Association between somatic RET mutations and clinical and genetic characteristics in patients with metastatic colorectal cancer.

BACKGROUND: Rearranged during transfection (RET) is a targetable oncogene. RET fusions have been reported in patients with metastatic colorectal cancer (mCRC). However, RET mutations in mCRC are less studied. Here, we aimed to characterize the clinical, pathological, and molecular landscape of RET-mutated mCRC.
METHODS: Five hundred and eighty-two patients were included in this study. Next-generation sequencing was performed to detect RET mutations and calculate tumor mutation burden (TMB). We compared the clinical, pathological, and molecular characteristics of mCRC cases with tumors that harbored somatic RET mutations (N = 16, 2.7%) or had wild-type RET (N = 566, 97.3%).
RESULTS: Males comprised the absolute majority of cases with RET mutations (15/16 [93.8%]) compared to their fraction among cases with wild-type RET (339/566 [59.9%]). Furthermore, all patients with RET mutations were younger than 60 years (16/16 [100%]), whereas such patients were less predominant in the group with wild-type RET (379/566 [67.0%]). Individuals with tumors positive for RET mutations more frequently exhibited mucinous histology (5/16 [31.2%] vs. 55/566 [9.7%]), exhibited a lower incidence of liver metastasis (4/16 [25.0%] vs. 335/566 [59.2%]), and higher incidence of peritoneal metastasis (9/16 [56.2%] vs.161/566 [28.4%]), expressed wild-type TP53 (8/16 [50.0%] vs.120/566 [21.2%]), and showed an increased frequency of MSI-high (6/16 [37.5%] vs. 18/566 [3.2%]). In those with microsatellite-stable mCRC, patients with RET mutations had a higher median TMB than patients with wild-type RET (9.4 vs. 6.7 mutations/Mb, respectively, p = 0.001). The median progression-free survival was similar in individuals with mutated and wild-type RET on the oxaliplatin-based regimen (7.1 vs. 8.7 months, p = 0.516).
CONCLUSIONS: Our study suggests that cases with RET mutations represent a separate mCRC subtype. Further studies are needed to evaluate the efficacy of RET inhibitors in mCRC patients with RET mutations.

BACKGROUND

The RET (rearranged during transfection) gene encodes a receptor tyrosine kinase that plays a key role in the activation of MAPK, PI3K, JAK, and PKA/C pathways. RET signaling is essential for the normal development and function of kidney, nervous system, and hematopoiesis. Aberrant RET activation is involved in various types of tumorigenesis, including medullary thyroid cancer (MTC), non‐small cell lung cancer (NSCLC), and colorectal cancer (CRC). , , Thus, RET has emerged as a therapeutic target in patients with aberrant RET activation. Several multikinase inhibitors with RET inhibitory activity have been tested in clinic, although the efficacy is limited. Two selective RET inhibitors, selpercatinib (LOXO‐292) and pralsetinib (BLU‐667), have shown promising anticancer activities and are currently widely evaluated in clinical trials. , , , ,

There are mainly two mechanisms underlying RET oncogenic activation. The first is chromosomal rearrangement, which causes fusion of RET with a partner protein. , The second mechanism is somatic or germline gain‐of‐function mutation. The mutations could disrupt the intramolecular disulphide bonds and further transform the structure of extracellular domain. The mutant protein leads to the ligand‐independent dimerization, confers constitutive RET signal activation, and promotes cancer cell proliferation, migration, and growth. RET fusions were observed in a small fraction (<1%) of metastatic CRC (mCRC), where their presence was associated with old age, right‐sided tumor origin, wild‐type RAS and BRAF, microsatellite instability (MSI)‐high tumors, and poor prognosis. However, the prevalence as well as clinical, pathological, and molecular features of somatic RET mutations in patients with mCRC are largely unknown. Understanding the clinical impact of RET mutations can help oncologists to properly translate this information into clinical practice. Kato et al. examined a cohort of 300 patients with CRC and identified two individuals with RET mutations. Such rare occurrence of RET mutations precluded their further analysis. Hence, this study aimed to evaluate the frequency and phenotypic characteristics of mCRC with somatic RET mutation in a larger patient cohort.

METHODS

The following patients were selected: (1) those diagnosed with mCRC based on pathology samples at Sun Yat‐sen University Cancer between 1 January 2015 and 31 March 2020; (2) those who had an available record of the next‐generation sequencing result performed at our institute. Those who (1) accepted chemotherapy or radiotherapy before next‐generation sequencing analysis or (2) had no available follow‐up information were excluded. The study was performed in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Sun Yat‐sen University Cancer (GZKJ2020‐02), and all patients provided informed consent. The detailed methods for next‐generation sequencing and tumor mutation burden (TMB) calculation are reported in a previously published paper and provided in the Supplementary Material.

We evaluated the association between the presence of somatic RET mutations and the following variables: age, gender, primary tumor location, differentiation, mucinous histology, T stage, N stage, time to metastasis, metastatic organs, TP53 mutations, APC mutation, RAS mutations, BRAF mutations, SMAD4 mutations, PIK3CA mutations, ERBB2 amplifications, and MSI status. Patients’ characteristics were compared by using the Chi‐squared test. TMBs were compared by using the Kruskal–Wallis test. Kaplan–Meier survival analysis and log‐rank test were performed to detect differences in progression‐free survival (PFS). The statistical tests were two‐tailed and the differences were considered statistically significant when p < 0.05.

RESULTS

Five hundred and eighty‐two patients were included in this study (Figure 1). The patients’ median age was 54 (range, 17–88) years, and 354 patients (60.8%) were male. Somatic RET mutations were identified in 16 (2.7%) patients. The genomic alterations of RET‐mutated tumors are shown in Table 1 and Figure 2.

FIGURE 1

Flow chart describing patient selection for this study

TABLE 1

Genomic alterations in RET‐mutated metastatic colorectal cancer

Patient ID	Genomic alterations
1	RET, EX3, R177W; RAS wild‐type; BRAF wild‐type; MSS
2	RET, EX3, R133C; KRAS A146T; BRAF wild‐type; MSS
3	RET, EX5, V351I; RAS wild‐type; BRAF wild‐type; MSI‐high
4	RET, EX6, N359T; RAS wild‐type; BRAF wild‐type; MSS
5	RET, EX7, T451M; KRAS G12D; BRAF wild‐type; MSS
6	RET, EX7, V485L; NRAS Q61L; BRAF wild‐type; MSI‐high
7	RET, EX10, E616del; KRAS G12D; BRAF wild‐type; MSI‐high
8	RET, EX10, E623G; KRAS G13D; BRAF wild‐type; MSS
9	RET, EX11, V642F; RAS wild‐type; BRAF wild‐type; MSS
10	RET, EX11, K662M; RAS wild‐type; BRAF wild‐type; MSS
11	RET, EX11, P695S; KRAS G13D; BRAF wild‐type; MSI‐high
12	RET, EX11, V706M; RAS wild‐type; BRAF wild‐type; MSS
13	RET, EX15, V871I; RAS wild‐type; BRAF wild‐type; MSI‐high
14	RET, EX15, V871I; KRAS G13D; BRAF wild‐type; MSI‐high
15	RET, EX15, S891L; RAS wild‐type; BRAF wild‐type; MSS
16	RET, EX15, S891L; RAS wild‐type; BRAF K601E; MSS

Abbreviations: MSI‐high: microsatellite instability‐high; MSS: microsatellite stable.

FIGURE 2

Frequent RET mutations detected in 582 patients with metastatic colorectal cancer

The presence of RET mutations was associated with male sex (15/16 [93.8%] patients with mutated RET vs. 339/566 [59.9%] patients with wild‐type RET), younger age (all 16 patients with mutated RET [100%] were ≤60 years old vs. 379 out of 566 [67.0%] patients with wild‐type RET), mucinous histology (5/16 [31.2%] patients with mutated RET vs. 55/566 [9.7%] patients with wild‐type RET), less liver metastasis (4/16 [25.0%] patients with mutated RET vs. 335/566 [59.2%] patients with wild‐type RET), and more peritoneal metastasis (9/16 [56.2%] patients with mutated RET vs. 161/566 [28.4%] patients with wild‐type RET, Table 2). Nine (12.3%) out of 73 male patients who were younger than 60 years of age and had peritoneal metastasis harbored RET mutations.

TABLE 2

Characteristics of patients with RET‐mutated metastatic colorectal cancer

Characteristics	RET mutation, N (%)	RET wild‐type, N (%)	p
Gender
Male	15 (93.8)	339 (59.9)	0.007
Female	1 (6.2)	227 (40.1)
Age
>60	0 (0)	187 (33.0)	0.002
≤60	16 (100)	379 (67.0)
Primary tumor location
Right colon	6 (37.5)	162 (28.6)	0.603
Left colon	6 (37.5)	198 (35.0)
Rectum	4 (25.0)	206 (36.4)
Differentiation
Well/moderate	11 (68.8)	416 (73.5)	0.784
Poor	5 (31.2)	150 (26.5)
Mucinous histology
Yes	5 (31.2)	55 (9.7)	0.018
No	11 (68.8)	511 (90.3)
T stage
1	0 (0)	2 (0.4)	0.415
2	0 (0)	21 (3.7)
3	5 (31.2)	266 (47.0)
4	7 (43.8)	161 (28.4)
N stage
0	6 (37.5)	123 (21.7)	0.029
1	0 (0)	164 (29.0)
2	5 (31.2)	146 (25.8)
Time to metastasis
Synchronous	9 (56.2)	398 (70.3)	0.269
Metachronous	7 (43.8)	168 (29.7)
Liver metastasis
Yes	4 (25.0)	335 (59.2)	0.009
No	12 (75.0)	231 (40.8)
Lung metastasis
Yes	2 (12.5)	159 (28.1)	0.257
No	14 (87.5)	407 (71.9)
Peritoneal metastasis
Yes	9 (56.2)	161 (28.4)	0.024
No	7 (43.8)	405 (71.6)
Distant nodes metastasis
Yes	3 (18.8)	73 (12.9)	0.452
No	13 (81.2)	493 (87.1)

Bold indicates statistical significance values p < 0.05.

As shown in Table 3, RET mutations were more frequent in patients with wild‐type TP53 and MSI‐high tumors (6/16 [37.5%] patients with mutated RET vs. 18/566 [3.2%] patients with wild‐type RET). No Lynch syndrome was observed in patients with RET mutation. There were no significant associations between RET mutation presence and APC, RAS, BRAF, SMAD4, PIK3CA, or ERBB2 status. In patients with microsatellite‐stable mCRC, those with RET mutations had a higher TMB than those with wild‐type RET (median TMB 9.4 and 6.7 Muts/Mb, respectively, p = 0.001). In patients with MSI‐high mCRC, those with RET mutations had a TMB similar to that of patients with wild‐type RET (median TMB 66.6 and 59.2 Muts/Mb, respectively, p = 0.923).

TABLE 3

Association between the presence of RET mutation and other molecular characteristics

Characteristics	RET mutation, N (%)	RET wild‐type, N (%)	p
TP53
Wild‐type	8 (50.0)	120 (21.2)	0.012
Mutated	8 (50.0)	446 (78.8)
APC
Wild‐type	4 (25.0)	194 (34.3)	0.595
Mutated	12 (75.0)	372 (65.7)
RAS
Wild‐type	9 (56.2)	280 (49.5)	0.622
Mutated	7 (43.8)	286 (50.5)
BRAF
Wild‐type	15 (93.8)	503 (88.9)	0.319
V600E mutation	0 (0)	49 (8.7)
No V600E mutation	1 (6.2)	14 (2.4)
SMAD4
Wild‐type	12 (75.0)	468 (82.7)	0.500
Mutated	4 (25.0)	98 (17.3)
PIK3CA
Wild‐type	13 (81.2)	481 (85.0)	0.721
Mutated	3 (18.8)	85 (15.0)
ERBB2
Amplification	16 (100)	548 (96.8)	1.000
No amplification	0 (0)	18 (3.2)
MSI status
MSS	10 (62.5)	548 (96.8)	<0.001
MSI‐high	6 (37.5)	18 (3.2)

Bold indicates statistical significance values p < 0.05.

Abbreviations: MSI‐high, microsatellite instability‐high; MSS, microsatellite stable.

The majority of the studied patients (413/582, 71.0%) were administered oxaliplatin‐based first‐line chemotherapy (FOLFOX or XELOX), whereas 63 (10.8%) received irinotecan‐based chemotherapy (FOLFIRI or XELIRI). Among the mCRC patients on the oxaliplatin‐based regimen, PFS was similar between patients with RET mutations and those with wild‐type RET (median PFS 7.1 vs. 8.7 months, respectively, p = 0.516, Figure 3).

FIGURE 3

Comparison of progression‐free survival between patients with RET wild‐type and RET‐mutated tumors who were administered oxaliplatin‐based first‐line chemotherapy

DISCUSSION

In our analysis of 582 patients with mCRC, we detected RET mutations in 16 (2.7%) individuals. Furthermore, RET mutations were more often present in younger male patients and they were associated more strongly with mucinous histology, weaker liver metastasis, and stronger peritoneal metastasis. Given that the incidence of RET mutations in mCRC patients is low, our study provides valuable information to potentially help physicians identify the enriched subgroups in which RET mutations could be more frequent.

Kato et al. studied RET aberrations in 4,871 patients with diverse malignancies. Three hundred patients with CRC were included, and only two of them (0.7%) had RET mutations. The tumor stages were unknown, and this might explain the low mutation rate of RET in that study. Villava et al. analyzed 197 patients and identified four patients with colorectal cancer and RET mutation. Oliveira et al. also identified four different somatic RET mutations in patients with colon cancer. However, the rare occurrence of RET mutations precluded further analysis in those studies. Recently, Zhao et al. retrospectively reviewed 3,272 patients with CRC and documented that the RET mutation rate was 3.39%, which was similar to the rate observed in the present study. RET fusions have been observed in a small subgroup of patients with mCRC. RET fusions were associated with older age, origin on the right side, wild‐type RAS and BRAF, MSI‐high tumors, as well as shorter survival. Our study makes several contributions to the current literature. First, we found the association between RET mutations and patients’ characteristics (sex, age, histology, and metastasis sites). These additional information revealed in our study suggests a potential enrichment strategy for further trials with targeted agents focused on the RET mutation. Second, we found RET mutations were enriched in younger and male patients. We did not observe significant correlations between the presence of RET mutations and primary tumor location and RAS or BRAF status, on the other hand. These results suggest that factors affecting the incidence of RET mutations are different from those associated with RET rearrangements, and RET‐mutated tumors did not share with RET fusions the same clinical and pathological characteristics. In our study, MSI‐high was associated with RET mutation. This may be caused by genome instability in MSI‐high tumors. Further studies are needed to evaluate the impact of RET mutation on response to immunotherapy in MSI‐high tumors.

To date, there are several commercially available inhibitors demonstrating activity against RET. Most of them are nonspecific inhibitors, such as cabozantinib, vandetanib, lenvatinib, sorafenib, sunitinib, and alectinib. Recently, two selective RET inhibitors have been introduced to the clinic. Pralsetinib has been demonstrated to inhibit the growth of tumors driven by various RET mutations and fusions in vivo, and it showed durable clinical responses in patients with RET‐altered NSCLC. , Selpercatinib also showed promising activity in patients with RET‐fused NSCLC and RET‐fused or mutated MTC. , , , Two patients with RET‐mutated mCRC were included in a clinical trial (NCT03037385), and the efficacy of pralsetinib is still under evaluation in these two patients.

Our study had few limitations, such as a limited sample size and retrospective bias. The identified RET mutations in this study could be driver aberrations or only neutral passengers. We could not distinguish between them owing to limited information regarding RET mutations in mCRC. Furthermore, because the follow‐up time was limited, overall survival could not be determined.

CONCLUSIONS

Despite these limitations, our study noted RET mutation in 2.7% patients with mCRC, and showed males younger than 60 years of age comprised the absolute majority of patients with RET mutations. These observations suggest that tumors with RET mutation represent a novel subtype of mCRC, and are different from tumors with RET fusion. Further studies are needed to evaluate the efficacy of RET inhibitors in patients with RET‐mutated mCRC.

ETHICS STATEMENT

All patients signed informed consent. The study was performed in accordance with the Declaration of Helsinki and was approved by the Institutional Review Board of Sun Yat‐sen University Cancer Center (GZKJ2020‐02).

CONFLICT OF INTEREST

The authors declare that they have no competing interest.

AUTHOR CONTRIBUTION

Concept and design: All authors. Acquisition, analysis, or interpretation of data: YZ Yang and WM Hu. Drafting of the manuscript: WZ He and LP Xia. Critical revision of the manuscript for important intellectual content: All authors. Statistical analysis: WZ He, WM Hu, and LP Xia. Obtained funding: WZ He. Administrative, technical, or material support: YZ Yang and WM Hu. Supervision: LP Xia.

Supplementary Material

Click here for additional data file.