Lessons learned from a review of paroxysmal nocturnal haemoglobinuria (PNH) requests: a report from the UK PNH Network

Paroxysmal nocturnal haemoglobinuria (PNH) is a rare life-threatening disease, with a UK prevalence of 15·9 per million (Hoffbrand et al, 2011). The UK has a nationally commissioned PNH service which works extensively to educate physicians about PNH. European prevalence is reported to be lower, probably due to lower screening patterns. Due to the rarity and non-specific symptoms, such as lethargy, erectile dysfunction or abdominal pain, patients present to different specialties and diagnosis may take several years. Prior to complement inhibition, median survival was 10–22 years (Hillmen et al, 1995; de Latour et al, 2008; Hill et al, 2010).

International flow cytometry (FC) recommendations advise PNH testing in at-risk populations by FC (Table 1; Borowitz et al, 2010). Early identification is essential to ensure appropriate management. The UK PNH Network, consisting of National PNH Service haematologists plus those with an interest in PNH, assessed the indications for PNH testing on sample receipt to laboratories against international recommendations.

Anonymised data from 10 central UK laboratories were provided to a pre-prepared proforma developed from the FC recommendations assessing clinical indications for testing, PNH clone size and, in the presence of a positive test, the final diagnosis. All patients who had undergone PNH testing for the first time, and a cohort who had undergone repeat PNH testing between January and December 2014 were included. Data were analysed with SAS Version 9.4 (SAS Institute Inc., Cary, NC, USA).

The proportion of PNH granulocytes ± monocytes was measured by FC. The lower limit of sensitivity for positive testing ranged from 0·1% to 0·01% glycosylphosphatidylinositol (GPI)-deficient cells; positivity was defined as the presence of PNH cells above the lower limit of detection for the laboratory.

A total of 1656 new PNH screens were analysed, of which 9·5% (158) were positive. 53% were male (878), 96·7% (1601) were aged >18 years. Median age was 56 years (range 18–99 years) in adults, and 13 years (range 2–17) in the paediatric population. Regional variation in test positivity ranged from 2·9% to 18·5%, with a median of 8·6%. Results of testing indications are shown in Table 2.

Of the patients with a positive PNH screen (n = 158), 37·5% had <10% PNH cells; 11·4% of PNH-positive patients had PNH cells >30%. Indications for screening in the 11·4% were: cytopenias (n = 6), haemolysis (n = 6), aplastic anaemia (n = 3), thrombosis (n = 1), no clinical details (n = 2).

A subgroup of 377 patients who had undergone repeat PNH testing were reviewed: 345 had positive PNH screens, all of whom were actively monitored, 72% were on eculizumab with the remaining 38% not requiring complement inhibition therapy. 25% of patients with symptomatic PNH had evolved from aplastic anaemia (AA), and 3% from myelodysplastic syndrome (MDS). Five of the retested patients who were previously positive were negative on re-testing. One had undergone bone marrow transplantation; the remaining four initially had a PNH clone <1%, reflecting the high sensitivity of the assay and dynamic nature of bone marrow function.

We report the largest UK analysis of PNH laboratory requests, which highlights several learning points. Similar to other studies, 9·5% of patients had a detectable PNH clone (Morado et al, 2016). There was regional variation in the numbers of positive results, which may relate to local physician experience of PNH and awareness of current guidance. Although PNH FC methodology is outwith the scope of this letter, variations are common (Fletcher et al, 2017). The lower limits of testing sensitivity also vary, which raises issues in patients with <0·1% GPI-deficient populations. This is unlikely to account for the wide variation in regional testing. Although PNH testing recommendations should be adhered to (Table 1), a small patient cohort will require testing at the discretion of the treating physician.

The AA cohort showed the highest detection rate of PNH screen positivity (42%), in keeping with published data (Killick et al, 2016). 10·9% of the cytopenia cohort had positive screens compared with 5–42% in the literature (Movalia et al, 2011; Morado et al, 2016). There was limited blood count information but the data highlights the importance of testing in unexplained cytopenia(s) especially when the underlying cause may be bone marrow dysfunction.

PNH positivity in low grade MDS is approximately 12% (Wang et al, 2009), however, compared to the literature (Morado et al, 2016), this study tested lower than expected numbers of MDS patients. It therefore appears that high numbers of MDS patients are not receiving appropriate PNH testing, thus ongoing education and prospective studies are advisable.

Arterial thrombosis is well documented in PNH, associated with high mortality (Hillmen et al, 1995). This analysis did not demonstrate PNH positivity in this patient group, and further studies exploring PNH incidence in thrombosis would be useful.

The majority of patients had PNH clones below 30%, in whom symptoms would be unusual, and a concurrent haematological diagnosis, such as AA or MDS, the main disease. There remains clinical distinction between patients with symptomatic PNH and those with smaller clones. It is essential these patients are monitored (Killick et al, 2016) and, as illustrated here, there remains a significant risk of clonal expansion with development of haemolytic/thrombotic PNH.

Suggested areas for improvement to PNH testing include a comment on bone marrow reports e.g., ‘Bone marrow in keeping with AA; advise PNH testing’. This could also be utilised in transfusion laboratories for Coombs-negative haemolytic anaemia. A PNH test request form developed from the international recommendations would also ensure all clinical information is provided, reducing inappropriate testing unless discussed.

Our study has limitations. As with all retrospective analyses, there are omitted data. The analysis does not assess whether patients considered at risk of PNH have undergone testing, rather this analysis assessed indications on sample receipt to the laboratory. A multi-departmental approach to capture all patients presenting to different specialties would resolve this.

In summary, we present the largest UK analysis of PNH laboratory testing indications. Whilst the analysis is encouraging, we summarise the lessons learned which include increasing MDS and thrombosis patient testing, and we encourage prospective studies in these areas. Adding statements to laboratory reports where appropriate should improve uptake of testing. We also highlight PNH clonal expansion risk in AA. Education remains integral to the PNH Service ethos, to maximise early diagnosis, ensure monitoring and reduce complications, thereby reducing morbidity and mortality.

Author contributions

MG: data collection, analysis, preparation and review of manuscript. AH: devised the concept and reviewed the data and manuscript. All authors: data collection, review of manuscript.

Funding

The UK PNH network receives an unrestricted educational grant from Alexion Pharmaceuticals, Inc. Alexion also provided independent statistical support.