Oncofusion

start / graph / comut-ptpn11-kmt2a-r

Geneweak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.Q10-genomedisputed

PTPN11 in KMT2A-r leukemia — named in every gene list, supported by almost no cases

also called PTPN11 in MLL-rearranged ALL · SHP2 co-mutation · PTPN11 co-lesion KMT2A-r

PTPN11 makes a protein called SHP2 that helps switch on the same growth pathway as RAS, and it appears in published lists of genes broken alongside KMT2A fusions. But the actual case count behind it is one or two patients at most, and the public version of the main dataset contains no PTPN11 mutation at all — so this is a name repeated more often than the data support.

The same thing, for a scientist

PTPN11 (SHP2) is listed among the non-recurrent kinase/PI3K/RAS pathway genes mutated in both the infant and the non-infant KMT2A-rearranged cohorts of the St. Jude Pediatric Cancer Genome Project. That is the entire primary evidence base in KMT2A-rearranged leukemia specifically: non-recurrent in both arms, i.e. a single case each, with no variant reported in the abstract and no functional validation of a PTPN11 allele in this disease. An independent query of the public curation of the same study returns zero PTPN11 mutations across 93 sequenced samples, a discrepancy this node reports rather than resolves. PTPN11 is a well-established leukemia driver in juvenile myelomonocytic leukemia and in pediatric AML generally, and it is plausible that citation of that broader role has inflated its apparent standing as a KMT2A-r co-lesion. This node is deliberately marked disputed.

Go deeper
What the primary source actually says1 deeper

The St. Jude/Washington University PCGP study is the only genome-wide dataset that names PTPN11 in KMT2A-rearranged leukemia, and it names it twice — both times as non-recurrent 1:

  • In the 22-case infant KMT2A-r whole-genome discovery cohort: recurrent mutations were found in KRAS (n = 4) and NRAS (n = 2), with non-recurrent mutations in FLT3, NF1, PTPN11 and PIK3R1.
  • In the 20-case non-infant KMT2A-r cohort (ages 7-19): recurrent mutations were found in FLT3 (n = 3), KRAS (n = 3) and NRAS (n = 3), with non-recurrent mutations in CBL, PIK3CD, PTPN11 and PPM1J.

'Non-recurrent' in this paper's usage means observed once. So the entire case base for PTPN11 in KMT2A-rearranged leukemia is, at most, one infant and one older child. No PTPN11 protein change is given in the abstract or the retrieved body text. PTPN11 was included among the 232 genes in the targeted validation panel — the panel covered all genes mutated in the discovery cohort plus other kinase/PI3K/RAS genes 1 — yet PTPN11 does not appear among the 21 genes or gene loci reported as recurrently mutated across the combined infant cohorts. That is a negative result at 47 infant KMT2A-r cases, and it is the most informative single fact in this node.

The database discrepancy

Querying the public curation of that same study returns, across 93 sequenced samples, KRAS mutations in 12 samples, NRAS in 8, FLT3 in 5, PIK3CA in 4, PIK3R1 in 3, NF1 in 1, and PTPN11 in zero 2.

Two readings are possible and this node does not choose between them. Either the PTPN11 variant(s) fell below the curation's inclusion filters (they may have been subclonal, or in the 454/Sanger-validated validation arm rather than the whole-genome MAF that cBioPortal ingests), or the paper's gene list includes a call that did not survive to the released mutation table. Either way the practical conclusion is the same: PTPN11 in KMT2A-r leukemia rests on one or two cases whose variants are not independently visible in the public data.

This is exactly the kind of claim that becomes a 'known co-mutation' through repetition. It should be cited as a single-case observation or not at all.

Where PTPN11's reputation actually comes from

PTPN11 is a genuine and important leukemia gene — just not, on present evidence, a KMT2A-r co-lesion.

In the TARGET pediatric AML cohort, PTPN11 mutations are present in a minority of cases; querying the public curation of that study returns PTPN11 mutations in 9 of 150 sequenced samples alongside NRAS in 22, FLT3 in 10 and KRAS in 9 3. That is a real, recurrent presence in childhood AML at large. Critically, this node could not link those PTPN11-mutant samples to KMT2A-rearranged cases: the structural-variant profile of that study returned no KMT2A fusion records, so the intersection between PTPN11 mutation and KMT2A rearrangement could not be computed 3. The 9/150 figure is therefore pediatric AML context, not a KMT2A-r co-mutation frequency, and must not be quoted as one.

The broader molecular landscape of pediatric AML from the same programme reports that fewer than 40 genes are mutated in more than 2% of cases, with recurrent mutations concentrated in GATA2, FLT3, CBL, MYC-ITD, NRAS, KRAS and WT1 4. PTPN11 is not among the genes that paper highlights. Its strongest disease association remains juvenile myelomonocytic leukemia and the RASopathy spectrum, which is a different clinical and mechanistic setting entirely.

The honest summary: SHP2 is a credible node in the pathway that KMT2A-r leukemia does co-opt, and a SHP2 inhibitor might well have activity in KMT2A-r blasts through the pathway rather than through the gene. But that would be a pathway argument (see comut-ras-pathway-kmt2a-r), not a PTPN11-genotype argument, and the distinction determines whether anyone should be sequencing PTPN11 to select patients. On present evidence, they should not.

What would change this

This node is one of the easiest in the beat to resolve, and the material exists.

  1. Read the PCGP supplementary mutation tables. The specific PTPN11 protein change, its mutant allele frequency, and which case carried it are in the paper's supplementary data, which this node did not retrieve. A single lookup would upgrade this from 'named, uncharacterised' to 'one case, allele X, MAF Y'.
  2. Query a KMT2A-r-specific cohort. The Interfant-99/06 124-case cohort reported recurrent mutations exclusively in KRAS and NRAS 5 — an independent negative for PTPN11 at n = 124. Combined with the PCGP negative at n = 47, the pooled evidence is that PTPN11 is mutated in well under 1-2% of infant KMT2A-r ALL.
  3. Separate the AML question. Whether PTPN11 is a co-lesion in KMT2A-r AML specifically — a different lineage where RAS-pathway mutations are more clonal and PTPN11 is a real player — is genuinely open and is not answered by any source found here. That is the version of this question worth pursuing.
Facts and figures
gene symbolPTPN11
proteinSHP2 (tyrosine-protein phosphatase non-receptor type 11)
hgnc IDHGNC:9644
entrez gene ID5781
uniprotQ06124
cytoband12q24.13
pathwayRTK-RAS-MAPK, upstream adaptor/phosphatase
evidence status in kmt2a rnamed but not substantiated
estimated frequency in infant kmt2a r all<1-2%, pooling the PCGP and Interfant negatives; no positive frequency estimate exists
genome build noteNo coordinates or variants asserted, because none could be verified.
negative results
PCGP combined infant KMT2…47.0PCGP combined infant KMT2A-r: 47cBioPortal all_stjude_201593.0cBioPortal all_stjude_2015: 93Interfant-99/06 infant B-…124.0Interfant-99/06 infant B-ALL: 124
case base in kmt2a r
{
  "pcgp_infant_kmt2a_r": "non-recurrent (1 case of 22 whole-genome)",
  "pcgp_non_infant_kmt2a_r": "non-recurrent (1 case of 20)",
  "variant_identity": "not reported in the retrieved text; in unretrieved supplementary tables",
  "functional_validation_in_kmt2a_r": "none",
  "cbioportal_all_stjude_2015_records": 0,
  "n_samples_queried": 93
}
negative results — the numbers behind the chart
[
  {
    "cohort": "PCGP combined infant KMT2A-r",
    "n": 47,
    "finding": "PTPN11 not among the 21 recurrently mutated genes/loci"
  },
  {
    "cohort": "Interfant-99/06 infant B-ALL",
    "n": 124,
    "finding": "recurrent mutations exclusively KRAS and NRAS"
  },
  {
    "cohort": "cBioPortal all_stjude_2015",
    "n": 93,
    "finding": "zero PTPN11 mutation records"
  }
]
context not kmt2a specific
{
  "target_pediatric_aml_ptpn11_mutated_samples": 9,
  "target_pediatric_aml_n_sequenced": 150,
  "warning": "pediatric AML at large; could NOT be intersected with KMT2A-rearranged status because the study's SV profile returned no KMT2A records",
  "strongest_disease_association_of_ptpn11": "juvenile myelomonocytic leukemia and the RASopathy spectrum, a different setting"
}
How this could bite us back
  • Biomarker error: sequencing PTPN11 to select patients for a SHP2- or MEK-directed agent in KMT2A-r leukemia would, on present evidence, select almost nobody and exclude the pathway-activated majority who lack any RAS-pathway mutation.
  • Compensation: SHP2 sits upstream of RAS, so SHP2 inhibition is bypassed entirely by downstream KRAS or NRAS mutations — which are present in roughly a third to a half of these cases and are the actual common lesions.
  • Essential-function toxicity: SHP2 is required for normal cytokine-receptor signalling and haematopoietic development; germline PTPN11 activation causes Noonan syndrome and germline loss is developmentally severe, so sustained SHP2 inhibition in infants under one year carries unquantified developmental risk.
  • Citation drift: this node's main hazard is bibliographic rather than biological. PTPN11's standing as a KMT2A-r co-mutation rests on repetition of a single-case gene list, and downstream reviews that report it without a denominator create a false impression of a recurrent partner.
Still open
  1. What is the actual PTPN11 variant in the one infant and one older child in the PCGP cohort, and what was its mutant allele frequency? This is a supplementary-table lookup that would settle whether PTPN11 belongs in this beat at all.
  2. Why does the public curation of the PCGP study contain no PTPN11 mutation when the paper's text names PTPN11 in both cohorts? Resolving this discrepancy is a prerequisite for citing PTPN11 in any KMT2A-r context.
  3. Is PTPN11 a co-lesion in KMT2A-rearranged AML specifically, as opposed to ALL? RAS-pathway mutations are more clonal in the myeloid setting and PTPN11 is a genuine pediatric AML gene, but no source located here computes PTPN11 frequency within KMT2A-r AML. This is the version of the question worth pursuing.
  4. Would a SHP2 inhibitor have activity in KMT2A-r blasts regardless of PTPN11 genotype? Given that the pathway is activated by expression signature even in RAS-wild-type cases, a genotype-agnostic SHP2 or MEK strategy is more defensible than PTPN11-based patient selection — and untested in this disease.
Every source behind this entry (5)
  1. E1weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.paper

    PTPN11 is listed among the non-recurrent (i.e. single-case) kinase/PI3K/RAS pathway mutations in both the 22-case whole-genome infant KMT2A-r ALL discovery cohort and the 20-case non-infant KMT2A-r cohort of the St. Jude PCGP study, and does not appear among the 21 genes or gene loci reported as recurrently mutated across the combined 47-case infant KMT2A-r cohort despite PTPN11-class genes being included in the 232-gene validation panel.

    The landscape of somatic mutations in infant MLL-rearranged acute lymphoblastic leukemias · Andersson AK, Ma J, Wang J, Chen X, Gedman AL, Dang J, et al. · Nature Genetics 47(4):330-7 2015
    PMID 25730765DOI
    What this evidence cannot show

    'Non-recurrent' means one case in each cohort. No PTPN11 variant is specified in the abstract or the retrieved body text; the specific allele and allele frequency sit in supplementary tables that were not retrieved for this node, so the mutation's identity and clonality are unknown. The claim that PTPN11 was specifically covered in the validation panel is an inference from the paper's stated panel-selection criteria (all genes mutated in the discovery cohort plus other kinase/PI3K/RAS genes), not from an explicit gene list in the retrieved text — treat the negative at n = 47 as probable rather than certain.

    Evidence that argues against this (1)
    • The public curation of the same study contains no PTPN11 mutation across 93 sequenced samples, so the two sources disagree on whether a PTPN11 variant exists in this dataset at all. cBioPortal for Cancer Genomics 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    A query of the public curation of the PCGP study returned zero PTPN11 mutations across 93 sequenced samples, in the same query that returned KRAS in 12 samples, NRAS in 8, FLT3 in 5, PIK3CA in 4, PIK3R1 in 3 and NF1 in 1.

    What this evidence cannot show

    A zero return from a curated database is weak evidence of absence: curation filters, tiering thresholds and which sequencing arm was ingested all differ from the source publication. This is reported as a discrepancy to be resolved, not as a refutation of the paper. The 93 samples also mix cohorts, so a negative here is not a clean negative for infant KMT2A-r specifically.

    Evidence that argues against this (1)
    • The source publication's own text names PTPN11 among mutated pathway genes in both its infant and non-infant KMT2A-r cohorts. Andersson AK et al. 2015
  3. E3weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.database

    In the public curation of the TARGET pediatric AML cohort, PTPN11 mutations are present in 9 of 150 sequenced samples, alongside NRAS in 22, FLT3 in 10 and KRAS in 9; the structural-variant profile of that study returned no KMT2A fusion records, so PTPN11 mutation could not be intersected with KMT2A rearrangement status.

    What this evidence cannot show

    This is pediatric AML at large, NOT KMT2A-rearranged AML. The 9/150 figure must not be quoted as a KMT2A-r co-mutation frequency. The intended intersection with KMT2A-rearranged cases failed because the study's structural-variant profile returned no KMT2A records for the query used, so the question this evidence item was meant to answer remains unanswered.

    Evidence that argues against this (1)
    • The entry states its own disqualifying limitation: the structural-variant profile of that study returned no KMT2A fusion records, so these 150 paediatric AML samples cannot be stratified by the variable the beat is about. Nine of 150 is a frequency in paediatric AML, which is a different denominator. Ensembl, EMBL-EBI 2026
  4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The molecular landscape of pediatric AML across nearly 1,000 Children's Oncology Group participants shows fewer than 40 genes mutated in more than 2% of cases, with recurrent mutations concentrated in GATA2, FLT3, CBL, MYC-ITD, NRAS, KRAS and WT1; PTPN11 is not among the highlighted recurrent genes, and a cluster of zinc-finger-encoding genes is convergently affected in KMT2A-rearranged cases.

    The molecular landscape of pediatric acute myeloid leukemia reveals recurrent structural alterations and age-specific mutational interactions · Bolouri H, Farrar JE, Triche T, Ries RE, Lim EL, Alonzo TA, et al. · Nature Medicine 24(1):103-112 2018
    PMID 29227476DOI
    What this evidence cannot show

    Absence from a list of highlighted genes is not the same as absence of mutation; this paper does not report a PTPN11 frequency either way. Cited to establish that PTPN11 is not a leading recurrent gene in pediatric AML as a whole, which weakens rather than settles the case for it as a KMT2A-r co-lesion.

    Evidence that argues against this (1)
    • Not appearing among the highlighted recurrent genes of a cohort is an editorial threshold, not a measurement. A gene at 2-5% frequency falls below such a cut-off while still being present in dozens of the nearly 1,000 participants, and absence from a highlights list cannot be converted into a frequency bound. Blood 2014
  5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In 124 infant B-ALL cases from Interfant-99/06, recurrent mutations were found exclusively in KRAS and NRAS — an independent negative result for PTPN11 recurrence at n = 124.

    Unraveling the cellular origin and clinical prognostic markers of infant B-cell acute lymphoblastic leukemia using genome-wide analysis · Agraz-Doblas A, Bueno C, Bashford-Rogers R, Roy A, Schneider P, Bardini M, et al. · Haematologica 104(6):1176-1188 2019
    PMID 30679323DOI
    What this evidence cannot show

    'Exclusively KRAS and NRAS' refers to genes reaching recurrence thresholds; a single PTPN11 case in this cohort would not contradict the statement. So this is a negative for PTPN11 as a recurrent lesion, not proof that no PTPN11 mutation occurred. Only 42 of the 124 cases were analysed genome-wide.

    Evidence that argues against this (1)
    • An exclusive-recurrence statement in 124 infant B-ALL cases is a negative result with a small denominator for anything rarer than about 3%, and PTPN11 in KMT2A-r is expected to be exactly that rare. This bounds the frequency loosely rather than establishing non-recurrence. Leukemia 2021
Connected to