PFM is associated with multiple genetic factors. Different genetic technologies provide new perspectives for the prenatal diagnosis of PFM. In our cohort, we conclude an optimal prenatal genetic diagnostic approach for PFMs. Considering cost-effectiveness, we recommend CNV testing as the primary genetic diagnostic method for isolated PFMs. A sequential genetic approach (karyotyping, CNV testing, and WES) should be performed in non-isolated PFMs. When the results of karyotyping and CNV testing are negative, additional WES testing may be required to further analyze the genetic etiology. In addition, we recommend WES as a first-tier genetic diagnostic tool when JS is suspected.

In isolated PFMs, our present detection rate of CNV testing (36.36%) is much higher than that of karyotyping and WES (both were 0.00%). Thus, we recommend CNV testing as the first-line cytogenetic diagnostic test for isolated PFMs. It is noteworthy that advancing gestation (> 24 weeks) is associated with a significant increase in the laboratory failure rate of karyotyping, especially in the third trimester. In this situation, CNV testing, instead of karyotyping analysis, can provide valuable genetic information, as some cases are diagnosed too late (for example, Cases 2, 10, 11, 13, and 14). We believe that WES is of limited value in the genetic diagnosis of isolated PFMs, although WES has been used in multiple structural abnormalities to provide a higher resolution in the identification of monogenic disorders. Two previous studies support our position. Tan’s study [16] reported that two fetuses with isolated PFM (one case of isolated DWM and one case of isolated BPC) had negative results with the use of WES. Another study recruited 268 fetuses with central nervous system malformations, and seven fetuses with isolated PFM (five cases with DWM and two cases with CH) had a negative genetic work-up based on WES [17]. However, Li et al. [18] studied seven patients with isolated cerebellar vermis defects and found two cases of DWM with a single gene defect, with a detection rate of 28.57%. However, one of the patients was found to have a VOUS, which should be excluded from the positive results. Besides, most of the patients in the study underwent proband-only WES, making the results difficult to verify. Therefore, the high detection rate has limited credibility. Given the high cost of WES, the low detection rate of P/LP variants, and the difficulty of variant analysis, we recommend that when CNV testing presents negative findings in isolated PFMs, WES may be not considered as a routine screening tool for isolated PFMs.

In non-isolated PFMs, we recommend a sequential genetic approach (karyotyping, CNV testing, and WES) for genetic diagnosis. Considering the cost-effectiveness, the sequential genetic approach was applied in our study; the result showed a high detection rate of 47.5% (19/40) of the genetic abnormalities. Compared to its application in isolated PFMs, WES can provide additional diagnostic value in non-isolated PFMs (0.00% vs. 40.91%). In a study of 34 cases of PFMs with multisystem organ abnormalities, Drexler et al. [19] estimated a diagnostic yield of over 50% using WES. Besides, the role of WES in the genetic diagnosis of PFMs associated with other malformations has also been demonstrated in a number of case reports [20,21,22]. We have shown that WES may provide additional diagnostic value in non-isolated PFMs. Therefore, a sequential genetic approach is the most appropriate diagnostic strategy for non-isolated PFMs.

In addition, our study revealed that a large percentage of our JS cases (5/6, 83.33%) had positive genetic variant using WES, which is consistent with the result that pathogenic variants in the associated genes account for ~ 60–90% of JS cases [23]. Based on our data, WES can be considered a first-tier genetic diagnostic approach for JS. Moreover, previous studies have demonstrated that WES can evaluate genotype–phenotype correlations in JS. The “molar tooth sign”, represented by the combination of CH, abnormally thick and horizontally oriented superior cerebellar peduncles, and/or a deep interpeduncular fossa, is a typical sign of JS [24]. When “molar tooth sign” is suspected, WES can be used directly as a first-line diagnostic tool instead of the sequential approach (karyotyping to CNV testing, and then WES) to save medical costs and improve the cost–benefit ratio.

Notably, three fetuses with isolated BPC in our study had negative genetic findings using prenatal sequential genetic testing (karyotyping, CNV testing, and WES) and were born with good outcomes, consistent with other studies [7, 8, 25]. A meta-analysis showed a very low rate of chromosomal anomalies (5.2%) in cases of isolated BPC [7, 8]. However, there is no study on the value of WES in cases of isolated BPC. Therefore, based on our data, we do not recommend routine use of WES in individuals with isolated BPC. However, the diagnostic strategy for non-isolated BPC remains cautious. It depends on the type and severity of associated malformations; for example, Case 10 and 13 had P CNVs with poor prognosis.

Before applying different strategies for the diagnosis of isolated and non-isolated PFMs, accurate prenatal diagnosis of PFMs using ultrasound is important. Our study revealed additional fetal anomalies in PFMs, including congenital heart defects (CHDs), corpus callosum malformations, and cortical hypoplasia. CHDs were the most common type of extracranial anomaly. Trisomy 13, Trisomy 18, Turner syndrome, 22q11 deletion syndrome, Noonan syndrome, Coffin–Siris syndrome 2, and Coffin–Siris syndrome 4 were observed in Cases 5, 6, 7, 8, 9, 12, 20, 21, and 23. Phenotypes of DWM, small cerebellar diameter, abnormal vermis, and clivus heights were common in these genetic syndromes [26,27,28,29,30,31]. The reason PFMs are often associated with CHD can be explained by two hypotheses. First, affected signals or pathways in aneuploidy or genetic syndromes may explain the mechanism underlying the co-occurrence of cardiac malformations and PFMs, such as Noonan syndrome [32,33,34]. Second, some studies have suggested that CHD exerts devastating effects on neurological development via blood circulation [32,33,34]. Therefore, once PFMs are detected, there should be a focus on the cardiovascular system to prevent misdiagnosis.

Our study has several strengths. First, our study is the first relatively large study to focus on prenatal genetic diagnostic strategy for PFMs. Second, our study first recommended optimal prenatal genetic diagnostic strategies for isolated and non-isolated PFMs. Third, our study recommends WES as a first-tier genetic diagnostic tool when JS is suspected.

Our study also has some limitations. Some cases were diagnosed > 24 weeks with failed karyotyping and some patients refused complete genetic testing, leading to the incomplete data. In addition, our conclusion on the diagnostic strategies for different subtypes of PFMs requires a prospective study with large sample size for verification and promotion.