Risk of serious infection among individuals with and without low count monoclonal B-cell lymphocytosis (MBL)

To the Editor

Monoclonal B-cell lymphocytosis (MBL), the precursor state to chronic lymphocytic leukemia (CLL), is one of the most common premalignant conditions in humans [1, 2]. MBL is subdivided into low count (LC) MBL and high count (HC) MBL depending on whether the absolute B-cell count is above or below 0.5 × 10⁹/L [3]. Although all individuals who develop CLL appear to pass through MBL [4], only a minority of individuals with MBL progress to develop CLL. The overwhelming majority (i.e., >98%) of individuals with MBL have LC MBL and never come to clinical attention [3, 5]. The clinical consequences of LC MBL are largely unknown. Although longitudinal studies indicate that the clonal B-cell population persists and expands over time in the majority of individuals with LC MBL [5, 6], progression to CLL is rare and believed to occur in ~1 in 500 LC MBL patients. Currently, it is unknown whether LC MBL has other clinical implications. Evidence suggests that some of the immune deficits observed in HC MBL are present in LC MBL, albeit to a lesser degree [5, 6].

Herein, we screened for MBL using highly sensitive flow cytometry in asymptomatic adults participating in the Mayo Clinic Biobank to evaluate the prevalence of MBL and its relationship to infectious complications and overall survival. The Mayo Clinic Biobank is a large scale bio-repository of adult patients assembled to study a wide array of health-related research studies [7]. With consent of participants and the approval of the Mayo Clinic IRB, we screened participating individuals from Olmsted County (the county in which Mayo Clinic is located), Minnesota, who were 40 years of age or older, no prior history of hematologic malignancy, and who had stored PBMC (samples from 7/14/2009 to 3/4/2016). Data on hospitalization with infection was abstracted from clinical records as previously published [4]. All hospitalizations occurring after the date of enrollment in the Biobank were ascertained through the earliest date of death, migration from Olmsted County, or 12/31/2017. Infections associated with hospitalization were considered “serious or life-threatening infections.”

Stored PBMC were used to screen for MBL using eight-color (CD38, CD45, Kappa, Lambda, CD19, CD23, CD5, and CD20) flow cytometry assay with the capacity to detect clonal B-cell events to the 0.005% level (1/20,000 events) consistent with other high sensitivity MBL screening approaches [1, 5]. For each sample, up to 500,000 PBMC events were captured for each individual. The pathologist and the technologist performing the MBL assays were blinded to clinical outcomes of all participants, and the medical record abstractor collecting data on infection and clinical outcome was blinded to the results of MBL screening assays.

For each individual, we computed the Charlson Comorbidity Index (CCI), which is a prognostic score based on 17 disease categories, using the comorbid health conditions at the time of sample date. Logistic regression was used to evaluate associations with MBL risk. Kaplan-Meier plots were generated to determine cumulative incidence. Cox regression was used to estimate hazard ratios (HR) and 95% confidence intervals (CI).

Results/discussion

A total of 1045 individuals were screened for MBL. Samples from 61 patients (6%) were not interpretable primarily due to having too few viable cells. Among the 984 (94%) remaining individuals, 119 (12%) had MBL including 106 (11%) with CLL-phenotype (CD5, CD19, CD20 (dim), CD23, kappa or lambda light chain restriction [dim]). Thirteen (1%) individuals had a B-cell clone of atypical (n = 7) or non-CLL phenotype (n = 6) detected. Five patients had both a CLL and a non-CLL [2] clone present.

The remainder of the analysis focused on the 106 MBLs with CLL phenotype and the 865 controls (i.e., those without any MBL clone). The demographic characteristics, medical history, and baseline complete blood count (CBC) results of these individuals are summarized in Table 1. The presence of CLL phenotype MBL was strongly associated with advancing age (p < 0.001) and sex (men 14%; women 8%; p = 0.003).

Table 1.

Demographics and characteristics of hospitalization for infections by group.

	All participants (n = 971)	CLL phenotype MBL (n = 106)	No MBL (n = 865)	Unadjusted p value	Age and sex-adjusted p value
Age, median (range)	62 (40–94)	70 (44–94)	62 (40–93)	<0.001	–
Age group, N (%)				<0.001	–
40–49	130 (13.4%)	4 (3.8%)	126 (14.6%)
50–59	298 (30.7%)	21 (19.8%)	277 (32.0%)
60–69	308 (31.7%)	30 (28.3%)	278 (32.1%)
70–79	186 (19.2%)	38 (35.8%)	148 (17.1%)
80+	49 (5.0%)	13 (12.3%)	36 (4.2%)
Sex, N (%)
Female	569 (58.6%)	48 (45.3%)	521 (60.2%)	0.003	–
Male	402 (41.4%)	58 (54.7%)	344 (39.8%)
Race, N (%)
White	927 (95.7%)	104 (98.1%)	823 (95.4%)	0.31	–
Asian	8 (0.8%)	0 (0.0%)	8 (0.9%)
African American	5 (0.5%)	0 (0.0%)	5 (0.6%)
American Indian/Alaskan Native	2 (0.2%)	1 (0.9%)	1 (0.1%)
Other and Mixed	27 (2.8%)	1 (0.9%)	26 (3.0%)
Missing	2	0	2
CBC, N, median (range)
Hemoglobin (g/dL)	N = 638, 13.7 (7.3–17.3)	N = 76, 14.1 (7.3–16.2)	N = 562, 13.6 (8.5–17.3)	0.45	0.50
Platelet count (×10(9)/L)	N = 638, 225 (51–599)	N = 76, 212 (84–453)	N = 562, 229 (51–599)	0.021	0.42
WBC (×10(9)/L)	N = 638, 6.2 (2.1–25.5)	N = 76, 6.2 (2.3–10.3)	N = 562, 6.2 (2.1–25.5)	0.052	0.011
ALC (×10(9)/L)	N = 529, 1.6 (0.4–5.0)	N = 68, 1.6 (0.4–4.0)	N = 461, 1.7 (0.5–5.0)	0.54	0.96
Co-morbid health conditions in baseline questionnaire, N (%)
Rheumatologic—any listed condition	364 (37.6%)	38 (36.2%)	326 (37.8%)
Arthritis (rheumatoid)	64 (6.8%)	5 (5.1%)	59 (7.1%)	0.45	0.25
Autoimmune disorder (lupus, scleroderma)	26 (2.7%)	2 (1.9%)	24 (2.8%)	0.99	0.72
Liver—any listed condition	34 (3.5%)	3 (2.9%)	31 (3.6%)
Hepatitis A, B, or C	25 (2.6%)	3 (2.9%)	22 (2.6%)	0.75	0.87
Infectious diseases
HIV (AIDS)	0 (0.0%)	0 (0.0%)	0 (0.0%)	–	–
Cancer—any nonhematologic	273 (28.2%)	44 (41.9%)	229 (26.5%)	<0.001	0.34
Thyroid cancer	8 (0.8%)	1 (1.0%)	7 (0.8%)
Lung cancer	12 (1.2%)	2 (1.9%)	10 (1.2%)
Breast cancera	44 (7.7%)	8 (16.7%)	36 (6.9%)
Esophageal cancer	4 (0.4%)	0 (0.0%)	4 (0.5%)
Pancreatic cancer	2 (0.2%)	0 (0.0%)	2 (0.2%)
Stomach cancer	4 (0.4%)	1 (1.0%)	3 (0.3%)
Colon or rectal cancer	12 (1.2%)	0 (0.0%)	12 (1.4%)
Liver cancer	2 (0.2%)	0 (0.0%)	2 (0.2%)
Uterine/endometrial cancera	10 (1.8%)	1 (2.1%)	9 (1.7%)
Cervical cancera	12 (2.1%)	2 (4.2%)	10 (1.9%)
Ovarian cancera	4 (0.7%)	1 (2.1%)	3 (0.6%)
Prostate cancera	41 (10.3%)	10 (17.5%)	31 (9.1%)
Testicular cancera	5 (1.3%)	0 (0.0%)	5 (1.5%)
Melanoma	35 (3.6%)	4 (3.8%)	31 (3.6%)
Nonmelanoma skin cancer	121 (12.6%)	16 (15.5%)	105 (12.3%)
Sarcoma	6 (0.6%)	1 (1.0%)	5 (0.6%)
Bone cancer	2 (0.2%)	1 (1.0%)	1 (0.1%)
Kidney cancer	1 (0.1%)	0 (0.0%)	1 (0.1%)
Urinary/bladder cancer	12 (1.2%)	3 (2.9%)	9 (1.0%)
Other cancer	13 (1.4%)	2 (1.9%)	11 (1.3%)
Neurologic—any listed condition	221 (22.9%)	25 (24.3%)	196 (22.7%)
Alzheimer’s disease	3 (0.3%)	0 (0.0%)	3 (0.3%)	0.99	0.99
Dementia	8 (0.8%)	2 (2.0%)	6 (0.7%)	0.20	0.52
Stroke (CVA)	19 (2.0%)	3 (2.9%)	16 (1.9%)	0.45	0.69
TIA (mini stroke)	29 (3.0%)	2 (2.0%)	27 (3.1%)
Cardiovascular—any listed condition	604 (62.4%)	72 (68.6%)	532 (61.6%)
Heart attack/myocardial infarction	61 (6.4%)	13 (12.7%)	48 (5.6%)	0.005	0.45
Coronary artery diseaseb	29 (8.7%)	8 (21.1%)	21 (7.1%)	0.010	0.19
Congestive heart failure	20 (2.1%)	1 (1.0%)	19 (2.2%)	0.71	0.17
Atrial fibrillation/arrhythmia	69 (7.2%)	12 (11.8%)	57 (6.7%)	0.059	0.64
High blood pressure (hypertension)	395 (41.0%)	49 (47.1%)	346 (40.2%)
High cholesterol (hyperlipidemia)	434 (45.2%)	46 (45.1%)	388 (45.2%)
Blood clots in a vein	41 (4.3%)	5 (5.0%)	36 (4.2%)
Respiratory—any listed condition	274 (28.4%)	35 (34.0%)	239 (27.7%)
Asthma	121 (12.6%)	10 (9.7%)	111 (13.0%)	0.34	0.47
Chronic obstructive pulmonary disease (COPD)	43 (4.5%)	8 (7.8%)	35 (4.1%)	0.12	0.32
Gastrointestinal—any listed condition	294 (30.4%)	34 (33.0%)	260 (30.1%)
Celiac disease	11 (1.1%)	0 (0.0%)	11 (1.3%)
Crohn’s disease or ulcerative colitis	16 (1.7%)	1 (1.0%)	15 (1.7%)
Endocrine—any listed condition	220 (22.8%)	24 (23.1%)	196 (22.8%)
Type 1 diabetes	8 (0.8%)	0 (0.0%)	8 (0.9%)	0.99	0.99
Type 2 diabetes	108 (11.3%)	12 (11.8%)	96 (11.2%)	0.87	0.53
Hyperthyroidism/hypothyroidism	124 (13.1%)	12 (11.9%)	112 (13.3%)
Family history leukemia or lymphoma any 1st degree relative
Family history lymphoma or leukemia	77 (9.0%)	12 (13.6%)	65 (8.5%)	0.11	0.051
Characteristics of Hospitalizations for Infections
Total Individuals with at least one hospitalization for infection following baseline date	97	20	77
Total hospitalizations for infection following baseline date	187	40	147
Site of infection for which hospitalizedc
Pneumonia	32	9	23
Cellulitis	33	3	30
Urinary tract	42	10	32
Blood stream	33	11	22
Otherd	47	7	40
8-year cumulative incidence estimate (95% confidence interval)
All Sites		23% (13–32%)	11% (9–13%)	<0.001
Specific Sites
Pneumonia		9% (4–18%)	2% (2–4%)	0.002
Blood stream		9% (4–18%)	2% (1–4%)	0.002
UTI		7% (3–16%)	3% (2–5%)	0.10
Cellulitis		2% (1–8%)	3% (2–5%)	0.79

^aSex specific frequencies.

^bOnly in versions 2 and 3 of questionnaire (i.e. large number of missing values).

^cSome patients experienced more than one hospitalization for infection At the individual patient level, 20 individuals with MBL and 77 individuals without MBL had at least one hospitalization for infection.

^d“Other” includes infection diagnostic/site categories such as colitis/diverticulitis, encephalitis/meningitis, hepatobiliary/cholangitis, influenza infection, osteomyelitis/septic joint, upper respiratory, thoracic/pleural.

No differences in medical history between MBLs and controls were noted after adjusting for age and sex (Table 1). Similarly, no evidence of any differences in the median Charlson Co-morbidity Index for individuals with (median 1, IQR 0–3) and without MBL (median 1; IQR 0–2) was observed after adjusting for age and sex (p = 0.35). No differences in hemoglobin (Hg), platelet count, or absolute lymphocyte count (ALC) were observed after adjusting for age and sex although the total leukocyte count was slightly higher in controls (Table 1).

Among the 68 MBLs with a clinical ALC within 1 year of the sample date, the median clonal B-cell count was 4.0 cells per uL (range 0.0002−1.1382 × 10⁹/L) and only two MBLs had a clonal B-cell count >0.5 × 10⁹/L. The median proportion of the B-cells that were clonal was 2.1% (range 0.2–96.7%). In aggregate, the clonal B-cells represented <50% of total B-cells in 98 of 106 (92%) MBL cases. Six of the 8 MBLs with a clonal B-cell count >50% had a CBC within 9 months of the sample date. Four of these six MBLs had clonal B-cell counts <0.5 × 10⁹/L (range 0.08 × 10⁹/L to 0.48 × 10⁹/L). The remaining two MBLs, both with clonal B-cell counts ≥85%, had clonal B-cell count of 0.63 × 10⁹/L and 1.1382 × 10⁹/L, respectively.

After a median follow-up of 8 years, 97 of the 971 individuals were hospitalized with infection for a total of 187 times (e.g., 57 were hospitalized with infection once, 18 were hospitalized twice, and 22 were hospitalized 3 or more times). Among these 187 hospitalizations for infection, the site of infection was urinary tract in 42 (22%), cellulitis in 33 (18%), blood stream/sepsis in 33 (18%), and pneumonia in 32 (17%). The site of infection was “other” in the remaining 47 (25%) hospitalizations (e.g., colitis/ diverticulitis, osteomyelitis/septic joint, upper respiratory, thoracic/pleural, influenza, hepatobiliary/cholangitis, encephalitis/meningitis). At least one hospitalization for infection was observed in 20 of 106 MBLs and 77 of 865 controls (p = 0.044 after adjusting for age and sex). Collectively, the 20 MBLs were hospitalized with infection 40 times while the 77 controls were hospitalized with infection 147 times (Table 1).

The unadjusted estimated 8-year cumulative incidence of any infection (MBL 23% [13–32%] vs. controls 11% [9–13%]; p < 0.001) as well as pneumonia (9% [4–18%] vs. controls 2% [2–4%]; p = 0.002) and blood stream/sepsis (9% [4–18%] vs. controls 2% [1–4%]; p = 0.002) were more common among MBL individuals (Table 1). No evidence of any differences in the 8-year cumulative incidence of urinary tract, cellulitis or other infection sites were observed. After adjusting for age and sex, the HR was 1.68 (CI: 1.01–2.79; p = 0.044; Fig. 1a). The age- and sex-adjusted 8-year cumulative incidence estimates of any infection was MBL 18% [7–28%] vs controls 12% [9–14%] (p = 0.004). After adjusting for age and sex, the HR for death was 1.55 (CI: 0.82–2.92; p = 0.18; Fig. 1b).

Fig. 1. Clinical outcomes of individuals with monoclonal B-cell lymphocytosis (MBL) compared to those without MBL.

a Cumulative incidence of hospitalization for infection among individuals with (n = 106) and without MBL (n = 865) adjusted for age and sex. b Overall survival among individuals with (n = 106) and without MBL (n = 865) adjusted for age and sex.

This prospective, cohort study provides critical new insights about LC MBL. Individuals with CLL-phenotype LC MBL appear to be at increased risk for serious infectious complications. This increased risk appears to primarily relate to an increased risk of pneumonia and blood stream infections/sepsis. Further, no evidence of any confounding due to underlying health conditions (e.g., diabetes, COPD, malignancy) explains this increased risk based on comprehensive co-morbidity assessment. To our knowledge, this is the first report of the relationship between LC MBL and non-CLL related clinical consequences specifically the risk of serious infections.

A plausible immunologic basis for our findings exists. A number of defects in humoral and cellular immunity predating exposure to chemotherapy exist in CLL patients including hypogammaglobinemia, B-cell and T-cell dysfunction, and defects in innate immunity [8, 9]. Preliminary data suggests that some of these defects also affect individuals with MBL [10–14]. We have found that the number of exhausted T-cells (CD8+, CD160+) is increased, and that the immunological T-cell synapse function is significantly impaired among individuals with HC MBL compared to age matched controls. Further, individuals with LC MBL are more likely to have co-existent clonal T-cell populations with restrictive T-cell specificity. Collectively, these observations suggest that individuals with MBL have detectable, but more limited, immune defects than those with CLL which may confer an increased risk for infection.

Few longitudinal studies of individuals with LC MBL have been reported to date. Our findings suggest that the risk of serious infections may be more important for individuals with LC MBL than their risk of progression to CLL or another lymphoproliferative disorder.

There is scant data evaluating whether personal medical history and/or co-morbid health conditions affects an individual’s risk of LC MBL [15]. Among the 971 individuals in this cohort, we found no correlation between baseline health conditions among those who did and did not have MBL after adjusting for age and sex. This observation would argue against an association between the health conditions assessed and the risk of developing LC MBL.

In summary, individuals with CLL-phenotype LC MBL appear to be at increased risk for serious infectious complications. If this observation is validated, it would suggest that 5–10% of adults over age 40 (6–12 million U.S. adults) have a largely unstudied clinical condition with important health implications.