133 Diagnostic utility of molecular and imaging biomarkers 2 HRAS, KRAS and NRAS mutations in indeterminate thyroid nodules Seventeen of the 20 studies reported the HRAS, KRAS and NRAS mutations separately: 268 RAS mutations occurred in 1,656 nodules, including 180 NRAS (67%), 54 HRAS (20%) and 34 KRAS (13%) mutations. Histopathological correlation of the individual RAS mutations was provided by 14 studies, including 1,554 indeterminate nodules and 176 (66%, 176/268) of these RAS mutations: 111 of the 180 NRAS (62%), 42 of the 54 HRAS (78%) and 23 of the 34 KRAS mutations (68%)[60, 75-77, 88, 93, 97-100, 102, 107, 109, 114, 118, 128, 350, 351]. Subgroup analysis was performed for each of the distinctive RAS mutations. Pooled performance estimates are presented (Table 13, Figure 19). Note that the estimated pooled performance for all RAS mutations in these 14 studies combined is similar to the results of the meta-analysis for all 20 studies. Individual HRAS, KRAS and NRAS mutation analyses demonstrate reciprocally similar specificities. They complement each other too, as they are mutually exclusive. Based on our pooled estimates and a 10.9% (180/1,656) occurrence rate, individual NRAS mutation analysis provides the most additional diagnostic value. It is presumed that the various RAS mutations are associated with slightly different types of cytology and histology, and consequently a different clinical course. For example, in another study KRAS was associated with oncocytic lesions and a lower malignancy rate than other RAS mutations [131]. Unfortunately, accurate analysis of such associations in our pooled population was hindered by missing histopathological data in many studies. Also, the different codons involved in the HRAS, KRAS and NRAS mutations were only distinguished in two studies [102, 128]. In general, point mutations in NRAS codon 61 and HRAS codon 61 are said to occur most frequently [43, 102]. In Korean studies, KRAS mutations often involve codon 61 instead of codon 12/13 [77]. RET/PTC rearrangement Our systematic literature search identified 17 studies that investigated the RET/PTC rearrangement in a total of 2,602 indeterminate thyroid nodules [60, 67, 69, 75, 76, 83, 87, 88, 93, 97, 99, 100, 102, 114, 118, 134, 351]. Baseline characteristics for all studies are reported in Table 14. Most studies investigated a combination of mutations or included the RET/PTC rearrangement in a gene mutation panel. In the aggregated 2,602 indeterminate nodules included in these studies, only 17 tested nodules positive for a RET/PTC1 or RET/PTC3 rearrangement (0.66%) (Table 15). Due to this very low frequency and the subsequent statistical limitations, meta-analysis of the studies was not possible. Histopathology was available in 76.3% (1,927/2,525) of the indeterminate nodules with a conclusive index test result, including 94% (16/17) of the nodules positive for a RET/PTC rearrangement (Table 16). Eight of these 16 (50%) harboured a malignancy: five PTC, two FVPTC and one unspecified thyroid carcinoma. The other eight were histopathologically benign (false positive), but their exact nature was unspecified.
RkJQdWJsaXNoZXIy MTk4NDMw