A total of 800 participants were recruited in 17 hematology units of 9 Italian regions (Figure 1). Data on 85 patients treated in Trento Hospital had already been published (Pilot Study)1, therefore were excluded from the present. Three other centres (n=82 patients) were also excluded because of too few patients, incongruent and/or data missing. To minimize the survivor bias, only naïve patients (without prior administrations of RTX) were considered in the analysis of ADR. Therefore, among 633 patients included in the study, we focused our analysis on 505 (79.8%) incident cases belonging to 13 centres (Figure 2)
The 505 patients (42% women) included in the study were affected by NHL (n=453), CLL (n=33) or one of the diseases included in the Italian Law 648/96 list (n=19). Patients had a median age at baseline of 66.8 years (interquartile range-IQR 57.5–73.9 years) and 41.6% had a performance status=0. The median follow-up of the patients was 317 days (IQR 217- 461 days) (Table 2).
During the study period the patient population received 3,681 infusions of rituximab, specifically 53% (n=1984) Tru, 24% (n=901) Rix, 19% (n=694) MabSC, 3% (n=102) MabIV. The mean number of infusions per patient was 7.3 (standard deviations 3.8).
Rituximab treatment
Out of the 505 naïve patients, 392 (78%) did not experience any switch, (“no switcher” group, n=257 on Tru and n=130 on Rix) and 113 (22%) were "switcher" patients (Table 1). Overall, 57% of patients were treated with a combination of RTX and chemotherapy (64% in “no switcher” patients and 32% in “switcher” patients), 160 (55.6%) patients of those had a diagnosis of aggressive NHL. Twenty-five patients were treated with RTX on monotherapy (23 “no switcher” patients), among these, 88% were affected by NHL.
In the first infusion, the majority of the 505 naïve patients started treatment with the biosimilars Tru (n. 298) and Rix (n. 139), while 67 patients started with the originator MabIV. The Sankey diagram (Figure 3) shows the RTX treatments of the 505 patients from the first to the eighth infusion, after which 40% of the patients were still in treatment. To note that 69% of "switcher" patients (n=78) switched to MabSC, the majority of these (73%) switched at the second infusion after initially receiving MabIV (40 patients) or Tru (15 patients) or Rix (2 patients) at the first infusion.
From the Sankey diagram (Figure 3), two types of switches can be highlighted: the first is the switch to MabSC, and the second is the switch among different IV RTX treatments. Comparing these two groups with the “no switcher” patients (Table 2), it becomes evident that the baseline characteristics (i.e., at the first infusion, which does not necessarily coincide with the infusion in which the switch occurs for “switcher” patients) are similar, except for the duration of the disease (for “no switcher” patients the median is 1 year, for “switcher” to MabSC is 1.5 years, and for “switcher” between IV Rituximab is 2 years), the type of diagnosis (the majority of “switcher” to MabSC - 63% - and “no switcher” - 47% - have aggressive NHL, while nearly half - 49% - of the “switcher” between IV Rituximab have unspecified NHL), the follow-up time (for “no switcher” patients, the median is nine and a half months, for “switcher” to MabSC, it is one year and three months, and for “switcher” between IV Rituximab, it is one and a half years), and the number of infusions (for “no switcher” patients, the average number is six, whereas “switcher” patients have an average of ten infusions)."
The characteristics of patients who discontinued RTX treatment before the 6th cycle are shown in table S2. For the majority of these patients (70%), the reason for interrupting treatment is not given, 21.8% interrupted the treatment forADR and 7.9% for other reasons (mostly due to the progression of the disease).
Adverse Reaction Related to rituximab
Overall, 85 (16.8%) patients reported ADRs to RTX. ADRs occurred more frequently in patients with indolent NHL (18.8%), compared to those with highly malignant NHL (15.6%), those with CLL (15.2%) and those with conditions treated under the law 648-1996 (15.8%); 21 ADRs occurred in 113 patients (18.8%) with unspecified NHL (Table 3).
The correlations between ADRs and the demographic and clinical-laboratory characteristics of the 505 naïve patients are shown in Table S1. The occurrence of ADRs is significantly associated with the age at diagnosis (p=0.05) and a higher neutrophil count (p<0.01). In the 85 patients that experienced at least one ADR during the study period, 124 RTX-related events were reported (1.5 events/patient). Among these, 16% were classified as grade 3-5. The most frequent reactions were General disorders and administration site conditions (n. 50, 8% serious) (Table 4).
In the study population, 16.8% of the patients experienced at least one ADR. This percentage varies significantly among the different treatment groups, with a higher incidence among the "no switcher" patients (20.7%), particularly among those receiving Rix (24.6%) and Tru (17.9%), and a lower incidence (3.5%) among the "switcher" patients (Table 3).
Among the 85 patients with at least one ADR to RTX in the 21 infusions, ADRs were observed most frequently during the first infusion (44 patients, 52%) and the second infusion (17 patients, 20%). Between the third and sixth infusions, fewer than 10 patients experienced ADRs, and from the seventh infusion onwards, it decreased to one. Figure 4 shows that the majority of ADRs occurred in patients undergoing treatment with Rix or Tru.
The frequency of ADRs also diminishes as the number of infusions increases. During the initial infusion, 9% of patients (44 out of 505) experienced ADRs, which drops to 3.4% in the second infusion with 17 out of 498 patients. From the third to the fifteenth infusion, the incidence remains below 2%. Notably, at the sixteenth infusion, there appears to be a slight increase, but this is attributed to the limited number of patients receiving treatment at that stage, resulting in a less stable incidence estimate (3.1%, 1 patient with ADR out of 32). (Figure S1).
In Figure S2 are reported the incidences of ADRs at each infusion, separately for "switcher" and "no-switcher" patients. Of the 85 ADRs, 4 ADRs occurred in "switcher" patients, whereas 81 in and "no-switcher" patients. In "switcher" patients the incidence does not reach 1% (n=112), while in "no-switcher" patients the incidence at the first infusion is 11.2% (n=393), it drops to 4.1% at the second infusion (n=386), remains below 2% until the eighth infusion (n=103), and in the ninth and sixteenth, it appears to increase again. However, as with the overall incidence, these fluctuations are probably due to the small number of patients receiving treatment from the ninth infusion onwards (n<43).