Hemolysis and Neurologic Impairment in PAMI Syndrome: Novel Characteristics of an Elusive Disease

Our patient was born at term after an uncomplicated pregnancy and showed only mild postnatal jaundice, which resolved without phototherapy. Moderate vesicoureteral reflux was diagnosed on prenatal ultrasonography and the child underwent multiple hospitalizations for recurrent pyelonephritis. Normochromic normocytic anemia was observed during his first hospitalization, at 3 months of age, and revealed a trend of increasing severity, becoming transfusion dependent by 6 months of age (∼10–15 mL/kg of concentrated red blood cells were transfused every 4–6 weeks to maintain hemoglobin [Hb] levels >8 g/dL). Worsening hepatosplenomegaly had first been noted at 5 months and biochemistry panels were suggestive of an ongoing hemolytic process. No skin or articular manifestations were reported.

After our evaluation, which occurred at 18 months of age, the physical examination was remarkable for general dystrophic features (Fig 1), weight below the third centile for age, dysmorphic features (coarse facial traits, frontal bossing, saddle-shaped nose, wide philtrum), slightly icteric skin, massive splenomegaly (lower pole at 8 cm below the left costal margin), moderate hepatomegaly, severe inguinal hernia, axial hypotonia, irritability, and marked developmental delay (not capable of sitting unassisted, no language development). Family history was noncontributory.

Multiple laboratory tests were performed: a complete blood count confirmed the presence of severe normochromic normocytic anemia (Hb 6.1 g/dL; mean cell volume 73.7 fL; mean cellular hemoglobin 25 pg; mean cellular hemoglobin concentration 30.9 g/dL; red cell distribution width 21.8%) with noncompensatory reticulocytosis (absolute reticulocyte count 140 000/uL; reticulocyte index 1.7) and adequate neutrophil, lymphocyte, and platelet counts; the biochemistry panel revealed mild-to-moderate hemolysis (indirect bilirubin 1.4 mg/dL; lactic dehydrogenase 1700 UI/L), although haptoglobin levels were within the range of normal; a direct Coombs test result was mildly positive; blood film evaluation revealed marked anisopoikilocytosis and multiple spherocytes; an acidified glycerol lysis test (AGLT) revealed increased erythrocyte osmotic fragility in multiple samples, whereas eosin-5′-maleimide–binding testing and ektacytometry excluded a diagnosis of membrane-protein defect; and sodium dodecyl sulfate–polyacrylamide gel electrophoresis revealed normal band 3 protein glycosylation. Bone marrow aspiration revealed a normocellular marrow with maturing trilineage hematopoiesis and moderate erythroid expansion with mild aspecific sign of dyserythropoiesis. Extensive lymphocyte subset analysis by flow cytometry was not consistent with autoimmune lymphoproliferative syndrome or other primary immuno-dysregulatory disorders. Specific testing to exclude lysosomal storage diseases were performed and results came back negative. A brain MRI revealed diffuse atrophy and ex vacuo enlargement of cerebral ventricles and subarachnoid spaces without meningeal enhancement (Fig 2A) and no specific anomalies at spectrometric analysis.

The child was later hospitalized for fever, clinical deterioration, cervical and inguinal lymphadenopathy, generalized but transient urticarial rash, worsening anemia, and inflammatory marker increases (C-reactive protein [CRP] up to 30 mg/dL; procalcitonin 0.8 ng/mL; ferritin up to 2000 ng/mL; interleukin-18 far beyond the upper normal limit). A urine sample collected during the fever peak revealed no leukocytes or nitrites and normal mevalonate levels. On further review of the patient history, we discovered that the previously reported recurrent episodes of “pyelonephritis” were in fact characterized by fever without a focus, repeatedly normal urinalysis, occasionally positive urine culture results, markedly increased CRP, normal procalcitonin levels, and unresponsiveness to broad-spectrum antibiotics. These episodes did not reveal any clear periodicity and lasted 10 to 14 days. Suspecting an autoinflammatory disease, we empirically started anakinra monotherapy at 3 mg/kg once daily with dramatic clinical improvement: fever and the cutaneous rash disappeared within 12 hours, the lymphadenopathies resolved, and CRP became undetectable in the ensuing week. Furthermore, over the subsequent 12 months, we observed a steady increase in Hb concentration paralleled by a spike in reticulocyte counts (that subsided when Hb stabilized at 12 g/dL), cessation of hemolysis and of transfusion support (Fig 3), and a considerable reduction of both spleen and liver size (Fig 4). Finally, our patient showed a gradual but significant improvement in his neurobehavioral profile, being able to walk with minimal assistance and use purposefully 7 bisyllabic words within 12 weeks of interleukin-1 (IL-1) blockade initiation; also, a brain MRI performed 6 months after therapy initiation revealed almost complete resolution of atrophy (Fig 2B). Anakinra has been well tolerated by our patient, with only mild transient urticarial skin reaction at the injection site.

A customized next-generation sequencing (NGS) panel with 41 genes known to be associated with autoinflammatory diseases⁶  identified a de novo heterozygous c.748G>A (p.E250K) pathogenic variant in the PSTPIP1 gene, later confirmed by Sanger sequencing.

PSTPIP1 is a cytoskeleton-associated adaptor protein encoded by the homonymous gene, which, by interacting with pyrin and thus helping in assembling the inflammasome, modulates the activation of neutrophils, phagocytes, and T-lymphocytes.^7,8  In PAMI, the mutation p.E250K determines a stronger interaction between PSTPIP1 and pyrin, leading to a complex dysregulation of the innate immune system with an uncontrolled release of cytokines (interleukin-1β and interleukin-18) and damage-associated molecular patterns (especially calprotectin), which in turn maintain a positive inflammatory feedback loop responsible for the clinical picture of patients with PAMI. The pathophysiological link between the dysregulation of the PSTPIP1-pyrin axis and the hematologic manifestations seen in PAMI (ie, neutropenia, pancytopenia) is yet to be elucidated, although a role for PSTPIP1 in the maturation of the hematopoietic stem cell has been proposed.⁹ 

Isolated hemolytic anemia has never been associated with PAMI syndrome, although this is a recently described syndrome whose incidence is probably underestimated. In our case, it is conceivable that the combination of a hyperactive hepatosplenic reticuloendothelial system (responsible for peripheral erythrocyte destruction) and a blunted marrow response, both driven by the uncontrolled inflammatory storm, led to the observed severe anemia. Indeed, hypersplenism, a reported cause of erythrocyte volume/surface ratio reduction,¹⁰  would also explain the decreased osmotic fragility revealed by AGLT. This hypothesis is strongly supported by the fact that anemia resolution and AGLT normalization paralleled the decrease in inflammatory markers and spleen volume after IL-1 blockade initiation.

In the reported case, PAMI syndrome was not initially included in the differential diagnosis given the isolated anemia in the absence of neutropenia, skin, or articular involvement, and testing for zinc or calprotectin serum levels was not pursued. Severe mevalonate kinase deficiency was suspected at first but was excluded because of the absence of mevalonate acid in the urine.¹¹  We also suspected an atypical cryopyrin-associated periodic syndrome, which can present with failure to thrive, neurologic involvement, and subtle dysmorphic features (ie, frontal bossing), although the absence of chronic urticarial rash would have been highly unusual.¹²  Given this suspicion, in the light of the tumultuous clinical deterioration of our patient, we were urged to start anakinra therapy to reach a diagnosis in an ex juvantibus fashion without waiting for genetic confirmation. Indeed, a good response to anti IL-1 blockers has already been reported in patients with PAPA and PAMI,⁸  although a partial response has also been described^1,3,7  especially regarding the hematologic feature of the disease. After the genetic diagnosis, we opted not to perform serum zinc or calprotectin evaluation because it would not have changed our clinical management.

To our knowledge, neurologic involvement has been previously linked to a mutation in PSTPIP1 only a single time, although in that case, the reported patient suffered from a ruptured posterior circulation aneurysm arising from a likely inflammatory vasculopathy.¹³  On the other hand, inflammatory encephalopathy is one of the hallmarks of neonatal-onset multisystem inflammatory disorder, the most severe entity on the cryopyrin-associated periodic syndrome spectrum.¹⁴  However, in neonatal-onset multisystem inflammatory disorder, neurologic impairment is usually the consequence of chronic aseptic meningitis with or without postinflammatory obstructive hydrocephalus.¹⁵  In our case, the observed psychomotor delay was likely secondary to severe persistent systemic inflammation given the improvement noted clinically and confirmed by imaging after anakinra. In fact, no inflammatory lesions were highlighted by fluid-attenuated inversion recovery, nor was diffusion restriction typical of ischemic damage uncovered by diffusion-weighted imaging, which would exclude inflammatory involvement of middle- to small-size vessels,¹⁶  and leptomeningeal contrast enhancement was absent, which would exclude a chronic meningitis pattern.

Finally, malformations are not a part of the clinical spectrum related to autoinflammatory syndromes. Our patient presented with inguinal hernia, which, along with developmental delay and hepatosplenomegaly, may be a manifestation of mucopolysaccharidoses (ie, Hurler and Hunter syndrome).¹⁷  Nonetheless, lysosomal storage disease testing results came back negative and given the relative frequency of inguinal hernias and vesicoureteral reflux in the general population, the absence of other features of syndromes possibly linked to that malformation pattern (eg, connective tissue disorders)¹⁸  and the fact that the uncovered genetic mutation was coherent with the rest of our patient’s clinical picture, we ascribed those findings to a chance association and did not pursue further genetic or chromosomal testing.

Our experience confirms that an open-minded approach combining clinical thoroughness and the use of customized NGS panels can favorably affect the diagnosis and management of complex and atypical cases. Our case also illustrates the broad spectrum of phenotypes associated with PAMI syndrome, which should be considered in any case of unexplained cytopenia coupled with autoinflammatory stigmata (ie, recurrent fever, persistently raised inflammatory markers). Testing promptly for elevated zinc and calprotectin serum levels and including PSTPIP in NGS panels for chronic cytopenias are both strategies that may help clinicians not to miss the diagnosis of this elusive disease.

AGLT

acidified glycerol lysis test

CRP

C-reactive protein

Hb

hemoglobin

IL-1

interleukin-1

NGS

next-generation sequencing

PAMI

PSTPIP1-associated myeloid-related proteinemia inflammatory

PAPA

pyogenic arthritis, pyoderma gangrenosum, and acne