Bone marrow necrosis and fat embolism syndrome: a near fatal complication in previously undiagnosed sickle beta + thalassaemia
- 1 Hematology and Oncology, NYU Langone Hospital—Long Island, Mineola, New York, USA
- 2 Geriatrics, NYU Langone Hospital—Long Island, Mineola, New York, USA
- 3 Pathology, NYU Langone Hospital—Long Island, Mineola, New York, USA
- Correspondence to Dr Nibash Budhathoki; nibash.budhathoki@nyulangone.org
Abstract
Prevalence of haemoglobin sickle-β+ thalassaemia (Hb S/β+thal) is variable with geography ranging from 0.2% to 10% among sickle cell patients. Clinical presentation of Hb S/β+thal patients depends on HbA level, with milder disease often going undiagnosed. However, rarely these patients can present with a fulminant vaso-occlusive crisis (VOC). Given VOC can present with non-specific symptoms, the diagnosis and treatment is often delayed. Here, we present a patient who initially developed altered mental status, pancytopenia and multiorgan failure due a critical VOC resulting in bone marrow necrosis and fat embolism. Subsequent workup confirmed that our patient had Sickle-β+ thalassaemia, which had gone undiagnosed, despite subclinical evidence of haemolysis on routine lab work for years. Following diagnosis and initiation of RBC exchange, he improved significantly and was discharged home. High index of suspicion and bone marrow biopsy is vital for early diagnosis and management of this rare condition.
Background
Sickle cell disease (SCD) is characterised by mutation in the gene encoding the β-globin subunit of haemoglobin. The mutation results in substitution of glutamic acid in sixth position of beta globin chain by valine.1 2 The mutation leads to production of abnormal sickle haemoglobin (HbSS). SCD has many genotypic and phenotypic variants with different clinical course.1 Beta-thalassaemia (β-Thal), results from a mutation leading to a deficit of normal beta globin production. Haemoglobin sickle beta thalassaemia (Hb S/βeta Thal), is seen in patients who possess both a sickle mutation and beta thal mutation. This condition was first described by Ezio Silvestroni and Ida Bianco in 1944.3
Hb S/β+thalassaemia prevalence varies with geography with prevalence ranging from 0.2% to 10% among sickle cell patients.4–10 Hb S/β+thalassaemia has variable clinical course ranging from mild to severe disease depending on amount of HbA.1 Rarely, patients with otherwise mild SCD may present with multiorgan failure with altered mental status, anaemia and thrombocytopaenia secondary to catastrophic bone marrow necrosis (BMN) with fat embolism syndrome as the initial presentation. Due to non-specificity of symptoms and rarity of the condition, diagnosis and treatment is often delayed leading to poor clinical outcomes.11 12 A high index of suspicion is vital for prompt diagnosis and early management of the condition prior to progression to permanent organ dysfunction or death.
Case presentation
The patient was 37-year-old vegetarian Italian American male who was brought to the emergency department with fever, altered mental status, low back pain, unsteady gait and urinary incontinence for 2 days. He was pale, icteric and confused with multiple small petechiae in the chest. He had stiffness of all four limbs at the time of presentation. A few days prior to the presentation, he reportedly went on a golfing trip with his friends when he developed fever and low back pain. He had history of chronic low-back pain and anaemia. His previous blood workup from 1998 showed haemoglobin of 105 g/L, mean corpuscular volume (MCV) of 78 fL and indirect hyperbilirubinaemia of 3.8 mg/dL. He had been self-medicating with cannabis for pain since the age of 15. His family history was positive for beta thalassaemia and Hashimoto’s thyroiditis in mother and vitamin B12 deficiency needing parenteral vitamin B12 in father. His father did not carry an official diagnosis for pernicious anaemia.
On the day of presentation, the patient was pancytopenic with haemoglobin of 71 g/L (MCV −78 fL), haematocrit 28.5, white cell count of (WCC) of 1500 and platelets of 43 000. His comprehensive metabolic panel was unremarkable except for indirect hyper bilirubinaemia with mild elevation of liver enzymes. His total bilirubin was 3.5 mg/dL (direct bilirubin - 0.8 mg/dL), aspartate transaminase (AST) was 69 IU/L, alanine transaminase (ALT) was 55 IU/L and alkaline phosphatase (ALP) was 407 IU/L. He had ongoing haemolysis with an elevated lactate dehydrogenase (3586 IU/L) and undetectable haptoglobin (<4 mg/dL). Peripheral smear showed true thrombocytopaenia, scant schistocytes (0–1 per high power filed), numerous nucleated red blood cells (RBC) along with target cells. Disseminated intravascular coagulation (DIC) profile was unremarkable with fibrinogen >700 mg/dL and normal coagulation profile (PT/INR- 11.7 s / 1.06, PTT- 23.6 s). Direct agglutinin test was negative. His vitamin B12 was found to be low 118 pg/mL. Drugs screening was positive for cannabinoids and oxycodone. His mother had given to him oxycodone the day prior to the presentation for low back pain.
He was admitted to intensive care unit (ICU) for further evaluation.
Investigations
Given severity of symptoms, our patient received an extensive evaluation including a brain magnetic resonance imaging (MRI). MRI showed evidence of subacute ischaemia and chronic microhaemorrhages. Computed Tomography (CT) scan of the abdomen and pelvis showed moderate splenomegaly, hepatomegaly, cholelithiasis and diffuse patchy sclerotic osseous abnormalities. Extensive workup for infectious aetiologies including bacterial, fungal and viral infections was done including lumbar puncture which were non-revealing. Rheumatological evaluation for possible autoimmune processes including vasculitis was non-revealing as well.
We performed bone marrow biopsy on day 2 of hospitalisation because of pancytopenia. On initial review of the bone marrow, the aspirate was entirely necrotic. Given bone marrow necrosis can occur in context of haematological malignancy, drug toxicity, infection, heterozygous haemoglobin S states or rarely autoimmune states or metastatic disease, an exhaustive workup was pursued. After learning about his mother’s history of beta-thalassaemia trait, we were also informed that his father had been screened for beta-thalassaemia trait prepregnancy. With this history, the patient’s presentation including prior lab findings and imaging findings, the findings were strongly suggestive of sickle cell process. This diagnosis was supported directly with visualisation of sickle cells within the bone marrow. Subsequently, sickle cell screening testing was done which was positive. After reviewing the imaging with radiology, we suspected that fat embolism resulting in small cerebrovascular accident was responsible for the patient’s altered mental status. Ultimately, the patient underwent RBC exchange on day 3 pending the results of the bone marrow core and haemoglobin electrophoresis given our concern for central nervous system (CNS) ischaemia. The patient responded remarkably to exchange transfusion with improvement of altered mental status, limb stiffness, improvement of his blood parameters and chest petechiae in following 2 days (hospitalisation days 4–5).
Official bone marrow biopsy smear was later reported to show sickle cells with BMN along with areas of trilieage haematopoiesis (figures 1–3). Bone marrow evaluation was otherwise unremarkable with no underlying malignancy or dysplasia. The haemoglobin electrophoresis was later reported on hospital day 5 as HbA2 5.3%, HbA 21.3%, HbS 67.8%, HbF 5.6 %, which was consistent with sickle cell beta + thal phenotype.
Bone marrow biopsy showing areas of necrosis (A) associated with sickle cell (B) and trilineage haematopoiesis with progressive maturation with relative erythroid hyperplasia (C).
Bone marrow aspirate showing areas of bone marrow necrosis.
Bone marrow aspirate showing sickle cells.
Differential diagnosis
We considered a broad differential given our patient’s presentation. Given that our patient presented with fever, altered mental status, anaemia and thrombocytopaenia, it was critical to strongly consider thrombotic thrombocytopenic purpura (TTP). However, lack of evidence for peripheral microangiopathy (no increased schistocytes) as well as his concurrent leucopaenia were not congruent with this diagnosis. Normal ADAMTS 13 level (73%) later confirmed that it was not TTP. Vitamin B12 deficiency with concurrent intramedullary haemolysis was another differential that was entertained, and may have contributed to his presentation.13 He was immediately started on vitamin B12 supplementation after he was determined to be deficient. However, vitamin B12 deficiency alone was unlikely to explain his clinical condition, particularly the areas of poor CNS perfusion on MRI diffusion weighted imaging. Haemophagocytic lymphohistiocytosis (HLH) was also considered as a differential diagnosis, even though this diagnosis would also be unlikely to result in findings seen on patient’s MRI. Laboratory testings were sent for HLH which were notable for a ferritin of 1420 ng/mL, however, interleukin-2 receptor was 331 and evaluation of bone marrow core did not show any evidence of haemophagocytosis. HLH was thus ruled out. Secondary aetiology including sepsis with concurrent meningitis and encephalitis was also on our differential which were ruled out with extensive infectious testing. These were felt to be unlikely to result in the degree of haemolysis observed in the absence of significant DIC.
Treatment
Given the morbidity and mortality associated with a delayed diagnosis of vaso-occlusive crisis (VOC) with fat embolism syndrome, the patient was managed with RBC exchange transfusion once our suspicion was high based on history, imaging and lab findings. Following the exchange transfusion, he had rapid clinical improvement with improvement of AMS, limb stiffness and improvement of overall well-being. For his low vitamin B12 level, he was treated with parenteral vitamin B12 supplementation. He did not require more sessions of exchange transfusion. His clinical condition and haematological parameters improved over next couple of days. He was then downgraded to regular medical floor on hospital day 7 and was discharged from hospital on day 10. At that time, his altered mental status had resolved and had partial improvement in haematological parameters.
Outcome and follow-up
He followed up in our outpatient haematology clinic in 1 week after the discharge. His blood workup showed WCC of 1.9, haemoglobin of 78 g/L, haematocrit 24 and platelets of 248 000 at the office. He received one unit of blood following which he did not require any transfusion. He continued to receive parenteral vitamin B12. His ferritin level, which was elevated at the time of presentation, was serially followed during the hospitalisation until it started to downtrend. Elevated ferritin was likely due to acute stress (ferritin being an acute phase reactant) and secondary to acute on chronic haemolysis. His ferritin during follow ups were 420 ng/mL (1 month after the discharge) and 271 ng/mL (4 months after discharge). Most recently, his ferritin level was 72 ng/mL (6/2020). We worked him up for possible pernicious anaemia as outpatient given his family history and low vitamin B12. Intrinsic factor antibody was negative which ruled out pernicious anaemia. He continued to follow up regularly, during which he continued to improve. We discussed further treatment options including prophylactic exchange transfusion versus hydroxyurea (after bone marrow recovery) versus stem cell transplant versus conservative management with pain medication going forward. He opted for conservative management with vitamin B12 supplementation. He subsequently moved to California, but reports intermittent painful crisis. His blood workup from his follow-up is shown in table 1.
Lab parameters during hospital course and follow-up
Presentation | D1 | Postexchange transfusion | Discharge | 4 months after discharge | 1 year after discharge | June 2020 | |
Hb/Hct | 74/28.5 | 54/17 | 106/31 | 86/25 | 109/34 | 112/33.4 | 110/35.9 |
WCC | 1.5 | 1.0 | 1.3 | 2.3 | 4.0 | 7 | 6.8 |
Platelets | 43 | 46 | 50 | 153 | 189 | 152 | 178 |
LDH | 3586 | 2517 | 1390 | 1136 | |||
Ferritin | 1420 | 1365 | 271 | 72 | |||
Haptoglobin | <4 | <4 | <4 | <4 | <4 | ||
ADAMTS 13 | 73.4 | ||||||
Vitamin B12 | 118 | 639 | >1200 | 469 | 798 | ||
Bilirubin (T) | 3.5 | 2.8 | 2.3 | 1.6 | 2.4 | 2.1 | |
AST/ALT | 69/55 | 52/43 | 30/24 | 25/19 | 15/13 |
-
ALT, alanine transaminase; AST, aspartate transaminase; Hb, haemoglobin; LDH, lactate dehydrogenase; WCC, white cell count.
After the acute episode, he had dramatic recovery with full improvement in his mental status, although he still has some subtle deficits in neurological functions including mild cognitive impairment, mild impairment in motor function and sensory deficits. He followed up with neurologist who recommended for neuro rehabilitation with cognitive exercises and to keep a journal to assist with daily activities and short term memory loss. He continues to have numbness in his lower lip and chin area and trouble with some of his motor skills that he used to be very proficient at which unfortunately has not improved to baseline despite physical therapy and occupational therapy. His last crisis was in June 2020. He notes precipitation of crisis on hot days (when temperature is above 37 °C) which is likely due to dehydration.
Discussion
VOC is a known complications in patients with SCD and its variants.14 Currently, VOC in sickle cell patients is thought to result from endothelial activation by sickled RBCs and/or inflammatory mediators leading to aggregation of sickled RBC, activated and adherent neutrophil and platelets causing the obstruction of blood vessels. Blockage of the vessel increases transit time for RBC leading to deoxygenation and subsequent polymerisation of haemoglobin propagating retrograde VOC and subsequent ischaemia. Tissue ischaemia results in further endothelial damage and creates a positive feedback loop leading to worsening of VOC.14 15 When VOC involves vital organs, it carries a high morbidity and mortality. Prompt diagnosis and early treatment significantly improves clinical outcome in these patients.11 16 17 This is particularly the case for intramedullary VOC resulting in BMN and subsequent fat embolism.11 12 16–19
In the reviews by Tsitsikas et al and Dang et al, the investigators reviewed cases of BMN and fat embolism reported in the world literature. They reported that majority of the patients who had extensive BMN with fat embolism syndrome had genotypes other than HbSS (66% vs 81%) with majority having HbSC genotype.16 20 Our patient had Hb S/β+thal genotype. Providers should therefore have high index of suspicion for BMN with fat embolism syndrome in patients with SCD who presents with worsening respiratory status, altered mental status and leukoerythorblastic smear with thrombocytopaenia especially those with HbSC and Hb S/β+. Tsitsikas et al in their review also reported an overall mortality of 64% with mortality reaching upto 91% in untreated patients.16 In this series, that 24% of patients were found to have Human Parvo Virus B19 infection potentially triggering the VOC.16 We had tested our patient for Human Parvo Virus B19 infection by serology and was found to be negative for both IgM (0.2) and IgG (0.8). Besides mortality, there is significant morbidity associated with delayed diagnosis. Case reports show that morbidity can range from permanent neurological deficits to being ventilator dependent.11 16 17 Timely intervention with exchange transfusion improves prognosis and clinical outcome as seen in our case and as reported by Adamski et al and Myers and Ipe.11 17 Myers and Ipe reported a patient with BMN with subsequent fat embolism precipitated by high altitudes in a young healthy man. Adamski et al in their case series described three cases, one with positive outcome compared with two with significant morbidity and mortality. Timely diagnosis and treatment was key in both cases where patients had great outcomes as was also seen in our case. Of note, peripheral smear evaluation did not show sickle cells on review by haematology service and pathology. This is not uncommon, and therefore it is important that we maintain an index of suspicion for sickle cell process even in absence of sickle cells on peripheral blood in patients who present with symptoms consistent with BMN and fat embolism syndrome. Our patient was admitted for total of 10 days, with 7 days of ICU stay and was discharged home. This was in contrast to other case series where there was prolonged hospital and ICU stay and poor clinical outcome due to delayed diagnosis.11 12
As was the case with our patient, early diagnosis and treatment can be particularly challenging when a patient does not carry a diagnosis of a haemoglobinopathy and presents with a fulminant life-threatening event. Our patient’s case underscore the value of obtaining a thorough history from the patient, as well as family history and historical labs when available.21
Retrospectively, his chronic low back pain was secondary to undiagnosed SCD, which was evident as osseous sclerotic lesions on his CT scan. The patient endorsed having painful episodes throughout his life starting around 10 or 11 years of age. He had terrible back pain, unusual for his age. Sometimes he felt it in his hands, knees, shoulders. Occasionally, he felt shortness of breath, blurred vision and chest pain. The pain was attributed to growing pains. His diagnosis as a child was missed likely due to his ethnicity and absence of family history of SCD.
We hypothesise that his acute VOC was likely precipitated by dehydration from golfing trip and alcohol use. His fulminant presentation may have been augmented by concurrent vitamin B12 deficiency in face of increased consumption due to acute haemolysis during VOC.
Learning points
-
Our case study reiterates the importance of corroborative information including family history, medical history, past blood work reports and demographic information in early diagnosis and management of the patients.
-
We strongly recommend considering bone marrow necrosis and Fat embolism syndrome as a differential diagnosis in patients with thrombotic thrombocytopenic purpura like picture, without significant evidence of microangiopathic haemolysis or when concurrent leucopaenia is present.
-
Providers should have high index of suspicion for bone marrow necrosis with fat embolism syndrome in patients with sickle cell disease who presents with worsening respiratory status, altered mental status and leucoerythorblastic smear with thrombocytopaenia especially those with HbSC and haemoglobin sickle-β+ thalassaemia.
-
Having an informed conversation with next-of-kin/family including shared decision making was critical in institution of empiric treatment in which ultimately resulted in the patient having a favourable outcome.
Patient’s perspective
I appreciate all you have done to diagnose me. For far too long I have had bad experiences with Drs and hospitals as no one was able to diagnose my condition. At a certain point in life I gave up hope that anyone was listening or even caring about the pain I was going through. I avoided going to the hospital at all costs because of that. Getting a diagnosis has helped considerably and has opened my mind to the times when I had previous attacks/ crisis. Being able to understand why they happened when they did is amazing to me. I am ready and willing to discuss my life with you and your team as to help further the science behind this disorder. Being able to read about this disorder has helped me pinpoint certain aspects that I have always known but now have more clarification on. I would like to help as to possibly give other people an earlier opportunity to ease their pain or at least manage it. I would like to continue to educate myself on this disorder as I'm living with it till the end. I believe you made mention of a possible case study. If you are not the correct person I should be contacting about this please guide me to who I would need to talk to. I again want to thank you for helping me get in touch with what has been happening with my body for so many decades. You have done an amazing job and I appreciate your help with this. Let’s get something started if you are interested in my perspective.
Footnotes
-
Twitter @Nibash_B
-
Contributors NB: manuscript preparation, literature review. ST: manuscript preparation, literature review. BR: pathology review, literature review. DM: manuscript preparation and review.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests None declared.
-
Patient consent for publication Obtained.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2021. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer