Here we presented a case of recurrent OC with NTRK1-TPM3 rearrangement. Additionally, the present case demonstrates the discrepancy between gene rearrangement detected by NGS and protein expression. This discrepancy may be a biomarker for predicting the ineffectiveness of entrectinib for cancers with NTRK rearrangement detected by NGS.
In the current case, NGS revealed NTRK1-TPM3 rearrangement and a missense variant of TP53. There are few approved therapies for TP53, although almost all cases of ovarian high-grade serous carcinoma (95%) have somatic TP53 variants.  On the other hand, NTRK fusions are oncogenic drivers and novel targets. Doebele et al. reported the safety and activity of entrectinib in adult patients with advanced or metastatic NTRK fusion-positive cancer across three clinical trials (ALKA-372-001, STARTRK-1 and STARTRK-2). In these trials, only one ovarian cancer patient was included. They showed that the objective response rate, which included complete response and partial response, was 57% (95% CI 43.2–70.8). The median duration of response was 10 months (95% CI 7.1 to not estimable) and the percentage of progressive disease (PD) was only 7%. However, the characteristics of cases with PD remained unclear in their report. 
In the present case, entrectinib was administered because NGS revealed NTRK1-TPM3 rearrangement and entrectinib was recommended after a discussion among experts. However, this novel target drug was ineffective. NTRK protein was not expressed, although IHC testing with a pan-Trk mAB clone [EPR17341] was performed. A previous study reported that gene fusions involving NTRK1, 2, and 3 and their partner genes result in a constitutive activation or overexpression of TRK receptors, potentially leading to oncogenesis.  Additionally, other reports have shown that pan-Trk IHC yielded a sensitivity of 75-95.2%, and a specificity of 92–100% and that the sensitivity of pan-Trk IHC for NTRK1 was 96.2%. [3, 8, 11, 12] Pan-Trk IHC is a reliable screening method for the detection of NTRK gene fusions based on this date. Moreover, pan-Trk IHC can rapidly assess malignancies which may harbor possible NTRK fusions in order to determine eligibility of patients for targeted therapy with TRK inhibitors.  However, it should be considered that there are NTRK rearrangements which are found to be negative by IHC, and can only be detected by NGS, such as in the present case.
Drilon et al. reported the efficacy of larotrectinib, which is a selective inhibitor of TRKA, TRKB and TRKC. In their study, six of an initial 55 patients showed primary resistance to larotrectinib. Three patients had tumor material available for central analysis, and in all three cases, pan-Trk IHC did not reveal the presence of TRK protein expression. This indicated that the rearrangements detected by NGS were false positives or that the identified fusion genes were not expressed at the protein level.  It is considered that entrectinib has the same characteristics as larotrectinib with regard to discrepancy between gene fusion and protein expression, as observed in the current case, and that this finding may be a key to predict the ineffectiveness of entrectinib for cancers with NTRK rearrangement detected by NGS.
To the best our knowledge, this is the first case report of OC with NTRK rearrangement. It is known that a small percentage of common adult cancers carry fusions of NTRK genes.  A large cohort study revealed that the frequency of NTRK gene fusions was 0.25% of general cancers. [2, 12] Therefore, physicians have few chances to experience this molecular characteristic. However, physicians should be aware of the pitfall that NTRK protein may not expression even if NGS reveals NTRK rearrangement.