Genetic features of neuroblastic tumors associated with opsoclonus-myoclonus syndrome opens up the possibility for detection in peripheral blood

KEYWORDS:

• Circulating tumor DNA
• neuroblastic tumors
• opsoclonus-myoclonus syndrome
• MYCN
• segmental chromosome alteration

Opsoclonus–myoclonus syndrome (OMS) is a rare paraneoplastic, postinfectious, or parainfectious or idiopathic acute neurological syndrome in children and adults. OMS is characterized by involuntary, omnidirectional eye movements (which may continue during sleep), and/or by myoclonus and ataxia of the limbs (‘dancing feet’), the trunk, and the eyelids [1,2]. A pathogenic role of the surface-binding autoantibodies against cerebellum and brainstem by cross-reactivity between onconeural antigens is suspected, although specific immunologic and genetic mechanisms are unclear [3–9].

In most patients, the neurological disorder precedes cancer detection. In children, at least 50% of OMS are associated with neuroblastic tumors (NT), whereas in adults, about 15% are associated with a tumor, mostly small cell lung cancer [2,3,5,10]. OMS of childhood is found in 2–3% of patients with NT with a median age of presentation of 18–22 months [6,10]. Multicenter collaborative studies such as the European consortium trial would help to determine more precisely the actual prevalence of OMS-associated NT [11]. In most OMS cases, the tumor is small or occult, with paravertebral location and presents normal catecholamine metabolites. Current investigative techniques and imaging procedures carry associated risks of, amongst others, repeated sedation and radiation on small children, and often provide inconclusive results [3,5,10,12]. NT in patients with OMS are localized, usually a ganglioNB or differentiating NB, have a tendency to cellular maturation and are associated with a good prognosis [5–7,10]. In 90% of OMS-associated NT, no metastases are found and the results of oncological treatment are usually very good [5,10]. It has been shown that concomitant tumors are usually aneuploid and have a single copy of the MYCN oncogene (no MNA), but few descriptions of other gains or losses of chromosome fragments (segmental chromosomal aberrations -SCA-), gains or losses of whole chromosomes (numerical chromosome alteration profile -NCA-), or mutational screening have been reported [13]. In NT without OMS, seven recurrent SCA (typical SCA) are currently regarded as prognostically meaningful, and analysis of chromosome 1p, 11q, gain of 17q and MYCN amplification (MNA) is used for patient stratification and subsequent therapy decision-making [14]. Other SCA are called atypical SCA. NT theranostics (a combination of diagnostics and therapy) is moving toward next-generation sequencing in order to detect actionable targets [15]. Disseminated tumor cells and circulating tumor DNA in cell-free plasma (DTC/ctDNA) are excellent sources for tumor genome analyses that have been recently used in a cancer theranostics and examination of cancer biology [16,17].

Most often, the oncological treatment consists of surgery only, and sometimes chemotherapy. The pharmacological treatment of OMS includes glucocorticoids, high doses of immunoglobulins, cytotoxic drugs, and rituximab. Half the patients have neurological sequelae after neurological and oncological treatment [12]. Although the long-term neurological outcome in patients with OMS is variable, the interval between the onset of OMS and the detection and initial therapy of NT tends to be longer in patients with neurological sequelae than in those without neurological sequelae [10,18].

Based on the clinical and pathological data reviewed and the characteristics collected when tumor material was referred to the Spanish Reference Center for NB Biological and Pathological studies (Department of Pathology, University of Valencia/INCLIVA), we support the findings that the majority of patients with OMS and NT very rarely have unfavorable prognostic features [1,6,13]. From the whole cohort of primary tumors diagnosed between 1997 and 2014, 15 presented OMS associated with NT, giving a frequency of 2.3% (15 out of 646). Essentially, all patients were children, mainly older than 12 months (60%, 9/15), almost equally distributed between the genders (7 males vs. 8 females) and diagnosed with nonmetastatic disease, especially for stage 1 (66.7%). All the patients are doing well and no progression or relapse has been reported so far after an average follow-up of 43 months.

Histologically, 80% of the tumors corresponded to poorly differentiated NB, the remaining were differentiating NB and ganglioNB (2 and 1 cases, respectively). This frequency is different to that reported in the literature, and may be due to intratumoral heterogeneity, a well-known fact in NT, and as such would not affect the generalizability of the results with respect to the genetic findings [6,7,19]. All tumors were triploid. As expected from the less aggressive course of the disease, none of the tumors presented homogenous MNA (all tumor cells showed more than four-fold increase in the MYCN signal number compared to the reference) by Fluorescence In Situ Hybridization/SNP array techniques. About 10 out of the 15 OMS-associated NT plus 264 out of 631 NT from the control cohort contained ≥60% of neuroblasts and it was possible to perform an SNP array to detect the presence of chromosome alterations. The OMS-associated NT profiles obtained were divided 50/50 between tumors with typical/atypical SCA (4/1) and tumors with NCA plus focal SCAs (3/5) (SCA 0.2–2 MB in size). The NT control cohort profiles obtained were divided into tumors with typical/atypical SCA (66.4%, 165/11), tumors with NCA (32%), and tumors with neither SCA nor NCA (1.5%). Typical SCA frequencies in OMS-associated NT for 1p-, +2p, or 11q− were 20% for each alteration (1 case each) and 40% for +17q (2 cases), while for atypical SCAs, +2pq, +5pq, or +16q, were 20% for each alteration (in 1 case each). Typical SCA frequencies in the NT control cohort were 51.7% (91 cases) for 1p−, 35.8% (63 cases) for +2p, 44.3% (78 cases) for 11q−, and 77.3% (136 cases) for +17q, while atypical SCAs were 0.6% (1 case) for +2pq, 1% (2 cases) for +5pq and 5.7% (10 cases) for +16q. Of OMS patients with SCA tumors, 40% (2/5) were older than 18 months compared to 62.3% (104/167) in the NT control cohort. Of OMS patients with SCA tumors, 60% (3/5) were stage 1 versus 13.8% (22/160) in the NT control cohort [19]. SCA cases presented either heterogonous MNA (hetMNA, coexistence of amplified as well as nonamplified tumor cells in the same tumor) or no MNA (3 and 2, respectively) and with less than 4 large SCA. hetMNA is an infrequent event with unknown prognosis reported in only 2.6% (28 cases) of our total cohort of NT [19]. In one hetMNA tumor, two amplifications at chromosome region 1p34.2 and 1p31.2 were detected.

We are currently setting up a protocol to analyze liquid biopsies of patients with NT with the aim of achieving an earlier diagnosis of this disease. This technique would provide a noninvasive method for screening for an occult NT in children with OMS with promising perspective and benefits especially for the group of patients with tumors with either large or focal SCA [18]. We found different frequencies of SCA profiles between OMS-NT and control cohort NT for patients older than 18 months (40% versus 62.3%) and for tumors with stage 1 (60% versus 13.8%) that help to broaden our present knowledge of their biology. Appropriate statistics to determine the significance of these differences and the different frequencies of hetMNA could not be performed due to the small size of the cohort. Most importantly, it was found that large and/or focal SCA, predominated (8 out of the 10 cases) within the diverse profiles presented in the analyzed cohort, and could therefore be detected in DTC/ctDNA). Detection of SCA in ctDNA from serum is achievable in NT patients [20], nonetheless the sensitivity, specificity, and reproducibility of the technique in the detection of the genetic alterations should be tested in OMS-associated NT, since low disease stage, small size, and the potential existence of intratumor genetic heterogeneity might be a hurdle. In cases where no such genetic alterations are detected, discarding an NT by standard procedures would still be necessary.

In conclusion, because (1) we found SCAs, amplified regions, acquired loss of heterozygosity events, and minor aberrations affecting <2Mb in NT-positive OMS, (2) the major challenges are the neurological sequelae, and (3) early detection and treatment of NT with OMS improve neurological outcome, we propose an international collaboration for genome analyses of tumor samples in addition to genome analyses in DTC/ctDNA. The exact diagnosis of OMS-associated NT with noninvasive blood test can be an excellent and cost-effective way of assessing the tumor genomic profile and accelerating adequate treatment as an important issue in childhood OMS clinical practice.

Declaration of interest

This work was supported by FIS [PI14/01008] and RTICC [RD12/0036/0020], grants from the ISCIII & FEDER (European Regional Development Fund), Spain. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Acknowledgement

The authors thank the Spanish Society of Pediatric Hemato-Oncology (SEHOP) for patient data management.