H	L	L	H	L	Yes	Yes	No	–	Alive (43)
3	M (10y)	No	Head and neck parameningeal (unfavorable, deep) 51	T2N1M1 IRS IV	H	L	H	H	H	Yes	Yes	No	28	Died (64)
4	F (16y)	No	Right lower limb (unfavorable, superficial) 155	T2N0M0 IRS III	H	L	H	L	H	Yes	No	Yes (R0)	–	Alive (45)
5	F (0)	No	Mediastinum (unfavorable, deep) 55	T2N0M0 IRS II	L	L	L	L	L	Yes	No	No data	–	Alive (16)
6	F (0)	No	Chest wall (unfavorable, superficial) 80	T2N0M1 IRS IV	H	L	L	H	L	Yes	No	Yes (R0)	–	Alive (31)
7	M (3y)	No	Pelvis minor (unfavorable, deep) 100	T2N0M0 IRS III	L	L	H	L	H	Yes	Yes	Yes (R0)	–	Alive (151)
8	F (13y)	Yes	Left axillary region (unfavorable, superficial) 86	T2N0M0 IRS III	H	H	H	H	L	Yes	Yes	No	18	Died (20)
9	M (17y)	Yes	Left lower limb (unfavorable, deep) 235	T2N0M0 IRS III	H	H	L	H	H	Yes	No	No	30	Died (56)
10	M (15y)	No	Left brachial plexus (unfavorable, deep) 25	T2N0M0 IRS II	L	H	L	H	H	Yes	Yes	Yes (R0)	39	Died (64)
11	F (14y)	Yes	Retroperitoneal region (unfavorable, deep) 191	T2N0M1 IRS IV	H	H	H	H	H	Yes	No	No	1	Died (8)
12	M (11y)	Yes	Pelvis minor (unfavorable, deep) 300	T2N1M1 IRS IV	L	H	H	H	H	Yes	No	No	1	Died (1)
13	M (9y)	Yes	Neck (favorable, deep) 29	T2N0M0 IRS III	H	H	H	H	L	Yes	Yes	Yes (R2)	9	Died (14)
14	M (10y)	No	Right chest (unfavorable. Deep) 40	T2N0M0 IRS III	H	L	L	L	L	Yes	Yes	Yes (R1)	–	Alive (89)
15	M (14y)	Yes	Pelvis minor (unfavorable, deep) 145	T2N1M0 IRS III	H	H	L	H	L	Yes	Yes	No	21	Died (44)
16	M (11y)	No	Left lower limb (unfavorable-superficial) 100	T2N0M0 IRS III	L	L	L	H	L	Yes	No	No	6	Alive (55)
17	M (10y)	No	Right chest (unfavorable, superficial) 150	T2N0M0 IRS II	H	L	L	L	H	Yes	No	Yes (R0)	–	Alive (126)
18	F (8y)	No	Neck and mediastinum (unfavorable, deep) 55	T2N0M0 IRS III	L	L	L	L	H	Yes	Yes	No	–	Alive (56)
19	F (11y)	No	Head and neck parameningeal (unfavorable, deep) 48	T2N0M0 IRS III	L	L	L	L	H	Yes	no	no	2	Alive (75)
20	F (2y)	Yes	Retroperitoneal region (unfavorable, deep) 65	T2N0M0 IRS III	H	H	H	H	H	Yes	No	No	3	Alive (169)
21	M (14y)	No	Head and neck parameningeal (unfavorable, deep) 30	T2N0M0 IRS III	H	H	H	H	H	Yes	No	No	4	Alive (81)

Table 1, continued

Case

Sex

NF1

Primary tumor location

TNM

Survi

Cyclin

p53

OPN

Fibro

First line therapy

Relapse free

Outcome –

(age)

max size in mm

IRS

vin

nectin

CTx

RTx

survival

overall survival

stages

(completeness)

(months)

F (8y)

Yes

Head and neck parameningeal (unfavorable, deep) 53

T2N0M0 IRS III

Yes

Died (23)

M (11y)

Limb (unfavorable, deep) 90

T2N0M1 IRS IV

Yes

Yes (R2)

Died (9)

M (8y)

Limb (unfavorable, superficial) 43

T1N0M0 IRS III

Yes

Yes (R1)

Alive (108)

M (1y)

Limb (unfavorable, superficial) 69

T1N0M0 IRS III

Yes

Yes (R1)

–

Alive (42)

F (2y)

Yes

Perineum (unfavorable, superficial) 30

T1N0M0 IRS II

Yes

–

Alive (114)

Abbreviations: NF1 – neurofibromatosis type 1; F – female; M – male; Unfavorable tumor sites: head and neck parameningeal region, bladder and prostate, inner organs, retroperitoneum, extremities, perineum and other; favorable tumor sites: head and neck non-parameningeal region, urogenital tract without bladder, prostate, orbit; T1 – primary tumor limited to the organ of origin, T2 – primary tumor extending beyond the organ of origin; N0 – no lymph nodes involvement, N1 – involvement of the lymph nodes; M0 – no distant metastases, M1 – distant metastases; IRS Intergroup Rhabdomyosarcoma Study; IRS stages: II – tumor resected macroscopically completely, but with microscopic residues, III – inoperable tumor feasible to macroscopically incomplete resection or diagnostic biopsy, IV – metastatic disease; H – high immunohistochemical expression of the marker, L – low immunohistochemical expression of the marker, CTx – chemotherapy, RTx – radiotherapy; completeness of the second look (SL) surgery: R0 – microscopically complete resection, R1 – macroscopically complete, but microscopically incomplete tumor resection, R2 – macroscopically incomplete tumor resection.

ue blocks obtained before systemic treatment and containing representative MPNST samples were included in this study. Histopathological reevaluation of microscopic slides was performed by two experienced pathologists. Subsequently, IHC staining was conducted on tissue microarrays (TMA) built using the commercially available Tissue-Tek ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ Quick-Ray ${}^{\text{TM}}$ TMA System kit, of the Sakura Finetek USA, Inc. Commercially available monoclonal antibodies against cyclin D1, p53, survivin, and fibronectin were employed to detect analyzed antigens. All markers were assessed semi-quantitatively, basing on the intensity of the reaction (on a scale of + to +++) and the percentage of positively staining cells (0–5, 6–25, 26–50, $>$ 50%). Combination of these two variables resulted in the final immunohistochemical score and, eventually, markers were subclassified to high (H) and low (L) expression categories in all analyzed tumors.

2.3 Statistics

Statistical analyses were performed with the use of Statistica software (version 13, Dell, licensed to the Medical University of Gdansk). Correlations between categorical variables were estimated with two-tailed Fisher’s exact test. The study outcomes were relapse-free survival (RFS) and OS. RFS and OS were calculated from the date of diagnosis to the date of relapse diagnosis or death from any cause, respectively. Surviving patients were censored at the date of the last follow-up. The demographic and clinical statistics included age at diagnosis, gender, the presence of the clinical set of symptoms fulfilling the diagnosis of NF1, tumor location, tumor size, regional lymph node status, and IRS stage. The Kaplan-Meier (KM) method was applied to create survival curves dependent on analyzed variables. The curves were subsequently compared using F-Cox test for significant differences in survival, which is dedicated to small groups. Log-rank test was performed to compare KM curves for OS for osteopontin, because of the lack of complete observations in low osteopontin subgroup. Uni- and multivariable proportional Cox regression models were constructed to explore significant prognostic markers. The proportional hazard assumption was checked by assessing Schoenfeld residual plots. The variables with $p<$ 0.1 in univariable models were chosen for multivariate analysis. Multivariable models predicting RFS and OS were created using backward selection to eliminate nonsignificant variables at $p<$ 0.05.

2.4 Ethical approval

The study was approved by the Independent Bioethical Committee of the Medical University of Gdansk, Poland (No. NKBBN/449/2013).

3. Results

A total of 26 patients (14 males/12 females) were enrolled in the current study with a median follow-up of 51 months. Basic demographic and clinical data are presented in Table 1. The median age at diagnosis was 10.5 years (mean 9.84 years). Ten patients (38%) suffered from NF1. Five patients (19.2%) presented with tumors feasible to macroscopically complete primary resection (IRS stage II) and five patients presented with distant metastases (IRS stage IV). The remaining 16 patients (61.6%) were diagnosed with local, but highly advanced, inoperable tumors (IRS stage III). Lymph nodes involvement was diagnosed in five patients, which was based on histopathologic confirmation in three and on ultrasound and computed tomography images in two. Like in adult sarcomas, a depth of the tumor was subcategorized into superficial (9 cases, 34.6%) and deep (17 cases, 65.4%). Table 1 includes also the detailed data on the components of the therapeutic modalities administered to the patients during the 1 ${}^{\text{st}}$ line oncological treatment.

The 5-year-RFS for the whole cohort was 35.6 $\pm$ 9.7% and the 5-year-OS – 62.2 $\pm$ 10.2%. At the time of the last follow-up, 15 patients (57.7%) were alive, including 2 patients with partial remission. Fifteen individuals experienced relapse and median time to relapse in this group was 7 months. RFS for the whole study group was 19.8 months. Of 11 dead patients (42.3%), 10 died due to a progression of the disease, and 1 succumbed to complications. The median interval between diagnosis and death was 20 months.

High IHC expressions of survivin, cyclin D1, osteopontin, fibronectin, and p53 were detected in 18 (69.2%), 13 (50%), 16 (61.5%), 16 (61.5%), and 13 (50%) tumors, respectively. The examples of the immunohistochemical staining towards particular markers were shown in Fig. 1A–E of our previous publication, investigating the possible roles of survivin, cyclin D1, osteopontin, fibronectin and p53 in predicting the response to neo-adjuvant chemotherapy in pediatric MPNST [8].

Figure 1.

Immunohistochemical markers significantly influencing the relapse-free survival of MPNST patients; $p$ values – F-Cox test.

To assess potential correlations between each IHC markers and clinical variables, the expression Fisher’s exact test was employed. A significant correlation was demonstrated between cyclin D1 and osteopontin ( $p=$ 0.004). Both these markers were associated with NF1 status ( $p=$ 0.041 and $p=$ 0.037, respectively). Expression of fibronectin was more prevalent in deeply located tumors ( $p=$ 0.046). None of the investigated IHC markers was associated with IRS stage, age at diagnosis, and tumor size.

Comparison of the KM curves for RFS revealed that patients with NF1 and high expression of cyclin D1, osteopontin, fibronectin, and p53 had significantly decreased RFS (Fig. 1). On the other hand, shorter OS was associated with positive nodal status, higher IRS stage, deep tumor location, the presence of distant metastases (IRS stage IV) and high expression of cyclin D1 and p53 (Figs 2 and 3). Mean size of the tumor was 99 mm and it was not found to be significantly associated with survival neither when the cut-off of 50 mm nor 100 mm was used.

Table 2

Uni- and multivariate Cox regression analysis for overall survival in pediatric MPNST patients

Characteristics	HR (95% CI)	$p$	HR (95% CI)	$p$
Age ( $<$ 10.5 vs $>$ 10.5)	0.380 (0.110–1.311)	0.126
Sex (female vs male)	0.832 (0.241–2.875)	0.771
IRS stage (II $+$ III vs IV)	0.210 (0.061–0.728)	0.014 ${}^{*}$
Tumor size ( $<$ 50 mm vs $>$ 50 mm)	0.482 (0.126–1.855)	0.289
Tumor depth (superficial vs deep)	0.177 (0.022–1.386)	0.099
LN metastases (presence vs absence)	0.283 (0.082–0.978)	0.046 ${}^{*}$	0.212 (0.055–0.817)	0.024 ${}^{*}$
NF1 (presence vs absence)	0.245 (0.071–0.848)	0.026 ${}^{*}$
Osteopontin (L vs H)	–	–
Fibronectin (L vs H)	0.732 (0.189–2.836)	0.652
Survivin (L vs H)	0.500 (0.108–2.314)	0.375
Cyclin D1 (L vs H)	0.073 (0.009–0.570)	0.013 ${}^{*}$	0.063 (0.008–0.505)	0.009 ${}^{*}$
p53 (L vs H)	0.295 (0.078–1.116)	0.072

Abbreviations: CI – confidence interval; HR – hazard ratio, LN – lymph nodes; NF1 – neurofibromatosis type 1; L – low expression; H – high expression; ${}^{*}$ – statistically significant.

Figure 2.

Clinicopathological variables significantly influencing the overall survival of MPNST patients; $p$ values – F-Cox test. Abbreviations: NF1 – neurofibromatosis type 1.

To further investigate the significance of these prognostic markers, the univariable and multivariable Cox proportional hazards models were built (Tables 2 and 3). Of the investigated IHC markers only survivin was not associated with survival in either uni- or multivariate analysis. Cyclin D1 and p53 were identified as independent variables associated with RFS, whereas cyclin D1 and nodal status were independently correlated with OS. Interestingly, well established clinical prognostic factors (e.g. IRS stage, NF1 status, size and location of the tumor) did not maintain their significance in multivariate analysis. Osteopontin was not included in the Cox models for OS prediction, due to the 100% survival rate of patients, whose tumors expressed low levels of osteopontin.

Table 3

Uni- and multivariate Cox regression analysis for relapse-free survival in pediatric MPNST patients

Characteristics	HR (95% CI)	$p$	HR (95% CI)	$p$
Age ( $<$ 10.5 vs $>$ 10.5)	0.374 (0.136–1.034)	0.058
Sex (female vs male)	0.827 (0.300–2.281)	0.714
IRS stage (II $+$ III vs IV)	0.483 (0.153–1.525)	0.215
Tumor size ( $<$ 50 mm vs $>$ 50 mm)	1.317 (0.474–3.654)	0.597
Tumor depth (superficial vs deep)	0.281 (0.079–0.991)	0.048 ${}^{*}$
LN metastases (presence vs absence)	0.587 (0.188–1.834)	0.360
NF1 (presence vs absence)	0.485 (0.181–1.300)	0.150
Osteopontin (L vs H)	0.133 (0.029–0.597)	0.009 ${}^{*}$
Fibronectin (L vs H)	0.391 (0.126–1.218)	0.105
Survivin (L vs H)	0.133 (0.029–0.597)	0.587
Cyclin D1 (L vs H)	0.105 (0.029–0.386)	0.001 ${}^{*}$	0.086 (0.021–0.352)	0.001 ${}^{*}$
p53 (L vs H)	0.379 (0.136–1.053)	0.063	0.310 (0.096–0.999)	0.050 ${}^{*}$

Figure 3.

Immunohistochemical markers significantly influencing overall survival of MPNST patients; $p$ values – F-Cox test.

4. Discussion

MPNST is an aggressive, rare malignancy and our knowledge about its biology and factors influencing its course and outcome are sparse. Importantly, there are only few cohort studies of pediatric MPNST published, making our understanding of MPNST occurring in children even more difficult [4, 13]. In the current study, we aimed to establish the prognostic significance of five markers assessed by immunohistochemistry in the tumor tissue at diagnosis: osteopontin, fibronectin, survivin, cyclin D1, and p53. We demonstrated that high immunohistochemical expression of cyclin D1 and p53 were independently associated with poorer outcomes in pediatric MPNST. Moreover, other investigated markers, such as osteopontin and fibronectin could serve as ancillary prognostic markers and their value should be validated in further studies.

Osteopontin is an extracellular matrix protein participating in various biological functions, including bone remodeling, immunoregulation, angiogenesis and regulation of apoptosis [14]. SPP1, a gene encoding osteopontin, is the most upregulated gene in MPNST compared to benign plexiform neurofibromas in patients with NF1, which explains the association between osteopontin expression and presence of NF1 in our cohort [2]. Osteopontin was a significant prognostic factor for RFS and OS in univariable models, which is consistent with studies conducted on adult soft tissue sarcomas [15, 16]. In our previous work, we revealed that low expression of osteopontin may be associated with chemo-sensitivity of MPNST [8]. Worse outcomes of tumors expressing osteopontin may result from its actions in cancer cells dissemination since all metastatic tumors in our cohort demonstrated positive IHC staining for this protein, however; this finding lacked statistical significance ( $p=$ 0.101). Osteopontin-induced upregulation of autophagy and inhibition of apoptosis may contribute to chemo-resistance in some cancers. It has been suggested that, by silencing of osteopontin expression, it is possible to enhance cytotoxic effects of chemotherapy and potentially diminish metastatic potential of the tumor [17].

Another matrix component, fibronectin, is frequently altered in cancer and participates in tumorigenesis in multiple aspects [12]. To our knowledge, this is the first attempt to evaluate the impact of fibronectin on prognosis of MPNST. Fibronectin was more abundant in deep-seated tumors, which emphasizes its role in cancer growth and expansion. Consequently, its low levels were found correlated with prolonged RFS of patients. This glycoprotein is not only a potential prognostic factor but also an emerging therapeutic target. Interestingly, Hemmerle et al. used the L19 antibody against extra-domain B of fibronectin to detect the tumor in a patient with MPNST by single-photon emission computed tomography imaging procedure [18]. The immunocytokine composed of that antibody and tumor necrosis factor (L19-TNF) is about to enter phase III clinical trials in combination with doxorubicin as a treatment for soft tissue sarcomas [19].

BIRC5 gene is another differentially expressed gene between benign plexiform neurofibromas and MPNST. It encodes survivin, which belongs to the inhibitor of apoptosis protein (IAP) family [20]. Survivin expression is frequently associated with aggressive tumor phenotype and worse outcomes in a variety of human malignancies, including MPNST and other soft tissue sarcomas [20, 21, 22]. Alaggio et al. showed that survivin expression, assessed either by IHC or real-time PCR, had a prognostic value in pediatric MPNST [4]. They also observed a correlation between high levels of survivin mRNA with high grade and high mitotic rate, emphasizing the potential role of survivin in cell division. However, we did not observe the influence of survivin expression on outcomes in our cohort. These discrepancies could be explained by different criteria used for IHC evaluation and the lack of molecular confirmation of survivin mRNA expression in our study, which might have impacted our calculations.

Disruption of the cell cycle regulation is one of the hallmarks of cancer. Loss of TP53 together with deletions of cell cycle progression inhibitor CDKN2A are important genetic abnormalities driving the development of MPNST [3]. The chemokine receptor CXCR4 and its ligand, CXCL12 promote cell cycle progression in NF1-associated MPNST via cyclin D1 induction [23]. We demonstrated that IHC tumor expression of both p53 and cyclin D1 were independent prognostic factors for RFS, whereas high expression of cyclin D1 was an independently associated with shortened OS. Many studies are consistent with our results and suggest the prognostic significance of these proteins in adult MPNST cohorts [24, 25, 26, 27]. However, some authors failed to confirm prognostic significance of p53 and cyclins [28, 29, 30]. Cyclin D1 has been recently investigated as a diagnostic marker in various small round blue cell tumors in children and adolescents. Especially, a diffuse nuclear expression of cyclin D1 could be used as an ancillary IHC marker in Ewing sarcoma and neuroblastoma/ganglioneuroblastoma [31]. In our cohort 50% of MPNST cases demonstrated high nuclear IHC staining of cyclin D1, thus it should be considered in differential diagnosis of poorly differentiated pediatric tumors expressing this protein. Importantly, cyclin D1 overexpressing cancers are potential candidates for targeting cyclin D1 and cyclin-dependent kinases, especially in advanced, metastatic cases resistant to standard therapy [31, 32]. On the other hand, aberrant p53 expression may drive cancer resistance to CTx and RTx [33]. Resistance to doxorubicin, a first-line cytotoxic drug in MPNST, is known to be associated with TP53 mutations in several malignancies [34]. Hence, the worse prognosis of MPNST patients with abnormal p53 expression may be partly associated with its adverse impact on treatment effects. Indeed, in our previous study, we demonstrated that high expression of cyclin D1 and p53 in pediatric MPNST patients is associated with worse response to neoadjuvant CTx.

Besides the above mentioned IHC markers, we also assessed the influence of multiple clinical variables on MPNST prognosis. Only age at diagnosis and presence of NF1 had an impact on RFS, whereas the presence of NF1, IRS stage, depth of tumor, and the presence of regional lymph node metastases influenced the OS. The latter factor maintained its significance in the multivariable model.

In the meta-analysis of $>$ 1800 patients, NF1-associated MPNSTs did not differ molecularly from sporadic MPNSTs and their disease-specific survivals were similar, whereas OS was shortened in NF1 patients, probably due to increased mortality from other causes [35]. In the study by Fan et al. tumor size, NF1 status, stage, MDM2 and p53 expression were correlated with survival, but none of them were identified as an independent prognostic factor [25]. One of the most interesting findings in our study was the characterization of lymph node metastases as independent variables correlated with OS. Data regarding the association between nodal status and survival in MPNST are lacking. Nevertheless, in the recent years, sentinel lymph node biopsy to identify pediatric sarcomas metastases (including MPNST) has been introduced in some institutions [36]. Proper lymph node staging in these tumors may provide crucial information about prognosis and adjust the therapy for the presence of more extensive disease [37].

4.1 Strengths and limitations of the study

Our study enrolled a high portion of children diagnosed and treated due to MPNST in Poland over the 21 years. These were all patients, in whom both full data on the clinical course and good-quality pathological material, were available. We managed to identify potentially useful, independent markers of survival when adjusted to various demographic and clinical factors. Nevertheless, this cohort remains relatively small, which might have disturbed statistical analyzes, since the hazard ratios in the proportional Cox regression models were highly dispersed within 95% confidence interval. Finally, we did not evaluate all of the potential confounders, that might have affected outcomes.

5. Conclusions

MPNST remains an ominous malignancy with the hardly predictable outcome. This is the first study aiming to assess prognostic significance of selected IHC markers for MPNST in Polish pediatric patients. We found that tumor expression of cyclin D1 and p53 may predict RFS, whereas cyclin D1 and regional lymph node status are independent variables associated with OS in children with MPNST. Osteopontin is another interesting marker which might become useful in clinical practice, however multicenter studies of larger cohort should be implemented to fully investigate its significance. The results of the study may contribute to the development of novel therapies targeted against overexpressed proteins improving the outcome of advanced inoperable MPNST in children.

Footnotes

Conflict of interest

The authors report no conflicts of interest in this work.

References

Zhou

Song

Zheng

Chen

Zhang

and Yang

, Prognostic roles for fibroblast growth factor receptor family members in malignant peripheral nerve sheath tumor, Oncotarget 7 (2016).

Thomas

L.E.

Winston

Rad

Mort

Dodd

K.M.

Tee

A.R.

McDyer

Moore

Cooper

D.N.

and Upadhyaya

, Evaluation of copy number variation and gene expression in neurofibromatosis type-1-associated malignant peripheral nerve sheath tumours, Hum. Genomics 9 (2015), 3.

Kim

Stewart

D.R.

Reilly

K.M.

Viskochil

Miettinen

M.M.

and Widemann

B.C.

, Malignant peripheral nerve sheath tumors state of the science: leveraging clinical and biological insights into effective therapies, Sarcoma 2017 (2017), 1–10.

Alaggio

Turrini

Boldrin

Merlo

Gambini

Ferrari

Dall’Igna

Coffin

C.M.

Martines

Bonaldi

De Salvo

G.L.

Zanovello

and Rosato

, Survivin expression and prognostic significance in pediatric malignant peripheral nerve sheath tumors (MPNST), PLoS One 8 (2013), e80456.

Porter

D.E.

Prasad

Foster

Dall

G.F.

Birch

and Grimer

R.J.

, Survival in malignant peripheral nerve sheath tumours: A comparison between sporadic and neurofibromatosis type 1-associated tumours, Sarcoma 2009 (2009), 1–5.

Gachiani

Kim

Nelson

and Kline

, Surgical management of malignant peripheral nerve sheath tumors, Neurosurg. Focus 22 (2007), E13.

Valentin

Le Cesne

Ray-Coquard

Italiano

Decanter

Bompas

Isambert

Thariat

Linassier

Bertucci

Bay

J.O.

Bellesoeur

Penel

Le Guellec

Filleron

and Chevreau

, Management and prognosis of malignant peripheral nerve sheath tumors: The experience of the French Sarcoma Group (GSF-GETO), Eur. J. Cancer 56 (2016), 77–84.

Karpinsky

Krawczyk

M.A.

Izycka-Swieszewska

Fatyga

Budka

Balwierz

Sobol

Zalewska-Szewczyk

Rychlowska-Pruszynska

Klepacka

Dembowska-Baginska

Kazanowska

Gabrych

and Bien

, Tumor expression of survivin, p53, cyclin D1, osteopontin and fibronectin in predicting the response to neo-adjuvant chemotherapy in children with advanced malignant peripheral nerve sheath tumor, J. Cancer Res. Clin. Oncol. (2018), 1–11.

Joerger

A.C.

and Fersht

A.R.

, The p53 pathway: Origins, inactivation in cancer, and emerging therapeutic approaches, Annu. Rev. Biochem. 85 (2016), 375–404.

10.

Qie

and Diehl

J.A.

, Cyclin D1, cancer progression, and opportunities in cancer treatment, J. Mol. Med. 94 (2016), 1313–1326.

11.

Karpinsky

Fatyga

Krawczyk

M.A.

Chamera

Sande

Szmyd

Izycka-Swieszewska

and Bien

, Osteopontin: its potential role in cancer of children and young adults, Biomark. Med. 11 (2017), 389–402.

12.

Wang

J.P.

and Hielscher

, Fibronectin: how its aberrant expression in tumors may improve therapeutic targeting, J. Cancer 8 (2017), 674.

13.

Carli

Ferrari

Mattke

Zanetti

Casanova

Bisogno

Cecchetto

Alaggio

De Sio

Koscielniak

Sotti

Treuner

and Treuner

, Pediatric malignant peripheral nerve sheath tumor: the italian and german soft tissue sarcoma cooperative group, J. Clin. Oncol. 23 (2005), 8422–8430.

14.

Weber

G.F.

, The metastasis gene osteopontin: a candidate target for cancer therapy, Biochim. Biophys. Acta 1552 (2001), 61–85.

15.

Bache

Kappler

Wichmann

Rot

Hahnel

Greither

Said

H.M.

Kotzsch

Würl

Taubert

and Vordermark

, Elevated tumor and serum levels of the hypoxia-associated protein osteopontin are associated with prognosis for soft tissue sarcoma patients, BMC Cancer 10 (2010), 132.

16.

Hahnel

Wichmann

Greither

Kappler

Würl

Kotzsch

Taubert

Vordermark

and Bache

, Prognostic impact of mRNA levels of osteopontin splice variants in soft tissue sarcoma patients, BMC Cancer 12 (2012), 131.

17.

Zhao

Chen

Alam

Cui

Suen

K.C.

Soo

A.P.

Eguchi

and Ma

, The role of osteopontin in the progression of solid organ tumour, Cell Death Dis. 9 (2018), 356.

18.

Hemmerle

Probst

Giovannoni

Green

A.J.

Meyer

and Neri

, The antibody-based targeted delivery of TNF in combination with doxorubicin eradicates sarcomas in mice and confers protective immunity, Br. J. Cancer 109 (2013), 1206–1213.

19.

Probst

Kopp

Oxenius

Colombo

M.P.

Ritz

Fugmann

and Neri

, Sarcoma eradication by doxorubicin and targeted TNF relies upon CD8+ t-cell recognition of a retroviral antigen, Cancer Res. 77 (2017), 3644–3654.

20.

Jaiswal

P.K.

Goel

and Mittal

R.D.

, Survivin: A molecular biomarker in cancer, Indian J. Med. Res. 141 (2015), 389–397.

21.

Kolberg

Høland

Lind

G.E.

Ågesen

T.H.

Skotheim

R.I.

Sundby Hall

Mandahl

Smeland

Mertens

Davidson

and Lothe

R.A.

, Protein expression of BIRC5, TK1, and TOP2A in malignant peripheral nerve sheath tumours – A prognostic test after surgical resection, Mol. Oncol. 9 (2015), 1129–1139.

22.

Würl

Kappler

Meye

Bartel

Köhler

Lautenschläger

Bache

Schmidt

and Taubert

, Co-expression of survivin and TERT and risk of tumour-related death in patients with soft-tissue sarcoma, Lancet 359 (2002), 943–945.

23.

Chen

Patel

Zhang

Chau

Cho

Lim

Lazar

A.J.

Creighton

C.J.

Bolshakov

McKay

R.M.

Lev

L.Q.

and Parada

L.F.

, CXCR4/CXCL12 mediate autocrine cell- cycle progression in NF1-associated malignant peripheral nerve sheath tumors, Cell 152 (2013), 1077–1090.

24.

Brekke

H.R.

Kolberg

Skotheim

R.I.

Hall

K.S.

Bjerkehagen

Risberg

Domanski

H.A.

Mandahl

Liestøl

Smeland

Danielsen

H.E.

Mertens

and Lothe

R.A.

, Identification of p53 as a strong predictor of survival for patients with malignant peripheral nerve sheath tumors, Neuro. Oncol. 11 (2009), 514–528.

25.

Fan

Yang

and Wang

, Clinical and molecular prognostic predictors of malignant peripheral nerve sheath tumor, Clin. Transl. Oncol. 16 (2014), 191–199.

26.

Holtkamp

Atallah

Okuducu

A.-F.

Mucha

Hartmann

Mautner

V.-F.

Friedrich

R.E.

Mawrin

and von Deimling

, MMP-13 and p53 in the progression of malignant peripheral nerve sheath tumors, Neoplasia 9 (2007), 671–677.

27.

Kim

J.R.

Moon

Y.J.

Kwon

K.S.

Bae

J.S.

Wagle

T.K.

Kim

K.M.

Park

H.S.

Lee

J.-H.

Moon

W.S.

Lee

Chung

M.J.

and Jang

K.Y.

, Expression of SIRT1 and DBC1 is associated with poor prognosis of soft tissue sarcomas, PLoS One 8 (2013), e74738.

28.

Cunha

Caruso

Faria

Silva

Pires

Geller

Lopes

and Moura-Neto

, Malignant peripheral nerve sheath tumors: clinicopathological aspects, expression of p53 and survival, Clinics 67 (2012), 963–968.

29.

Watanabe

Oda

Tamiya

Kinukawa

Masuda

and Tsuneyoshi

, Malignant peripheral nerve sheath tumours: high Ki67 labelling index is the significant prognostic indicator, Histopathology 39 (2001), 187–197.

30.

Zhang

Chang

F.-Y.

Zhou

W.-Y.

and Yang

J.-L.

, The prognostic value of C-X-C motif chemokine receptor 4 in patients with sporadic malignant peripheral nerve sheath tumors, Chin. J. Cancer 36 (2017), 80.

31.

Magro

Salvatorelli

Alaggio

D’Agata

Nicoletti

Di Cataldo

and Parenti

, Diagnostic utility of cyclin D1 in the diagnosis of small round blue cell tumors in children and adolescents, Hum. Pathol. 60 (2017), 58–65.

32.

Asghar

Witkiewicz

A.K.

Turner

N.C.

and Knudsen

E.S.

, The history and future of targeting cyclin-dependent kinases in cancer therapy, Nat. Rev. Drug Discov. 14 (2015), 130–146.

33.

Czapiewski

Kunc

and Haybaeck

, Genetic and molecular alterations in olfactory neuroblastoma: implications for pathogenesis, prognosis and treatment, Oncotarget 7 (2016).

34.

Hientz

Mohr

Bhakta-Guha

and Efferth

, The role of p53 in cancer drug resistance and targeted chemotherapy, Oncotarget 8 (2017), 8921–8946.

35.

Kolberg

Holand

Agesen

T.H.

Brekke

H.R.

Liestol

Hall

K.S.

Mertens

Picci

Smeland

and Lothe

R.A.

, Survival meta-analyses for > 1800 malignant peripheral nerve sheath tumor patients with and without neurofibromatosis type 1, Neuro. Oncol. 15 (2013), 135–147.

36.

Dall’Igna

De Corti

Alaggio

and Cecchetto

, Sentinel node biopsy in pediatric patients: the experience in a single institution, Eur. J. Pediatr. Surg. 24 (2014), 482–487.

37.

Wagner

L.M.

Kremer

Gelfand

M.J.

Sharp

S.E.

Turpin

B.K.

Nagarajan

Tiao

G.M.

Pressey

J.G.

Yin

and Dasgupta

, Detection of lymph node metastases in pediatric and adolescent/young adult sarcoma: Sentinel lymph node biopsy versus fludeoxyglucose positron emission tomography imaging-A prospective trial, Cancer 123 (2017), 155–160.

Immunohistochemical assessment of cyclin D1 and p53 is associated with survival in childhood malignant peripheral nerve sheath tumor

Abstract

BACKGROUND:

OBJECTIVE:

METHODS:

RESULTS:

1. Introduction

2. Material and methods

2.1 Patients

2.2 Immunohistochemistry

Table 1 The clinical data and expression pattern of immunohistochemical markers

2.4 Ethical approval

3. Results

4.1 Strengths and limitations of the study

5. Conclusions

Footnotes

Conflict of interest

References

Table 1
The clinical data and expression pattern of immunohistochemical markers