Background

Leukemia is the most common cancer in children and the involvement of the central nervous system (CNS) increases the risk of CNS relapse [1]. Risk factors of CNS leukemia include infant onset, T-cell acute lymphoblastic leukemia (T-ALL), hyperleukocytosis, and chromosomal abnormalities including KMT2A rearrangements, BCR-ABL1, and TCF3-PBX1 [1, 2]. These features as well as a traumatic tap at initial diagnosis also increase the risk of CNS relapse in ALL [1,2,3].

CNS status at initial diagnosis is classified as follows: CNS1, no leukemic cells in cerebrospinal fluid; CNS2, white blood cell (WBC) counts < 5/µL with leukemic cells; CNS3, WBC counts ≥ 5/µL with leukemic cells or neurological symptoms with imaging findings by computed tomography (CT) or magnetic resonance imaging (MRI) [4]. MRI is helpful to diagnose CNS leukemia in patients with neurological symptoms [5, 6]. The percentage of CNS3 leukemia with neurological symptoms and imaging findings in newly diagnosed ALL was estimated to be less than a few percent [1, 5].

Residual masses present in the CNS after induction or consolidation therapy are considered treatment failure (TF) [7]. TF affects later therapies, such as the decision to withdraw from protocol therapy and whether patients proceed to allogeneic hematopoietic cell transplantation (HCT) [7, 8]. The percentage of pediatric patients with induction failure (IF), defined by leukemic cells in bone marrow (BM) or extramedullary site after induction, was 2.4% (1041/44,017 cases) [8]. Among the 1041 cases, CNS status was available for 684 cases. Of these, CNS leukemia accounted for 6% (44/684) [8]. However, the ratio and prognosis of patients with isolated CNS residual masses by imaging have not been defined.

Here, we report a patient with T-ALL who developed multiple intracranial masses at diagnosis and had remaining residual lesions detected by follow-up MRI at the end of early intensification. We decided treatment strategy based on minimal residual disease (MRD) status in BM, assessment of spinal fluid, and whether 11C-methionine (MET) uptake in these CNS residual lesions.

Case presentation

A 9-year-old boy was introduced to our hospital with somnolence, slurred speech, and petechiae. A complete blood count showed marked hyperleukocytosis (WBC 760 000/μL). Flow cytometric analysis showed positivity for CD1a, CD2, CD4, CD5, CD7, CD8, cytoplasmic CD3, and terminal deoxynucleotidyl transferase. G-banding revealed a normal karyotype and multiplex polymerase chain reaction (PCR) analysis did not detect fusion genes. Taken together, we diagnosed him with T-ALL. MRI revealed multiple intracranial masses with strong signal heterogeneity consisting of hemorrhage and leukemic infiltrations, which were consistent with definition of CNS3 (Fig. 1A and B).

Fig. 1
figure 1

Multiple central nervous system infiltrations with hemorrhages on magnetic resonance images (MRI). A and B T2-weighted and T2*-weighted image (T2*WI) MRI shows heterogeneous signals with leukemic masses and hemorrhages. Edema is observed around multiple lesions

Full size image

We started pre-phase therapy consisting of prednisolone (PSL) in accordance with the Japan Leukemia/Lymphoma Study Group T11 protocol (Fig. 2A) [9]. The number of lymphoblasts steadily decreased to 26 (< 1000)/μL in peripheral blood on day 8, and he was considered a PSL good responder (PGR) (Fig. 2A) [9]. There was no tumor lysis syndrome under the supportive therapy including hydration and rasburicase administration. His initial symptoms including somnolence and slurred speech gradually improved a week after starting corticosteroid. On day 8, we conducted initial intrathecal therapy (IT). Lumbar puncture (LP) showed a WBC count of 2/µL with mild cellular atypia by conventional cytospin. We completed the induction phase without intracranial hemorrhage. However, the patient had grade 3 hyponatremia due to cerebral salt wasting syndrome, and required additional sodium chloride supplementation by infusion for two weeks. On day 29, he also had grade 3 muscle weakness lower limb due to vincristine or corticosteroid myopathy and skipped vincristine once.

Fig. 2
figure 2

Treatment course, serial MRI, and 11C-methionine (MET) positron emission tomography (PET). A The Japan Leukemia/Lymphoma Study Group (JPLSG) T11 high-risk protocol with treatment response and CNS evaluation by imaging studies. B T2-weighted MRI showed residual multiple lesions at the end of induction. C At the end of early intensification, T2-weighted MRI shows two representative residual lesions at the same position as in (B). No significant MET uptake was detected. Scale bar indicates the standardized uptake value (SUV). D MRI images were captured two months after maintenance therapy. BL, blast; BMA, bone marrow analysis; CNS, central nervous system; CRT, cranial radiation therapy; HR, high risk; MET-PET, 11C-methionine positron emission tomography; MRD, minimal residual disease; NEL, nelarabine; IT, intrathecal; PSL, prednisolone

Full size image

After day 18 of induction therapy, LP showed no evidence of malignant cells by cytospin. We did not perform flow cytometry analysis for spinal fluid in this case. The minimal residual disease (MRD) determined by PCR to monitor BM responses showed a reduction in lymphoblasts below 0.001% after induction therapy (Fig. 2A). However, MRI revealed multiple CNS masses (Fig. 2B). PCR-MRD for BM samples was maintained below 0.001%; however, MRI still showed multiple CNS lesions post early intensification (Fig. 2C).

To avoid invasive approaches, including needle biopsy or surgical resection, which might cause delaying further intensification therapies, we used 11C-methionine positron emission tomography (MET-PET) to investigate the activity of the residual lesions. MET-PET revealed no significant uptake, suggesting the low possibility of viable leukemic cells in residual regions (Fig. 2C). Thus, we classified him as belonging to a high-risk group, but not a very high-risk group requiring HCT, and he received continued chemotherapy including intensive IT, nelarabine, and cranial radiation therapy (CRT) in accordance with the T11 protocol [9]. The follow-up MRI revealed remission immediately after the maintenance therapy (Fig. 2D). At 18 months from the cessation of maintenance therapy and 46 months from onset, he has maintained a first remission and can attend regular school without neurological sequelae.

Discussion and conclusions

We experienced a patient with T-ALL who showed multiple residual lesions in the CNS at the end of early intensification. PGR, good PCR-MRD status in BM, and no specific uptake of MET in the CNS allowed a continuation of chemotherapy with CRT, and the avoidance of HCT.

A meta-analysis of hematologic tumors, including pediatric to adult cases with intracranial sarcomas, reported 82 cases from 1999 to 2019 [10]. The report included 66 cases of AML, 10 cases of CML, 5 cases of APL, and 4 cases of ALL. Among them, three patients with ALL were relapsed cases [10]. Only two reports showed residual intracranial lesions after the start of treatment in primary ALL, both of which were Philadelphia chromosome positive ALL [11, 12]. The prognosis and appropriate treatment for cases of isolated residual CNS masses with excellent treatment responses in the non-CNS region have not been established because of the rarity of these cases.

The prognosis of IF in T-ALL was worse than that of B-ALL, with 10-year survival rates of 28 ± 3% and 41 ± 3%, respectively [8]. Historically, radiation therapy is beneficial for high-risk groups with T-ALL. The omission of prophylactic CRT in the T-ALL and CNS3 groups has not been established. Patients with CNS3 in the T-ALL group were ineligible in the recent AALL0434 study [13]. Taken together, chemotherapy with CRT was appropriate for our patient, even with negative MET uptake and good responses in the BM.

18F-fluorodeoxyglucose (FDG) is involved in glucose transport in tumors and noncancerous lesions including infections, inflammation, and the brain, especially the grey matter. Instead, MET has greater specificity for brain tumors using large amino acid transporter 2 [14]. MET-PET was developed to detect brain tumors including gliomas [14,15,16]. Compared with FDG, MET has an equivalent uptake in children with Hodgkin lymphoma and non-Hodgkin lymphoma at diagnosis and follow-up [17]. Interim PET has been feasible for evaluating T-lymphoblastic lymphoma [18]. Furthermore, MET-PET has been used to determine treatment responses in CNS lymphoma [19, 20]. Seo-Yeon et al. showed three response patterns correlated with prognosis, including low-, intermediate-, and high-risk, based on the interim tumor-to-normal tissue (T/N) ratio by MET uptake [19]. The interim T/N ratio in our patient was equivalent to the low-risk category, suggesting the low possibility of viable tumors in residual lesions.

However, there were limitations to this study. We could not perform MET-PET at onset due to his poor initial condition. If the lesion is small-scale or has low uptake of MET, it is difficult to distinguish the viability of residual tumor, which should be judged comprehensively in conjunction with other modalities. Thus, other clinical responses should be taken into account when decision-making. Finally, careful follow-up is required for our patient.

In conclusion, we experienced a patient with T-ALL and multiple residual lesions in the CNS at the end of early intensification. Further retrospective or prospective studies are required to determine the frequency of cases with intracranial residual lesions after induction therapy.