Detection of mutant TET2 in myeloid malignancies other than myeloproliferative neoplasms: CMML, MDS, MDS/MPN and AML

TET2 (TET oncogene family member 2) is a putative tumor suppressor gene located at chromosome 4q24.¹ Delhommeau et al.¹ were the first to report the occurrence of TET2 mutations in myeloproliferative neoplasms (MPN); the authors discovered frameshift, nonsense and missense TET2 mutations in approximately 14% of 181 patients with JAK2V617F-positive polycythemia vera, essential thrombocythemia or primary myelofibrosis (PMF). We subsequently confirmed this observation in a recent study of 239 patients and reported mutational frequencies of ∼16, ∼5 and ∼17% in polycythemia vera, essential thrombocythemia and PMF, respectively.² In the latter study, TET2 mutations were seen in both JAK2V617F-positive (∼17%) and negative (∼7%) cases and their frequency was similar in chronic and advanced phase disease.² In another recent Leukemia paper,³ we reported an even higher TET2 mutational frequency of 29% in systemic mastocytosis (SM), a disease that is also formally classified as MPN.⁴ In this study, we looked for the presence of TET2 mutations in myeloid malignancies other than MPN, including chronic myelomonocytic leukemia (CMML), myelodysplastic syndrome (MDS), ‘MDS/MPN’ and acute myeloid leukemia (AML). We have also included some cases with MPN, unclassifiable (MPN-U).⁴

After approval by the Mayo Clinic institutional review board, study patients were selected on the basis of availability of stored bone marrow cells for DNA extraction and mutation analysis. Diagnoses were established on the basis of 2001 World Health Organization (WHO) criteria.⁵ Mutation screening for JAK2V617F (reverse transcription-PCR) was performed using bone marrow-derived cells, according to the earlier published methods.⁶ High-throughput DNA sequence analysis was used to screen for TET2 mutations in bone marrow-derived DNA as described earlier.² A total of 50 patients were included in this study: 15 with CMML, 16 with MDS, 7 with secondary AML, 5 with de novo AML, 3 with MDS/MPN and 4 with MPN-U. The 15 patients with CMML included 10 with CMML-1 and 5 with CMML-2. The 16 MDS cases included 5 with refractory anemia/cytopenia (RA)/refractory cytopenia with multilineage dysplasia (RCMD), 5 with RA with ring sideroblasts (RARS)/RCMD-RS, 4 with RA with excess blasts (RAEB)-1/2 and 2 with del(5q). The seven secondary AML cases arose from antecedent MDS in five cases (including two therapy-related) and in one case each from MDS/MPN or CMML. The five de novo AML cases included AML-M7 (n = 3), AML-M6 and AML-M3. The three MDS/MPN cases included one patient with RARS and thrombocytosis (RARS-T).

A total of 13 TET2 mutations were identified in 11 patients including 8 nonsense and 5 frameshift; these mutations mostly involved exons 4 (n = 8) and 12 (n = 3) (Table 1). TET2 mutations were seen in 3 (20%) of 15 CMML patients; mutational frequencies were ∼13 and ∼29% in our earlier studies with classic BCR-ABL1-negative MPN² and SM,³ respectively. As illustrated in Table 2, there were no overt differences in the clinical manifestations between TET2 mutated and unmutated CMML patients but a formal statistical comparison was not attempted because of the low numbers involved. However, it is reasonable to mention, without making any conclusions, that all three TET2-mutated CMML patients showed less than 5% bone marrow blasts and one of them developed SM 2 years later. This was also true for the single TET2-mutated MDS patient, out of 16 (6%), who had RARS. Higher proportions of TET2-mutated patients were seen with secondary AML (three of seven patients) and MPN-U (two of four patients), but a much larger study is needed to confirm this trend and clarify its relevance. One of the latter TET2-mutated patients with MPN-U also showed MPLW515L. One patient each with AML-M3 and MDS/MPN also had mutant TET2.

Table 1. TET2 mutation details in 11 patients with myeloid malignancies other than myeloproliferative neoplasms^a.

Diagnosis	Exonb	Nucleotide change	Consequence	Mutation type	Age (years)	Sex	JAK2 V617F	BM blast ≥5%	BM cytogenetics
CMML-1	4	1316_1317insA	Frameshift	Frameshift	76	M	Neg.	No	Normal
CMML-1	11	486 G>T	E1490X	Nonsense	56	F	Neg.	No	Normal
CMML-1	4	2946 C>T	Q916X	Nonsense	60	M	Neg.	No	Normal
	12	1010_1011insA	Frameshift	Frameshift
MDS-RARS	4	1173 C>T	Q325X	Nonsense	63	M	Neg.	No	Normal
MDS/MPN	9	264_265insA	Frameshift	Frameshift	67	M	Pos.	No	add(2)(p13)
MPN-U	4	1022_delC	Frameshift	Frameshift	52	F	Pos.	No	Normal
MPN-Uc	12	1136 C>T	Q1825X	Nonsense	75	M	Neg.	No	Normal
Post-MDS AML	4	2424 C>T	Q742X	Nonsense	72	F	Neg.	Yes	Complexd
Post-MDS AML	4	1261 C>G	S354X	Nonsense	66	M	Neg.	Yes	Normal
	4	2019 G>T	G607X	Nonsense
Post-CMML AML	4	1040_1041insT	Frameshift	Frameshift	85	M	Neg.	Yes	Complexe
AML-M3	12	767 C>T	Q1702X	Nonsense	68	M	Neg.	Yes	t(15;17)(q22;q21)

Abbreviations: AML, acute myeloid leukemia; AML-M3, acute promyelocytic leukemia; BM, bone marrow; CMML, chronic myelomonocytic leukemia; MDS, myelodysplastic syndrome; MPN, myeloproliferative neoplasm; MPN-U, MPN unclassifiable; Neg., Negative; Pos., Positive; RARS, refractory anemia with ringed sideroblasts.

^aMPN-U is classified under the MPN category according to the World Health Organization classification system.

^bExon counting in this study starts with exon 4 and ends with exon 12 and might be different than that used by other investigators where it starts with exon 3 and ends with exon 11. In other words, what is assigned as exon 4 in this study is referred to as exon 3 by others and so forth.

^cPatient also had MPLW515L mutation.

^d73–127,X,−X,−X,+4,del(5)(q13q33),+6,−7,+8,+8,+9,+11,+13,−14,add(15)(p13),−17,+18,+19,+20,+21,+21[15]/46,XX,del(5)(q13q33)[5].

^e45,XY,add(1)(q32),add(3)(q12),−5,−6,add(7)(q32),−10, add(11)(p13),+13,add(13)(p11.2)×2,−22,+2mar[7]/46,idem,+add(13)(p11.2)[13].

Table 2. Clinical and laboratory features of 16 patients with chronic myelomonocytic leukemia stratified according to their TET2 mutational status.

	All patients	TET2 mutated	TET2 unmutated
N (%)	15 (100)	3 (20)	12 (80)
Males, n (%)	6 (40)	2 (67)	4 (33)
Age in years, median (range)	68 (49–87)	60 (56–76)	68 (49–87)
Splenomegaly, n (%)	8 (53)	0	8 (67)
Transfusion dependent, n (%)	3 (20)	1 (33)	2 (17)
Hemoglobin (g per 100 ml), median (range)	10.7 (6.4–13.4)	11.5 (6.4–12.2)	10.5 (8.8–13.4)
Leukocyte count (× 10/l), median (range)	14.4 (4.3–102.3)	8.6 (7.1–29.1)	18.3 (4.3–102.3)
Lymphocyte count (× 109/l), median (range)	2.7 (0.3–11.2)	1.6 (1.5–3.5)	2.7 (0.3–11.2)
Monocyte count (× 109/l), median (range)	3 (0.2–13.4)	1.5 (0.2–6.7)	3.2 (0.9–13.4)
Platelet count (× 109/l), median (range)	117 (18–726)	69 (18–158)	150 (28–726)
LDH (U/l), median (range)	218 (118–499)	216 (165–228)	233 (118–499)
BM cellularity, median (range)	90 (55–100)	80 (70–90)	93 (55–100)
BM blast % ≥5, n (%)	7 (47)	0	7 (58)
Abnormal karyotype, n (%)	6 (40)	0	6 (50)
JAK2V617F, n (%) n = 4	1 (7)	0	1 (8)

Abbreviations: BM, bone marrow; LDH, lactate dehydrogenase.

This study illustrates the ubiquitous nature of TET2 mutations across a spectrum of acute and chronic myeloid malignancies. The type of TET2 mutations and the specific exons involved in this study were also largely similar to those seen in patients with MPN or SM.^2,3 Because of the relatively small number of patients included in this study, it is important not to overplay the observed disease-specific mutational frequencies. Furthermore, accurate interpretation of TET2 mutational frequency requires accounting for age; in our earlier report involving patients with BCR-ABL-negative classic MPN (n = 239),² overall mutational frequency was 23% in patients ≥60 years of age versus 4% in younger patients (P<0.0001).² In another related study,³ the corresponding figures in SM (n = 42) were 39 and 13% (P = 0.07). Another interesting observation from this and our earlier reported related studies^2,3 is the fact that mutant TET2 can coexist with other pathogenetically relevant (or potentially relevant) mutations including PML-RARA (this study), MPLW515L (this study), JAK2V617F² or KITD816V.³

The above-mentioned observations raise several questions: (i) Considering the ubiquitous nature of mutant TET2 in myeloid malignancies, is it possible that it is also present in lymphoid or other neoplasms? (ii) In instances where mutant TET2 coexists with other mutations, does it predate or postdate their emergence? (iii) Within the context of a specific disease, does the presence of mutant TET2 affect phenotype, prognosis or treatment response? (iv) Is mutant TET2, in those who carry the mutation, a better indicator of clonal response to treatment? (v) What exactly is the pathogenetic contribution of mutant TET2 in myeloid malignancies? It is relatively easy to address the first four questions, but we encourage patience before making any clinical associations until one sees a consistent pattern across studies from different centers. The question that is both most important and difficult to address concerns the precise pathogenetical contribution of mutant TET2 in myeloid neoplasms, especially in view of its occurrence across different molecular profiles.

The type of TET2 mutations seen so far are mostly frameshift or nonsense, and are therefore inactivating. Furthermore, Delhommeau et al.¹ have shown that both copies of the gene were affected in a subset of TET2-mutated MPN patients, and that mutant TET2 was associated with loss-of-heterozygosity/somatic deletion at its 4q24 chromosomal location, suggesting a tumor suppressor function for the wild-type allele. They also showed the occurrence of mutant TET2 in both JAK2V617F-positive and negative clones from otherwise JAK2V617F-positive MPN patients, suggesting that TET2 mutations antedated JAK2V617F during the evolution of the malignant clone, at least in certain instances. In the presence of mutant TET2, as opposed to its absence, the coexisting JAK2V617F-positive clone constituted a higher proportion of the CD34+/CD38– cell fraction and showed enhanced engrafting in NOD-SCID mice. Taken together, these observations would suggest that TET2 mutations, although ubiquitous, are not simply passenger mutations and that their acquisition by the malignant clone might enhance its stem cell-like properties without necessarily affecting its differentiating capacity. Additional laboratory studies are needed to further clarify the biological consequence of these mutations.