Optimal disease management and health monitoring in adults with sickle cell disease

With improvements in health care over the past 3 decades, the majority of children with sickle cell disease (SCD) now survive to adulthood, but significant early mortality remains; the mean life expectancy is still around the fourth to fifth decade. As a result, SCD has evolved into a chronic and debilitating disorder with an accumulation of end-organ dysfunction as patients age. Here, we consider management of the adult patient from the point of transition from pediatric care, with focus on the care of older adults in whom the results of a lifetime of SCD on chronic organ dysfunction interact with the comorbidities of normal physiological aging, leading to complex clinical situations. In some ways, SCD can be considered an accelerated aging process with outcomes of end organ damage (eg, end-stage renal failure [ESRF], neurological impairment, chronic cardiorespiratory disease) that are manifested a decade or so earlier relative to people without SCD. There has been little study of the outcomes of older adults with SCD and, of course, these outcomes may change (and improve) over time as the effects of early treatment (eg, hydroxyurea [HU]) and improved comprehensive care become apparent.

Patient A is a 22-year-old woman with HbSS recently transitioned from pediatric care where she has received comprehensive care since birth. At her first visit in the adult clinic, it was noted that she was started on HU at the age of 6 years and since then has experienced pain crises needing hospital admission about once a year; a silent infarct was detected on a routine magnetic resonance imaging (MRI) brain scan at the age of 7 years; and she has a history of mild asthma, controlled on inhaler therapy, and microalbuminuria. Following this initial review, she was offered regular outpatient follow-ups, including blood monitoring for HU therapy and comprehensive annual reviews. She is worried that she is likely to die early because she has family members with SCD who died in their 40s. She wants to know how to keep well and what treatments are available.

At the first consultation, the patient should be introduced to the adult multidisciplinary team, and the provision of emergency and outpatient care should be discussed. Important issues of management include the development of an acute pain plan (ie, a management plan with the types and dosages of analgesic agents and appropriate adjuvants). The patient carries this plan in paper or electronic form, and it can be shared with the emergency care provider, if needed. Indications for attendance at the Emergency Room (ER) or local Infusion Center (if available) should be discussed; “red-flag” symptoms include severe uncontrolled pain, temperature > 38°C, chest pain, or neurological symptoms. The patient should be given the contact details for the sickle cell specialist team for clinical queries.

The importance of long-term comprehensive care and the role of the outpatient follow-up should be discussed in the context of the multiorgan complications associated with SCD and the screening that will be offered. This first consultation also offers an opportunity to discuss work and education, psychological and welfare support, reproductive issues, and therapeutic options. These discussions can be repeated at subsequent annual visits. It is often helpful to have a clinical pro forma, either paper-based or electronic, to ensure that all elements of the comprehensive review are captured and to guide junior and nonmedical staff in completing the review (Table 1).

Patient A attends intermittently over the next 10 years; she is working full-time as an administrator. She reattends clinic because she has not been feeling so well over the last 12 months, having had 4 hospital admissions with pain crisis. For the past 18 months, she has been having chronic pain in her right hip and worsening shortness of breath on exertion. Further investigations reveal proteinuria of 100 mg/mmol. Radiograph of the hip is normal, but hip MRI reveals avascular necrosis. Oxygen saturations on air are 91%, and she is referred for further investigation.

However mild their symptoms appear to be, all patients with SCD are continuously undergoing some degree of end-organ deterioration. The rate of the development of chronic disease complications is modulated by the SCD genotype (patients with HbSC and HbS/mild β thalassemia exhibit less deterioration than do those with HbSS and HbS/severe β thalassemia), genetic background, lifestyle, including compliance with medication, health care access, and environment.

Pain is the most common symptom for which patients seek medical attention. There are considerable differences in the frequency and intensity of acute sickle pain between adults and children; in children, acute pain is episodic with shorter hospital stays,¹  whereas adults are likely to have more frequent acute pain, and return to the pain-free state takes longer.²  With age, pain becomes increasingly chronic. A daily diary study in adults (16 years and older) showed that 55% reported pain on >50% of the days,³  whereas a similar study in children showed that the patients reported pain on only 8.4% of diary days.⁴  Initially, patients usually try to manage their acute pain at home; however, if pain escalates, attendance at infusion clinics instead of the ER is often preferred because pain relief is provided by specialist providers in a timely fashion, and this can prevent hospital admissions or reduce the length of a hospital stay.⁵  In older adults, differentiating acute pain from chronic pain can be very challenging, and overlap is often seen. The complexity of chronic pain in adults is further illustrated by the persistence of pain, despite successful reversal of sickle hematology after hematopoietic stem cell transplant (HSCT) in a subgroup that had chronic pain pretransplant.⁶  Chronic pain is debilitating, and its management is often frustrating for provider and patient alike; in some cases, opioid-induced hyperalgesia may contribute to the chronic pain.⁷  It is very important that the adult with SCD is educated that chronic pain may become a part of their daily life and that they should work with a multidisciplinary team that includes a pain-management specialist, psychological support, and physiotherapy to develop adequate coping mechanisms.

All patients with SCD suffer some degree of renal dysfunction manifesting as glomerular hyperfiltration and hyposthenuria (the inability to concentrate urine) from early childhood and progressing to microalbuminuria (albumin/creatinine ratio [ACR] persistently > 40 mg/g), which is seen in 20% of children, as well as in >60% of adults older than 46 years.⁸  Renal damage is insidious; at the minimum, renal function should be monitored annually with ACR. Once ACR is persistently >40 mg/g, this should be replaced by the protein/creatinine ratio, which is more accurate at higher levels of proteinuria. Proteinuria and the less common presentation of nephrotic syndrome (seen in ∼4% of patients) are associated with poor outcome and increased mortality.⁹  Recent cohort studies have confirmed a decline in estimated glomerular filtration rate (eGFR) over time in adults with SCD,¹⁰  and this decline is associated with markers of disease severity (eg, anemia, recurrent pain episodes). In a small number of patients (4-12%), renal function progressively declines, leading to ESRF requiring renal-replacement therapy.⁹  Almost 1 in 4 adults older than 60 years of age have chronic kidney disease (CKD) (stage III-IV), and ESRF has been identified as the cause of death in 45%.¹¹  Once patients with SCD develop ESRF and require dialysis, their long-term outcome is poor compared with non-SCD patients on dialysis.¹² 

Further investigation is indicated for those patients with significant proteinuria or worsening renal function (eGFR decrease >10% per year or eGFR < 60 mL/min per 1.73 m²), and this would include immunology for lupus nephritis, virus serology, myeloma screen, and renal tract ultrasound, with subsequent referral to a specialist nephrologist. The decision for renal biopsy should be made after discussion with a renal specialist, and it should take into account the patient’s wishes. Biopsy should be considered if there are any indications that this is not sickle cell nephropathy (SCN).⁸ 

Observational data suggest that the proteinuria in patients with sickle nephropathy responds to angiotensin-converting enzyme inhibitors (ACEis) or angiotensin receptor blockers (ARBs). ACEi may also slow loss of kidney function in SCD,¹³  although this needs confirmation in a randomized trial. Small observational trials suggest that HU may reduce albuminuria, although there is no evidence that it improves long-term outcomes.¹⁴ 

We consider ACEi or ARB treatment when the protein/creatinine ratio is persistently >50 mg/mmol (500 mg/g); it should be introduced cautiously because of the risk for postural hypotension and hyperkalemia, which are both more common in SCD patients. We also discuss HU alongside ACEi/ARB treatment, particularly in those patients with rapidly deteriorating renal function.⁸  Systemic hypertension and diabetes, common comorbidities that can accelerate deterioration of renal impairment, should be rigorously controlled (maintaining a blood pressure < 130/80 mmHg). Worsening anemia is often apparent in patients with CKD as a result of the declining ability of the kidney to synthesize erythropoietin; erythropoiesis-stimulating agents (ESAs) can be used in this situation in combination with HU. Hemoglobin (Hb) target should be <10 g/L to reduce the risk of vaso-occlusion, stroke, and venous thromboembolism. Once patients are approaching stage V CKD, renal-replacement therapy (hemodialysis or peritoneal dialysis) and early referral for transplantation should be discussed with them.⁸  These points are summarized in Table 2.

Bone problems, such as osteopenia/osteoporosis, chronic arthritis, and osteonecrosis (or avascular necrosis [AVN]), are common in SCD; the frequency and severity increase with age.¹⁵  AVN is most common in the hip joints, but it can also occur in the shoulder, knees, feet, and back, can often affect multiple joints, and manifests as intermittent and progressive pain. A review of discharge databases in California found that 22% of patients had a diagnosis of AVN (based on imaging) of the femoral head that was made at a median age of 27 years.¹⁶  Imaging findings ranged from early changes of diffuse sclerosis and cysts to collapse of the femoral head and acetabular osteoarthritic changes. If the plain radiograph is normal, an MRI should be considered because it has the highest sensitivity and specificity for AVN.¹⁷  AVN of the hips is progressive, and >75% of hips with early disease will progress to collapse within 10 years.¹⁸  AVN in nonweight-bearing joints is thought to progress more slowly; a longitudinal study of patients with symptomatic osteonecrosis of the humeral head showed that 86% progressed to collapse, and 61% needed surgery over the 15- to 24-year follow-up.¹⁹ 

Management of AVN is challenging given the limited evidence for standardized guidelines in most surgical procedures in SCD. Early (noncollapsed) disease of the femoral head may respond to conservative treatment with physical therapy, pain management and injection with local anesthetic, although progression is inevitable. A randomized trial showed that core decompression was not superior over physical therapy,²⁰  but this study included a mix of patients with early and radiographic changes; there may be a role for this intervention in carefully selected younger adults with early noncollapsed disease. Once there is evidence of femoral head collapse, prognosis is poor, and the only effective management is total hip arthroplasty (THA). There have been some early studies with injection of stem cells (osteoblasts) into the avascular lesions in the femoral head, but these need further investigation.²¹ 

Early studies showed high rates of failure and revision surgery of THA, but modern techniques with cementless ceramic prosthetic devices and joint management by an experienced surgeon and hematologist have improved outcomes of THA in SCD, prompting earlier intervention.^22,23  Patients should be referred for consideration of THA once they have a collapsed hip with pain impacting on their daily lives regardless of age.

Hypoxia and impaired exercise tolerance are common in adults with SCD and are associated with an increased risk for death; the etiology is multifactorial due to chronic damage to the respiratory and cardiovascular systems.²⁴  With aging, sickle cell patients accumulate a combination of standard categories of pulmonary disease: pulmonary arterial hypertension, chronic thrombotic pulmonary hypertension (PH), pulmonary fibrosis due to parenchymal damage from recurrent acute chest syndrome (ACS), reactive airway disease, and others. “Chronic sickle lung disease” has been used as an umbrella term for these pulmonary complications; however, because they are not unique to SCD, the use of this term should be discontinued.

Airway hyperreactivity is seen in >70% of pediatric patients and is also apparent in adults, albeit at a lower rate. Adults with SCD are more likely to show a restrictive defect on lung function testing, which was apparent in >70% in 1 study, in addition to evidence for impaired gas diffusion.²⁵  These defects worsen with age and are associated with interstitial lung disease and fibrosis or with PH and with a history of ACS. PH (defined as mean pulmonary arterial pressure > 25 mmHg at rest by right heart catheterization), which affects 6% to 11% of adults with SCD, should certainly be considered a cause of hypoxia. Transthoracic echocardiography (ECHO) is a useful noninvasive screening test for PH; a tricuspid regurgitant jet velocity (TRV) > 2.5 ms (present in ∼30% of adults with HbSS) is associated with an increased risk for death and a higher incidence of PH.²⁶  The role of regular routine screening using ECHO in adults with SCD is still not clear; the recent American Society of Hematology (ASH) guideline recommended against screening ECHOs, although clinicians should have a low threshold for ECHO in any patient with symptoms suggestive of PH (dyspnea, chest pain, or exertional hypoxia). Venous thromboembolism occurs in young and older adults with SCD; it is common during episodes of ACS and as a cause of death. Pulmonary thrombosis can cause hypoxia as a result of ventilation-perfusion mismatch and can lead to PH. A recent study found very high rates of pulmonary thrombosis in patients with SCD and PH, as well as that ventilation-perfusion scanning was more effective in identifying pulmonary thrombosis than were computed tomography (CT) pulmonary angiograms.²⁷  An increased incidence of obstructive sleep apnea (OSA) and nocturnal hypoxia as causes of sleep disordered breathing is reported in adults. Routine screening for sleep disordered breathing in the asymptomatic adult is not indicated, but any patient with daytime hypoxia, a history of ACS, priapism, or symptoms of OSA should undergo an overnight sleep study.

For our patient with low saturations and symptoms of dyspnea, investigations should include pulmonary function testing, overnight sleep study, ECHO, and chest CT (consider computer tomography pulmonary angiogram [CTPA] or ventilation-perfusion). She should be referred to a cardiologist/pulmonologist with an interest in SCD or to a joint clinic with a hematologist. A TRV > 2.9 ms on ECHO merits further examination with cardiac catheterization to delineate the cause for further specific treatment. In those with an intermediate TRV of 2.5 to 2.9 ms, the N-terminal pro b-type natriuretic peptide (NT-proBNP) and 6-minute walk distance are useful additional noninvasive tests.²⁶  Those with abnormalities on overnight sleep study should be treated accordingly (eg, continuous positive airway pressure for those with OSA and overnight oxygen therapy for those with nocturnal hypoxia). Additionally, we should consider introducing HU.

HU has been the only available drug treatment for the prevention of complications in SCD for >30 years. Following the outcome of the BABY-HUG study, which treated unselected asymptomatic infants with HU,²⁸  guidelines from the United States and United Kingdom recommend that young children with HbSS should be offered HU, regardless of clinical severity, to reduce sickle cell complications.^29,30  Although there is no comparable randomized trial in asymptomatic adults, HU should be discussed with all adults with HbSS. Key considerations when discussing HU therapy with the patient are summarized in Table 3. We also note that the doses of HU tolerated by patients decrease with increasing age, with an increased risk for myelosuppression, which may develop even in patients on a stable dose of HU for many years. We recommend a starting dose of 15 mg/kg HU in adults (compared with 20 mg/kg in children) and regular monitoring of full blood count.

The majority of randomized trials of long-term blood transfusion involve children rather than adults, but observational evidence for indications for blood transfusion in adults has been reviewed and is summarized in Figure 1.^29,31,32 

When standard therapy (HU, regular blood transfusion, Endari, and supportive care) has failed in limiting disease progression, other interventions, such as fully matched or haploidentical HSCT, gene therapy, or experimental drug therapy, should be considered, even in the older adult with SCD.³³  Although outcomes are poorer in those older than 16 years of age,³⁴  a reduction in the intensity of conditioning and improved peri-transplant care outcomes have expanded allogeneic HSCT as a treatment option for adults with preexisting organ damage.^35,36  The ASH guidelines suggest that HSCT should be considered in adults with a history of stroke, silent stroke, abnormal transcranial doppler, recurrent ACS, or frequent pain if they have an HLA-matched sibling, although there is a low certainty of evidence for these recommendations. For those without a matched sibling donor, data on outcomes using haploidentical donors are more limited.³⁷  Gene addition using the antisickling β-globin vector containing the HbAT87Q mutation (bluebird bio) for SCD is still in its early stages, but preliminary reports have revealed therapeutic expression of the antisickling HbAT87Q with reduction in the frequency of acute sickle pain and an increase in Hb.³⁸ 

Pharmacy-grade l-glutamine (Endari) was approved by the U.S. Food and Drug Administration in July of 2017 for use in SCD patients older than 5 years of age to reduce the frequency of acute pain, and it came at a time when HU has been the sole disease-modifying agent for SCD for >30 years. Its mechanism is unclear, but it is thought to improve the NAD/NADH intracellular ratio, enhancing the redox potential within the red cells. l-Glutamine is also thought to decrease adhesion of sickled red cells to the endothelium and, thus, the likelihood of vasoocclusive crisis (VOC). In the 1 phase 3 study, which included 152 patients in the treatment group and 78 patients in the placebo group, there was a significant reduction in acute pain (25%), hospitalization (30%), and ACS (8 vs 23%) in the treatment group.³⁹  The high drop-out rate (36% l-glutamine arm, 24% placebo arm) makes it difficult to interpret the results. There are concerns about increased mortality with l-glutamine use among critically ill patients⁴⁰ ; therapeutic vigilance is needed if it is prescribed for SCD patients who are at risk for renal impairment, hepatic dysfunction, or multiorgan failure. Endari is supplied in 5-g packets of powder, and the treatment dose is 1 to 3 packets twice a day. Currently, it is reasonable to consider l-glutamine a second option in patients who cannot tolerate or fail to respond to HU.⁴¹  Other agents currently in phase 3 clinical trials include antiselectins, crizanlizumab,⁴²  rivipansel,⁴³  and small molecule antisickling agents (eg, GBT440 [voxelotor]).⁴⁴ 

Patients with SCD can become iron loaded secondary to regular transfusion therapy or with repeated intermittent transfusions. Although liver iron loading should be assessed every 1 to 2 years by liver MRI (R2, T2*, or R2*), compared with serial monitoring of ferritin levels alone in all patients on regular transfusions, the ASH guideline recommends against adding routine screening for cardiac iron content by T2* MRI. However, cardiac MRI should be considered in the subgroup with evidence of very high liver iron (liver iron concentration > 15 mg/g dry weight), evidence of end organ damage due to transfusional iron overload, or evidence of cardiac dysfunction. Automated red cell exchange causes less iron overload than simple transfusion or manual exchange transfusion, and if the ferritin level is <1000 ng/mL, liver MRI screening is unlikely to be needed (ASH guidelines). Patients with an increased liver iron concentration should be treated with iron chelation therapy.

In view of her frequent acute pain episodes, Patient A was restarted on HU. She was started on an ACEi for treatment of her proteinuria and was referred to the orthopedic team for her AVN and consideration of hip replacement surgery. Her ECHO was normal, pulmonary function testing showed a mild restrictive defect, and chest CT showed changes consistent with parenchymal damage secondary to SCD. An overnight sleep study revealed nocturnal hypoxia, and the patient was started on overnight oxygen therapy.

Patient A remained well for the next 10 years. She was considering pregnancy with her partner and asked for advice about conception and pregnancy. At the age of 42 years, she presented acutely with a left hemiparesis and was transferred to an acute stroke unit where she was treated with thrombolytic therapy and, subsequently, an exchange blood transfusion. MRI showed an acute ischemic stroke on the background of silent infarcts and she was referred to a neurologist for discussion of ongoing care.

Pregnant women with SCD are at increased risk for preeclampsia, pregnancy-induced hypertension, thromboembolism, and cesarean section and have increased rates of sickle complications (eg, acute pain crisis, acute chest crisis, infection). They experience increased rates of miscarriage, premature labor, and restriction of fetal growth.⁴⁵  These risks can be reduced by optimization of care prior to conception and scrupulous joint care during the antenatal, perinatal, and postnatal periods. We recommend that women are seen regularly and jointly by an obstetrician and hematologist and, if admitted, should be under the joint care of a specialist obstetrician and hematologist.

Women should be advised to take folic acid prior to conception and should be offered partner screening. If the couple is at risk for having a child with SCD, genetic counseling should be offered, including options for preimplantation genetic diagnosis and invasive prenatal diagnosis. Current evidence does not allow firm recommendations on the role of prophylactic transfusion throughout pregnancy, but certainly transfusion should be offered for symptomatic anemia or acute sickle complications. In view of the increased thrombotic risk, thromboprophylaxis should be prescribed postpartum and considered in the third trimester.⁴⁶ 

Historical data show high rates of acute ischemic stroke in adults (11% by 20 years and 24% by 45 years), but there are fewer data on the rates of new or recurrent strokes in the era of childhood stroke screening and the increasing use of HU. Unlike in the pediatric SCD population, in which strokes are almost invariably due to sickle-related vasculopathy, adults with SCD often have other comorbidities that contribute to stroke risk, including hypertension, diabetes, hyperlipidemia, and renal dysfunction; investigation for these causes is mandatory for any adult SCD patient presenting with stroke. Acute hemorrhagic stroke is also seen at increased rates in adults with SCD, in part because of the high rate of cerebral aneurysms (reported in up to 11% of adults), which can rupture.⁴⁷ 

The main dilemma in acute stroke management in adults with SCD is the role of exchange blood transfusion vs the need for early thrombolysis. It is important that the algorithms for managing acute stroke recognize the relevance of a diagnosis of SCD and consider the need for acute transfusion therapy. In children there is observational evidence that early treatment with exchange transfusion is associated with improved outcomes, and it is probable that these data will also apply to adults with sickle-related stroke. Equally, sickle vasculopathy may not be the cause of acute ischemic thrombotic stroke in adults, and thrombolytic therapy may be more appropriate. A recent study showed that coexistent SCD does not affect the outcome of thrombolytic therapy in adults with SCD⁴⁸  who, therefore, may benefit from both thrombolysis and exchange transfusion. Nevertheless, discussion with a stroke physician is necessary.

Children with SCD who have experienced a stroke are at a high risk for recurrent stroke, and this risk is reduced with long-term transfusion therapy. It is likely that this is also true in adults and that transfusion will be most beneficial in those with no traditional risk factors for stroke and with previous silent infarcts or evidence of sickle vasculopathy on magnetic resonance angiography (MRA). We recommend a discussion among the neurologist, the hematologist, and the patient with outcomes of poststroke work-up to decide on the need for long-term transfusion.

Necessary investigations include imaging with MRI/MRA to look for evidence of sickle-related vasculopathy and previous silent infarcts, ECHO to look for patent foramen ovale (increased incidence in SCD), and traditional stroke risk factors, including lipids, diabetic screening, electrocardiogram.

Silent cerebral infarctions (SCIs) are common in SCD; they increase with increasing age and contribute to the neurocognitive decline.⁴⁹  There is evidence that intervention with regular transfusion therapy in children with SCI prevents progression of silent infarcts or the development of overt stroke,⁵⁰  but there is no evidence that such an intervention prevents progression of neurocognitive decline or overt stroke in adults with SCI. The ASH guidelines recommend that adults have a 1-time MRI screen for SCIs to identify individuals with an increased risk for cognitive impairment and to facilitate occupational therapy and resource input, although there was a low certainty of evidence for this recommendation. If SCIs are found on MRI scan, we recommend regular follow-up scanning and intervention with HU or transfusion. It is prudent to continue with ongoing transfusion therapy if started in childhood; transfusion therapy should also be considered if the SCIs are progressive, and HU cannot be used. We also recommend that adults with evidence of SCI, particularly if progressive, be referred for neuropsychological assessment and additional appropriate support. Clinicians should be asking about memory concerns as part of the comprehensive annual review and referring for MRI scanning and neuropsychological testing, if indicated.⁵¹ 

Further investigations in patient A found no traditional risk factors for stroke; she was normotensive with normal lipid studies, had no evidence of cardiac disease, and her ECHO was normal. She was advised to continue on long-term blood transfusion therapy as secondary stroke prevention. She was monitored regularly for iron overload. Annual brain MRI/MRA was performed to ensure that there was no progression of neurological disease. Patient A remained well for the next decade, being seen regularly for automated exchange blood transfusion, until an acute admission with confusion and acute kidney injury. A provisional diagnosis of “sickle crisis” was made, but subsequent investigation revealed a blood glucose of 32 mmol/L, leading to a new diagnosis of insulin-dependent diabetes. She was treated appropriately and discharged on insulin therapy, as well as the ongoing transfusion therapy.

As adults with SCD grow older, they are likely to have multiple aging comorbidities, adding to the cumulative complications of a chronic disease that affects multiple organ systems. Thus, it is imperative that any coexisting cause is identified accurately and treated appropriately. Given the impairment in multiple organs, an adult with SCD also tends to have a medication regimen that increases in complexity and needs careful consideration and appropriate dosage adjustments. Cognitive dysfunction that increases with age impacts adherence to the prescribed drug regimen and management plans. Key messages in managing the adult with SCD are summarized in Table 4. As more adults with SCD live longer, it is time to shift the traditional care practice of SCD as a single disease to a more patient-centered framework in a multidisciplinary team, with a primary care physician/hematologist coordinating care of the patient. Additionally, due diligence should be paid to routine screening tests for diabetes and cancers (Figure 2).

This work was supported by the Intramural Research Program of the National Heart, Lung, and Blood Institute, National Institutes of Health (S.L.T.).

Contribution: S.L.T. and J.H. cowrote the article.

Swee Lay Thein, Sickle Cell Branch, National Heart, Lung and Blood Institute, The National Institutes of Health, Building 10-CRC, Room 6S241, 10 Center Dr, Bethesda, MD 20892; e-mail:sl.thein@nih.gov.