To B or not to B maintained?

In this issue of Blood, Bhoj et al explore the longevity of plasma cells (PCs) and humoral immunity in the absence of B cells. This is done by studying subjects experiencing B-cell aplasia after CD19-directed chimeric antigen receptor (CAR) therapy.¹ 

Rapid immune protection against previously encountered antigens depends on the presence of circulating antibodies. These antibodies are secreted by antigen-specific PCs, a product of the germinal center reaction after infection or vaccination. Despite a maximum half-life of ∼1 month for IgG, this humoral protection can last for the entire lifetime of the host. Thus, PCs must be either particularly long-lived or constantly replenished from the B-cell pool, or both.

It has been >40 years since McMillan and colleagues first suggested that long-lived PCs residing in the bone marrow are the primary source of circulating IgG in humans.²  Understanding the PC compartment in humans is essential to fully appreciate the mechanisms of how long-lasting humoral immunity is formed and maintained. Still, this population remains largely elusive in humans. Conversely, in mice, the concept of long-lived PCs is widely accepted. There is evidence that long-lived PCs are maintained independently of both antigen and replenishment from the B-cell pool.^3,4  To date, the best evidence supporting a long-lived PC population in humans is the demonstration that PCs persist for at least 6 months after B-cell depletion therapy.^5,6  Recent reports of bone marrow–resident CD19^–CD38^hiCD138⁺ PCs with a prosurvival phenotype and specificity for historic antigens^6,7  also provide evidentiary support.

In this issue, Bhoj et al provide evidence for the persistence of long-lived CD19^– PCs in the complete absence of CD19⁺CD20⁺ B cells in humans. The authors studied 16 subjects (4 adults and 12 pediatric patients) included in clinical trials for a CD19-directed CAR-based adopted T-cell therapy (CTL019) for the treatment of several B-cell malignancies. CD19 is expressed not only on tumor cells, but also on all normal B cells from the early precursor stages through maturation until terminal differentiation. Therefore, profound B-cell aplasia and hypogammaglobulinemia⁸  were expected adverse effects of CTL019 treatment. Consequently, patients undergoing this treatment offer a unique opportunity to study the long-term maintenance of CD19^– PCs and preexisting humoral immunity in humans without the interference of recruitment from the B-cell pool.

Bhoj et al confirm previous findings of a CD19^–CD38^hiCD138⁺ subset of PCs^6,7  in the bone marrow. Using flow cytometry, they further show that, as expected, CD19^–, but not CD19⁺, PCs are spared after CTL019 treatment. Immunohistochemical analysis of bone marrow biopsies taken post–CTL019 treatment confirmed the results from flow cytometry. By both of these techniques, they showed complete loss of CD19⁺ and CD20⁺ cells in all patients except one. In several of the subjects, there was a CD19^–CD138⁺ PC population preserved for as long as 25 months after CTL019 treatment. These results demonstrate that CD19⁺CD20⁺ B cells are not required for maintenance of the long-lived PC population. It is worth noting that although CD138⁺ PCs were present in the bone marrow biopsies from all 4 adult subjects after CTL109 treatment, these cells were detectable in only 4 of the 12 pediatric patients. Interestingly, these 4 pediatric patients were between 17 and 21 years old, and thus were among the oldest included in the pediatric cohort, indicating that it takes time to form a detectable CD19^– PC population. Mei et al previously showed that CD19^– PCs are completely lacking in infants for up to at least 7 months.⁶  Thus, the long-lived CD19^– PC subset appears to be formed throughout childhood and adolescence, not reaching numbers above a detectable limit in bone marrow biopsies until between 15 and 20 years of age. This model emphasizes the importance of immune priming by vaccination during these years for making up a pool of protective circulating antibodies. In fact, Bhoj et al show that, even in a state of chronic B-cell aplasia, the continuous secretion of antigen-specific IgG and IgA remains. Although the cohort was limited in size, these data nevertheless convincingly demonstrate the existence of a long-lived PC population important to maintaining humoral immunity to antigens encountered early in life.

During the course of the study, the authors also had the opportunity to analyze postmortem tissue samples from a CTL019-treated subject. They reported that during chronic B-cell aplasia, CD19^– PCs are not restricted to the bone marrow, as previously thought,^6,7  but are also present in lymph nodes and in the mucosa of the gastrointestinal tract. Supporting this observation was a study of human gut biopsies,⁹  as well as a recent study in mice showing that there is indeed a survival niche in the intestinal lamina propria for long-lived PCs generated in mucosal responses.¹⁰ 

With this study, Bhoj et al provide evidence for a long-lived CD19^– PC population that is maintained independently of B cells. These findings may also be of direct clinical importance when B-cell depletion is used to treat autoimmune diseases or any disease with antibody-mediated pathology. That is, the pathogenic autoantibodies may be a product of long-lived PCs and will consequently persist despite the ensuing B-cell aplasia. Future studies on larger cohorts will likely provide firmer conclusions and greater insight into PC biology. These studies could demonstrate whether loss of CD19 expression reflects commitment to the long-lived PC population, or if CD19⁺ PCs can also be long-lived. Finally, little is known about the biological triggers that drive long- vs short-lived PC differentiation or development as a B-memory cell vs as a PC.