The current study described the molecular genetic analysis of patients with lateralized overgrowth. By utilizing deep sequencing of the genes related to the PIK3CA/AKT/mTOR and RAS/MAPK pathways, we could expand the known genetic variants of this condition. Moreover, by comparing the VAF in the affected tissues to that in the normal tissues, we could detect mutations with low levels of mosaicism.
Most of our patients were identified as having PROS, but KRAS, PTEN, MAP2K3, GNAQ, TBC1D4, and TEK mutations were also identified. Careful physical examination of dysmorphic features can help to identify a recognizable pattern in some cases, but differential diagnosis is challenging in other cases because of the variety of overlapping phenotypes between the disorders with lateralized overgrowth. The degree of mosaicism varies widely from individual to individual, and the diverse phenotypes also makes it difficult to diagnose. In these cases, genome testing and methylation studies can enable differential diagnosis. To improve the genetic detection rate, the affected tissue, such as vascular malformation tissue, should be obtained. In addition, as the mutation exists in a mosaic pattern, deep sequencing using massively parallel sequencing techniques needs to be applied. Therefore, targeted gene panel testing with deep sequencing is preferential to whole exome sequencing for the genetic diagnosis of lateralized overgrowth. The overall molecular diagnostic yield of OS has been reported as up to 45% in the affected tissues [20]. In the present study, a customized panel was used to sequence as many genes as possible that are related to the PIK3CA/AKT/mTOR and RAS/MAPK pathways (the mean depth of coverage was 878 reads per base) and thereby enhance the overall genetic detection rate. Consequently, we were able to identify genetic defects in all of our 14 patients. Methylation defects may be the cause in some patients, so methylation studies should be considered if the causative gene is not identified by exome sequencing.
The levels of mosaicism can be as low as < 5% in the affected tissues of patients with PROS [21]. In the present study, additional deep-sequencing tests with blood samples were helpful in diagnosing seven patients with low levels of mosaicism. Germline filtering also help to identify germline mutations, as in patient 10 of our study [22].
The association between the genetic variants and overgrowth in the patients in the present study is supported by multiple observations. All the variants found in the present study have not been reported in the normal population but, importantly, have been identified in the affected tissues of the enrolled patients. Furthermore, using multiple in silico prediction tools and per the ACMG guidelines, these variants are predicted to alter the protein function [14]. Besides PIK3CA, variants related to the RAS/MAPK pathway were identified in KRAS and MAP2K3 mutations in three patients. Mosaic Ras/MAPK variants are related to vascular malformation [18]. The MAP2K3 variant found in our study has not been reported, but considering the cell proliferation function and correlation with tumorigenesis of MAP2K3 [23], it could have a relation with overgrowth. The germline PTEN mutation (in patient 10) causes epidermal nevus and mild vascular malformation of the buttock and thigh that is associated with PTEN related Proteus syndrome. The high VAF of 0.7 from the blood sample in this patient might have resulted from partial deletion or second hit of the wild-type PTEN allele [24]. The GNAQ mutation in patient 12 explains the capillary malformation involving OS observed in Sturge–Weber syndrome [25–27]. The TEK mutation in patient 13 can cause multiple sporadic venous malformations [19, 28, 29]. Although the association between the TBC1D4 mutation and lateralized overgrowth in patient 14 is not clear, the same somatic mutation has been reported in large intestinal adenocarcinoma tissue [30]. TBC1D4 is a GTPase-activating protein that functions downstream of AKT and seems to regulate the proliferation of multiple cell types [31–33].
Although no significant improvement was observed in the affected areas following propranolol treatment, some patients experienced improvement of pain, range of motion, cutaneous symptoms, inflammation (ESR), and quality of life. Notably, rapid worsening of symptoms was observed in some patients after discontinuation of propranolol.. Recent research suggests that propranolol can inhibit activity of the PI3K and AKT pathways [9, 10]. Propranolol may regulate the cytokines related to angiogenesis and subtly reduce the burdens caused by vascular malformations. In this study, the ESR level decreased after administration of propranolol, which is thought to be the consequence of decreased inflammatory vascular lesions.
The mTOR inhibitor sirolimus has been clinically administered at low doses and showed a modest reduction in OS symptoms. However, 72% of participants had at least one adverse event; therefore, risk-benefit evaluations must be carefully considered [34]. The PIK3CA inhibitor alpelisib has been undergoing clinical trials for several PIK3CA-dependent tumors and PROS [4, 5, 35]. The use of alpelisib has been approved for certain advanced or metastatic breast cancer [36], but only a few reports exist regarding its application in patients with PROS. The first evidence for the use of alpelisib in patients with PROS showed promising efficacy and no substantial side effects [4]. There was an improvement in vascular tumor size, congestive heart failure, lateralized overgrowth and scoliosis [4]. The two patients with PROS in the present study participating in the alpelisib MAP [37] experienced a reduction in the degree of hypertrophy after administration for 1 year. There were no side effects in these patients following an 18-months trial of alpelisib. Nonetheless, some common side effects of alpelisib include hyperglycemia, diarrhea, nausea, fatigue, stomatitis, and pneumonitis, and patients should therefore be carefully monitored [36]. PI3K/AKT/mTOR inhibitors would also affect the metabolism of healthy cells [3], and the need for life-long therapy with these inhibitors raises concerns of unknown side effects. Further research on the treatment of patients with PROS with alpelisib is required to dispel these concerns.
Several limitations should be addressed in the present study. Although most of the identified mutations have also been found in vascular malformation tissues of other literatures, more evidence is needed to prove their association with overgrowth. The low levels of mosaicism in some patients means that the causality of the mutation remains elusive. The alleles with mosaic mutations could be present in only a subset of cells, and the causative variant can easily be missed without a precise technical procedure. The depth of sequencing was relatively low to detect all the potential mosaic variants in this study. Achieving a much higher reading depth is needed to increase the low frequency fraction variant detection sensitivity of mosaicism and avoid misinterpretation [38, 39]. As an open-label, non-randomized, observational study, the objective evaluation of the efficacy of propranolol was impossible. Although alpelisib showed promising efficacy, compared to a fixed dose of 250 mg in an adult patient, its pediatric dosage has not been established. To prove the efficacy of alpelisib in children, an appropriate dosage should be determined and administrated according to body weight.
In conclusion, customized panel-gene deep-sequencing enhanced the genetic diagnosis in patients with lateralized overgrowth syndrome. Targeted therapy considering genetic causes would be the leading therapeutic strategy of overgrowth syndrome in the future.