Inotuzumab Plus Ponatinib Tied to Sinusoidal Obstruction Syndrome in Ph-Positive Leukemia

Hematologists in Japan are warning of a previously underappreciated liver hazard hiding inside one of modern leukemia medicine’s most potent drug pairings. In a report published in the journal eJHaem, physicians describe two patients in their eighties who developed sinusoidal obstruction syndrome, also known as veno-occlusive disease, after treatment with the antibody-drug conjugate inotuzumab ozogamicin combined with the tyrosine kinase inhibitor ponatinib — without ever undergoing the bone marrow transplantation that is usually blamed for the condition. Both patients fell ill after only two courses of inotuzumab ozogamicin, and neither showed the classic warning sign of jaundice. The authors state that this is the first report to warn of a potentially high risk of sinusoidal obstruction syndrome when these two agents, both capable of damaging the cells that line small blood vessels, are used together — a finding with immediate implications for the treatment of relapsed and refractory Philadelphia-chromosome-positive lymphoblastic malignancies around the world.

The two cases involved Philadelphia-chromosome-positive disease, an aggressive form of blood cancer in which an exchange of genetic material between chromosomes creates the BCR::ABL1 fusion gene, an abnormal enzyme that drives white blood cells to multiply relentlessly. Treatment of these malignancies rests on tyrosine kinase inhibitors, drugs designed to shut down the fusion protein’s signaling. Ponatinib is among the most powerful of these agents, able to control the disease even when the tumor carries the T315I mutation, a structural change that renders most other inhibitors useless. Inotuzumab ozogamicin works differently. It is an antibody-drug conjugate: a laboratory-made antibody against CD22, a molecule carried by malignant B cells, chemically tethered to calicheamicin, a bacterial toxin that inserts itself into DNA and breaks it apart. Once the antibody docks onto CD22 and is engulfed by the cell, the toxin is released inside, making the conjugate a precision-guided weapon against relapsed or refractory disease.

Sinusoidal obstruction syndrome is, at its core, a plumbing failure inside the liver. Blood is filtered through sinusoids, a dense lattice of low-pressure capillaries lined by a single layer of fenestrated endothelial cells. When toxins injure this lining, the cells round up, detach and slough into the vessel lumen, where, together with debris and activated clotting, they obstruct the narrow sinusoids and the small central veins beyond them. Blood backs up, pressure in the portal system climbs, fluid leaks into the abdomen as ascites, and the liver swells and becomes tender. The syndrome was first characterized after hematopoietic stem cell transplantation, where high-dose conditioning with radiation and cytotoxic drugs routinely injures the sinusoidal endothelium. It has since been linked to radiation, several chemotherapy agents, calicheamicin-based antibody-drug conjugates, immunosuppressants and even plants containing pyrrolizidine alkaloids. Its textbook features — painful hepatomegaly, weight gain above five percent, rising bilirubin and jaundice — were enshrined in the older Baltimore and modified Seattle diagnostic criteria.

Here lies the biological paradox that caught the team’s attention. Normal liver cells do not express CD22, so inotuzumab ozogamicin should, in theory, ignore the liver entirely. Instead, researchers believe the conjugate is taken up non-specifically by liver sinusoidal endothelial cells, exposing them to calicheamicin’s DNA-damaging assault. Tyrosine kinase inhibitors add a second insult: laboratory studies have shown that dasatinib, ponatinib and nilotinib each exert distinct toxic effects on vascular endothelial cells, impairing wound healing, survival, proliferation, migration and the integrity of the junctions that hold the vessel lining together. Endothelial cells already wounded by one drug may have little reserve left to withstand another. Neither patient in this report had undergone transplantation, received radiation conditioning or carried any other classic risk factor for the syndrome — only the drug combination itself, suggesting that the liver’s quiet capillaries can fail when two endothelium-damaging agents are stacked on top of one another.

The first patient, an 85-year-old woman, had been diagnosed with chronic myeloid leukemia in lymphoid blast crisis. After reduced-dose CHOP chemotherapy she began ponatinib at 30 milligrams daily and achieved a deep molecular response. Concern about cardiovascular events later prompted a dose reduction to 15 milligrams, but her disease relapsed roughly a year and a half after the reduction, and inotuzumab ozogamicin was started on a 28-day cycle while ponatinib was resumed at the higher dose, held only on the days the conjugate was infused because concomitant use of the two drugs is off-label. After a single cycle, her major BCR::ABL1 transcript fell below the level of detection. Yet following two courses of inotuzumab ozogamicin, ascites and whole-body edema set in, and her weight climbed by 8.8 kilograms, a 14.5 percent gain. Liver enzymes and C-reactive protein rose slightly, but bilirubin stayed normal and her liver was not painful. Computed tomography revealed moderate-to-severe ascites and a patent umbilical vein, a collateral channel that reopens when pressure in the portal system rises. An ultrasound-based HokUS-10 score of 6 supported a diagnosis of probable sinusoidal obstruction syndrome, made two months after her last conjugate dose. Diuretics and prednisolone at 0.5 milligrams per kilogram slowly cleared the fluid, and ponatinib could be safely restarted eight months after diagnosis.

The second patient, an 82-year-old man with Philadelphia-chromosome-positive acute lymphoblastic leukemia, followed a harder road. Induction therapy with dasatinib and prednisolone restored his blood counts, but measurable residual disease, detected with a sensitive assay for the minor BCR::ABL1 transcript, remained positive. Three cycles of the immunotherapy blinatumomab failed to extinguish it, so treatment was switched to inotuzumab ozogamicin, with ponatinib continued on the same intermittent schedule used in the first case. After his seventh dose, ascites and edema appeared and he gained 6.2 kilograms, an 11 percent increase. Again, his bilirubin never rose. But computed tomography showed mild ascites and a mosaic pattern of liver enhancement, a radiologic signature of sinusoidal dilatation, and a transjugular liver biopsy — chosen because the usual percutaneous route is unsafe in patients with ascites and low platelet counts — delivered a definitive answer: fibrosis around the central and portal veins, degenerating hepatocytes and marked congestion packed with lymphocytes and hemosiderin-laden macrophages. This was proven sinusoidal obstruction syndrome. He received defibrotide, a drug that protects the vessel lining, at 6.25 milligrams per kilogram every six hours for five weeks, alongside diuretics. His symptoms eased only slowly, and ponatinib was cautiously restarted three months after diagnosis as his residual disease crept upward.

What makes the report clinically significant is how easily both cases could have been missed. Under the conventional Baltimore and modified Seattle criteria, a significant rise in bilirubin or painful hepatomegaly is essentially the price of admission, and neither patient paid it. The European Society for Blood and Marrow Transplantation issued refined criteria in 2023 that sort adult disease into probable, clinical and proven categories and deliberately drop the requirement for hyperbilirubinemia, allowing ultrasound and elastography findings to count toward a diagnosis. One such tool is HokUS-10, a scoring system that awards points for ten sonographic parameters, including enlargement of the liver’s left and right lobes, gallbladder wall thickening, a dilated portal vein, a visible paraumbilical vein, the volume of ascites, slowed or reversed portal flow and a stiff hepatic arterial waveform. In post-transplant patients it achieves 100 percent sensitivity and 95.8 percent specificity at a cut-off score of 5. The first patient’s score of 6 fit neatly; the second patient’s score of 3 fell short, underscoring why histology, pressure measurements and liver stiffness assessment remain essential when suspicion persists despite an unremarkable ultrasound.

The epidemiology strengthens the suspicion that the combination, rather than inotuzumab ozogamicin alone, is at fault. In the randomized phase 3 INO-VATE trial, which established the conjugate’s efficacy in relapsed or refractory acute lymphoblastic leukemia, patients who received inotuzumab ozogamicin without a subsequent transplant developed sinusoidal obstruction syndrome at a rate of only 3.0 percent, with a median treatment duration of 8.9 weeks across roughly three cycles. The cumulative doses given to the two patients in the new report were not unusually high, yet both developed the syndrome after just two courses. Clues from other trials point in the same direction. A phase 2 study of dasatinib combined with inotuzumab ozogamicin-based induction for newly diagnosed Philadelphia-positive leukemia had to amend its protocol mid-study, moving dasatinib to a sequential schedule after cases of sinusoidal obstruction syndrome appeared. Intriguingly, a phase 1/2 trial pairing the conjugate with bosutinib reported no cases despite truly concomitant administration — consistent with laboratory evidence that tyrosine kinase inhibitors differ sharply in how much endothelial damage they inflict. Ponatinib, the new report suggests, may sit near the dangerous end of that spectrum.

The authors are careful about the limits of a two-case report. They cannot exclude that inotuzumab ozogamicin alone caused the injury, independent of ponatinib, and no generalized conclusions can be drawn from so small a sample. They also concede, in hindsight, that the conjugate might have been stopped earlier in the first patient once her molecular remission had been confirmed. Still, the practical lessons are difficult to ignore. Both episodes of sinusoidal obstruction syndrome were only moderate in severity and neither was life-threatening, but recovery consumed months, during which treatment for the underlying cancer had to be suspended — a costly delay in diseases where residual leukemia can rebound quickly. The team urges clinicians to weigh treatment duration and drug combination carefully, to avoid pairing inotuzumab ozogamicin with endothelium-damaging kinase inhibitors where alternatives exist, and to watch for atypical presentations in which jaundice never arrives and only weight gain, edema and subtle imaging changes herald the syndrome.

The report arrives at a moment when modern cancer therapy is increasingly assembled from combinations of precisely engineered molecules, each tested thoroughly on its own and far less thoroughly together. The liver’s sinusoidal endothelium, a living filter only one cell thick, may be one of the quiet places where such combinations exact their toll. As new regimens pairing antibody-drug conjugates with kinase inhibitors move from concept into clinical trials for Philadelphia-chromosome-positive leukemia, the team’s warning amounts to a request for vigilance: track patients’ weights, scrutinize ultrasound waveforms and question unexplained ascites even when bilirubin is calm. Both of the first two patients survived their liver injuries and eventually returned to leukemia therapy, but only after months of lost time. The next patient may not be recognized so quickly unless physicians learn to look for a liver disease that no longer announces itself with yellow eyes.