1. Introduction
However, this emerging cellular immunotherapy faces multiple challenges in the treatment of malignancies. On the one hand, most of the patients treated with CAR T-cells suffer from cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), hematological toxicity, and grade 3/4 side effects that can threaten the life of patients
[8][9][10]; on the other hand, short persistence of CAR T-cells in vivo is related with poor efficacy
[11]. Therefore, some optimization schemes for CAR T-cells, such as multiple target treatment, discovery of new targets, and CAR construct optimization, have been widely studied to reduce adverse events and improve CAR T-cell efficacy
[12][13][14][15][16]. However, the clinical application strategies of CAR T-cells, which have an essential role in the overall treatment process, have rarely been discussed. After determining that a patient is suitable for CAR T-cell therapy, it is necessary to determine the lymphatic clearance scheme and dosing strategy according to the patient’s basic condition, to help the patient obtain the best blood environment and maximize the utilization of CAR T-cells
[17]. Furthermore, for refractory and relapsed hematological malignancies, the combined application of multiple treatments may lead to longer event-free survival for patients
[18][19]. The implementation of combined strategies to increase the persistence or antitumor activity of CAR T-cells has become a research hotspot. In addition, the safe and effective implementation of a strategy is inseparable from the supervision of the disease and careful comprehensive management, which will promote patients obtaining complete remission (CR) and disease-free survival.
2. Lymphodepletion Regimen
After infusing into the blood, CAR T-cells specifically recognize the target antigen through the single-chain variable fragment (scFv) binding domain, which requires a suitable immunomodulatory environment for T cell survival and proliferation
[20][21][22]. Lymphodepletion followed by CAR T-cell infusion remodels the immune environment for creating enough T cell pool, decreasing immunosuppressive cell populations, increasing the availability of cytokines and thus promoting lymphocyte proliferation and survival, and increasing tumor antigen presentation by inducing cell death
[23][24][25]. Cyclophosphamide (Cy) and fludarabine (Flu) are two major lymphodepletion compositions, known as potent immunosuppressive agents. In one case, the overall response rate (ORR) without lymphodepletion was only 25%, while ORR rose to 58% after lymphodepletion that had made changes in cytokine profiles and had increased CAR T-cell durability
[26][27]. However, how the lymphodepletion regimen impacts CAR T-cell efficacy is still unclear.
Generally, lymphodepletion with cyclophosphamide and/or fludarabine is implemented on -5, -4, and -3 days before CAR T-cell injection to create an immunosuppressive environment that provides CAR T-cell expansion in vivo
[28][29]. Moreover, the chemotherapy pretreatment regimen adopted in most trials is a dose of fludarabine generally maintained at about 30 mg/m
2·d or 25 mg/m
2·d; the dose of cyclophosphamide is 200 to 3000 mg/m
2·d
[7][11][30][31][32], which is considered to be a relatively safe dose. The co-administration of cyclophosphamide and fludarabine is viewed as adequate lymphodepletion. Studies have shown that a low incidence of CR and high rate of relapse may be associated with inadequate CAR T-cell persistence and expansion in vivo and cell-mediated immunosurveillance
[17][33][34]. Cyclophosphamide and fludarabine-mediated lymphodepletion reshape the blood environment, prolong CAR T-cell survival time, regulate the release of activating cytokines, and improve the event-free survival time of patients
[33][35][36].
Studies have shown that the effect of lymphodepletion and the reactivity of subsequent CAR T-cell therapy is dose-dependent
[11]. A total of 17 patients received high-dose cyclophosphamide of 3 g/m
2·day (HD-Cy), while 8 patients received low-dose cyclophosphamide at 1.5 g/m
2·day (LD-Cy), with fludarabine at 25 mg/m
2·day for 3 days. Results manifested that CAR T-cell expansion in the HD-Cy group was remarkably higher than that in the LD-Cy group. The disease response rate of the former was 94%, while the latter was only 38%
[11]. This indicates that deeper lymphodepletion leads to higher clinical response due to thoroughly eliminating immunosuppression and reducing cytokine rejection of CAR T-cells. In addition, high-dose Cy increased the level of IL-7 and MCP-1 and was associated with CR and/or progression-free survival (PFS)
[35][37]. However, the mechanism through which HD-Cy leads to a favorable cytokine profile has yet to be determined
[11][38]. Nevertheless, in a previous study targeting diffuse large B-cell lymphoma (DLBCL), patients who received modulated chemotherapy with low a dose of cyclophosphamide (500 mg/m
2) and fludarabine (30 mg/m
2) had a higher level of cytokines, particularly IL-5 and IL-15, which are critical for T cell proliferative cytokines
[29][39].
The addition of fludarabine to cyclophosphamide as a lymphodepleting regimen before CD19-directed CAR T-cell therapy significantly improved outcomes in patients treated with CAR T-cells
[40]. Over 90% of patients who received fludarabine-containing lymphodepletion achieved CR and had no relapse, while 58% of patients without fludarabine suffered a relapse. In patients with CD19+ relapse, a loss of CAR T-cell in blood was detected by PCR analysis
[33]. Incorporation of fludarabine into the lymphodepletion regimen, lower concentration of LDH, and higher platelet count before the lymphodepletion regimen containing fludarabine may lead to minimal residual disease-negative complete remission (MRD
- CR)
[26]. Compared with the lower cumulative fludarabine exposure during lymphodepletion, patients in the higher group had an 11-month improvement in leukemia-free survival, and the CD19+ recurrence rate within 1 year decreased from 100% to 27.4%
[41]. However, when the disease burden is high, the effect is weakened, so the dose of fludarabine should be formulated according to the patient’s disease burden and designed individually
[42]. Interestingly, for DLBCL patients who relapsed after CAR T-cell therapy, pre-existing CAR T-cells were revitalized after the second lymphatic clearance containing fludarabine and created an anti-tumor effect with a predominant grade 2 CRS
[43]. This case indicates that CAR T-cells could be reactivated under circumstances by disturbing the immune equilibrium. CAR T-cell activation and cytotoxicity depend on the immune homeostasis of the blood environment, and the blood environment after lymphatic clearance seems merely to maintain for a period time, which may be one of the reasons why patients can only achieve short-term remission. At the same time, to obtain satisfactory clinical efficacy, the degree of lymphatic clearance needs to be proportional to the patient’s disease burden; thus, a universal single dosage is not suitable
[11][33].
3. Dosing Regimen of CAR T-Cells
Before lymphocytes are removed, T cells will be isolated from the peripheral blood mononuclear cells (PBMC) of patients as raw materials for subsequent manufacturing of CAR T-cells
[44]. The preparation of CAR T-cells is about 13 days, including T cells expanding for 9 days; another study shortens the preparation time to 7 days with similar effectiveness
[45][46]. Surprisingly, Gracell Bio in China has developed a fabrication technique of FasTCAR, which shortens the preparation time of CAR T-cells to 1 day
[47]. FasTCAR technology transforms the activation, transduction, and expansion steps into a single “concurrent activation-transduction” step, eliminating the time for in vitro expansion of CAR T-cells
[48]. After obtaining CAR T-cells and achieving the standards of identity, potency, sterility, and adventitious agents, the infusion strategy of CAR T-cells can be determined by clinicians and manufacturers based on phase I clinical trials identifying the maximally tolerated dose (MTD)
[49].
In CD19-directed CAR T-cell products (axicabtagene ciloleucel, lisocabtagene maraleucel, tisagenlecleucel, brexucabtagene autoleucel) that have been proved by FDA, the single infusion dose is about 1–3 × 10
6/kg, except that tisagenlecleucel is 0.1–6×10
8 viable cells, of which 5×10
8 viable cells are more suitable for DLBL and relapsed or refractory follicular lymphoma (FL)
[3][4][6][7][28][29][50][51][52]. In adults, some CD19-targeted products in the treatment of different diseases in the dose range of 2 × 10
6/kg showed higher ORR and CR and lower incidence of high-grade CRS and neurological event (NE) side effects. Determination of a safe/efficacious dose range is based on understanding dose-response/exposure/safety analyses. A dose-escalation test showed that 1 × 10
6/kg is the highest single-injection tolerated dose for R/R ALL and non-Hodgkin’s lymphoma (NHL) patients; in this condition, all toxicities, including grade 4 CRS, are reversible
[53]. Moreover, a lower CAR T-cell dosage of 1 × 10
5/kg is effective and safe for treating r/r B-ALL
[54]. It can be seen that the clinical results cannot intuitively find the relationship between the dose and the efficacy of CAR T-cells, except that high doses show dose-dependent toxicity. This may be due to the initial CAR T-cell viability and different immunosuppressive gene expression of the tumor microenvironment
[55].
A single injection of low-dose CAR T-cells is considered safe and effective, but not with durable remission. Some researchers have studied one-time injection or segmental injection of low-dose or high-dose CAR T-cells
[32][38]. A single-arm, open-label study verified the difference between these two strategies in efficacy. The trial is divided into a high-dose single infusion (5 × 10
8/kg), low-dose single/fractionated infusion (5 × 10
7/kg), and high-dose fractionated infusion (5 × 10
8/kg); the numbers of patients were 6, 9, and 20, respectively
[56]. The high-dose fractionated group showed the best effect, with a complete remission rate of 90% and a two-year overall survival rate of 73%, while in the high-dose single infusion group, patients suffered from refractory CRS complicated with culture-positive sepsis
[56]. The low rate of adverse events of the high-dose fractionated group indicates that fractionated infusion of a high dose may alleviate immune response. Characteristics including age, disease phenotype, chemotherapy regimen received before infusion, and whether to undergo stem cell transplantation are all related to dosage selection, indicating that it is difficult to set the dosing interval to enable long-term complete remission
[57]. Before receiving CAR T-cell therapy, these patients had experienced more than two lines of systemic therapy; however, at the same therapeutic dose, there were differences in the patients’ physical condition, immune homeostasis of the blood environment, disease burden, and cytokine pool after lymphatic depletion. A scoring system is set up to determine various aspects of the patient’s physical state, thereby determining the corresponding dose standard within a certain scoring range. The dosing regimen of CAR T-cells is individualized. However, studies of the clinical response of patients to CAR T-cells could provide clinicians with data for developing a personalized dosage strategy.
4. Combination Strategies
Anti-CD19 CAR T-cells show a strong short-term CR in the treatment of refractory and relapsed B-cell diseases. Its limitations include the inability to achieve long-term disease-free survival and risk of recurrence due to a lack of sustained expansion of CAR T-cells
[55], and patients with a high tumor burden treated with allogeneic hematopoietic stem cell transplantation (allo-HSCT) could not achieve complete remission
[58]. CAR T-cells or allo-HSCT achieved considerable short-term efficiency when used alone; researchers can explore combination to achieve long-term disease-free survival.
Allo-HSCT, as a kind of consolidation strategy, has shown a powerful capability for R/R B-cell lymphoma
[58][59][60][61][62][63]. With a quick bridge to allo-HSCT after CAR T-cell therapy, the one-year event-free survival (EFS) was up to 73% compared with 50% for CAR T-cells alone, and the rate of severe grade 2–4 acute graft-versus-host disease (aGVHD) was less than 23.1% compared with 41.6% of allo-HSCT alone
[7][64][65][66]. Moreover, the timing of allo-HSCT treatment affects patient prognosis. It was shown that the period within three months after treatment is considered to be a suitable time window for consolidation therapy
[65]. MRD
- CR patients receiving allo-HSCT showed detectable CAR T-cells for 21 months. Jiang H et al.
[66] found that allo-HSCT was an independent prognostic factor for EFS and relapse-free survival (RFS) in MRD
- CR patients. For patients with CD19-negative relapse, allo-HSCT may be desirable as a consolidated method because of immune reconstitution
[54]. From another point of view, CAR T-cell therapy has obtained a larger range of patients for allo-HSCT
[66]. In addition, CAR T-cells can be used as maintenance therapy after allo-HSCT therapy. Shi et al. found that CAR T-cells combined with lenalidomide maintenance therapy significantly prolonged PFS in patients with multiple myeloma. Lenalidomide increased the antitumor activity and persistence of CAR T-cells, which might be due to the fact that lenalidomide can maintain the long-term immune surveillance of CAR T-cells in remission, and this combined strategy promotes T cells toward the less terminally differentiated phenotype
[67]. RNA sequencing and assays for transposase-accessible chromatin indicated that lenalidominde can alter T-helper response, cytokine production, T cell activation, cell-cycle control, and cytoskeletal remodeling
[68][69].
CAR T-cells combined with ibrutinib, a kind of bruton’s tyrosine kinase inhibitor, improved CAR T-cells engraftment, tumor clearance, and survival in human xenograft models of resistant acute lymphocytic leukemia
[70]. Ibrutinib is a crucial component of a first-line treatment option for chronic lymphocytic leukemia (CLL), and it showed superior PFS and CR in progressive CLL patients who were too frail to receive aggressive therapy
[71]. Continuous use of ibrutinib leads to tolerance and tumor recurrence, and fortunately, CAR T-cells show obvious therapeutic efficacy and durability after ibrutinib failure
[72]. Therefore, the combination of these two therapies may create an advance for r/r CLL patients. Compared with CLL patients treated with CAR T-cells without ibrutinib, CLL patients treated with CAR T-cells combined with ibrutinib were associated with lower CRS severity and lower serum concentrations of CRS-associated cytokines, which is consistent with pre-clinical results
[73]. Moreover, one-year overall survival and PFS probabilities of a combination of CAR T-cells with ibrutinib were 86% and 59%, respectively, while anti-CD19 CAR T-cell therapy was only 38% and 50%
[74]. After more than 1 year of ibrutinib therapy, the efficiency of CAR T-cell engraftment, tumor clearance, and survival were improved
[70].
CAR T-cell therapy with immune checkpoint inhibitors has now moved to the clinical stage. A pre-clinical study demonstrated that CAR T-cell therapy combined with PD-1 blockade showed better tumor control, and the persistence of CAR T-cells was up to 21 days in mice
[75]. Furthermore, neuroblastoma and metastatic melanoma with CAR T-cell therapy combined with PD-1 blockade indicated that PD-1 blockade can enhance cytokine production and reduce activation-induced cell death (AICD) of CAR T-cells after repeated antigen stimulation
[17][76].
Moreover, the effect of the first-line chemotherapy for CLL patients can be strengthened by combining it with CAR T-cell therapy. After regular chemotherapy treatment with pentostatin, cyclophosphamide, and rituximab, patients who received CAR T-cells ranging from 3 × 10
6 to 3 × 10
7 CAR T-cells/kg were observed to have more than 28 months CR and modest CRS neurotoxicity
[34]. In addition, the above methods combined with CAR T-cell therapy are moving to the clinical stage, and other molecular medicines such as immunomodulatory drugs, utomilumab, hypomethylating agents, Phosphoinositide 3-kinase (PI3K) inhibition, γ-secretase inhibitors, and fas blockade, are part of widely ongoing pre-clinical studies
[77]. These combination strategies represent a new treatment paradigm that leverages the ability of genetically modified T cells to target and destroy tumor cells.
5. Occurrence and Management of Side Effects
Anti-CD19 CAR T-cell therapy led to cytokine release syndrome, hematological toxicity, and ICANS attributed to inflammatory cytokines and chemokines released by immune cells, which are the main obstacles after CAR T-cells were administrated into the blood system via intravenous injection
[78]. The average onset time of CRS is 3 days after injection, and it lasts for about 7 days
[79]. When the neurological event occurs, it is generally within 8 weeks and lasts for 28 days
[79]. Blood events such as thrombocytopenia and neutropenia may exist throughout the whole disease event. The incidence of CRS in the milestone trial is as high as 77%, the average incidence of neurological events is more than 30%, and blood events are more than 40%
[7]. Although corresponding support and drug treatments were performed during the treatment process, the final mortality rate of patients is more than 40%, accompanied by disease progression and aggravating effects of side effects
[6][7][80]. Therefore, it is meaningful to review the occurrence and management of side effects. Moreover, early detection and timely control of adverse events also play important roles in the treatment of disease.
Generally, grade 1/2 CRS, mainly including fever, fatigue, nausea, mild hypoxia, or headache, is not life-threatening and is relieved after symptomatic and supportive treatments performed with close monitoring of the functional status of various organs and the level of cytokines in the blood. Grade 3/4 CRS is life-threatening to patients and is closely related to the subsequent neurotoxic events
[10]. Grade 3/4 CRS is accompanied by more severe hypoxia, which requires ventilator support if necessary, and ICU care is needed in some severe cases
[81]. In 2020, Pierre Sesques et al.
[76][79] did a retrospective study on commercial products such as axicabtagene ciloleucel and tisagenlecleucel and found that grade 3/4 CRS and neurological events were reduced to 8% and 10% compared with the key trial in 2017, which is 11% and 32%, respectively. The lower rate of grade 3/4 side effects is due to earlier and more systematic interventions.
When CRS occurs, the levels of IL-2, IL-4, IL-6, IL-10, TNF-α, and IFN-γ in the blood are significantly increased; in particular, IL-6 is parallel to the severity of CRS. Compared with grade 1/2 CRS, grade 3/4 CRS showed a rapid increase in IL-6 in the early stage, and the patients’ fever was high and lasted for a significant period. These clinical symptoms can be used as an early prediction of severe CRS and can provide treatment strategies for adequate treatment in the later period
[10][82][83]. On the other hand, the rapid expansion of infused CAR T-cells is also related to CRS
[9]. At the same time, the concentration of the above cytokines also significantly increased with the expansion of CAR T-cells, which is associated with not only injection protocol but also the kind of costimulatory domain of CAR T-cells
[16][31]. The IL-6 antagonist tocilizumab is the first-line treatment for CRS; however, it is usually used when supportive treatment does not work. If necessary, it can be used in combination with the immunosuppressant dexamethasone. Generally, it can provide relief after one week of treatment. However, when the tumor burden and the number of recurrences in the past are high, severe CRS requires hemodialysis or plasma exchange to overcome difficulties
[10][29].
In addition, grade 3/4 CRS is related to the increased risk of subsequent infections, especially bloodstream infections that are difficult to distinguish from grade 3/4 CRS. Bloodstream infections are similar to CRS symptoms, but the treatment methods for the two are different. Researchers can determine the infection through imaging methods and laboratory tests, such as computed tomography, bacterial examination, magnetic resonance imaging, C-reactive protein (CRP), procalcitonin, etc. When it is impossible to distinguish, clinicians should use antibiotics based on their experience
[84][85]. Moreover, the incidence of disseminated intravascular coagulation (DIC) increases in patients with grade 3/4 CRS. As a result, early and proper interventions targeted to CRS-related coagulopathy greatly contribute to the control of side effects in CAR T-cell therapy
[82].
Neurotoxic events occurred in the study as compared with the high incidence of CRS, but with the optimization of CAR T-cells, the incidence of CRS and neurological events decreased
[32][86]. ICANS often occurs with grade 3/4 CRS. ICANS includes encephalopathy, confusion, delirium, tremor, restlessness, lethargy, and epilepsy in severe patients (Grade 3)
[7]. There is no specific treatment method, other than relying on supportive treatment. Specific antibodies targeting cytokines related to neurotoxic events may be found in the future. Studies have shown that neurotoxicity is related to CAR T-cells crossing the blood–brain barrier and triggering an immune response in the brain. IFN-γ and IL-6 in the cerebrospinal fluid are nearly 40 times higher than those in the serum
[87]. At the same time, the qPCR of the CAR expression indicates that it is in the cerebrospinal fluid, and the CAR copy number is three times higher
[87]. Thus, the detection of components in cerebrospinal fluid can be used as a method to predict neurotoxicity
[83][85][87]. Due to individual differences, CRS, hematological toxicity, and neurotoxicity may occur in patients at any time after CAR T-cell treatment. Identifying the characteristics of side effects and determining the treatment may be helpful for the positive progress of CAR T-cell therapy.