The Association of Anemia and Patient Reported Outcomes in the New Era of PNH Treatment

Background: Paroxysmal Nocturnal Hemoglobinuria (PNH) is a rare acquired hematopoietic stem cell disorder characterized by hemolytic anemia and possibly complicated with thromboembolism and other cytopenias. Symptoms such as fatigue and dyspnea have been attributed to intravascular hemolysis rather than to the extent of anemia. Anti-complement treatment in symptomatic hemolytic PNH patients is associated with improved survival and quality of life (QoL). To date, cancer-specific patient-reported outcome measures (PROMS), such as EORTC QLQ-C30, have been used in PNH patients, with fatigue being identified as a major PRO issue. However, many other impacts of PNH are unassessed by the cancer PROM. A novel PNH-specific PROM, QLQ-AA/PNH, consisting of 54 items, is undergoing validation. In this observational multicenter international study, the primary endpoint is to evaluate differences in EORTC QLQ-C30 scores between patients without versus those with intravascular hemolysis and, as a secondary endpoint, the same differences in QLQ-AA/PNH scores. Other secondary endpoints include reliability and intraclass correlation by test-retest methodology of QLQ-AA/PNH and factors associated with QoL and symptoms.

Materials and Methods: Patients of age ≥ 18 years with PNH from Italy and UK have been included. Exclusion criteria are active bone marrow failure, history of treatment for cancer within the last 5 years and clinically significant comorbidities that impact PROMs. Patients completed both EORTC QLQ-C30 and QLQ-AA/PNH on the day of the scheduled treatment at the clinic and after 1-2 weeks. Demographics and disease features were collected at baseline. Intravascular hemolysis is defined by serum lactate dehydrogenase (LDH) ≥ 1.5 times upper limit of normal.

Results: From May 13, 2021, and March 12, 2024, 82 PNH patients (male 43, female 39) of mean age 50 (SD 15) years were included. Ongoing treatment consisted of eculizumab (N=24), ravulizumab (N=32), investigational products (N=20), and supplements or other therapy (N=6). At the time of this analysis, thirteen patients (16%) suffered from hemolysis. Mean Hb level was 11.0 (SD 1.7, range 7.9 – 15.4) g/dL. EORTC QLQ-C30 scores were comparable to the normal European and US population. Noteworthy, mean fatigue score by EORTC QLQ-C30 was 27.9 (SD 21.9; normal reference = 29.5, SD 25.5). Cronbach alpha estimates of the QLQ-AA/PNH at the first assessment (t1) and at retest (t2) were above 0.8 for all domains except for the social support (t1=0.26, t2=0.20) and body index (t1=0.26, t2=0.11). The intra-class correlation coefficient for all domains ranged from 0.70 to 0.95. The correlation between the mean score of t1 and t2 by Spearman’s correlation coefficient ranged from 0.45 to 0.92, with QLQ-AA/PNH domains PNH illness intrusiveness, PNH physical functioning, PNH emotional functioning and PNH stigmatisation showing the strongest correlations. Convergent validity with Spearman’s rank correlation test showed high correlations between the following domains of QLQ-AA/PNH: emotional function (EF), cognitive function (CF), physical function (PF) and role function (RF) with the corresponding domains of the EORTC QLQ-C30 [ EF (r_s=0.69), CF (r_s=0.60), PF (r_s=0.72) and RF (r_s=0.68); p<0001]. Neither the EORTC QLQ-C30 nor the QLQ-AA/PNH correlated with residual intravascular hemolysis (p=ns). Noteworthy there were significant correlations between Hb levels and all domains of the QLQ-AA/PNH, except for the social support. Between cases had Hb <10 g/dL (N=19) and the remaining cases with higher levels , there were significant differences in almost all domain scores of the EORTC QLQ-C30: PF (78.9 vs 89.6, p=0.005), RF (71.0 vs 89.9, p<0.0001), CF (75.4 vs 89.7 p<0.001), SF (73.7 vs 92.7, p<0.0001)] and in the QLQ-AA/PNH stigmatisation (56.1 vs 84.7, p=0.009), illness intrusiveness (63.7 vs 83.7, p=0.006), infection (66.7 vs 87.1, p=0.002) and other symptoms (76.9 vs 88.2, p=0.009).

Conclusions: In this cohort of patients with PNH with ongoing treatment, PROs were similar to the normal European and USA population. While impaired PRO was not associated with the extent of residual intravascular hemolysis (as LDH), we demonstrate that the degree of anemia (as Hb level) impacts on outcome in several domains, other than fatigue, underscoring the opportunity of switching to resolution of anemia to further improve outcome in affected areas.