Characteristics of patients and available data on surface marker expression

Of the 5034 patients with MC disorders available for analysis in the data set of the ECNM registry, 2531 patients with SM were identified in whom at least one CD marker (CD2, CD25, or CD30) was analyzed in/on MC by flow cytometry and/or IHC.

Most of these patients had ISM. In fact, within the group of ISM, marker studies were performed in 2008 patients (79.6% of all ISM patients). This was followed by patients with advSM (458 patients, 77.6% of all cases with advSM) and SSM (65 patients, 77.3% of all patients with SSM). The 458 patients with advSM were diagnosed to have ASM (n = 110), SM-AHN (n = 307), or MCL (n = 41).

Demographic and laboratory variables of patients are summarized in Table 1. Patients with ISM were younger with a median age of 47 years compared to advSM (median age of 64 years). There was a preponderance of male sex in patients with advSM (284 male patients, 62% of patients with advSM; 174 female patients, 38% of patients with advSM) compared to ISM (893 male patients, 44.5% of patients with ISM; 1115 female patients, 55.5% of patients with ISM). Regarding data availability of expression of CD2, CD25 and CD30 in/on MC in patients with SM, CD2 was available in 75.8% of cases, CD25 in 94.6%, and CD30 in 23.5% of SM patients (Table 1).

Table 1 Summary of demographic and laboratory variables of patients with MC disease with at least one analyzed CD marker.Expression of CD2 is associated with non-advanced SM, whereas lack of CD2 expression is often found in advSM

Results of both CD2 and CD25 by flow cytometry and/or IHC were available in 2225 patients (44.25%).

In patients with SM in whom MC expressed CD25, lack of expression of CD2 was recorded in 290 of 1522 patients with ISM (19.05%), in 14 of 45 patients with SSM (31.11%) and in 134 of 294 patients with advSM (45.58%). Among these cases with advSM and CD25 in/on MC, lack of CD2 in/on MC was found in 24/67 patients (35.82%) with ASM, 102/201 (50.75%) with SM-AHN, and 8/26 (30.77%) in patients with MCL. On the other hand, expression of CD2 in MC was found in 1232 of 1522 patients (80.95%) with ISM, in 31 of 45 patients (68.89%) with SSM and in 160 of 294 patients (54.42%) with advSM. In fact, in most patients with ISM, MC were found to express CD2 and CD25, whereas in patients with SSM and advSM, MC often failed to display CD2 but still expressed CD25 (Fig. 1, Table 2).

Fig. 1: Distribution of different patterns of expression of CD2 and/or CD25 in/on MC in various groups of patients.

Expression of CD2 and CD25 in or on bone marrow MC was determined by flow cytometry and/or immunohistochemistry in 2225 patients (44.25%) with mastocytosis and available results of CD2 and CD25 expression, including cutaneous mastocytosis (CM, n = 221), indolent systemic mastocytosis (ISM, n = 1595), smoldering systemic mastocytosis (SSM, n = 48), and advanced systemic mastocytosis (advSM, n = 313). In most patients with ISM, MC expressed both CD2 and CD25. However, whereas in a considerable number of patients with SSM and advSM, MC did not express CD2 but still expressed CD25. In very few patients with SM, MC were found to lack CD2 and CD25. MC mast cells.

Table 2 The number of patients with available results of CD2 and CD25 expression in patients with different entities of MC diseases is depicted in Fig. 1.Lack of CD2-expression in MC is indicative of a poor prognosis in SM

Survival analysis was performed in 1183 patients with available results of expression of CD2 and CD25 in MC as well as of survival time data. Survival analysis was performed in patients alive with a date of last FU.

When examining all patients with SM (non-advanced SM and advSM), we found a significantly reduced OS in patients with CD2-negative MC expressing CD25 compared to patients in whom MC expressed both CD2 and CD25. Moreover, a significantly shorter OS was found when comparing SM patients with CD2-negative MC with patients in whom MC expressed CD2 but did not co-express CD25 (p < 0.0001) (Fig. 2).

Fig. 2: Overall survival (OS) in patients with SM stratified by CD2 and CD25 expression patterns in MC.

OS was analyzed in a univariate analysis in 1183 patients with SM in whom FU and survival data were reported and results by flow cytometry and/or immunohistochemistry for CD2 expression and CD25 expression on/in MC were available. Four groups of patients were examined based on the pattern of CD2 and CD25 expression on/in MC: CD2-/CD25+ MC, CD2+/CD25+ MC, CD2+/CD25- MC; and CD2-/CD25- MC. The probability of OS in these 4 groups of patients was determined according to the method of Kaplan and Meier. Patients with CD2-negative MC expressing CD25 had a significantly reduced OS compared to patients with MC expressing both, CD2 and CD25. Patients with CD2-negative MC had a significantly shorter OS compared to patients in whom MC expressed CD2 without co-expressing CD25 (p < 0.0001). The p value refers to the comparison of all survival curves as assessed by log-rank test. SM systemic mastocytosis, FU follow-up, MC mast cells.

In a next step, stratification of CD2 expression analysis was done according to the three different disease entities ISM, SSM and advSM. There was a statistically significant reduction in OS in patients with ISM and advSM with CD2-negative MC compared to patients with CD2-positive MC (p = 0.018, and p = 0.012, respectively). Although there was a trend to a reduced OS in patients with SSM and CD2-negative MC, this difference did not reach statistical significance because of the small number of patients (Supplementary Figs. S1a–c). In fact, because of the small number of patients in all three entities of advSM, ASM, SM-AHN and MCL, a separate statistical analysis would interfere with power and would not provide clinically meaningful results.

Next, we examined PFS and EFS and both were considerably influenced by CD2 expression in/on MC in ISM and advSM (Supplementary Figs. S2, S3, Supplementary Figs. S5, S6). When examining all patients with SM, there was a significantly reduced PFS in patients with CD2-negative MC expressing CD25 compared to patients with co-expression of CD2 and CD25 in/on MC (p < 0.0001) (Supplementary Fig. S4). Also, the EFS was significantly reduced in patients with MC not expressing CD2 but expressing CD25 compared to MC with co-expression of CD2 and CD25 (p < 0.0001) (Supplementary Fig. S7).

Looking at different MC disease entities, this significant difference in PFS and EFS influenced by CD2-expression was present in ISM, whereas in SSM – although there was a trend—and in advSM this difference in PFS and EFS was not statistically significant.

In a final step, we applied multivariate analysis using CD2 expression, CD25 expression, age and sex of patients. In these studies, the predictive power of lack of CD2 expression was maintained regarding OS, PFS and EFS for the total cohort of all patients with SM as well as in patients with ISM. However, in SSM and advSM the predictive power of lack of CD2 expression regarding OS, PFS and EFS was lost (Supplementary Tables S2a–d; S3a–d and S4a–d).

We also employed prognostic markers that were available in at least 10% of all patients, including age, extramedullary involvement, albumin, LDH and BST, in patients with ISM and lack of CD2 expression of MC in a univariate and multivariate analysis, using these prognostic variables as non-imputed data. We found that lack of CD2 expression of MC was an independent prognostic variable concerning OS, PFS and EFS in this cohort of patients with ISM (Supplementary Tables S5a–c).

Death in patients with ISM was significantly less disease related compared to patients with SSM and advSM, although power of statistical analysis is limited due to small number of cases (Supplementary Table S6).

Lack of CD2 expression in MC is associated with extramedullary involvement in SM

Malignant expansion of mobilized MC in advSM often involves extramedullary organs, such as the spleen, liver, and/or lymph nodes. Therefore, we asked whether CD2-negativity in/on MC in SM is associated with an extramedullary spread of these cells and/or with splenomegaly, hepatomegaly, and/or lymphadenopathy, defined as palpable lymph nodes or lymph nodes >2 cm in size in sonography or computed tomography. Logistic regression analysis, taking into account the absent expression of CD2, expression of CD25, age and sex, with the outcome of extramedullary involvement, revealed that patients with CD2-negative MC have a 2.63 higher odds ratio of extramedullary involvement compared to patients with CD2-positive MC (Table 3). In fact, in SM patients with CD2-negative MC, 31.9% had extramedullary involvement, whereas in patients with CD2-positive MC extramedullary involvement was only found in 14.9% (Fig. 3).

Table 3 Analysis of extramedullary involvement in a univariate and multivariate analysis, taking into account CD2/CD25 expression patterns in MC, age at inclusion and sex.Fig. 3: Percentage of patients with SM and extramedullary involvement according to the CD2 expression status in MC.

The two bars reflect the percentage of patients with SM according to the CD2 expression status in MC. In SM patients with CD2-negative MC, 31.9% had extramedullary involvement, whereas in SM patients with CD2-positive MC extramedullary involvement was only found in 14.9%. This difference was statistically significant (p < 0.001). SM systemic mastocytosis, MC mast cells.

We also examined patients with non-advanced SM and advSM separately regarding the impact of lack of CD2 expression in/on MC on extramedullary involvement. In most patients with non-advanced SM (n = 25, 6.5%), there was only one extramedullary organ effected, whereas in a majority of patients with advSM (n = 25, 30.1%) all three organs, spleen, liver and lymph nodes, were affected.

Associations between expression of aberrant CD2 and CD25 in MC with other clinical and laboratory parameters

Logistic regression models were created to examine the association between CD2 expression and CD25 expression with clinical parameters in SM, adjusted for age and sex (Supplementary Tables S7, S8). CD2-negativity of MC was associated with lower rates of allergies (Odds Ratio (OR): 0.53; 95% CI 0.42, 0.67; p < 0.001), constitutional/cardiovascular symptoms (OR: 0.65; 95% CI: 0.50, 0.84; p = 0.001), and SM-related osteopathy (osteopenia or osteoporosis) (OR: 0.62; 95% CI: 0.49, 0.79; p < 0.001). By contrast, we found no associations between CD2-negativity of MC and the presence of pruritus, blistering/bullae, or gastrointestinal symptoms in our SM patients. CD25-positivity did not show significant associations with any of the aforementioned clinical parameters (Supplementary Figs. S8–S13).

We also performed univariate and multivariate analyses to detect possible associations between expression of CD2 and CD25 in/on MC and SM-related laboratory parameters, such as mutational status with KIT D816V mutation, other activating KIT mutation, mutations in genes other than KIT in molecular genetic analysis, and chromosomal aberrations in cytogenetic analysis. There was a statistically significant association between CD2-negative MC and mutations in genes other than KIT (p = 0.011) with an OR of 7.54. Ultimately, there was a positive association between CD2-negative MC and an abnormal karyotype with conventional cytogenetic analysis, although this did not reach statistical significance (p = 0.4) (Supplementary Table S9).

There was an association between CD2-negativity of MC and basal serum tryptase. The median basal serum tryptase was 40.55 µg/l (Interquartile range (IQR): 21.00, 102.00) and 29.00 µg/l (IQR: 17.00, 63.40), in 531 patients with CD2-negative MC and in 1441 patients with CD2-positive MC, respectively (p < 0.001) (Supplementary Fig. S14). Patients with lack of CD2 expression of MC had a significantly higher median serum tryptase levels compared to patients in whom MC were found to express CD2 (Supplementary Fig. S15). The Wilcoxon-signed-rank test did not reveal a statistically significant difference between patients with CD25-positive (median: 32.40 µg/l, IQR: 19.00, 78.95) and CD25-negative (median: 31.00, IQR: 15.38, 61.48) MC (p = 0.109).

The expression or lack of CD30 in MC is not associated with advSM and does not interfere with the prognostic impact of CD2-negativity of MC in SM patients

Results of expression of all three markers, CD2, CD25, and CD30, by flow cytometry and/or IHC were available in 506 patients (10.05%). This includes 42 patients with CM and ten with “mastocytosis in the skin”. In these patients with availability of all three mentioned markers, expression of CD30 on MC was tested in 364/2008 patients with ISM, 13/65 patients with SSM, and 77/458 patients with advSM (Fig. 4, Table 4). Comparing ISM with advSM, the proportion of cases with CD2-, CD25+ and CD30+ MC was considerably lower in patients with ISM. On the other hand, the percentage of patients in whom MC displayed the aberrancy pattern CD2+, CD25+, CD30- or the pattern CD2+, CD25+, CD30+ was considerably higher in patients with ISM compared to patients with advSM. There was no difference in the percentage of cases where MC exhibited the aberration profile CD2+, CD25+, CD30+ when comparing ISM, SSM, and advSM.

Fig. 4: Distribution of different patterns of expression of CD2, CD25 and CD30 in/on MC in various groups of patients.

Expression of CD2, CD25, and CD30 in or on bone marrow MC was determined by flow cytometry and/or immunohistochemistry in 506 patients (10.95%) with mastocytosis, including cutaneous mastocytosis (CM, n = 42), “mastocytosis in the skin” (MIS, n = 10), indolent systemic mastocytosis (ISM, n = 364), smoldering systemic mastocytosis (SSM, n = 13), and advanced systemic mastocytosis (advSM, n = 77). In patients with ISM, MC expressed both CD2 and CD25, with or without CD30 expression, in a considerably higher number compared to patients with advSM. However, in patients with ISM there was a lower percentage of patients not expressing CD2 but still expressing CD25 and CD30, compared to SSM and advSM. In very few patients with SM, MC were found to lack CD2 and CD25, regardless of CD30 expression. MC mast cells.

Table 4 Number of patients with available results of CD2, CD25 and CD30 expression in patients with different entities of MC diseases depicted in Fig. 4.

Analyzing patients with CD30 expression only without analysis of CD2 and CD25 expression at the same time, there was CD30 expression in 224/428 patients with ISM (52.34%), in 12/18 patients with SSM (66.67%), and in 67/111 patients with advSM (67.68%). The difference between these groups of MC disorders was statistically significant (p = 0.014).