Immunohistochemical markers of potential utility in identifying POLE-mutant endometrial carcinomas: An assessment of autocrine motility factor (AMF) and autocrine motility factor receptor (AMFR)

Abstract

POLE status determination is necessary for the molecular classification of endometrial carcinomas (EC). However, this determination is only achievable by molecular techniques, which are not available in many practice settings. A previously published study reported elevated AMF/GPI and AMFR/gp78 levels in POLE-mutant EC. We examined the relationship between POLE status and AMF and AMFR expression. Our study included 55 molecularly classified EC, assessed for AMF and AMFR immunohistochemically. Staining intensity was scored 0 (negative), 1 (weak), 2 (medium), 3 (strong), extent was scored 0 (0 %), 1 (1–25 %), 2 (26–50 %), 3 (51–75 %), 4 (76–100 %), with those parameters summed for the final score for each case. The molecular subtypes POLE mutant, mismatch repair-deficient, no specific molecular profile, p53 abnormal had mean AMF scores of 5.909, 4.643, 5.000, 4.667, respectively. The POLE-mutant subtype had a significantly higher average AMF score than POLE wild-type (POLEwt) group (p = 0.003). Using POLE mutant status as an end-point, ROC analysis showed that an AMF immunohistochemical score of 6 and above had an 81.8 % sensitivity, 61.4 % specificity, AUC of 0.708 (95 % CI, 0.565–0.851). POLE-mutant subtype had higher prevalence of a score of 6 and above than the POLEwt group (9/11 vs 17/44 cases, p = 0.010). An AMF score 6 and above increased the likelihood of being POLE-mutant by a factor of 10.496 (95 % CI, 1.592–69.212). Similarly, the POLE-mutant subtype had higher prevalence of AMFR scores of 5 and above than the POLEwt group (p = 0.023). AMF may offer some promise in identifying POLE-mutant EC with relatively high sensitivity, but suboptimal specificity indicates that it cannot be applied alone in practice. Additional studies are required to determine whether AMF can be combined with other markers to more optimally identify POLE mutant EC.