In 2017, the Food and Drug Administration (FDA) granted the first approvals of two CAR-T therapies for the treatment of adult patients with relapsed or refractory diffuse large B-cell lymphoma (DLBCL) after two or more lines of systemic therapies, including DLBCL not otherwise specified, primary mediastinal large B-cell lymphoma, high-grade B-cell lymphoma, and DLBCL arising from follicular lymphoma, and for patients up to 25 years of age with B-cell precursor acute lymphoblastic leukaemia (ALL) that is refractory or in second or later relapse [1,2].

The registration studies enrolled patients who were previously treated and considered of poor prognosis, with some having had prior stem cell transplantation. Yet, the results that were obtained from the studies (albeit single-arm studies) were quite encouraging and resulted in high responses in those difficult-to-treat patient populations [3]. The genetically engineered T cells, which are injected into the patient’s body, result in engraftment and rapid proliferation. Subsequently, each CAR-T cell attacks many cancer cells. This unique mechanism of action, with the involvement of the patient’s own immune system, resulted in a very high response rate despite the late disease stages and poor prognoses [4,5]. Furthermore, CAR-T therapy was used to bridge other therapeutic strategies with further improvement of responses. With the high response rates and durable responses with long-term follow-up [6], it was hoped that the responses would be for long durations with the possibility of a cure for these patients. Unfortunately, the relapse rate was around 30–60% for patients treated with CAR-T therapy [7]. With limited treatment options for patients who relapse after CAR-T therapy, it is of paramount importance to understand the mechanism of resistance in those patients in order to develop a subsequent effective therapy.

Patients who relapsed following CAR-T therapy fall into two main groups of relapse: CD19-positive cells or CD19-negative cells [6]. The CD19-positive cells relapse occurs when the response to CAR-T therapy is not sufficient, with minimal effect for CAR-T cells and only transient B-cell aplasia. With flow cytometry, CD19 can be detected on the surface of the cells. Different animal studies attempted to identify the cause of this weak response to CAR-T therapy. Some of the factors that were proposed are the quality of CAR-T cells synthesised from children and young adults compared to those from older adults. For younger patients, the quality of the CAR-T cells was better, which was reflected in the tendency to receive better responses (longer event-free survival times) in the younger age groups versus older age groups. Studies have shown that the starting T-cell phenotype of CAR-T cell manufacturing is crucial to patients’ prognosis. A study conducted by Gardner et al. enrolled 43 children and young adults [7]. They looked for factors that contributed to a maintained response among those who responded, achieved, and maintained a minimal residual disease (MRD)-negative state beyond 63 days, versus the non-responders who did not achieve that [7,8]. The analysis of apheresis obtained from both groups showed that in non-responders, the percentage of CD8+ T cells, which expressed lymphocyte-activated gene-3 (LAG-3) and PD-1, had significantly increased, whereas in the responders group, the number of CD4+ CAR cells and CD8+ CAR cells was significantly higher at the time of peak implantation. They also found that the number of CD8+ T cells expressing tumour necrosis factor-α (TNF-α) was much less in the non-responder group compared to the responder group. This led to the assumption that the increase in LAG-3-expressing cells (LAG-3+ T cells) and decrease in the cytokines after stimulation leads to the production of ineffective CAR-T cells, or ones with low potency resulting in a CD19-positive relapse. This is in comparison to the responders who have less LAG-3-expressing cells and an increase in cytokines after stimulation, thus having a better response [9]. The presence of LAG-3+ T cells is a biomarker for T-cell exhaustion. T-cell exhaustion refers to a state where the T cells produced are dysfunctional with the decrease in the effectors and increase in the expression of the inhibitory receptors. This phenomenon is induced by chronic stimulation in patients with cancer. Chronic stimulation leads to the production of T cells that are usually deficient in their intrinsic activity. Accordingly, CAR-T cells derived from patients with cancer will be defective in function, which can lead to poor prognosis and relapse. The mechanism of CAR-T cell exhaustion is still not clearly understood. It was suggested that the CARs on the CAR-T cells could form a cluster that is formed independently of the antigen stimulation, which leads to the development of tonic CAR-CD3ζ signalling that can cause CAR-T cell exhaustion. Additionally, the endogenous T-cell receptor signal of CAR-T cells caused by a specific antigen stimulation can result in T-cell exhaustion [6,9].

The CD19-negative relapses represent up to 20% of the relapsed CAR-T cell treated patients. With CD19-negative relapse, CD19 is not detected on the surface of the cells. This occurs due to mutation in exons 1–13 of the CD19 gene [10]. Exons 1–4 carry the codes for the extracellular domains, and exons 5–13 carry those of the transmembrane domain. Studies have shown that patients who show CD19-negative relapse have developed mutations in exons 2–5 in the form of one insertion or deletion [11]. The percentage of CD19-negative cells can be measured by flow cytometry, which is also able to assess the frequency of allelic mutations and the percentage of cells where there is a homozygous loss and biallelic mutations, which are major reasons for the loss of the targeted epitope in the membrane of CD19, with the subsequent escape from the effect of anti-CD19 CAR-T. Called antigen escape, this leads to the development of resistance to CD19-directed CAR-T therapies.

There are other mechanisms that were described as possible causes for the occurrence of CD19 gene mutation and antigen escape, such as alternative splicing with the subsequent development of CD19-negative resistance [12]. These resistance mechanisms are not unique to CAR-T therapy. They have been observed in other tumour types, e.g., breast cancer tissues show splicing of exon 16 of HER2 leading to resistance to trastuzumab. Likewise, melanoma tissues show splicing of BRAF (V600E), which results in dimerisation and consequently resistance to vemurafenib. Splicing of mRNA (e.g., in SF3B1, SRSF2) also occurred in haematological tumours such as myelodysplastic syndrome (MDS) and chronic lymphocytic leukaemia (CLL) resulting in resistance to treatment [13].

Both approved CAR-T cells cause the condition called cytokine release syndrome (CRS), which includes a set of life-threatening or even fatal reactions to the infusion of CAR-T therapy. Generally, CAR-T therapy should not be administered to patients who have active infections or inflammatory disorders.

Treatment with CAR-T cells causes rapid activation of T cells and the release of high levels of cytokines. The main cytokines involved in CRS are IL-6, IL-10, and interferon (IFN)-Υ. IFN-Υ, when secreted, causes the release of other cytokines such as IL-6, IL-10, and TNF-α. Its release is also accompanied by fever, chills, headache, malaise, and other symptoms occurring with CRS. IL-6 also is considered a key cytokine in the pathophysiology of CRS since it is highly elevated in cases suffering from CRS. It is also responsible for the occurrence of many of the severe symptoms such as vascular leakage and disseminated intravascular coagulation (DIC). The occurrence of life-threatening CRS mandates immediate treatment with tocilizumab or tocilizumab and corticosteroids. [1,2,14]. Tocilizumab, an anti-IL-6 receptor antagonist, is the standard for CRS management, although the optimal timing of administration is still unclear [14,15] (Figure 1). Management is also depending on the severity of the symptoms. The commonly used scheme for the assessment of the severity of CRS is the one developed by the National Cancer Institute (NCI).

CRS can affect different body systems and organs. Other medical specialties such as neurology, nephrology, cardiology in addition to the treating haematologist are required for the effective management of patients with severe CRS. Intensive care facilities should also be available when needed. The early and appropriate management of severe cases of CRS generally can lead to a positive outcome [15].

CRS can affect different organs, and the symptoms can range from mild to severe and debilitating [16]. Respiratory symptoms that are common with CRS can range from cough and hypoxia to acute respiratory distress syndrome, which can be seen on lung images as opacities affecting large areas of the lung and can be bilateral [16,17,18].

Cardiac and circulatory manifestations can be in the form of hypotension and can progress to capillary leak syndrome with subsequent peripheral and pulmonary edemas and signs of cardiac failure [19].

Neurotoxicity is also a serious type of toxicity that can occur following treatment with CAR-T therapy [20]. It can be in the form of mild confusion symptoms and headaches to a severe form of hemiparesis, cranial nerve palsies, and even seizures [21]. As a result of the relatively common incidence of neurotoxicity and the seriousness of it, the term CAR-T cell-related encephalopathy syndrome (CRES) is now used [22]. Despite the common incidence of CRES (being the second most common toxicity associated with CAR-T therapy) and the severity of the symptoms, neither the exact pathophysiology [23] nor the best way to treat it is clear; thus, the management of it may need to be carried out in specialised care units [24,25] (Figure 2).

Investigators have attempted to decrease the CRS signs/symptoms of CAR-T therapy while maintaining its positive therapeutic effect. Frey et al. conducted a study on 35 adults with relapsed/refractory (R/R) ALL who received CAR-T therapy [26]. Patients were enrolled in 1 of 3 dosing cohorts. The first cohort was a low-dose cohort where nine patients received a low dose of therapy either as a single dose or fractionated dose. In the second cohort, six patients received CAR-T therapy as a high single dose, and the third cohort had 20 patients enrolled and received high dose therapy either as a single dose or fractionated dose. Any fractionated dose was given on 3 consecutive days with 10%, 30%, and 60% given on days 1, 2, and 3, respectively. This schedule allowed for the assessment of CRS occurrence, taking necessary measures if they occur, and trying to avoid fatal complications.

The nine patients in the low single or fractionated dose had a complete remission (CR) rate of 33% with manageable toxicity. In the six patients who received the high single dose, three had fatal CRS complications, and the other three achieved CR. The results in the third cohort, where patients received a fractionated high CAR-T dose, were impressive with a CR rate of 90%, an event-free survival rate (EFS) of 49.5% (95% CI: 21–73%), and a 2-year overall survival (OS) rate of 73% (95% CI: 46–88%). Superior results were achieved for the patients who were dosed with a high fractionated dose of CAR-T cells, whereas the toxicity was manageable with only one patient out of the 20 patients having CRS grade 4 and no grade 5 CRS. The fractionation of CAR-T over 3 days allowed for early detection of clinical symptoms of CRS and withholding of consequent doses. Accordingly, in the third cohort, 7 out of the 20 patients received all three doses, nine patients received only one dose, and four patients received two doses. Only two patients in this cohort did not achieve CRS, and both only received two doses. The authors concluded that CAR-T cell fractionation allowed for early dose modification and consequently optimising the safety of patients with R/R ALL who received CAR-T therapy without compromising the efficacy of treatment [26].

Gardner et al. tested the early intervention of tocilizumab plus or minus corticosteroids with the first occurrence of mild CRS symptoms in a study of 43 patients. The first 23 patients received tocilizumab with or without corticosteroids if there was an occurrence of any grade 4 toxicity not meeting the study’s defined dose-limiting toxicity (DLT) and lasting >48 h, or there was a DLT lasting >48 h, and in either situation, the CRS was not effectively managed by medication. Corticosteroids were also given to any non-haematologic toxicity >grade 3 that was considered a result of CAR-T infusion, lasting for >48 h, and not controlled despite active medical intervention. The subsequent 20 patients received concomitant tocilizumab and corticosteroids every 6–12 h for persistent mild CRS symptoms according to the requirement of a protocol modification. The investigators reported that despite the patients being treated with tocilizumab and steroids, there was no increase in the incidence of infection. Likewise, the efficacy was not affected by the early treatment with tocilizumab and steroids or the expansion and engraftment of T cells. They also reported a decrease in severe CRS in the 20 patients who had early intervention and concomitant tocilizumab and corticosteroids, versus the first 23 patients who had their CRS managed as above (15% vs. 20%) [27].

In conclusion, the therapeutic effect of CAR-T therapy is well established and has been proven in a wealth of studies; however, the fact that there is a high percentage of patients suffering from relapse and CRS presents serious complications for treatment with CAR-T therapy. It represents huge challenges for the treating physicians and the patients. Ongoing research activity is trying to address the mechanism of relapse and the best way to administer CAR-T therapy in order to reduce the incidence of CRS [28,29].

The unique mechanism of CAR-T therapy in addition to the high positive results mentioned above also triggered wide development programmes assessing this novel therapy in different haematologic malignancies as well as solid tumours.

The previously discussed studies in different haematologic malignancies represent a huge opportunity for patients who are suffering from those diseases. The fact that most of these studies are early phase studies with promising results and that there are planned Phase III studies represents great research opportunities for the scientific community and a chance for patients to be treated with a highly effective CAR-T therapy. These opportunities come with some challenges; the CRS is one of them. Some ongoing studies try to minimise this side effect while maintaining the superior outcome seen with CAR-T therapy [26]. Cost still represents a challenge to treatment with CAR-T therapy. Treatment with CAR-T cells is still not affordable for many patients in the approved indications. With the potential of more indications to come, the cost of treatment, either for the patients or any health care system, should be thoroughly evaluated, which also includes the cost of hospitalisation during treatment with CAR-T therapy and possible costs for the management of side effects [75].

Treatment with CAR-T therapy has started a new concept for the treatment of pa- tients with advanced ALL and DLBCL. The search for subsequent therapy for those who failed treatment with CAR-T therapy, together with the huge ongoing and planned stud- ies in different haematologic malignancies, opens a new era of research and treatment with gene/targeted therapy. Although CAR-T therapy is currently approved for ad- vanced/heavily pre-treated patient populations, the research activities will potentially be conducted in earlier disease settings. All the ongoing and planned activities lead to one conclusion: this is not the end of the chapter; it is the beginning of a new chapter or even chapters in our fight against different haematologic malignancies.