Mechanisms of Resistance To CA

Reviews
Mechanisms of resistance to CAR

T cell therapy
Nirali N. Shah1* and Terry J. Fry 2,3
Abstract | The successes with chimeric antigen receptor (CAR) T cell therapy in early clinical trials
involving patients with pre-B cell acute lymphoblastic leukaemia (ALL) or B cell lymphomas have
revolutionized anticancer therapy, providing a potentially curative option for patients who are
refractory to standard treatments. These trials resulted in rapid FDA approvals of anti-CD19 CAR
T cell products for both ALL and certain types of B cell lymphoma — the first approved gene
therapies in the USA. However, growing experience with these agents has revealed that
remissions will be brief in a substantial number of patients owing to poor CAR T cell persistence
and/or cancer cell resistance resulting from antigen loss or modulation. Furthermore, the initial
experience with CAR T cells has highlighted challenges associated with manufacturing a
patient-specific therapy. Understanding the limitations of CAR T cell therapy will be critical to
realizing the full potential of this novel treatment approach. Herein, we discuss the factors that
can preclude durable remissions following CAR T cell therapy, with a primary focus on the
resistance mechanisms that underlie disease relapse. We also provide an overview of potential
strategies to overcome these obstacles in an effort to more effectively incorporate this unique
therapeutic strategy into standard treatment paradigms.
In April 2012, following successes first experienced approximately 30–50% of patients who achieve remis-
with CD19-directed chimeric antigen receptor (CAR) sion with anti-CD19 CAR T cells will have disease
T cell therapy in adults with follicular lymphoma or relapse, the majority within 1 year of treatment9,10.
chronic lymphocytic leukaemia (CLL) 1–3, the first Relapses will not be unique to agents targeting CD19,
child with acute lymphoblastic leukaemia (ALL) was as the initial clinical experience with other CAR targets,
infused with anti-CD19 CAR T cells4,5. Despite devel- such as CD22, indicates that relapse will be a common
oping a life-threatening toxicity now referred to as and recurring challenge15. Furthermore, approximately
cytokine-release syndrome (CRS), this child’s remarka- 10–20% of patients fail to enter remission after receiving
ble recovery and ongoing remission (now persisting for anti-CD19 CAR T cell therapy7–10. Loss or modulation
>6 years), alongside concurrent developments in other of the target antigen15–17 and/or a lack of CAR T cell per-
paediatric and adult patients, provided the first realiza- sistence18, as well as product manufacturing failures19,20
tion of the potential of CAR T cell therapy to induce (Fig. 1), are among the more commonly cited impedi-
durable remissions. Subsequently, unprecedented suc- ments to effective CAR T cell therapy. Furthermore,
cesses in early phase trials of anti-CD19 CAR T cell similar successes with CAR T cells have not yet been
1
Pediatric Oncology Branch,
for the treatment of relapsed and/or refractory CD19- achieved in diseases beyond B cell leukaemia and lym-
Center for Cancer Research,
National Cancer Institute,
expressing B cell malignancies, with many patients phoma. As antigen-directed CAR T cells become more
National Institutes of Health, achieving long-term remission and potentially cure6–10, widely used, understanding the limitations of CAR T cell
Bethesda, MD, USA. have led to FDA approval of two distinct anti-CD19 therapy and overcoming these obstacles will be crucial
2
Department of Pediatrics, CAR T cell products11–13 for the treatment of both to harnessing the full potential of this highly effective
University of Colorado B cell ALL and diffuse large B cell lymphoma and have treatment modality.
Anschutz Medical Campus,
revolutionized the field of anticancer immunotherapy. In this Review, we summarize the rapidly evolv-
Aurora, CO, USA.
Furthermore, with >100 trials of CAR T cells ongoing ing knowledge regarding the barriers to durable CAR
3
Center for Cancer and Blood
Disorders, Children’s Hospital
globally14, comparable high rates of remission induction T cell-induced remissions. With a primary focus on
Colorado, Aurora, CO, USA. have now been demonstrated to be possible with CAR identifying mechanisms of emerging resistance to CAR
*e-mail: [email protected] T cells targeting antigens other than CD19 (ref.15). T cell therapy associated with antigen escape, we also dis-
https://doi.org/10.1038/ As more patients are treated and longer follow-up cuss antigen-positive relapses associated with poor CAR
s41571-019-0184-6 data are becoming available, we are realizing that T cell persistence and address limitations associated with
372 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

Reviews
Key points an anti-CD22 CAR T cell product was successfully man-

ufactured for 55 of 56 patients enrolled on a clinical trial
• Chimeric antigen receptor (CAR) T cell immunotherapy is a highly effective form of performed at the NCI; the median CD3+ cell count in
adoptive cell therapy, as demonstrated by the remission rates in patients with B cell these patients was 567 cells/μl (range 145–2,144 cells/μl),
acute lymphoblastic leukaemia or large B cell lymphoma, which have supported FDA and the median ALC was 775 cells/μl (range 230–4,620
approvals.
cells/μl) (N.N.S. and T.J.F., unpublished observations).
• A complete understanding of the limitations of CAR T cell therapy will help to identify The likelihood of obtaining a sufficient number of T cells
crucial areas requiring further research to improve patient outcomes.
is certainly an important consideration, and additional
• Factors that can preclude durable remissions following CAR T cell therapy include data are needed in this area.
CAR T cell manufacturing issues, limited CAR T cell expansion and/or persistence,
Characteristics of the apheresis product beyond cell
various resistance mechanisms and toxicities.
quantity can also affect the ability to successfully manu-
• Various intuitive strategies to overcome these obstacles are being investigated in
facture a CAR T cell product or the quality of the man-
order to optimize this unique therapeutic strategy and expand the indications for
treatment.
ufactured CAR T cell product. Previous treatment with
chemotherapy regimens, for example, those containing
clofarabine or doxorubicin, has been implicated in the
product manufacturing, nonresponsiveness or incom- generation of quantitatively insufficient or poor-quality
plete responses and toxicities. Furthermore, we will CAR T cell products7,21. Additionally, clinical data sug-
highlight areas of active research on potential strategies gest that prior treatment with cyclophosphamide and
to overcome the limitations of CAR T cell therapies and cytarabine selectively reduces early lineage T cells that
novel indications for these treatments. are associated with productive CAR T cell expansion22.
Much of the data on CAR T cell manufacturing have
Barrier 1: failure to achieve remission been generated in patients with ALL or CLL; however,
Potential benefit from CAR T cells first requires that a results published in the past 2 years indicate that man-
patient has access to the therapy and that a CAR T cell ufacturing CAR T cell products from patients with
product is successfully manufactured, infused and effec- high-risk solid tumours might be particularly challeng-
tively mediates a cytotoxic response, ideally resulting in ing19,21. Reasons for this finding remain under investiga-
complete remission. Remarkably, these prerequisites tion, but one potential explanation relates to the higher
have proved to be feasible in most patients with ALL, quantity of circulating myeloid-derived suppressor cells
although they still present potential barriers to effec- often found in patients with solid tumours than in those
tive CAR T cell therapy in some patients and, therefore, with haematological malignancies23,24. Implementation
provide opportunities for improvement. of CAR T cell manufacturing in patients with haema-
tological malignancies beyond B cell leukaemias might
Product manufacturing. Given the aggressive nature also reveal further barriers to successful cell harvesting
of most cancers for which treatment is indicated, the in certain patient populations and/or disease-specific
successful manufacture and infusion of CAR T cells in contexts. Early collection of T cells in patients identi-
a timely manner remain barriers to the implementation fied as having a high risk of relapse, or before treatment
of effective CAR T cell therapy. The first hurdle in this of those with relapsed disease, might improve the quality of
process is the collection of T cells to be used to generate a the apheresis product and thus the resulting CAR T cells.
CAR-expressing product. Unfortunately, limited data are Other details of the CAR T cell manufacturing process
available on the actual percentage of patients with ALL are also likely to affect qualitative aspects of the resulting
for whom a sufficient number of T cells cannot be har- product and thus subsequent clinical responses. Current
vested, as patients are often excluded from consideration manufacturing protocols incorporate a T cell expan-
for CAR T cell therapy, typically as a result of pre-existing sion step typically induced by activation through the
lymphopenia that can result from prior cytotoxic therapy. T cell receptor (TCR)25,26. Both anti-CD3 and anti-CD28
Even if T cells are collected successfully, patients might bead-based expansion and anti-CD3 antibody expansion
not receive therapy, for example, in the event of a product protocols are used in the generation of the FDA-approved
manufacturing failure. Notably, in one phase I study in anti-CD19 CAR T cell products27,28. The majority of
which the baseline absolute lymphocyte counts (ALCs) ongoing trial protocols also use one of these approaches
of paediatric candidates for treatment with anti-CD19 for CAR T cell expansion. To date, the superiority of one
CAR T cell therapy were analysed8, the median ALC was method of T cell expansion has not been established.
1,228 cells/μl (range 168–4,488 cells/μl). This finding Expansion protocols for CAR T cells also typically include
indicates that patient selection is highly skewed towards the use of cytokines, such as IL-2, IL-7 and/or IL-15;
those with ALCs within the normal physiological range, again, the superiority of one method has not been estab-
which might result in the exclusion of patients who are lished, but differential effects of these cytokines are likely
heavily pretreated or receiving ongoing therapy. While to qualitatively affect the resultant CAR T cell product29.
groups at many centres are currently incorporating a Almost all CAR T cell products are generated using
minimum ALC or CD3+ cell count to proceed with both CD4+ T cells and CD8+ T cells, and both of these
apheresis, improvements in CAR T cell manufacturing cell populations probably contribute to therapeutic effi-
will probably lower the limits for cell collection. Our cacy. The ratio of CD4+ to CD8+ T cells can vary sub-
experience at the National Cancer Institute (NCI) sug- stantially between patients and might also affect the CAR
gests that manufacturing of a CAR T cell product is fea- T cell product, although differences in efficacy between
sible with a CD3+ cell count of ≥150 cells/μl. Specifically, products with different ratios have not been established.
NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 373

Reviews
n
dulatio
igen mo
Ant CA
Rt
ox
ici
Relapse tie
s
Neurotoxicity
CD19
Antigen
Severe CRS
loss or down-
CD22 modulation
e
ur
ail Solid
Cannot
Rf
tumours
Un
harvest
CA
me
enough
T cells
t ne
eds
T cells
Time
CAR T cells Limited CAR
do not expand T cell persistence Lymphoma
(in vitro or in vivo) in vivo? subtypes
CNS disease
CAR T cell
Fig. 1 | Limitations to durable remissions after CAR T cell therapy. This figure summarizes several different limitations
to achieving a durable remission with chimeric antigen receptor (CAR) T cell therapy. First, CAR T cell failures have several
causes: for some patients, the CAR T cell product cannot be successfully manufactured or the generated CAR T cells do
not expand sufficiently (either during manufacturing in vitro or after administration in vivo); in other patients, the problem
of limited persistence in vivo is a potential mechanism underlying disease relapse. Second, antigen modulation — depicted
by the loss or downregulation of CD19 and/or CD22 on malignant B cells — enables antigen escape as a mechanism
of resistance to CAR T cell therapy, which can also be a problem in non-B cell malignancies, including solid tumours.
Third, the characteristic toxicities of CAR T cell therapy — primarily severe cytokine-release syndrome (CRS) and/or
neurotoxicity — can be fatal, thus abolishing the potential for therapeutic benefit in a small proportion of patients.
Furthermore, data on the effect of therapeutic interventions for CRS on the durability of CAR remission remain unknown.
Finally, unmet needs include some disease contexts that are a focus for ongoing research efforts to optimize the clinical
utility of CAR T cell therapies. For example, although anti-CD19 CAR T cells can provide substantial benefit for adults with
lymphoma, the complete remission rates are lower than those achieved in patients with leukaemia. Additionally, the
outcomes of CAR T cell therapy in paediatric patients with lymphoma and in patients with central nervous system (CNS)
involvement remain an area of ongoing investigation. Notably, such therapies currently have limited efficacy in patients
with solid tumours, and approaches to optimize response are being explored.
On the basis of preclinical data demonstrating improved a successful product 35. Preferential manufacturing
potency30, in some trials, separate CD4+ and CD8+ CAR approaches involving these unique T cell populations are
T cell products have been generated for delivery to patients under development. In a study of patients with CLL who
in a defined ratio8,31. Iterative modifications to the manu- received an anti-CD19 CAR T cell product36, responders
facturing process have the potential to improve the effi were found to have a CAR T cell population that was
cacy of CAR T cells, which will probably be of greater enriched for expression of memory-related genes, com-
importance as CAR T cells are tested beyond B cell malig- pared with that of nonresponders, and had higher num-
nancies. Novel approaches to CAR T cell manufacturing, bers of CD27+PD-1−CD8+ CAR T cells expressing high
such as use of closed-system automated devices, incor- levels of the IL-6 receptor, thus leading to better tumour
porating the aforementioned foundations are just begin- control; this cell phenotype was more predictive of a
ning to be implemented32,33; clinical trials using these response than other disease and patient characteristics36.
approaches and further analysis of how these techniques Other cellular components in the apheresis product
will impact future CAR T cell therapy are warranted. can affect CAR T cell manufacturing; this might be a par-
Even if a CAR T cell product can be manufactured ticular challenge given patient heterogeneity and the pres-
successfully, the starting T cell phenotype has been ence of circulating blasts and nonmalignant cells, such as
demonstrated to be an important determinant of sub- myeloid-derived suppressor cells, which can inhibit T cell
sequent clinical activity. Selection of T cells with spe- growth37. A report of the incidental CAR transduction of
cific phenotypes, for example, central memory or stem B cells, which conferred resistance to subsequent CAR
cell-like memory T cells, before manufacturing22,34,35 T cell therapy38, highlights the importance of having a
or manipulation of the manufacturing conditions to purified starting product. Strategies to optimize the input
skew CAR T cell production towards a particular T cell apheresis product before CAR T cell manufacturing, gen-
population might improve the likelihood of generating erally through T cell selection and enrichment, have been

Reviews
developed by several groups19,37,39. Indeed, at the NCI, we CAR T cell products. At present, standard parameters that
have observed that, by further enhancing the purity of the define T cell potency, such as markers of T cell exhaus-
starting T cell population and reducing the presence of tion, have been disappointing in terms of predicting clin-
additional inhibitory cells, the positive selection of CD4+ ical efficacy55,56. Given the high response rates with CAR
T cells and CD8+ T cells using a magnetic bead-based T cell therapy in patients with leukaemia and thus the
approach can result in successful manufacturing of CAR small numbers of nonresponders, systematically evalu-
T cell from an apheresis product following failed manu- ating and establishing parameters associated with a lack
facturing using a protocol with elutriation-based enrich- of responsiveness are difficult — particularly considering
ment for lymphocytes. Furthermore, we have found that that the underlying causes are likely to be multifactorial
this manufacturing change can also increase the potency and not fully attributable to product variables alone.
of the resulting CAR T cell product40. Guidelines for the In patients with CLL, however, among whom response
development of commercial CAR cell products have been rates have been substantially lower than those in patients
developed and are likely to evolve as our knowledge and with ALL or lymphoma, Fraietta et al.36 were able to iden-
experience grow41,42. tify favourable product characteristics, such as enrichment
CAR construct design is another parameter that for IL-6–STAT3 signatures and an elevated frequency of
probably affects the characteristics of a CAR T cell prod- CD29+CD45RO−CD8+ T cells before CAR T cell gener-
uct and subsequent in vivo behaviour of the modified ation. Furthermore, pre-infusion anti-CD19 CAR T cell
cells, including their kinetics of expansion and duration products comprising polyfunctional T cells subsets,
of persistence. The majority of CAR T cell products being defined on the basis of cytokine and chemokine expression
tested in clinical trials are second-generation agents, profiles, have been associated with an improved response
meaning that they contain both a TCR stimulatory in patients with lymphoma compared with products with-
domain (typically derived from the T cell surface glyco- out such polyfunctional activity57. Additional data will be
protein CD3 ζ-chain (CD3ζ)) and a single co-stimulatory required to establish desirable attributes of CAR T cells,
domain. The current FDA-approved products con- but the optimal characteristics might differ depending on
tain either a CD28 or a 4-1BB (also known as CD137) the CAR construct and the malignancy being targeted.
co-stimulatory domain. The effect of the co-stimulatory Generating a CAR T cell product using cells from
domain on response rates has not been systematically healthy donors is an alternative strategy to circum-
evaluated, although preclinical data43 and observations venting issues of poor CAR T cell quality. Several
in patients44,45 indicate that this aspect of CAR design groups have now tested donor-derived CAR-based
markedly affects the persistence of the cell product. treatment strategies both preclinically58 and in patients
Other details of CAR design, such as specific character- with post-transplant disease relapse using T cells har-
istics of the antigen-binding domain, the presence and vested directly from their original allogeneic stem cell
structure of an extracellular hinge region and features of donors59–61. The results of these early studies demonstrate
the transmembrane domain, might also affect CAR T cell the feasibility of this approach, with only a low frequency
attributes, but definitive data defining the effects of these of high-grade graft-versus-host disease (GVHD)59–61.
design details have not yet been generated. Manufacturing CAR T cells from third-party donors
The current FDA-approved CAR T cell products and might enable the development of universal, off-the-
those used in the majority of clinical trials exploit retro shelf products and is another method to overcome the
viral vectors for insertion of the CAR gene construct problem of quantitatively insufficient or poor-quality
into the genome of T cells46,47. Both gamma-retroviral CAR T cell products62,63. Generating such products will
vectors and HIV-derived lentiviral vectors have been probably require additional genetic modification to
successfully used for this purpose. Detailed descrip- limit CAR T cell rejection and/or GVHD but also offers
tions of these vectors can be found elsewhere25,48,49 and opportunities to overcome resistance mechanisms, such
are beyond the scope of this Review. Certain aspects of as ‘fratricide’ reported with T cell antigen-targeted CAR
vector design are likely to affect the efficiency of gene T cell products64. Other novel strategies to enhance
transfer and, potentially, the activity of the resulting CAR-based therapeutic strategies and provide an alter-
CAR T cell product, although these aspects have not native off-the-shelf product include CAR-engineered
been systematically evaluated. RNA transfection has also nature killer (NK) cells; indeed, preclinical and early
been used to deliver the CAR gene construct, resulting in phase clinical studies of such products are underway65–67.
transient CAR expression (because the CAR gene is not Additional potential benefits of incorporating NK cells
inserted into the genome of the T cells50). Such transient into a CAR-based cell therapy include a more favoura-
CAR-expressing T cell products have only short-lived ble adverse-effect profile than CAR T cell therapy alone,
benefits in preclinical models, with no durable responses potentially owing to differential cytokine responses
observed51. Finally, non-viral vector-based introduction between CAR NK cells and conventional CAR T cells;
of the CAR gene into the T cell genome has been suc- however, further studies are needed to identify other
cessfully achieved using a transposon–transposase sys- features that determine the toxicity profiles of these two
tem52,53, and clinical trials using such products have been types of CAR-modified cell products68,69.
reported, with the data demonstrating the safety and
feasibility of this approach and providing preliminary Infusion of CAR T cells. Clearly, failure to infuse a CAR
evidence of CAR T cell activity54. T cell product in a timely manner before patients develop
One important issue in the field of CAR T cell research progressive disease or disease-related complications can
is establishing reliable criteria to define the potency of also preclude successful treatment. In the registration

Reviews
trial of the FDA-approved anti-CD19 CAR T cell product of patients to access CAR T cell therapy. With the FDA
tisagenlecleucel in children and young adults with ALL10, approval of two different anti-CD19 CAR constructs11,12,
17 (18%) of 92 patients who had an apheresis product the most dramatic improvement in access reflects a very
collected did not receive a CAR T cell infusion for rea- steady increase in the number of centres across the USA
sons including tisagenlecleucel-related product issues that are qualified to administer these products: at the
in 7 patients (primarily from poor cell growth), death time of the submission of this manuscript, 73 individ-
before CAR T cell infusion in 7 patients (attributable to ual centres were listed as Kymriah Treatment Centers73
disease progression in 4 patients and to infection-related authorized to administer tisagenlecleucel, with a similar
complications in the others) and development of interval number of authorized axicabtagene ciloleucel treatment
adverse events (fungal disease or GVHD) that rendered centres74. Thus, the ability of eligible patients to receive
the patient ineligible to receive the CAR T cell infusion. an FDA-approved CAR T cell product at their local can-
In this study10, the median time to cell infusion was cer centre is increasing, thereby facilitating incorpora-
45 days (range 30–105 days). Similarly, in a study per- tion of these therapies into their individual treatment
formed at the Memorial Sloan Kettering Cancer Center9, plan. Additionally, the international approvals of the
11 (14%) of 78 adult patients with ALL who underwent same CAR T cell products in Europe75 and Canada76 fur-
apheresis did not proceed with cell manufacturing (the ther contribute to increasing access for patients in need
majority sought alternative treatments), and 13 (19%) of and to reducing the regional disparities — although such
the 67 patients for whom CAR T cell manufacturing was limitations continue to exist77, at least partially owing to
pursued did not undergo infusion owing in part to prod- the fact that candidate treatment centres must have the
uct failures, disease progression or complications ren- capacity and capability to adhere to product manufactur-
dering them ineligible to receive therapy. With 20–30% ing and administration protocols and safely manage the
of candidates ultimately not being infused with CAR patient. Establishing standardized guidelines for CAR
T cells, future strategies to shorten manufacturing times T cell therapy, determining which centres are most able
would be predicted to improve the likelihood that eligible to deliver this therapy and exploring the role for accred-
patients will receive an infusion and thereby increase the itation of sites are all subjects of ongoing research efforts
number of patients who benefit from CAR T cell ther- aimed at improving access to this treatment modality.
apy. Commercialization of CAR T cell therapy and more Additionally, costs and insurance coverage are persis-
readily accessible manufacturing programmes would tent barriers to expanding patient access to CAR T cell
hopefully decrease the delays in time to infusion. therapy78–80. Cost analyses and optimization of manufac-
turing strategies to reduce costs are needed as this ther-
CAR T cell activation and expansion. Clinical studies apy moves forward. In this regard, developing position
have revealed that the dose of CAR T cells required for statements supporting the utility of CAR T cell therapy as
effective therapy is remarkably small, with the current part of current treatment paradigms will probably facili-
dosing regimen of 0.2–5.0 × 106 transduced CAR T cells tate a clearer path towards addressing issues of payment
per kg or a total of 0.1–2.5 × 10 8 transduced CAR and insurance coverage. Many patients will continue to
T cells per infusion, although activation and expo- seek enrolment on clinical trials in order to circumvent
nential expansion of the cells following infusion are limitations in access to FDA-approved constructs, but ini-
essential6–10,70,71. As discussed, the quality and inher- tial access to those FDA-approved agents would be more
ent T cell phenotype of the CAR T cell product can desirable than inclusion in a trial of an experimental agent
affect post-infusion CAR T cell behaviour. In addition, — the benefits which are unlikely to have been established.
recipient-related factors are also important in CAR T cell Additionally, eligibility criteria for receipt of the
expansion. For example, disease (and thus antigen) bur- FDA-approved agents might be considered a limitation
den can positively influence the degree of cell expan- to patient access. For example, active central nervous
sion, which in turn might increase the risk and severity system (CNS) involvement remains an exclusionary
of CRS7,8. Lymphodepletion also seems to be important criterion among patients with B cell lymphoma; how-
for CAR T cell expansion, with evidence indicating ever, clinical trials to study the safety and efficacy of
that certain chemotherapeutic agents, such as fludara- CAR T cell therapy in this population and others are
bine, might be more effective in this regard31. However, essential, with the ultimate aim of extending the treat-
fludarabine has been implicated as a potential contribu- ment indications to all populations that could potentially
tor to CAR T cell-associated neurotoxicity, although the derive benefit.
timing of classical fludarabine-associated neurotoxicity
differs from that of CAR T cell-related toxicities, and the Barrier 2: disease relapse
safety of lymphodepletion with this agent is generally Disease relapse following anti-CD19 or anti-CD22 CAR
supported72. Whether CAR T expansion can be achieved T cell therapy can occur in up to 50% of patients with
without the need for cytotoxic chemotherapy remains to pre-B cell ALL by 12 months after infusion8–10,15 in two
be determined, and this requirement will probably vary major patterns: early relapse of antigen-positive leukae-
between disease subtypes. mia or later relapse typically associated with antigen loss
(Fig. 1). An increased understanding of the mechanisms
Access to CAR T cell therapy. Despite the FDA approv- underlying poor persistence of and/or resistance to
als of CD19-directed CAR T cell products over the past CAR T cells and identifying patients with the highest
2 years11,12, access to these novel therapies remains limited. likelihood of relapse will be crucial to optimizing CAR
Nevertheless, ongoing efforts are improving the ability T cell therapy.

Reviews
Antigen-positive relapse. Early ALL relapse, typically recurrent stimulation can reactivate and numerically
within the first few months after successful induction of expand the CAR T cells and prevent antigen-positive
remission, is often associated with limited CAR T cell relapse (NCT03186118). More broadly, the use of arti-
persistence and/or transient B cell aplasia, which sug- ficial antigen-presenting cells provides a potentially
gests a loss of active CAR T cell-mediated surveillance off-the-shelf approach for optimizing adoptive T cell
of leukaemia6. Determinants of CAR T cell persistence immunotherapy by increasing the therapeutic efficacy
remain to be fully determined but, in addition to inher- and persistence of the infused T cells95–97. Shifting the
ent T cell quality (which might be patient-dependent CAR T cells towards a central memory or stem cell-like
and context-dependent81) and initial T cell phenotype memory phenotype is another unique method of
(including the proportion of CD4+ versus CD8+ T cells82), enhancing therapeutic responses and cell persistence35,36.
include the co-stimulatory domain built into each unique Combining CAR T cell therapy with immune-
CAR construct83, with preclinical reports indicating checkpoint inhibitors or other immunomodulatory ther-
that CD28 co-stimulatory domain-containing CARs apies provides a synergetic approach to optimizing the
tend to persist less well than those containing a 4-1BB rate, depth and durability of clinical responses98. Evidence
co-stimulatory domain45,84; clinical experience is con- for increased PD-1 expression in CAR T cells during the
sistent with these data6,7,85. In the aforementioned piv- time from infusion to peak expansion has been demon-
otal clinical trial of tisagenlecleucel10, 4-1BB-based CAR strated in clinical samples85, and preclinical data support
T cells persisted in the blood for a median duration of a role of PD-1–PD-L1 blockade in improving the effec-
168 days (range 20–617 days), generally with concurrent tiveness of CAR T cell therapy99,100. Clinical testing of this
B cell aplasia in patients who remained in remission. By strategy with FDA-approved immune-checkpoint inhib-
contrast, the median duration of CD28-based anti-CD19 itors and CAR T cell products has provided anecdotal
CAR T cell persistence has been ~30 days, and these cells evidence of improved persistence101; future trials to test
are rarely detected beyond 3 months7,86. The better per- such approaches are under development.
sistence of the 4-1BB domain-containing CAR T cells The persistence of anti-CD19 CAR T cell has an
might result, in part, from the reduced propensity for important role in ongoing surveillance and seems to
T cell exhaustion induced by tonic CAR signalling when be important for durable remission in patients with
co-stimulation is mediated by a 4-1BB versus a CD28 ALL; however, whether cell persistence is absolutely
domain87. Efforts to calibrate the CAR activation poten- necessary to maintain durable remissions achieved
tial in order to optimize durability of response and bal- with all CAR T cell products remains unclear. To date,
ance effector versus memory T cell expansion88, among the longest follow-up data on CD28-based anti-CD19
other approaches89,90, are underway. Other co-stimulatory CAR T cell therapies in patients with ALL have been
domains and multiple co-stimulatory domains have also reported by Park and colleagues9. In this study, CAR
been used in CAR constructs and will probably affect T cell persistence beyond induction of remission
CAR T cell persistence. Additionally, further analysis of was rarely detected, yet the median event-free survival was
observations in patients with a notable expansion and/or 6.1 months overall and was 10.6 months in patients with
persistence of the CAR T cells suggest a role for targeted a low disease burden (<5% bone marrow blasts)9. This
genomic integration of the CAR construct to enhance duration of CAR T cell persistence is generally shorter
persistence. For example, the anecdotal experience of than that of the 4-1BB-based constructs, although these
clonal expansion in a patient with CLL revealed that findings nevertheless demonstrate a capacity for dura-
TET2 disruption results in alterations in CAR T cell biol- ble remission beyond the CAR T cell persistence, raising
ogy, leading to enhanced potency and a central memory additional questions about the determinants of remis-
phenotype91. Similarly, specific integration of the CAR sion duration. Notably, the durability of remission in
gene into the TCRα constant (TRAC) locus of the T cell patients with B cell lymphoma has been relatively similar
genome using CRISPR–Cas9 editing machineries results with both the 4-1BB-based and the CD28-based prod-
in better antitumour responses than those observed with ucts71,102, thus suggesting that the relevance of CAR T cell
conventionally transduced CAR T cells in preclinical persistence might vary by cancer type. Careful monitor-
models92. Advances in gene-editing technologies93 com- ing of outcomes following anti-CD19 CAR T cell ther-
bined with enhanced understanding of determinants of apy for B cell malignancies and following treatment with
CAR T cell potency and knowledge of determinants CAR T cells targeting other antigen and/or tumour types
of therapeutic responsiveness will be crucial to the design of will be important to fully understand the impact of cell
the next generation of CAR T cell technologies36,94. Thus, persistence on the durability of remissions. Regardless,
future directions to improve CAR persistence will rely the ability for CAR T cells to persist in the long term and
heavily on understanding the optimal T cell biology in thereby mediate ongoing surveillance against disease
relation to CAR T cell functionality and subsequent opti- relapse is clearly a potential advantage of these agents
mization of CAR T cell design to promote persistence over other targeted immunotherapies, such as bispecific
(when persistence is desired). or conjugated antibodies, which generally have a limited
Strategies to improve persistence independently capacity to induce a long-term durable remission103–106.
of CAR T cell design and manufacturing are being Relapse with antigen-positive disease presents a
tested in the clinic; for example, administration of potential opportunity for re-treatment with CAR T cells,
T cell-antigen-presenting cells (T-APCs) designed to acti- although re-infusion strategies to treat antigen-positive
vate anti-CD19 CAR T cells at regular intervals follow- relapse occurring with loss of CAR T cell persistence
ing remission induction in order to determine whether have unfortunately had limited success. Gardner et al.8

Reviews
Table 1 | Summary of antigen loss or modulation detected in published clinical trials of CAR T cell therapy
Target Publication Number of Number Number of Median time Comments
antigen patients of CRs patients to antigen loss
treated (%) with antigen or modulation
modulation (%) (months)
CD19 Lee et al.7 21 14 (67) 2 (14) ~6 10 of 12 patients who were
MRD-negative after CAR
T cell therapy subsequently
underwent HSCT
Maude et al.6 30 27 (90) 4 (15) ~3 None
Gardner et al.8 43 40 (93) 7 (18) ~3 11 of 40 subsequently
underwent HSCT
Park et al.9 53 44 (83) 4 (9) Unknown None
Maude et al. 10
75 61 (81) 15 (25) Unknown None
CD22 Fry et al.15 21 12 (57) 7 (58) ~3 None
CAR , chimeric antigen receptor ; CRs, complete responses; HSCT, haematopoietic stem cell transplantation; MRD, minimal
residual disease.
administered a second infusion of anti-CD19 CAR T cells Antigen loss or modulation as a mechanism of immune
to ten children and young adults with ALL. Eight of these escape. Target antigen modulation is one of the clear-
patients had loss of CAR T cell persistence, only two of est mechanisms of disease relapse following successful
whom had CAR T cell expansion after re-infusion and remission induction using CAR T cells (Table 1) and has
only one of whom had a complete response8. The two also been described with other targeted immunothera
remaining patients, who had detectable CAR T cells, were peutic approaches, such as bispecific T cell-engager
re-infused for persistent or relapsed CD19+ disease, but (BiTE) antibody constructs110–112 or monoclonal anti-
neither had a substantial CAR T cell re-expansion, B cell bodies113. Established mechanisms leading to loss of
aplasia or anti-leukaemic effect8. Lee et al.7 described three CD19 expression include alternative splicing, which
patients who received a second infusion of CAR T cells for generates CD19 isoforms with disruption of the target
residual or recurrent CD19+ ALL at 2–5.5 months after epitope and/or reduced cell surface expression114,115, and
the initial infusion, and none had an objective response. interruption in the transport of CD19 to the cell sur-
Maude et al.6 used a strategy with repeated re-infusions face116; other pathways leading to antigen loss are under
to counter early loss of CAR T cells and associated B cell active study. Complete antigen loss, however, might not
recovery, which led to subsequent persistence of CAR be necessary for the development of resistance to ini-
T cells101. Turtle et al.31 reported on five patients with tially effective CAR T cell therapy — even diminution of
ALL who received an anti-CD19 CAR T cell re-infusion, antigen expression can be sufficient. For example, with
none of whom had re-expansion, persistence or anti-CD22 CAR T cell therapy, a simple quantitative
anti-leukaemia activity. Additionally, CAR-specific T cell decrease in CD22 cell surface expression or antigen den-
responses were detected in all five patients, thus suggest- sity in the leukaemic population was adequate to evade
ing an immune-mediated rejection of CAR T cells upon the CAR T cells, thus enabling leukaemic relapse despite
repeat dosing31. The same group reported similarly poor ongoing CD22 positivity, with individual variation seen
outcomes following re-infusion of patients with B cell in the threshold antigen density that conferred relapse or
non-Hodgkin lymphoma, but use of an intensified lym- resistance15. Notwithstanding, a minimum threshold of
phodepletion regimen containing fludarabine in addition antigen expression is likely to be needed for functional
to cyclophosphamide improved the re-infusion response, and/or preserved CAR T cell activity — a concept that
as well as improving initial CAR T cell expansion and per- has been realized in preclinical models with targeting of
sistence107. This concept of intensified lymphodepletion both CD20 in B cell malignancies117 and ALK in neuro-
has similarly been used in re-infusion strategies for other blastoma118, in which CAR T cell cytolytic activity and
CAR-based therapies, with improved clinical outcomes108. cytokine production were increased with higher levels
Alternative strategies to optimize the re-infusion approach of antigen expression. Along these lines, preclinical
include the use of a different CAR construct or even a experience with an EGFR-targeted CAR T cell therapy
CAR targeting a different antigen. The use of humanized suggests therapeutic potential that relies specifically on
CD19 CARs to overcome immune-mediated rejection the differential antigen density of EGFR in tumours ver-
of murine-derived anti-CD19 CAR constructs is one sus nonmalignant tissues to limit on-target, off-tumour
such strategy being tested in the clinic (NCT02374333), toxicities while maintaining anticancer activity119.
with early data suggesting that this approach can induce Antigen loss following effective CAR therapy has
complete remission109. Alternative targeting strategies been best described in patients with ALL114,115 but is cer-
(for example, with anti-CD22 CAR T cells) might also be tainly not limited to this disease120. Findings of preclini-
effective in patients with disease relapse after prior CAR cal121 studies in solid tumour models and clinical studies
T cell therapy15. Thus, multiple-infusion strategies to treat in patients with glioblastoma122,123 have similarly impli-
or prevent disease relapse remain an active and important cated antigen modulation as a potential pitfall undermin-
area of investigation. ing the efficacy of CAR T cell therapy. In the design of

Reviews
future studies and novel CAR constructs, recognition Despite the appreciation of antigen modulation as a
of the role of antigen density in antitumour responses mechanism of immune escape, we currently have a lim-
and identification of mechanisms leading to disrupted ited ability to predict which patients have a high risk of
target expression will be needed in order to optimize developing antigen-modulated relapsed disease, beyond
CAR T cell responses. those with identification of pre-existent antigen-negative
The additive effect of prior targeted immunothera- subclones, those with established heterogeneity in anti-
pies might further increase the complexity of immune gen expression or those who have received prior immu-
evasion after CAR T cell therapy. For example, both notherapies targeting the same antigen. Minimal residual
the anti-CD19 BiTE blinatumomab and the anti-CD22 disease (MRD) monitoring strategies predicated on flow
antibody–drug conjugate inotuzumab ozogamicin are cytometry will be essential to identifying the existence
FDA-approved therapies indicated for treatment of of pretreatment subpopulations that might not be fully
ALL, and loss of response with emergence of CD19– amenable to therapies targeting a single antigen, with a
(refs111,112) or CD22– (ref.113) escape variants has been greater onus on actively screening for antigen-negative
reported in patients treated with these immunothera- disease; thus, more complex flow cytometric gating
pies. Thus, such agents might render future therapy with methods and other improvements in the immuno
CAR T cells targeting the same antigens less effective or phenotypic characterization of single cells will be required,
decrease the durability of responses by increasing the particularly when the disease burden is low. Tracking of
risk of antigen-negative relapse. Notably, receipt of prior leukaemic clones over time with adjunctive PCR and/or
therapy with blinatumomab was an exclusion criterion molecular assessments of VDJ immunoglobulin heavy
of the pivotal registration trial of tisagenlecleucel in chain (IgH) rearrangements might improve the predic-
patients with ALL10. However, whether such avoidance tion of disease relapse by identifying cell populations that
of targeted immunotherapy before the administration of are not fully eradicated by CAR T cell therapy or are not
CAR T cells is necessary to achieve durable long-term as easily identified using flow cytometry. Accordingly,
responses remains to be determined. in their phase I study of anti-CD19 CAR T cell therapy,
Independent of treatment-related antigen loss or mod- Gardner et al.8 found that 27 (67.5%) of 40 patients who
ulation, inherent tumour heterogeneity also has a role achieved an MRD-negative remission according to flow
in predisposition to emergence of an antigen-negative cytometry assessment had a malignant clone identified
clone. CD19 has been considered to be ubiquitously concurrently using next-generation sequencing. The
expressed on all pre-B cell ALL clones, with development majority of these patients (17 of 27; 65%) subsequently
of antigen-negative subclones upon CD19-targeted treat- achieved a molecular complete remission by day 63,
ment; however, more-detailed analysis of pre-therapy although several did not, suggesting the ongoing pres-
CD19 expression is necessary, as rare patients have malig- ence of leukaemic disease that might ultimately be the
nant cells with CD19 negativity or partial expression at harbinger of future relapse.
diagnosis124. Indeed, we now have a greater apprecia- Given the propensity of antigen modulation as a
tion that pre-existing CD19– subclones can be present mechanism for evasion of effective immunotherapy,
at diagnosis115, with data from some studies indicating CAR constructs incorporating multi-antigen target-
the possibility that the malignant B cell progenitors are ing are being developed to address inherent tumour
CD19–, particularly in patients with BCR–ABL1 ALL125. heterogeneity and thus decrease the risk of leukaemic
CD22, although also expressed in a high percentage of relapse. Preclinical data supporting the multi-targeted
pre-B cell ALL cells, has a well described heterogeneity approaches include the use of tandem anti-CD19–
in surface expression, particularly in infants with KMT2A CD20 CAR constructs131, combinatorial anti-CD19
(MLL)-rearranged ALL, in whom CD22– ALL cell sub- and anti-CD123 (also known as IL-3Rα) strategies132 and
populations are more frequently detected124,126–128, lead- CD19 and CD22 targeting using a single CAR construct
ing to the emergence of CD22– or CD22dim populations targeting both antigens or two unique CARs targeting
following CD22-targeted therapy15,127. each antigen individually133, and several clinical trials
Lineage switching is another mechanism for evading of these strategies are underway (Table 2). Follow-up
CAR T cells. This phenomenon was appreciated before data on these protocols will help to determine whether
the era of targeted therapies in the context of infant, dual-antigen-targeted approaches are adequate to pre-
KMT2A-rearranged leukaemia subtypes, which often vent disease relapse or whether additional combinatorial
present as mixed-lineage leukaemias, with patients hav- strategies for targeting more than two antigens will be
ing transformation to AML following ALL-specific ther- necessary to make CAR T cell therapy curative. In the
apy or vice versa129. In both the preclinical and the clinical development of combinatorial multi-antigen-targeted
setting, an analogous phenomenon with emergence of strategies, ensuring an effective response to each anti-
myeloid subtypes following CD19-directed immuno- gens is essential: preferential targeting of one antigen
therapy has now been described in ALL with or without over another could lead to a predilection for a functional
KMT2A rearrangement16,17,110,125. Similarly, targeting of response to only a single antigen without obviating the
FLT3 with CAR T cells in preclinical ALL models was problem of antigen-negative relapse. Developing func-
found to induce a reversible B cell to T cell lineage switch tional multi-targeted constructs, however, is no easy task
(while effectively avoiding transformation to FLT3+ and is highly dependent on preclinical testing to identify
myeloid lineage leukaemias)130. Whether this resistance biologically active constructs with equivalent capacity to
mechanism is active in non-leukaemic malignancies target coincident antigens simultaneously, as has been
remains to be determined. nicely described in the work of Qin and colleagues133.

Reviews
Table 2 | Active clinical trials of multi-antigen CAR T cells in the USA and UK
Target Disease Age CAR construct signalling Treatment centre ClinicalTrials.gov
antigens group domains reference number
(years)
CD19 and ALL and NHL 1–26 CD3ζ−4-1BB (combinatorial Seattle Children’s NCT03330691 (PLAT-05)
CD22 approach with anti-CD19 Hospital (Seattle, WA,
CAR T cells, anti-CD22 CAR USA)
T cells and co-transduced
anti-CD19 and anti-CD22
CAR T cells)
ALL 1–30 CD3ζ−4-1BB Lucile Packard NCT03241940
Children’s Hospital,
Stanford University
(Palo Alto, CA, USA)
ALL and ≥18 CD3ζ−4-1BB Stanford University NCT03233854
DLBCL (Palo Alto, CA, USA)
ALL and NHL 3–30 CD3ζ−4-1BB National Cancer NCT03448393
Institute (Rockville,
MD, USA)
ALL 1–24 CD3ζ–OX40 (CD19) and Great Ormond Street NCT03289455
CD3ζ−4-1BB (CD22) Hospital (London, UK)
CD19 and NHL and CLL 18–70 CD3ζ−4-1BB Medical College NCT03019055
CD20 of Wisconsin
(Milwaukee, WI, USA)
ALL , acute lymphoblastic leukaemia; CAR , chimeric antigen receptor ; CD3ζ, T cell surface glycoprotein CD3 ζ-chain; CLL , chronic
lymphoblastic leukaemia; DLBCL , diffuse large B cell lymphoma; NHL , non-Hodgkin lymphoma.
The role of consolidative therapy after CAR T cell- only 2 (14.2%) had disease relapse135. Park et al.9 reported
induced remission. Given the concern over antigen- that 17 (38.6%) of 44 patients with ALL proceeded to
positive or antigen-negative disease relapse, the need to allo-HSCT after attaining complete remission with
consolidate CAR T cell-induced remissions needs to be CD28-based anti-CD19 CAR T cell therapy; after a
considered. The first successful clinical applications median follow-up duration of 29 months from CAR
of CAR T cell therapies have been in patients with T cell administration, 12 (70.5%) of these 17 patients
haematological malignancies, for whom allogeneic had died or had disease relapse after transplantation. Of
haematopoietic stem cell transplantation (allo-HSCT) the 26 patients who did not proceed to transplantation,
is a validated curative option; thus, consolidation treat- 17 (65%) had relapsed or died9. Among 32 patients who
ment with allo-HSCT is a very relevant topic of discus- obtained MRD-negative complete remission, those who
sion, particularly when no prior transplantation has proceeded to allo-HSCT had no difference in event-free
been performed. In initial reports from Davila et al.86 survival or overall survival compared with those who did
and Lee et al.7,134, a high proportion of adults and chil- not (P = 0.64 and P = 0.89, respectively)9. Thus, the role
dren or young adults with ALL, respectively, who entered of allo-HSCT following CAR T cell therapy needs to be
remission after treatment with CD28-based anti-CD19 better defined but is likely to be of greater importance
CAR T cells proceeded to allo-HSCT, with overall with the use of shorter-acting CAR T cell products134, for
improved outcomes. Specifically, in an analysis combin- example, CD28-based CAR T cells in patient with pre-
ing data from trials of anti-CD19 and anti-CD22 CAR B cell ALL. Regardless, a reduced reliance on consolida-
T cell products performed at the NCI, 25 patients sub- tive therapy to achieve cure will be an important goal of
sequently underwent allo-HSCT, and this was the first efforts to optimize CAR T cells.
allo-HSCT for 19 patients134. Using a competing risks
analysis (risk of relapse versus transplantation-related Barrier 3: CAR T cell-related toxicity
mortality), the 24-month cumulative incidence of No article on CAR T cell therapy would be complete
post-allo-HSCT relapse in all 25 patients and in the without a discussion of the related toxicities, particularly
19 patients who underwent HSCT for the first time CRS; however, several in-depth reviews of the current
was 13.5% (95% CI 3.2–32.1%) and 11.3% (95% CI state-of-the-art knowledge and management of CAR
1.7–31.1%), respectively134. Summers et al.135 reported T cell-related toxicities have been published in this jour-
a trend towards improved leukaemia-free survival in nal and elsewhere over the past few years136–142. Notably,
patients without a history of allo-HSCT who under- efforts are underway to establish uniform multicentre
went consolidative allo-HSCT following anti-CD19, grading scales and, in this regard, the original CRS grad-
4-1BB-based CAR T cell therapy compared with those ing scale proposed by Lee et al.141 has now been updated
who did not (P = 0.057)135. Specifically, among 64 evalu- and published as the American Society for Blood and
able patients, 17 had no prior history of allo-HSCT, 3 of Marrow Transplant consensus guidelines143. Algorithms
whom chose not to proceed to allo-HSCT, and 2 (66.7%) and treatment approaches to optimize the safety of CAR
of these 3 patients subsequently had disease relapse; T cell therapies, including early intervention strategies144,
among the 14 patients who proceeded to allo-HSCT, and the incorporation of suicide genes and other genetic

Reviews
engineering strategies to reduce CAR T cell toxicity are CAR T cell response, persistence and relapse in lym-
under investigation145. phoma. With the FDA approval of axicabtagene cilo-
Of particular relevance to the discussion herein, leucel and subsequently tisagenlecleucel, CAR T cell
severe or even fatal CRS and other toxicities are barri- therapies provide a highly effective approach to the
ers to durable CAR T cell-induced remissions in some treatment of relapsed and/or refractory large B cell lym-
patients and, in this regard, in addition to established phoma; however, unique challenges exist in this setting,
literature and guidelines in development, we offer two including lower remission rates (complete remission
additional areas for consideration. rates of 30% and 54% in the pivotal trials of axicabtagene
First, one must be cognizant of the fact that the liter- ciloleucel71 and tisagenlecleucel102, respectively, versus
ature on CRS management is predominantly based on 67–93% with anti-CD19 CAR T cells in patients with
the experience gained in trials of anti-CD19 CAR T cell ALL; Table 1) and a limited understanding of the mech-
therapy trials. As novel antigens are targeted, therefore, anisms of relapse. Antigen loss can be a mechanism of
it will be important to recognize that not all cases of CRS lymphoma relapse after CAR T cell therapy120,146 but
will be the same and, as such, the appropriate interven- seems to occur less often than in patients with ALL
tion strategies might differ. For example, CRS observed (Table 1). Notably, reports on the frequency of antigen
after anti-CD22 CAR T cell therapy seemed to be less loss in patients with B cell lymphoma are limited, prob-
severe than that seen with anti-CD19 CAR T cells15, ably owing to less frequent sampling of lymphomatous
without the implementation of any early intervention disease than of leukaemic disease, but it remains an
strategies; however, novel toxicities, such as clinically rel- important consideration for patients with lymphoma
evant coagulopathy, occurred with CD22 targeting that relapse after CAR T cell therapy. In addition, CAR
required a different approach to treatment (N.N.F. and T cell persistence might not be as necessary for a dura-
T.J.S., unpublished observations). When targeting solid ble response in patients with lymphoma as it seems to
tumours with CAR T cell-based strategies, the poten- be in those with ALL. With the seemingly shorter-acting
tial off-tumour, on-target toxicities might be less toler- CD28 domain-containing anti-CD19 CAR T cells, the
able than the prolonged B cell aplasia associated with longest follow-up data in patients with B cell lymphoma
CAR T cells targeting B cell antigens, thus warranting have been reported by Kochenderfer and colleagues147.
close consideration. In patients with diffuse large B cell lymphoma, specif-
Second, as the treatment of CRS and/or other tox- ically, these authors found that most patients had last
icities becomes better established, evaluating the measurable CAR T cell detection at <6 months after
effects of such treatment, and particularly the effects infusion, with 4 of 5 complete responses continuing
of pre-emptive or prophylactic strategies to mitigate (at 56, 51, 44 and 38 months) despite recovery of the non
CRS, on anticancer activity will be equally important. malignant B cell population147. Moreover, despite the
Currently, the limited available data indicate that early FDA approvals for adults with large B cell lymphomas,
intervention strategies can effectively reduce the sever- the experience with CAR T cell therapies in paediatric
ity of CRS without compromising peak CAR T cell lymphomas is scant, and responses have been limited,
expansion or functional persistence144. However, these for reasons that are not fully understood.
data come from a single trial using a single CAR con- Ongoing optimization of responses to CAR T cells
struct and, therefore, might not be broadly applicable. in patients with lymphoma, using various aforementioned
Furthermore, the effect of prolonged steroid use in those strategies that are being incorporated into the treat-
who develop severe CRS on CAR persistence needs to ment of ALL, is underway. Testing of these approaches
be further studied across various constructs. In addition, uniformly in paediatric patients with lymphoma,
determining the ideal time to implement CRS interven- an area of unmet need (Fig. 1), remains a priority for
tion strategies while balancing the potential for an opti- future research.
mal anticancer response will be particularly challenging,
compounded by the complexity of novel CAR con- CAR T cell strategies beyond B cell targeting. CD19
structs. Future studies are needed to establish the opti- is generally ubiquitously expressed on the majority of
mal strategies for intervention to negate toxicities and B cells, and targeting of both nonmalignant and malig-
their impact on CAR persistence. A trial investigating nant CD19+ cells is a therapeutic strategy with acceptable
the optimal timing of anti-IL-6 therapy with tocilizumab safety, whereas on-target, off-tumour toxicity might be
for the treatment of CRS associated with anti-CD19 more prohibitive in other cancers. With regard to AML,
CAR T cell therapy is underway (NCT02906371). for example, broad targeting of myeloid lineage cells, par-
ticularly if sustained, is not acceptable nor can it be made
Barrier 4: moving beyond leukaemia more feasible using supportive interventions as simple as
Given the successes to date in heavily pretreated patients, immunoglobulin replacement in patients with prolonged
earlier use of CAR T cells has considerable potential to B cell aplasia. In such settings, use of a less persistent
change the treatment paradigm for children and adults CAR T cell product, together with the administration
with relapsed and/or refractory or high-risk B cell malig- of rescue stem cells from an allogeneic donor following
nancies. Indeed, efforts are underway to understand how an effective CAR T cell response, might be a reasonable
to best incorporate these therapies into upfront regi- strategy. A first-in-human study of myeloid cell target-
mens in order to improve overall outcomes. However, ing using anti-CD123, CD28 domain-containing CAR
substantial barriers exist when the use of CAR T cells is T cells revealed promising anti-leukaemia activity with-
extended to cancers beyond ALL. out prolonged myelosuppression148. Clinical testing of

Reviews
myeloid cell targeting with a CD28-based anti-CD33 is another promising approach162. Anti-HER2, 4-1BB
CAR T cell product is planned149. domain-containing CAR T cell-based targeting of
Going beyond haematological tumours, solid medulloblastoma has also demonstrated efficacy in both
tumours present a new set of unique challenges to the mouse and non-human primate models, with good toler-
development effective CAR T cell therapies, as reviewed ance of intraventricular administration163. Most notably,
elsewhere150–153. One key issue with solid tumours is HER2-specific CAR T cells have been tested in a phase I
that the inherent tumour heterogeneity is likely to be dose-escalation study involving patients with progres-
a substantial barrier to identifying an optimal target, sive glioblastoma, and preliminary results demonstrated
and relatedly, antigen loss will probably be a key factor the safety and feasibility of this approach and provided
precluding curative remissions. Indeed, antigen density an early sign of activity by virtue of partial remission in
(as mentioned above regarding ALK in neuroblastoma one patient and disease stabilization in several others164.
cells118) and inherent tumour heterogeneity in antigen Similarly, EGFRvIII-directed CAR T cells were found to
expression in solid tumours (for example, mesothelin154, be safe and feasible in patients with glioblastoma but,
HER2 (ref.155) or MUC1 (ref.156) expression in non-small- as seen in patients with ALL, antigen loss and tumour
cell lung cancers) can limit the therapeutic potential of heterogeneity were major determinants of immune
agents that target a single antigen and raise concerns escape123. Thus, optimizing CAR T cell approaches
over differing on-target, off-tumour toxicities. Among and testing novel delivery mechanisms for CAR T cells
the CAR T cell-based strategies that are furthest in targeting CNS disease remains an active area of research.
development, products targeting the disialoganglioside Ultimately, many challenges must be overcome in
GD2 have demonstrated antitumour activity in patients the optimization of CAR T cell therapies for diseases
with neuroblastoma, leading to complete remissions157, beyond ALL, but much progress has been made in the
thus emphasizing the potential for CAR T cell therapy past several years. Lessons learned from B cell-targeted
of solid tumours. strategies will provide a foundation for moving this field
In addition, overcoming the immunosuppres- of research forward, but a cautious approach is needed to
sive tumour microenvironment, which might render understand and attempt to mitigate novel CRS-related
adoptively transferred T cells inactive, will be particu- toxicities and off-tumour, on-target toxicities that might
larly relevant to effective CAR T cell therapy for solid otherwise result in the premature abandonment of
tumours. Several approaches to overcome this issue promising new CAR T cell-based strategies for treatment
are under evaluation, including optimization of the of solid tumours.
preparative regimen before CAR T cell infusion150,158.
Armoured CAR T cells that constitutively secrete Conclusions
pro-inflammatory cytokines as a mechanism to over- CAR T cell-based therapy is among the most promising
come local immunosuppression are a novel concept. anticancer therapies of all time. Many challenges remain,
Similarly, incorporation of immune-checkpoint inhi- however, as we strive to make remissions induced by this
bition into CAR T cell-based strategies, by various therapy durable for all treated patients. CAR T cell manu-
combinatorial or built-in approaches, might improve facturing considerations — from patient selection to the
responses98; the findings of several preclinical studies characteristics of the infused product and the potential
support the effectiveness of this approach, warranting for subsequent cell expansion — reflect one particular
further clinical translation98,152,159,160. aspect of efforts to improve outcomes, and advances in
Targeting the CNS tumours demands careful con- cell manufacturing will certainly make these products
sideration of optimal delivery mechanisms for CAR accessible for a greater number of patients. For those with
T cell therapy. A case report from 2016 described the response, confirming CAR T cell persistence and mon-
use of CAR T cells targeting IL-13 receptor subunit itoring for antigen loss are necessary for relapse predic-
α2 (IL-13Rα2) in a patient with glioblastoma, which tion and prevention strategies; however, antigen-negative
incorporated multiple infusions over a 7-month period disease relapse is currently difficult to treat, and even
via two intracranial delivery routes (into the tumour antigen-positive relapsed disease might not respond to
resection cavity followed by infusions into the ventricu- CAR T cell re-infusion. Akin to multi-agent chemother-
lar system)122. The patient had an impressive clinical apy, multi-antigen-targeting strategies might address
response with complete regression of all intracranial and these mechanisms of relapse and, therefore, might be
spinal tumours that lasted for 7.5 months122. Strategies a pathway to more durable remissions. Optimizing this
to optimize IL-13Rα2-targeted strategies include therapeutic approach in patients with non-B cell malig-
designing a second-generation 4-1BB co-stimulatory nancies is the next frontier for research in this field of
domain-containing CAR product using enriched cen- immunotherapy, but many obstacles need to be over-
tral memory T cells; in orthotopic glioblastoma models, come, and close attention to the design of novel CAR
this product was effective irrespective of corticosteroid constructs will be informative in identifying further bar-
treatment (a mainstay of glioblastoma therapy) and riers, such as those in solid tumours. In this fast-paced
had higher efficacy with intraventricular infusions field of research, a shift in focus to overcome the obsta-
than intravenous delivery (and perhaps also than intra- cles identified (Fig. 1) will be imperative to make this
tumour administration in the multifocal disease set- promising therapy more broadly available, more effective
ting)161. Anti-GD2 CAR T cell therapy for histone H3 and potentially curative for all those treated.
lysine 27-methylated (H3K27me) diffuse midline glio-
mas, a uniformly fatal type of paediatric CNS tumour, Published online 5 March 2019

Reviews
1. Kalos, M. et al. T cells with chimeric antigen receptors 25. Levine, B. L., Miskin, J., Wonnacott, K. & Keir, C. 49. Golumba-Nagy, V., Kuehle, J. & Abken, H. Genetic
have potent antitumor effects and can establish Global manufacturing of CAR T cell therapy. modification of T cells with chimeric antigen receptors:
memory in patients with advanced leukemia. Mol. Ther. Methods Clin. Dev. 4, 92–101 (2017). a laboratory manual. Hum. Gene Ther. Methods 28,
Sci. Transl Med. 3, 95ra73 (2011). 26. Wang, X. & Riviere, I. Clinical manufacturing of CAR 302–309 (2017).
2. Porter, D. L., Levine, B. L., Kalos, M., Bagg, A. & T cells: foundation of a promising therapy. Mol. Ther. 50. Riet, T. et al. Nonviral RNA transfection to transiently
June, C. H. Chimeric antigen receptor-modified T cells Oncolyt. 3, 16015 (2016). modify T cells with chimeric antigen receptors for
in chronic lymphoid leukemia. N. Engl. J. Med. 365, 27. Tumaini, B. et al. Simplified process for the production adoptive therapy. Methods Mol. Biol. 969, 187–201
725–733 (2011). of anti-CD19-CAR-engineered T cells. Cytotherapy 15, (2013).
3. Kochenderfer, J. N. et al. Eradication of B-lineage cells 1406–1415 (2013). 51. Panjwani, M. K. et al. Feasibility and safety of
and regression of lymphoma in a patient treated with 28. Kochenderfer, J. N. et al. Construction and preclinical RNA-transfected CD20-specific chimeric antigen
autologous T cells genetically engineered to recognize evaluation of an anti-CD19 chimeric antigen receptor. receptor T cells in dogs with spontaneous B cell
CD19. Blood 116, 4099–4102 (2010). J. Immunother. 32, 689–702 (2009). lymphoma. Mol. Ther. 24, 1602–1614 (2016).
4. Grupp, S. A. et al. Chimeric antigen receptor-modified 29. Gargett, T. & Brown, M. P. Different cytokine and 52. Monjezi, R. et al. Enhanced CAR T cell engineering
T cells for acute lymphoid leukemia. N. Engl. J. Med. stimulation conditions influence the expansion and using non-viral Sleeping Beauty transposition from
368, 1509–1518 (2013). immune phenotype of third-generation chimeric minicircle vectors. Leukemia 31, 186–194 (2017).
5. Rosenbaum, L. Tragedy, perseverance, and chance — antigen receptor T cells specific for tumor antigen 53. Singh, H., Huls, H., Kebriaei, P. & Cooper, L. J.
the story of CAR-T therapy. N. Engl. J. Med. 377, GD2. Cytotherapy 17, 487–495 (2015). A new approach to gene therapy using Sleeping
1313–1315 (2017). 30. Sommermeyer, D. et al. Chimeric antigen Beauty to genetically modify clinical-grade T cells to
6. Maude, S. L. et al. Chimeric antigen receptor T cells receptor-modified T cells derived from defined target CD19. Immunol. Rev. 257, 181–190 (2014).
for sustained remissions in leukemia. N. Engl. J. Med. CD8 + and CD4 + subsets confer superior antitumor 54. Kebriaei, P. et al. Phase I trials using Sleeping Beauty
371, 1507–1517 (2014). reactivity in vivo. Leukemia 30, 492–500 (2016). to generate CD19-specific CAR T cells. J. Clin. Invest.
7. Lee, D. W. et al. T cells expressing CD19 chimeric 31. Turtle, C. J. et al. CD19 CAR-T cells of defined 126, 3363–3376 (2016).
antigen receptors for acute lymphoblastic CD4 + :CD8 + composition in adult B cell ALL patients. 55. de Wolf, C., van de Bovenkamp, M. & Hoefnagel, M.
leukaemia in children and young adults: a phase 1 J. Clin. Invest. 126, 2123–2138 (2016). Regulatory perspective on in vitro potency assays for
dose-escalation trial. Lancet 385, 517–528 32. Zhang, W., Jordan, K. R., Schulte, B. & Purev, E. human T cells used in anti-tumor immunotherapy.
(2015). Characterization of clinical grade CD19 chimeric Cytotherapy 20, 601–622 (2018).
8. Gardner, R. A. et al. Intent-to-treat leukemia remission antigen receptor T cells produced using automated 56. Xu, J., Melenhorst, J. J. & Fraietta, J. A. Toward
by CD19 CAR T cells of defined formulation and dose CliniMACS Prodigy system. Drug Des. Devel Ther. 12, precision manufacturing of immunogene T cell
in children and young adults. Blood 129, 3322–3331 3343–3356 (2018). therapies. Cytotherapy 20, 623–638 (2018).
(2017). 33. Zhu, F. et al. Closed-system manufacturing of CD19 57. Rossi, J. et al. Preinfusion polyfunctional anti-CD19
9. Park, J. H. et al. Long-term follow-up of CD19 CAR and dual-targeted CD20/19 chimeric antigen receptor chimeric antigen receptor T cells are associated with
therapy in acute lymphoblastic leukemia. N. Engl. T cells using the CliniMACS Prodigy device at an clinical outcomes in NHL. Blood 132, 804–814
J. Med. 378, 449–459 (2018). academic medical center. Cytotherapy 20, 394–406 (2018).
10. Maude, S. L. et al. Tisagenlecleucel in children and (2018). 58. Ghosh, A. et al. Donor CD19 CAR T cells exert potent
young adults with B-cell lymphoblastic leukemia. 34. Sabatino, M. et al. Generation of clinical-grade CD19- graft-versus-lymphoma activity with diminished
N. Engl. J. Med. 378, 439–448 (2018). specific CAR-modified CD8 + memory stem cells for the graft-versus-host activity. Nat. Med. 23, 242–249
11. US Food & Drug Administration. FDA approves CAR-T treatment of human B cell malignancies. Blood 128, (2017).
cell therapy to treat adults with certain types of large 519–528 (2016). 59. Brudno, J. N. et al. Allogeneic T cells that express an
B cell lymphoma. FDA https://www.fda.gov/ 35. Blaeschke, F. et al. Induction of a central memory and anti-CD19 chimeric antigen receptor induce remissions
newsevents/newsroom/pressannouncements/ stem cell memory phenotype in functionally active of B-cell malignancies that progress after allogeneic
ucm581216.htm (2017). CD4( + ) and CD8( + ) CAR T cells produced in an hematopoietic stem-cell transplantation without
12. US Food & Drug Administration. FDA approval automated good manufacturing practice system for causing graft-versus-host disease. J. Clin. Oncol. 34,
brings first gene therapy to the United States. the treatment of CD19( + ) acute lymphoblastic 1112–1121 (2016).
FDA https://www.fda.gov/newsevents/newsroom/ leukemia. Cancer Immunol. Immunother. 67, 60. Chen, Y. et al. Donor-derived CD19-targeted T cell
pressannouncements/ucm574058.htm (2017). 1053–1066 (2018). infusion induces minimal residual disease-negative
13. US Food & Drug Administration. FDA approves 36. Fraietta, J. A. et al. Determinants of response and remission in relapsed B cell acute lymphoblastic
tisagenlecleucel for adults with relapsed or refractory resistance to CD19 chimeric antigen receptor (CAR) leukaemia with no response to donor lymphocyte
large B cell lymphoma. FDA https://www.fda.gov/Drugs/ T cell therapy of chronic lymphocytic leukemia. Nat. infusions after haploidentical haematopoietic stem
InformationOnDrugs/ApprovedDrugs/ucm606540.htm Med. 24, 563–571 (2018). cell transplantation. Br. J. Haematol. 179, 598–605
(2018). 37. Stroncek, D. F. et al. Myeloid cells in peripheral blood (2017).
14. Tang, J., Hubbard-Lucey, V. M., Pearce, L., O’Donnell- mononuclear cell concentrates inhibit the expansion 61. Kochenderfer, J. N. et al. Donor-derived CD19-
Tormey, J. & Shalabi, A. The global landscape of of chimeric antigen receptor T cells. Cytotherapy 18, targeted T cells cause regression of malignancy
cancer cell therapy. Nat. Rev. Drug Discov. 17, 893–901 (2016). persisting after allogeneic hematopoietic stem cell
465–466 (2018). 38. Ruella, M. et al. Induction of resistance to chimeric transplantation. Blood 122, 4129–4139 (2013).
15. Fry, T. J. et al. CD22-targeted CAR T cells induce antigen receptor T cell therapy by transduction of a 62. Georgiadis, C. et al. Long terminal repeat
remission in B-ALL that is naive or resistant to CD19- single leukemic B cell. Nat. Med. 24, 1499–1503 CRISPR-CAR-coupled “universal” T cells mediate
targeted CAR immunotherapy. Nat. Med. 24, 20–28 (2018). potent anti-leukemic effects. Mol. Ther. 26,
(2018). 39. Fesnak, A., Lin, C., Siegel, D. L. & Maus, M. V. 1215–1227 (2018).
16. Jacoby, E. et al. CD19 CAR immune pressure induces CAR-T cell therapies from the transfusion medicine 63. Poirot, L. et al. Multiplex genome-edited T cell
B-precursor acute lymphoblastic leukaemia lineage perspective. Transfus. Med. Rev. 30, 139–145 (2016). manufacturing platform for “off-the-shelf” adoptive
switch exposing inherent leukaemic plasticity. 40. Shah, N. N. et al. CD4/CD8 T-cell selection enhances T cell immunotherapies. Cancer Res. 75, 3853–3864
Nat. Commun. 7, 12320 (2016). CD22 CAR-T cell transduction and in-vivo CAR-T (2015).
17. Gardner, R. et al. Acquisition of a CD19-negative expansion: updated results on phase I anti-CD22 64. Cooper, M. L. et al. An “off-the-shelf” fratricide-resistant
myeloid phenotype allows immune escape of CAR dose expansion cohort. Blood 130, 809 (2017). CAR-T for the treatment of T cell hematologic
MLL-rearranged B-ALL from CD19 CAR-T cell therapy. 41. Vormittag, P., Gunn, R., Ghorashian, S. & Veraitch, F. S. malignancies. Leukemia 32, 1970–1983 (2018).
Blood 127, 2406–2410 (2016). A guide to manufacturing CAR T cell therapies. Curr. 65. Daher, M. & Rezvani, K. Next generation natural killer
18. Mueller, K. T. et al. Cellular kinetics of CTL019 in Opin. Biotechnol. 53, 164–181 (2018). cells for cancer immunotherapy: the promise of
relapsed/refractory B cell acute lymphoblastic 42. Perica, K., Curran, K. J., Brentjens, R. J. & Giralt, S. A. genetic engineering. Curr. Opin. Immunol. 51,
leukemia and chronic lymphocytic leukemia. Blood Building a CAR garage: preparing for the delivery of 146–153 (2018).
130, 2317–2325 (2017). commercial CAR T cell products at Memorial Sloan 66. Tang, X. et al. First-in-man clinical trial of CAR NK-92
19. Stroncek, D. F. et al. Elutriated lymphocytes for Kettering Cancer Center. Biol. Blood Marrow cells: safety test of CD33-CAR NK-92 cells in patients
manufacturing chimeric antigen receptor T cells. Transplant. 24, 1135–1141 (2018). with relapsed and refractory acute myeloid leukemia.
J. Transl Med. 15, 59 (2017). 43. Kawalekar, O. U. et al. Distinct signaling of coreceptors Am. J. Cancer Res. 8, 1083–1089 (2018).
20. Ceppi, F. et al. Lymphocyte apheresis for chimeric regulates specific metabolism pathways and impacts 67. US National Library of Medicine. ClinicalTrials.gov
antigen receptor T cell manufacturing in children memory development in CAR T cells. Immunity 44, https://clinicaltrials.gov/ct2/show/NCT03579927
and young adults with leukemia and neuroblastoma. 380–390 (2016). (2019).
Transfusion 58, 1414–1420 (2018). 44. June, C. H. & Sadelain, M. Chimeric antigen receptor 68. Quintarelli, C. et al. CD19 redirected CAR NK cells are
21. Das, R. K., Storm, J. & Barrett, D. M. T cell therapy. N. Engl. J. Med. 379, 64–73 (2018). equally effective but less toxic than CAR T cells. Blood
dysfunction in pediatric cancer patients at diagnosis 45. van der Stegen, S. J., Hamieh, M. & Sadelain, M. 132 (Suppl. 1), 3491 (2018).
and after chemotherapy can limit chimeric antigen The pharmacology of second-generation chimeric 69. Hofer, E. & Koehl, U. Natural killer cell-based cancer
receptor potential. Cancer Res. 78 (Suppl), 1631 antigen receptors. Nat. Rev. Drug Discov. 14, immunotherapies: from immune evasion to promising
(2018). 499–509 (2015). targeted cellular therapies. Front. Immunol. 8, 745
22. Singh, N., Perazzelli, J., Grupp, S. A. & Barrett, D. M. 46. Cornetta, K. et al. Absence of replication-competent (2017).
Early memory phenotypes drive T cell proliferation in lentivirus in the clinic: analysis of infused T cell 70. US Food & Drug Administration. Package insert —
patients with pediatric malignancies. Sci. Transl Med. products. Mol. Ther. 26, 280–288 (2018). Kymriah. FDA https://www.fda.gov/downloads/
8, 320ra3 (2016). 47. Cornetta, K. et al. Screening clinical cell products for UCM573941.pdf (2018).
23. Zhang, H. et al. Fibrocytes represent a novel MDSC replication competent retrovirus: the National Gene 71. Neelapu, S. S. et al. Axicabtagene ciloleucel CAR T-cell
subset circulating in patients with metastatic cancer. Vector Biorepository experience. Mol. Ther. Methods therapy in refractory large B-cell lymphoma. N. Engl.
Blood 122, 1105–1113 (2013). Clin. Dev. 10, 371–378 (2018). J. Med. 377, 2531–2544 (2017).
24. De Veirman, K. et al. Myeloid-derived suppressor cells 48. Qin, D. Y. et al. Paralleled comparison of vectors for 72. Lowe, K. L. et al. Fludarabine and neurotoxicity in
as therapeutic target in hematological malignancies. the generation of CAR-T cells. Anticancer Drugs 27, engineered T cell therapy. Gene Ther. 25, 176–191
Front. Oncol. 4, 349 (2014). 711–722 (2016). (2018).

Reviews
73. Novartis. Kymriah treatment center locator. Kymriah 97. Butler, M. O. & Hirano, N. Human cell-based artificial 119. Caruso, H. G. et al. Tuning sensitivity of CAR to EGFR
https://www.us.kymriah.com/treatment-center-locator/ antigen-presenting cells for cancer immunotherapy. density limits recognition of normal tissue while
(2018). Immunol. Rev. 257, 191–209 (2014). maintaining potent antitumor activity. Cancer Res. 75,
74. Kite Pharma. Where can Yescarta be received? 98. Yoon, D. H., Osborn, M. J., Tolar, J. & Kim, C. J. 3505–3518 (2015).
Yescarta https://www.yescarta.com/treatment-centers Incorporation of immune checkpoint blockade into 120. Shalabi, H. et al. Sequential loss of tumor surface
(2018). chimeric antigen receptor T cells (CAR-Ts): antigens following chimeric antigen receptor T cell
75. European Medicines Agency. First two CAR-T cell combination or built-in CAR-T. Int. J. Mol. Sci. 19, therapies in diffuse large B cell lymphoma.
medicines recommended for approval in the European E340 (2018). Haematologica 103, e215–e218 (2018).
Union. EMA https://www.ema.europa.eu/en/news/ 99. Cherkassky, L. et al. Human CAR T cells with 121. Krenciute, G. et al. Transgenic expression of IL15
first-two-car-t-cell-medicines-recommended- cell-intrinsic PD-1 checkpoint blockade resist improves antiglioma activity of IL13Ralpha2-CAR
approval-european-union (2018). tumor-mediated inhibition. J. Clin. Invest. 126, T cells but results in antigen loss variants. Cancer
76. Novartis Pharmaceuticals Canada Inc. Novartis 3130–3144 (2016). Immunol. Res. 5, 571–581 (2017).
receives Health Canada approval of its CAR-T cell 100. Gargett, T. et al. GD2-specific CAR T cells undergo 122. Brown, C. E. et al. Regression of glioblastoma after
therapy, Kymriah™ (tisagenlecleucel)i. Newswire potent activation and deletion following antigen chimeric antigen receptor T-cell therapy. N. Engl.
https://www.newswire.ca/news-releases/ encounter but can be protected from activation- J. Med. 375, 2561–2569 (2016).
novartis-receives-health-canada-approval-of-it induced cell death by PD-1 blockade. Mol. Ther. 24, 123. O’Rourke, D. M. et al. A single dose of peripherally
s-car-t-cell-therapy-kymriah- 1135–1149 (2016). infused EGFRvIII-directed CAR T cells mediates
tisagenlecleuceli-692581041.html (2018). 101. Li, A. M. et al. Checkpoint inhibitors augment CD19- antigen loss and induces adaptive resistance in
77. June, C. H., O’Connor, R. S., Kawalekar, O. U., directed chimeric antigen receptor (CAR) T cell therapy patients with recurrent glioblastoma. Sci. Transl Med.
Ghassemi, S. & Milone, M. C. CAR T cell in relapsed B-cell acute lymphoblastic leukemia. Blood 9, eaaa0984 (2017).
immunotherapy for human cancer. Science 359, 132 (Suppl. 1), 556 (2018). 124. Piccaluga, P. P. et al. Surface antigens analysis reveals
1361–1365 (2018). 102. Schuster, S. J. et al. Primary analysis of Juliet: significant expression of candidate targets for
78. Shah, G. L., Majhail, N., Khera, N. & Giralt, S. a global, pivotal, phase 2 trial of CTL019 in adult immunotherapy in adult acute lymphoid leukemia.
Value-based care in hematopoietic cell transplantation patients with relapsed or refractory diffuse large Leuk. Lymphoma 52, 325–327 (2011).
and cellular therapy: challenges and opportunities. B-cell lymphoma. Blood 130, 577 (2017). 125. Nagel, I. et al. Hematopoietic stem cell involvement in
Curr. Hematol. Malig. Rep. 13, 125–134 (2018). 103. Kantarjian, H. M. et al. Inotuzumab ozogamicin versus BCR-ABL1-positive ALL as a potential mechanism of
79. Caffrey, M. With approval of CAR T-cell therapy comes standard therapy for acute lymphoblastic leukemia. resistance to blinatumomab therapy. Blood 130,
the next challenge: payer coverage. Am. J. Manag. N. Engl. J. Med. 375, 740–753 (2016). 2027–2031 (2017).
Care https://www.ajmc.com/journals/evidence-based- 104. Gokbuget, N. et al. Blinatumomab for minimal 126. Raponi, S. et al. Flow cytometric study of potential
oncology/2018/february-2018/with-approval-of-ca residual disease in adults with B cell precursor acute target antigens (CD19, CD20, CD22, CD33) for
r-tcell-therapy-comes-the-next-challenge-payer-c lymphoblastic leukemia. Blood 131, 1522–1531 antibody-based immunotherapy in acute
overage (2018). (2018). lymphoblastic leukemia: analysis of 552 cases.
80. Bach, P. B. National coverage analysis of CAR-T 105. Kantarjian, H. et al. Blinatumomab versus Leuk. Lymphoma 52, 1098–1107 (2011).
therapies — policy, evidence, and payment. N. Engl. chemotherapy for advanced acute lymphoblastic 127. Shah, N. N. et al. Characterization of CD22 expression
J. Med. 379, 1396–1398 (2018). leukemia. N. Engl. J. Med. 376, 836–847 (2017). in acute lymphoblastic leukemia. Pediatr. Blood
81. Kotani, H. et al. Aged CAR T cells exhibit enhanced 106. Martinelli, G. et al. Complete hematologic and Cancer 62, 964–969 (2015).
cytotoxicity and effector function but shorter molecular response in adult patients with relapsed/ 128. Chevallier, P. et al. Simultaneous study of five
persistence and less memory-like phenotypes. Blood refractory philadelphia chromosome-positive candidate target antigens (CD20, CD22, CD33, CD52,
132, 2047 (2018). B-precursor acute lymphoblastic leukemia following HER2) for antibody-based immunotherapy in B-ALL:
82. Gardner, R. et al. Starting T cell and cell product treatment with blinatumomab: results from a phase II, a monocentric study of 44 cases. Leukemia 23,
phenotype are associated with durable remission of single-arm, multicenter study. J. Clin. Oncol. 35, 806–807 (2009).
leukemia following CD19 CAR-T cell immunotherapy. 1795–1802 (2017). 129. Mitterbauer-Hohendanner, G. & Mannhalter, C.
Blood 132, 4022 (2018). 107. Turtle, C. J. et al. Immunotherapy of non-Hodgkin’s The biological and clinical significance of MLL
83. Fesnak, A. D., June, C. H. & Levine, B. L. Engineered lymphoma with a defined ratio of CD8 + and abnormalities in haematological malignancies. Eur. J.
T cells: the promise and challenges of cancer CD4 + CD19-specific chimeric antigen receptor- Clin. Invest. 34 (Suppl. 2), 12–24 (2004).
immunotherapy. Nat. Rev. Cancer 16, 566–581 modified T cells. Sci. Transl Med. 8, 355ra116 130. Chien, C. D. et al. FLT3 chimeric antigen receptor T cell
(2016). (2016). therapy induces B to T cell lineage switch in infant
84. Zhao, Z. et al. Structural design of engineered 108. Shalabi, H. et al. Intensification of lymphodepletion acute lymphoblastic leukemia. Cancer Res. 78 (Suppl),
costimulation determines tumor rejection kinetics and optimizes CAR re-treatment efficacy. Blood 130 1630 (2018).
persistence of CAR T cells. Cancer Cell 28, 415–428 (Suppl. 1), 3889 (2017). 131. Schneider, D. et al. A tandem CD19/CD20 CAR
(2015). 109. Maude, S. L. et al. Efficacy of humanized CD19- lentiviral vector drives on-target and off-target antigen
85. Kochenderfer, J. N. et al. Chemotherapy-refractory targeted chimeric antigen receptor (CAR)-modified modulation in leukemia cell lines. J. Immunother.
diffuse large B cell lymphoma and indolent B cell T cells in children and young adults with relapsed/ Cancer 5, 42 (2017).
malignancies can be effectively treated with refractory acute lymphoblastic leukemia. Blood 128, 132. Ruella, M. et al. Dual CD19 and CD123 targeting
autologous T cells expressing an anti-CD19 chimeric 217 (2016). prevents antigen-loss relapses after CD19-directed
antigen receptor. J. Clin. Oncol. 33, 540–549 (2015). 110. Zoghbi, A., Zur Stadt, U., Winkler, B., Muller, I. & immunotherapies. J. Clin. Invest. 126, 3814–3826
86. Davila, M. L. et al. Efficacy and toxicity management Escherich, G. Lineage switch under blinatumomab (2016).
of 19-28z CAR T cell therapy in B cell acute treatment of relapsed common acute lymphoblastic 133. Qin, H. et al. Preclinical development of bivalent
lymphoblastic leukemia. Sci. Transl Med. 6, 224ra25 leukemia without MLL rearrangement. Pediatr. Blood chimeric antigen receptors targeting both CD19
(2014). Cancer 64, e26594 (2017). and CD22. Mol. Ther. Oncolyt. 11, 127–137 (2018).
87. Long, A. H. et al. 4-1BB costimulation ameliorates 111. Mejstrikova, E. et al. CD19-negative relapse of 134. Shalabi, H. et al. Chimeric antigen receptor T-cell
T cell exhaustion induced by tonic signaling of pediatric B cell precursor acute lymphoblastic (CAR-T) therapy can render patients with ALL into
chimeric antigen receptors. Nat. Med. 21, 581–590 leukemia following blinatumomab treatment. Blood PCR-negative remission and can be an effective bridge
(2015). Cancer J. 7, 659 (2017). to transplant (HCT). Biol. Blood Marrow Transplant.
88. Feucht, J. et al. Calibration of CAR activation potential 112. Jabbour, E. et al. Outcome of patients with relapsed/ 24, S25–S26 (2018).
directs alternative T cell fates and therapeutic potency. refractory acute lymphoblastic leukemia after 135. Summers, C. et al. Long term follow-up after
Nat. Med. 25, 82–88 (2018). blinatumomab failure: no change in the level of CD19 SCRI-CAR19v1 reveals late recurrences as well as a
89. Sadelain, M., Brentjens, R. & Riviere, I. The basic expression. Am. J. Hematol. 93, 371–374 (2018). survival advantage to consolidation with HCT after
principles of chimeric antigen receptor design. Cancer 113. Bhojwani, D. et al. Inotuzumab ozogamicin in pediatric CAR T cell induced remission. Blood 132, 967 (2018).
Discov. 3, 388–398 (2013). patients with relapsed/refractory acute lymphoblastic 136. Hay, K. A. et al. Kinetics and biomarkers of severe
90. Maus, M. V. & June, C. H. Making better chimeric leukemia. Leukemia https://doi.org/10.1038/s41375- cytokine release syndrome after CD19 chimeric
antigen receptors for adoptive T cell therapy. Clin. 018-0265-z (2018). antigen receptor-modified T cell therapy. Blood 130,
Cancer Res. 22, 1875–1884 (2016). 114. Sotillo, E. et al. Convergence of acquired mutations 2295–2306 (2017).
91. Fraietta, J. A. et al. Disruption of TET2 promotes the and alternative splicing of CD19 enables resistance to 137. Neelapu, S. S. et al. Chimeric antigen receptor T cell
therapeutic efficacy of CD19-targeted T cells. Nature CART-19 immunotherapy. Cancer Discov. 5, therapy — assessment and management of toxicities.
558, 307–312 (2018). 1282–1295 (2015). Nat. Rev. Clin. Oncol. 15, 47–62 (2018).
92. Eyquem, J. et al. Targeting a CAR to the TRAC locus 115. Fischer, J. et al. CD19 isoforms enabling resistance to 138. Teachey, D. T. et al. Identification of predictive
with CRISPR/Cas9 enhances tumour rejection. Nature CART-19 immunotherapy are expressed in B-ALL biomarkers for cytokine release syndrome after
543, 113–117 (2017). patients at initial diagnosis. J. Immunother. 40, chimeric antigen receptor T cell therapy for acute
93. Jung, I. Y. & Lee, J. Unleashing the therapeutic 187–195 (2017). lymphoblastic leukemia. Cancer Discov. 6, 664–679
potential of CAR-T cell therapy using gene-editing 116. Braig, F. et al. Resistance to anti-CD19/CD3 BiTE in (2016).
technologies. Mol. Cells 41, 717–723 (2018). acute lymphoblastic leukemia may be mediated by 139. Porter, D., Frey, N., Wood, P. A., Weng, Y. & Grupp, S. A.
94. Maus, M. V. Immunology: T cell tweaks to target disrupted CD19 membrane trafficking. Blood 129, Grading of cytokine release syndrome associated with
tumours. Nature 543, 48–49 (2017). 100–104 (2017). the CAR T cell therapy tisagenlecleucel. J. Hematol.
95. Hasan, A. N., Selvakumar, A. & O’Reilly, R. J. Artificial 117. Watanabe, K. et al. Target antigen density governs the Oncol. 11, 35 (2018).
antigen presenting cells: an off the shelf approach for efficacy of anti-CD20-CD28-CD3 zeta chimeric antigen 140. Gust, J. et al. Endothelial activation and blood-brain
generation of desirable T-cell populations for broad receptor-modified effector CD8 + T cells. J. Immunol. barrier disruption in neurotoxicity after adoptive
application of adoptive immunotherapy. Adv. Genet. 194, 911–920 (2015). immunotherapy with CD19 CAR-T cells. Cancer
Eng. 4, 130 (2015). 118. Walker, A. J. et al. Tumor antigen and receptor Discov. 7, 1404–1419 (2017).
96. Turtle, C. J. & Riddell, S. R. Artificial antigen- densities regulate efficacy of a chimeric antigen 141. Lee, D. W. et al. Current concepts in the diagnosis
presenting cells for use in adoptive immunotherapy. receptor targeting anaplastic lymphoma kinase. and management of cytokine release syndrome.
Cancer J. 16, 374–381 (2010). Mol. Ther. 25, 2189–2201 (2017). Blood 124, 188–195 (2014).

Reviews
142. Mahadeo, K. M. et al. Management guidelines for 152. Watanabe, K., Kuramitsu, S., Posey, A. D. Jr & June, C. H. 162. Mount, C. W. et al. Potent antitumor efficacy of
paediatric patients receiving chimeric antigen receptor Expanding the therapeutic window for CAR T cell anti-GD2 CAR T cells in H3-K27M( + ) diffuse midline
T cell therapy. Nat. Rev. Clin. Oncol. 16, 45–63 therapy in solid tumors: the knowns and unknowns of gliomas. Nat. Med. 24, 572–579 (2018).
(2018). CAR T cell biology. Front. Immunol. 9, 2486 (2018). 163. Nellan, A. et al. Durable regression of
143. Lee, D. W. et al. ASBMT Consensus Grading for 153. DeRenzo, C., Krenciute, G. & Gottschalk, S. medulloblastoma after regional and intravenous
cytokine release syndrome and neurological toxicity The landscape of CAR T cells beyond acute lymphoblastic delivery of anti-HER2 chimeric antigen receptor T cells.
associated with immune effector cells. Biol. Blood leukemia for pediatric solid tumors. Am. Soc. Clin. J. Immunother. Cancer 6, 30 (2018).
Marrow Transplant. https://doi.org/10.1016/j. Oncol. Educ. Book 38, 830–837 (2018). 164. Ahmed, N. et al. HER2-specific chimeric antigen
bbmt.2018.12.758 (2018). 154. Kachala, S. S. et al. Mesothelin overexpression is a receptor-modified virus-specific T cells for progressive
144. Gardner, R. et al. Decreased rates of severe marker of tumor aggressiveness and is associated with glioblastoma: a phase 1 dose-escalation trial. JAMA
CRS seen with early intervention strategies for CD19 reduced recurrence-free and overall survival in Oncol. 3, 1094–1101 (2017).
CAR-T cell toxicity management. Blood 128, 586 early-stage lung adenocarcinoma. Clin. Cancer Res.
(2016). 20, 1020–1028 (2014). Acknowledgements
145. Perales, M. A., Kebriaei, P., Kean, L. S. & Sadelain, M. 155. Heinmoller, P. et al. HER2 status in non-small cell lung The work of the authors is supported in part by the Intramural
Reprint of: building a safer and faster CAR: seatbelts, cancer: results from patient screening for enrollment Research Program, the National Cancer Institute and the NIH
airbags, and CRISPR. Biol. Blood Marrow Transplant. to a phase II study of herceptin. Clin. Cancer Res. 9, Clinical Center.
24, S15–S19 (2018). 5238–5243 (2003).
146. Yu, H. et al. Repeated loss of target surface antigen 156. Situ, D. et al. Expression and prognostic relevance of Author contributions
after immunotherapy in primary mediastinal large B MUC1 in stage IB non-small cell lung cancer. Med. Both authors made substantial contributions to all stages of
cell lymphoma. Am. J. Hematol. 92, E11–E13 (2017). Oncol. 28 (Suppl. 1), 596–604 (2011). the preparation of this manuscript.
147. Kochenderfer, J. N. et al. Long-duration complete 157. Louis, C. U. et al. Antitumor activity and long-term
remissions of diffuse large B cell lymphoma after fate of chimeric antigen receptor-positive T cells in Competing interests
anti-CD19 chimeric antigen receptor T cell therapy. patients with neuroblastoma. Blood 118, 6050–6056 The authors declare no competing interests.
Mol. Ther. 25, 2245–2253 (2017). (2011).
148. Budde, L. et al. Remissions of acute myeloid leukemia 158. Long, A. H. et al. Reduction of MDSCs with all-trans Publisher’s note
and blastic plasmacytoid dendritic cell neoplasm retinoic acid improves CAR therapy efficacy for Springer Nature remains neutral with regard to jurisdictional
following treatment with CD123-specific CAR T cells: sarcomas. Cancer Immunol. Res. 4, 869–880 (2016). claims in published maps and institutional affiliations.
a first-in-human clinical trial. Blood 130 (Suppl), 811 159. Yeku, O. O., Purdon, T. J., Koneru, M., Spriggs, D.
(2017). & Brentjens, R. J. Armored CAR T cells enhance Reviewer information
149. Yang, L. et al. Preclinical efficacy of CD33 chimeric antitumor efficacy and overcome the tumor Nature Reviews Clinical Oncology thanks S. Grupp and other
antigen receptor T cell immunotherapy in childhood microenvironment. Sci. Rep. 7, 10541 (2017). anonymous reviewer(s) for their contribution to the peer
acute myeloid leukemia. Pediatr. Blood Cancer 65 160. Avanzi, M. P. et al. Engineered tumor-targeted T cells review of this work.
(Suppl.), O-100 (2018). mediate enhanced anti-tumor efficacy both directly
150. Schmidts, A. & Maus, M. V. Making CAR T cells a solid and through activation of the endogenous immune Disclaimer
option for solid tumors. Front. Immunol. 9, 2593 system. Cell Rep. 23, 2130–2141 (2018). The content of this publication does not necessarily reflect the
(2018). 161. Brown, C. E. et al. Optimization of IL13Ralpha2- views of policies of the Department of Health and Human
151. Morgan, M. A. & Schambach, A. Engineering CAR-T targeted chimeric antigen receptor T cells for Services, nor does mention of trade names, commercial products
cells for improved function against solid tumors. Front. improved anti-tumor efficacy against glioblastoma. or organizations imply endorsement by the US Government.
Immunol. 9, 2493 (2018). Mol. Ther. 26, 31–44 (2018).

Reproduced with permission of copyright owner. Further reproduction
prohibited without permission.

Mechanisms of Resistance To CA

Uploaded by

Copyright:

Available Formats

Mechanisms of Resistance To CA

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mechanisms of Resistance To CA

Uploaded by

Copyright:

Available Formats

Reviews

Mechanisms of resistance to CAR

372 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

Key points an anti-CD22 CAR T cell product was successfully man-

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 373

374 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 375

376 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 377

378 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 379

380 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 381

382 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 383

384 | JUNE 2019 | volume 16 www.nature.com/nrclinonc

NatUrE REviEws | CliNicAl ONcology volume 16 | JUNE 2019 | 385

You might also like