Oncofusion

start / graph / target-cd22-directed-therapy-kmt2a-r

Targetstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q6b-therapeutics2

CD22-directed therapy (inotuzumab ozogamicin) — the second B-lineage antigen, with high response and poor durability

also called inotuzumab ozogamicin · Besponsa · CD22 CAR-T · ITCC-059

CD22 is a second marker on B-cell leukemia that can be targeted when CD19 stops working. A CD22-targeted antibody carrying a chemotherapy warhead puts most children back into remission, but only a minority are alive a year later.

The same thing, for a scientist

Inotuzumab ozogamicin is a CD22-directed antibody-drug conjugate carrying calicheamicin. In the paediatric ITCC-059 phase 1 (EudraCT 2016-000227-71), 25 heavily pretreated children with relapsed/refractory CD22-positive ALL had an overall response rate after course 1 of 80% (95% CI 59-93), with MRD-negative CR in 84% of responders (60-97), but 12-month overall survival of only 40% (95% CI 25-66). The recommended phase 2 dose was 1.8 mg/m² per course, the same as in adults. Sinusoidal obstruction syndrome occurred in 2 of 25 after subsequent chemotherapy. Crucially, the trial enrolled children aged 1 year or older — there is no infant-specific dataset.

Go deeper
The numbers and the gap between them1 deeper
EndpointValue
Overall response after course 180% (95% CI 59-93)
MRD-negative CR among responders84% (95% CI 60-97)
12-month overall survival40% (95% CI 25-66)
Recommended phase 2 dose1.8 mg/m² per course (0.8, 0.5, 0.5 mg/m² on days 1, 8, 15)
Hepatic sinusoidal obstruction syndrome2 of 25, after subsequent chemotherapy

Nine of 25 patients went on to HSCT or CAR-T after inotuzumab 1.

An 80% response rate with 40% one-year survival is the signature of a bridging agent, not a curative one. That is a legitimate and valuable role — in a disease where high-risk infants died before reaching transplant in Interfant-06, a drug that reliably produces MRD-negative remission for long enough to get to a cellular therapy is doing real work. It should be described that way rather than as a treatment.

The sinusoidal obstruction syndrome problem is specific and serious

Calicheamicin conjugates cause hepatic veno-occlusive disease / sinusoidal obstruction syndrome, and the risk is amplified by subsequent conditioning for HSCT. In a population where the whole point of the drug is to bridge to transplant, the toxicity and the indication are in direct conflict. Two of 25 children in ITCC-059 developed SOS after subsequent chemotherapy 1. In an infant with an immature liver, this is the toxicity that would define the risk-benefit, and there are no infant data.

Why CD22 matters strategically in KMT2A-rearranged disease

CD19 is now being spent early: blinatumomab is moving into frontline infant therapy, and CD19 CAR-T follows at relapse. CD22 is the only other established B-lineage antigen with an approved agent and clinical CAR-T constructs. It is therefore the field's insurance policy.

The insurance is partial. CD22-directed therapy still requires the blast to be B-lineage. The dominant KMT2A-specific escape route, myeloid lineage switch, defeats CD19 and CD22 equally. Adding CD22 to the sequence buys another line of therapy against antigen-loss relapse; it buys nothing against lineage switch. Any strategy that treats CD22 as the answer to CD19 failure without first establishing which failure mode occurred is misallocating the child's remaining options.

Facts and figures
modalityantibody-drug conjugate (calicheamicin payload); CD22 CAR T cells also in clinical development
statusInotuzumab ozogamicin approved for relapsed/refractory B-ALL in adults and, in some regions, in children aged 1 year and older; no infant indication and no infant data
eudract2016-000227-71
pediatric eligibletrue
infant datanone
recommended phase2 dose1.8 mg/m² per course (0.8, 0.5, 0.5 mg/m² days 1, 8, 15)
rolebridge to HSCT or CAR-T rather than definitive therapy
target genes
[
  "CD22"
]
key results
{
  "orr_pct": 80,
  "orr_ci95": [
    59,
    93
  ],
  "mrd_neg_cr_pct": 84,
  "mrd_ci95": [
    60,
    97
  ],
  "os_12m_pct": 40,
  "os_ci95": [
    25,
    66
  ],
  "n": 25
}
key ae
[
  "hepatic sinusoidal obstruction syndrome (2/25, after subsequent chemotherapy)",
  "grade 3-4 adverse events in 23/25",
  "hepatotoxicity",
  "myelosuppression"
]
resistance
[
  "CD22 antigen loss or density reduction",
  "myeloid lineage switch (defeats CD22 as completely as CD19)"
]
How this could bite us back
  • Hepatic sinusoidal obstruction syndrome from a calicheamicin conjugate, amplified by the transplant conditioning the drug is meant to enable — the toxicity and the indication are in direct conflict.
  • Lineage switch defeats CD22 exactly as it defeats CD19; adding a second B-lineage antigen does not address the genotype-specific escape route.
  • Antigen loss: CD22 density reduction rather than complete loss is a described escape from CD22-directed therapy and is harder to detect on routine flow.
  • No infant data at all: extrapolating dose and hepatic safety from children aged 1 and above to a 3-month-old with an immature liver is exactly the kind of extrapolation that has caused deaths in this population.
  • Late effects: hepatic injury in infancy has lifelong consequences and no survivor cohort exists for this agent in this age group.
Still open
  1. Is inotuzumab safe and effective in infants under one year? The one paediatric trial excluded them, and this is the population that most needs a bridge to transplant.
  2. What is the CD22 surface density on KMT2A-rearranged infant blasts relative to older children? CD22 density predicts ADC and CAR-T efficacy and the infant distribution is not published.
  3. Does prior inotuzumab increase SOS risk specifically in infants proceeding to conditioning? The conflict between the drug's role and its toxicity is sharpest exactly where it is most needed.
  4. In a child relapsing after CD19-directed therapy, what fraction of relapses are CD19-negative B-lineage (treatable with CD22) versus lineage-switched (not)? This determines whether CD22 is a real insurance policy or a false one.
Every source behind this entry (1)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the paediatric ITCC-059 phase 1 of inotuzumab ozogamicin, 25 children aged 1 to under 18 years with multiply relapsed/refractory CD22-positive ALL had an overall response rate after course 1 of 80% (95% CI 59-93), MRD-negative complete response in 84% of responders (95% CI 60-97), 12-month overall survival of 40% (95% CI 25-66), a recommended phase 2 dose of 1.8 mg/m² per course, and hepatic sinusoidal obstruction syndrome in 2 patients after subsequent chemotherapy; 9 patients proceeded to HSCT or CAR-T.

    PMID 33067614DOI
    What this evidence cannot show

    Phase 1, n=25, single-arm, heavily pretreated and heterogeneous. Eligibility was age 1 year or older, so INFANTS UNDER 1 YEAR WERE NOT STUDIED and no infant dosing, safety or efficacy data exist. No KMT2A-rearranged subgroup analysis is reported. Twenty-three of 25 patients had grade 3-4 adverse events. The 40% 12-month survival includes deaths after subsequent HSCT/CAR-T and cannot be attributed to inotuzumab either way.

    Evidence that argues against this (2)
    • The 80% response rate did not reproduce in the larger prospective paediatric trial. Children's Oncology Group AALL1621, a single-arm phase 2 of single-agent inotuzumab in 48 evaluable patients aged 1-21 with R/R CD22-positive B-ALL, reported CR+CRi of 58.3% (two-sided 90% CI 46.5-69.3) - roughly 22 points below ITCC-059 - and 6 of 21 patients (28.6%) transplanted after inotuzumab developed grade 3 sinusoidal obstruction syndrome. AALL1621 also found that partial CD22 expression and lower CD22 site density predicted non-response. O'Brien MM et al. 2022
    • In the only randomised test of the agent, the remission advantage did not convert into a survival advantage. INO-VATE randomised 326 adults with R/R ALL to inotuzumab or standard intensive chemotherapy: CR/CRi was 80.7% versus 29.4% (p<0.001) but median overall survival was 7.7 versus 6.7 months with HR 0.77 (97.5% CI 0.58-1.03, p=0.04, not meeting the prespecified significance boundary), median remission duration was only 4.6 months, and veno-occlusive disease of any grade occurred in 11% versus 1%. This is the strongest published argument that high CD22-directed response rates are a bridging phenomenon rather than a cure. Kantarjian HM et al. 2016
Connected to