Lactate dehydrogenase versus haemoglobin: which one is the better marker in paroxysmal nocturnal haemoglobinuria?

Abstract

As an experienced group who all managed patients with paroxysmal nocturnal haemoglobinuria (PNH) before the availability of complement inhibition therapies, we felt compelled to respond to the article ‘How we(’ll) treat paroxysmal nocturnal haemoglobinuria: diving into the future’ by Risitano and Peffault de Latour.1 PNH often develops on the background of a bone marrow failure syndrome, most commonly aplastic anaemia, but also myelodysplasia. The most significant complications of PNH arise from a failure to regulate terminal complement activity on the surface of blood cells. In PNH, uncontrolled terminal complement activation on PNH granulocytes, monocytes, platelets and red cells leads to thrombosis, organ damage and intravascular haemolysis (IVH).2, 3 IVH, as measured by lactate dehydrogenase (LDH), is the best measure for severity of disease in the vast majority of patients. A high LDH reflects high disease activity and risk of thrombosis, renal impairment, pulmonary hypertension and death.4 The only patients in whom LDH is not a good measure are those with a low proportion of PNH Type III red cells. These patients are, however, still at risk if they have higher proportions of PNH Type II red cells, monocytes or granulocytes. Thrombosis is the leading cause of mortality in PNH. The four-year survival rate for patients with PNH and thrombosis at presentation is approximately 40%.5, 6 Thrombosis in PNH relates to terminal complement activity on red cells (IVH) and, very importantly, on white cells and platelets resulting in their activation.7-9 Therefore, control of terminal complement activity is most important for preventing thrombosis in PNH and is measured by LDH.10 The introduction of eculizumab, a terminal complement inhibitor, transformed patient lives — significantly reduced thrombosis risk, decreased or eliminated transfusion requirements, improved renal function and pulmonary pressures, improved quality of life, and improved survival for patients.10-16 With this therapy, a high bar has been achieved in the management of patients with PNH. Elevated levels of LDH are a biomarker of IVH, and monitoring and reducing LDH levels is vital in the effective management of PNH.16 Patients with PNH experiencing IVH with an LDH > 1·5 × ULN (upper limit of normal), are at risk of thrombosis and death. An LDH > 1·5 × ULN will detect 96% of patients suffering a thrombosis. This explains why the primary aim in the management of patients with PNH is to achieve an LDH < 1·5 × ULN.4 Sustained terminal complement blockade is essential in the safe management of patients with this disease. While it is true that eculizumab will not improve bone marrow function, terminal complement inhibition is sometimes needed to prevent thrombosis, even in patients with severe bone marrow failure.17 Patients with significant bone marrow failure and PNH can be safely, and importantly, effectively managed with eculizumab, concomitantly with therapies for the underlying bone marrow failure if needed.18-20 We are concerned that the proposed response classification by Risitano and Peffault de Latour based on haemoglobin does not emphasise the central role of terminal complement activity in the development of the serious complications in PNH. A low haemoglobin may lead to mild to moderate symptoms of dyspnoea and fatigue but is seldom life-threatening. Risitano and Peffault de Latour propose a ‘Major response’ as one in which LDH remains >1·5 × ULN. This continues to place the patient at risk of thrombosis, organ damage and symptoms. Risitano and Peffault de Latour also propose ‘Partial’, ‘Minor’ and ‘No response’ based only on low haemoglobin levels with no consideration of LDH. Disease classifications deserve clinical validation and should not simply propose arbitrary cut-offs. For example, what is the evidence that a 57-year-old male with a haemoglobin of 128 g/l and an LDH < 1·5 × ULN (‘Good response’) has an inferior outcome compared to a similar patient with a haemoglobin of 131 g/l (‘Complete response’)? Once the primary goal of terminal complement inhibition has been achieved, we agree that if patients have a symptomatic anaemia this needs to be investigated and managed. There are many causes of anaemia in PNH.19 Extravascular haemolysis was managed supportively for the last 15+ years and survival still matched closely to a healthy population.15 Ideally, treatment of PNH should lead to control of terminal complement with a normal to near-normal haemoglobin. Many are exploring if complement inhibitors upstream of C5 can safely enhance treatment for PNH patients without losing the essential terminal complement control. Improving quality of life is important but we should not compromise the high bar of disease control — control of the terminal complement system that prevents thrombosis. In summary, we agree with Risitano and Peffault de Latour that for symptomatic patients with anaemia on a C5 inhibitor, improvements are needed, be it correcting haematinic deficiencies, improving bone marrow function or reducing extravascular haemolysis. However, we should be cautious to remember the importance of controlling terminal complement. The proposed classification has not been validated and does not specifically take thrombosis, the leading cause of death from PNH, into consideration.