Discussion
In certain cases of congenital fetal neck tumors, establishing a definite prenatal diagnosis might be challenging or even impossible. Despite all advances in imaging techniques available today it remains challenging to delineate the underlying cause of such etiologically heterogeneous tumors in utero. Both, the infant and physicians, as well as the parents, will benefit from the most accurate clinical and, if possible, molecular pathological classification of the etiology of the tumor in order to be able to plan further interdisciplinary treatment. If the final diagnosis in this case presented had been known earlier, a number of therapeutic approaches could have been avoided.
During fetal life tumors of the neck tend to impair fetal swallowing, often resulting in polyhydramnios.8 Another serious thread potentially arises from local compression phenomena.6 Lymphatic malformations of the fetal neck with lymphangiomas constitute the most prevalent fetal neck mass.1,2,4,8 In the last two decades there were a number of attempts to classify these lesions according to their etiological origin8,10, morphological appearance5,10, histological features5,7 or postnatal course.10All these different classifications can coexist and their characteristics occur side by side. A common and generally accepted classification of lymphatic malformations is merely based on histological criteria. There are four distinct histological types of lymphangiomas consisting of capillary or cavernous lymphangioma, cystic hygroma and vasculolymphatic malformation.7 However, Eivazi et al. stressed that this traditional histological classification is obsolete as with increasing frequency, it has been shown, that anterolateral lymphangiomas receive their growth stimuli by activating missense mutations characterizing regional-located tumors in PROS.10–12 Axt-Fliedner et al. mentioned, that theories of origin of lymphatic malformations include the failure in connection between the lymphatic and the venous system and that they could be suitable to explain the origin of lymphangiomas embryologically.5,8 However, they also emphasized, that in cases with late presenting, regional-located, anterolateral lymphangiomas, this theory has to be questioned.8 It therefore might be that late-presenting lymphangiomas beyond 30 weeks of pregnancy rather arise as a result of direct alteration in DNA from a post-zygotic PIK3CA mutation.
Depending on time of post-zygotic cell division in embryogenesis, either somatic or, even rarer, constitutional mutations are the result.6,9,15 As one of the most frequent hot-spot missense mutations, c.1633G>A;p.Glu545Lys was identified in the helical domain of PIK3CA.6,16 Hot spot mutations, activating, gain-of-function variants, show significantly elevated biological and biochemical activities.14,16 These are often associated with higher numbers of transformed cell foci suggesting more rapid cell proliferation.17
In 2015, diagnostic criteria have been defined by Keppler-Noreuil et al. to standardize entities, that originate from PIK3CA mutations, and to simplify the diagnostic work-up.14 Accordingly, in the present case, the regional-located and sporadic overgrowth of predominantly lymphatic tissue of the fetal neck, already detected prenatally and classified as a lymphangioma, combined with the molecular biological detection of the somatic mutation in the PIK3CA gene with low‐level mosaics, clearly point to the definitive diagnosis of a disorder within the spectrum of PROS.
Due to their inexorable growth throughout all tissue layers, diagnosis and treatment of lymphangiomas within PROS disorders is often difficult.3,4,8 However, prenatal ultrasound remains the primary diagnostic method of choice in the detection of congenital tumors. Additional volume ultrasound, combined with color Doppler interrogation, may act as an adjunct in establishing the most likely diagnosis as it potentially provides valuable information regarding the spatial relationship to surrounding cervical structures and general extent of PROS disorders.19 The value of an additional fetal MRI to accomplish the diagnostic work-up and to give further essential information, particularly in unfavorable maternal conditions (high Body-Mass-Index) and in complex fetal tumors with multiloculated masses in a variety of affected anatomical regions has been highlighted by a number of publications.8,19,20 Combined with information from 3D volume ultrasound, it can increase the accuracy of prenatal diagnosis including location, architecture of tissue, volume or intracranial or thoracic spread.18,19
Histopathologically, in the present case, the cyst wall-coating cells presented themselves without expression of D2-40, a lymphatic endothelial marker, but showed clear CD31 expression, characterizing blood vascular endothelial cells, so that the finding was classified most likely as a cavernous hemangioma due to the complex histologic pattern (Fig. 3). But considering the clinical aspects, a lymphangioma should be assumed. The D2-40 negativity does not exclude a lymphangioma. A distinct histological or immunohistochemical differentiation between a lymphangioma and a hemangioma was not possible. Due to the histological expression pattern (CD31+/D2-40), it is most likely, that endothelial cells harbor the PIK3CA mutation.13,21
All case reports published so far highlight the complexity of prenatal diagnosis of somatic mosaic mutations.9,18 In 2014, Keppler-Noreuil et al. also defined testing eligibility criteria for somatic PIK3CA mutations.17 Adequate tissue sampling is already a challenge postnatally – and, above all, as a highly invasive procedure, prenatally.17,18 Maybe cultured amniocytes obtained by amniocentesis for prenatal diagnosis might be promising and are less invasive.9,17,18
Using NGS, the sensitivity of genetic testing has improved significantly and molecular diagnosis of somatic overgrowth has become feasible.15–17 By now, mosaic levels can be detected with a variant allele frequency (VAF) down to 1 %.16In the tissue received in this case from multiple biopsies, the activating missense mutation-hotspot c.1633G>A;p.Glu545Lys with low-grade somatic mosaic and a VAF of 6.5 % was detected. This is one of the most frequent mutation hotspots in the PIK3CA gene, causing somatic overgrowth in PROS. Only with the change from the pure morphologic description of lymphangiomas to the recognition of their molecular causes, they can be categorized into the spectrum of PROS diseases.
The identification of a component of the PI3K-AKT-mTOR signaling pathway can partly explain the effect of the mTOR-inhibitor Sirolimus, which resulted temporarily in a minimal size reduction. The targeted application of a PIK3-inhibitor could be more promising, even as a serious alternative treatment to surgical debulking. Finally, obtaining tissue for histopathological and molecular genetic examination to confirm the definitive diagnosis is mandatory.18
From a therapeutic perspective, debulking procedures were the most promising approaches yet.3–6,8,13 However, in the future, a shift from surgical debulking to personalized, targeted pharmacological intervention has been proposed.15,18Pharmacological therapy approaches are sclerosing therapy of fetal neck lymphangiomas with Bleomycin or Picibanil (OK-432), maybe prenatal. In addition, various molecule inhibitors, targeting different components of the PI3K-AKT-mTOR signaling pathway, are under clinical investigation: mTOR-inhibitors like Sirolimus (SRL or Rapamycin) or Everolimus (RAD-001) and recently, PIK3-inhibitors like Alpelisib (BYL719). Due to the increasing understanding of molecular biology of PROS disorders, targeted therapies are highly promising and less invasive. Moreover, it can improve quality of life of affected patients.2,4–6,10,13,22
Despite all efforts, there is a high risk of recurrence.4,10 Fetal prognosis of congenital tumors of the neck depends largely on the nature of lesion, on their location and size, the affected surrounding structures as well as the presence of other anomalies.7,20