The five key things you need to know to manage adult patients with sickle cell disease

This article describes five different clinical issues, for which a strong evidence base doesn't exist but which play a role in the daily management of patients with sickle cell disease (SCD). Using case vignettes, important issues in patient management will be presented and discussed. Pain is the major complication of SCD and we will focus on the complexities of chronic pain pathogenesis and management. In addition, the role of neurocognitive dysfunction, renal disease, and venous thromboembolism will be reviewed. Finally, this article will highlight the risks associated with the inappropriate use of red cell transfusions, especially in the setting of an uncomplicated vaso-occlusive crisis.

The most common genotypes of sickle cell disease are hemoglobin SS disease (those who are homozygous for the hemoglobin S mutation), hemoglobin S β⁰-thalassemia which is clinically similar to SS disease and so these two are referred to as sickle cell anemia. Other common genotypes include the compound heterozygotes hemoglobin SC disease and hemoglobin S β⁺-thalassemia.

KJ is a 25-year-old woman with hemoglobin SC disease who has been admitted every month for treatment of vaso-occlusive crisis pain. In addition, she is in the emergency department weekly for pain management. She reports that her pain is in her legs, back, and arms. She is on extended release morphine 60 mg three times a day and uses oxycodone 15 mg for break through pain, and typically takes this 4 times a day. All of her pain medication is provided by a pain management clinic. You see her in the emergency department where she is complaining of 4/10 pain and is insisting on being admitted because of her pain which she tearfully reports never goes away.

Pain, both acute and chronic, is the most common manifestation of SCD, yet the evidence base to manage pain in SCD is lacking. Much of the current literature centers around the prevention and management of acute painful episodes with little if any data on the mechanism or management of chronic pain. It is no wonder then, that both providers and patients are frustrated with the management of pain.¹  Although there remains a paucity of treatments for chronic pain, a better understanding of the pain experience of patients with SCD, as well as the potential mechanisms of this pain may assist providers in relating with their patients in a more productive and satisfying way for both provider and patient.

The pain experience of individuals with SCD was best described by Smith et al²  who examined 232 individuals, 16 years and older who kept daily diaries of their pain. His group analyzed >30 000 days of data and found that people with SCD reported pain in >50% of those days. There were 2 enlightening findings from this study, one is that people with SCD suffer from a lot of pain on a daily basis and the second important finding was that most of the time patients are managing this pain at home. This study showed that patients sought out care for their pain on only 3.5% of days. Data from the Cooperative Study of Sickle Cell Disease demonstrated that chronic pain impairs health status and quality-of-life more than any other disease-related complication.³  Even adults who do not have chronic pain will experience periods of worsened pain for which they do not seek care, nor define as a vaso-occlusive crisis.

A daily diary study done in children, in contrast to the study in adults, showed that in only 8.4% of diary days patients reported sickle pain.⁴  This increasing burden of pain as patients age adds to the difficulties that already exist in the transition from pediatric to adult care. Whereas younger patients may have an episodic pain experience, adults likely will have more frequent pain and rapid return to a pain-free state is not guaranteed. Adults may require adjustment to living with chronic pain and accepting that pain will be a daily part of their lives. It is incumbent on those who care for young patients, especially, to educate them about this and help them come to terms with chronic pain. These conversations can be difficult, but essential. Without adequate coping mechanisms, patients, such as the one described, end up on escalating doses of opioids. The choice to initiate chronic opioid therapy is difficult as one must consider the long-term implications of this therapy. One must take into account the increased dose of opioid this patient may require in 10 years if she is already on a 270 mg/day dose of morphine equivalents, and whether this therapy should be considered futile if her current dose does not relieve her pain. As pain and its management is a constant and integral part of the care of anyone with sickle cell disease, it is essential that hematologists have some understanding of the most common therapies and challenges in the treatment of pain. Because of the complexity of managing pain, in what is typically a young population with chronic pain complicated by episodes of excruciating acute pain, the ideal is to integrate a pain management specialist in to the sickle cell team. Such a specialist can enhance the skills of the entire sickle cell team in managing pain and be relied on to assist with the more complex patients. An approach to this young woman's pain must begin with an understanding of the cause of her pain.

The mechanisms of chronic pain are poorly understood. A number of factors ranging from genetic to behavioral likely regulate the pain response in people with SCD and include interactions between the nervous, endocrine, and immune systems.⁵  The role of the immune system is of particular interest in understanding pain due to SCD and the role of inflammation in acute painful episodes is currently being exploited through clinical trials investigating new treatment modalities. That ongoing inflammation may play a role in chronic pain is suggested by the identification of elevated levels of inflammatory cytokines and substance P, the neuropeptide modulator of inflammation and nociception, in SCD.⁶  Chronic pain resulting from SCD is now hypothesized to be partly a disorder of central sensitization and peripheral neural sensitization. Central sensitization is a complex phenomenon, with contributions from plasticity induced by excitatory amino acids via NMDA receptors⁷  and neural changes induced by inflammatory mediators and other nociceptive chemical messengers. To make things more complicated, our group and others have shown that opioid induced hyperalgesia may contribute to chronic pain via central sensitization.⁸  Using pain testing in the laboratory, our group has shown that individuals with high levels of central sensitization, as measured by Quantitative Sensory Testing, had on average significantly higher clinical pain and more vaso-occlusive crises than those with low levels of central sensitization.⁹  Interestingly, preliminary data from this same laboratory shows that subjects taking hydroxyurea were generally less sensitive to pain testing in the laboratory, exhibiting a profile more similar to healthy controls suggesting the possibility that hydroxyurea might play a role in chronic pain.

In addition to the role central sensitization plays in the pain phenotype of adults with SCD, there are the better understood and explainable causes of chronic pain, the most common of which is avascular necrosis (AVN). Avascular necrosis occurs when the vasculature is occluded by sickled erythrocytes at distal portions of a bone near a joint where collateral circulation is inadequate.¹⁰  The hip joint is the most common site of AVN. The overall prevalence of AVN in SCD is ∼10%, but in people with sickle cell anemia it is ∼50% by age 33. A small randomized study showed no benefit of hip decompression surgery over physical therapy in improving hip function.¹¹  The natural history of AVN is one of increased loss of bone, loss of function, and increased pain leading many patients to require hip replacement. The timing of hip replacement has not been extensively studied in this patient population. Such studies need to take into account the impact that AVN has on quality-of-life and hospital utilization. Several studies have shown an increased prevalence of AVN in patients who are high utilizers of acute care services,^12,13  suggesting that the pain of AVN contributes significantly to morbidity in this disease.

Initial evaluation of pain must begin with an investigation of the cause of the pain, such common pain conditions as degenerative disk disease, osteoarthritis, headache disorders, and others need to be evaluated as each may have very specific management strategies. For those with AVN, physical therapy or surgical intervention may be needed. Back pain, too may be amenable to physical therapy or epidural injections of anesthetic or steroids. Maximal available therapy for the underlying SCD must be instituted such as hydroxyurea or the use of chronic transfusion therapy to decrease crisis frequency. Once identifiable and treatable causes of the pain have been exhausted, pharmacologic and behavioral therapy may be needed.

As there are no studies in SCD of therapies that are used to manage chronic pain in other disorders, such as SNRIs and anticonvulsants, a multimodal approach must be taken following established guidelines, such as those from the Institute for Clinical Systems Improvement¹⁴  or The American Society of Regional Anesthesia and Pain Medicine.¹⁵  Patients with SCD will often need opioid therapy to manage their chronic pain. When providers choose to use opioids, the goals of that treatment must be clearly defined; opioid contracts should be used and frequent reassessments of functional status must be made. When the treatment goals are not met, one must consider the possibility that the patient's pain is opioid refractory and providers should avoid escalating doses of opioids.

Despite the lack of available data to manage chronic pain, there are numerous treatment options available. The first step toward improving outcomes for patients is an understanding of the severity and complexity of the pain and this must be understood by both patient and provider.

JP is a 26-year-old male with sickle cell anemia who comes to clinic accompanied by his mother. You last saw JP a month ago at which time he had been discharged from the hospital after an episode of acute chest syndrome. You had started him on hydroxyurea and he was supposed to return for laboratory tests two weeks later but he didn't have those drawn. He reports during the visit that he hadn't taken the hydroxyurea daily as he thought that he was only supposed to take it when having a painful crisis. In the office, he and his mom begin to fight about him not obtaining his GED and his mom voices concern that he is not really trying. An assessment of JP's neurocognitive status using the Montreal Cognitive Assessment shows that he scores well below published norms.

Patients who struggle to adhere with their provider's recommendations can be very frustrating to care for. Although there may be a number of causes of non-adherence, one major consideration in adults with SCD should be whether neurocognitive dysfunction is playing a role. In a retrospective study of children with SCD (genotypes SS, Sβ-thalassemia and SC), 27% had experienced a stroke and 13% a silent cerebral infarct. Those patients without evidence of CVA functioned normally on tests of cognitive ability and achievement, but children with cerebral vasculopathy had impaired cognitive function and comparatively lower scores on verbal and performance IQ scales.¹⁶  Further research has shown that ∼40% of children with hgb SS disease will have a silent cerebral infarct.¹⁷  Understanding of the impact of silent cerebral infarcts on children's neurocognitive function is a relatively recent discovery; it is therefore highly likely that many adults have had silent events that have been undiagnosed and it would be astonishing if the effect of silent CNS events in childhood did not lead to impaired function in adulthood. It is unknown if the risk of silent infarcts persists in adults or what the consequences of these events might be. There are few studies evaluating neurocognitive function in adults, the most comprehensive of these done by Vichinsky et al¹⁸  was a cross-sectional study of asymptomatic adults with sickle cell anemia (SCA). This study showed that mean WAIS-III performance IQ scores of patients with SCA were significantly lower than that of controls and that 33% of subjects performed >1 SD below the population mean. In this study anemia was associated with poorer neurocognitive function in older patients. The etiology of these findings is unknown. An additional study by this same group found that individuals with SCA exhibited thinner frontal lobe cortex and reduced basal ganglia and thalamus volumes compared with healthy controls. These structural changes were associated with lower Performance IQ.¹⁹ 

Despite the lack of data in adults, recognition that patients with SCD may have significant neurocognitive dysfunction leading to poor educational attainment and difficulties with employment, as well as challenges understanding and complying with medical advice may help in providing tools so that adults can be more productive. Our group has begun screening all patients for neurocognitive dysfunction using the Montreal Cognitive Assessment. This screening tool has been widely used and has a total possible score of 30 points; a score of <26 was used in the original study to identify mild cognitive impairment.²⁰  However, the MoCA performs differently in certain demographic subgroups and a wide range of mean scores has been reported.²¹  In our first 100 patients screened, the mean score was 24.5, with 46% of the population having a score below the original cutoff for mild cognitive impairment of 26, and with 25% scoring a 22 or below.²²  Although much more research needs to be done to validate screening tools, providers should be mindful that the prevalence of neurocognitive dysfunction in adults is likely quite high and often undiagnosed. A low threshold of suspicion for neurocognitive dysfunction should be used for referral for formal neurocognitive testing. One concern is the lack of insurance coverage for testing in adults, so evaluation prior to transition should be considered. Once identified, referral to neurocognitive rehabilitation services should be considered. Lack of access for adults to formal neurocognitive testing and rehabilitation will need to be addressed to meet the needs of this patient population. As a first step, patient, family, and provider understanding of the potential for neurocognitive dysfunction provides knowledge that can help alleviate frustration and move toward productive and informed solutions to medical non-adherence.

ES is a 35-year-old male with sickle cell anemia who presents to the emergency department with excruciating pain in his right first metatarsophalangeal joint. ES rarely has crises and has not seen a hematologist in years. He works fulltime as a lawyer for the State Department and notes that he saw his primary physician for a full physical 3 months prior and was told he was doing well. He knows that his baseline hemoglobin runs around 6.0 gm/dL. On exam, the joint is swollen, erythematous, and tender. Aspiration of synovial fluid reveals intracellular monosodium urate crystals and a serum uric acid level is 10.0 mg/dL. The patient's bun and creatinine are 15 mg/dL and 0.9 mg/dL, respectively. His hemoglobin is 6.0 gm/dL and his absolute reticulocyte count is 200.0 K/cu mm. His urinalysis shows 2+ protein on dipstick.

Gout is usually diagnosed in the 6^th or 7^th decade of life.²³  The likely cause of this patient's gout is renal dysfunction. Early renal dysfunction is often missed when serum creatinine and typical measures of GFR are used to assess function in people with SCD.¹⁰  The earliest renal complication in people with SCD is hyposthenuria, or the inability to concentrate the urine, which progresses with age. In addition to hyposthenuria increased renal tubular secretion of creatinine also starts in childhood.²⁴  Due to this, serum creatinine values in SCD do not rise until significant renal impairment occurs (GFR of 30 mL/min or less).¹⁰  Relying on serum creatinine levels will lead providers to miss significant renal disease. Renal dysfunction in SCD is very common with progression to chronic kidney disease in up to 18% of patients.¹⁰  One feature that should lead a provider to investigate for renal dysfunction is hypertension. Patients with SCD have been noted to have lower mean blood pressures than the general African American population,²⁵  so even mild elevations in blood pressure should trigger an investigation of renal function. Along with the increased secretion of creatinine, patients with SCD will also hypersecrete uric acid.²⁴  The development of gout should therefore lead to a thorough investigation of renal function as it suggests that renal impairment is preventing the hypersecretion of uric acid and causing the increase in serum levels. In addition, this patient may be at greater risk for renal disease and gout because of his high red cell turnover as reflected by his high reticulocyte count. Although he has had high red cell turnover his whole life, he now has gout; this suggests that in addition to increased production of uric acid, there is now an issue with renal secretion. In a review of 90 patients who were enrolled in a Natural History protocol at the Clinical Center of the National Institutes of Health, 13 had gout. Those patients with gout had significantly higher creatinine levels than those without gout.²⁶ 

As using creatinine to screen for renal dysfunction in this population will miss significant renal impairment, current recommendations from the NHLBI guidelines are to screen annually with routine urinalysis to evaluate for proteinuria starting at the age of 10 years. If the result is positive, the guidelines recommend performing a first morning void urine albumin-creatinine ratio and if abnormal, consult with a renal specialist. It is important to remember that the standard calculations of glomerular filtration rate can't be used reliably in patients with SCD and in the acute setting, an increase of 0.3 mg/dL in creatinine should prompt an avoidance of nephrotoxic agents.¹⁰ 

In this patient who is discovered to have significant proteinuria on urine dipstick, further investigation can be done by sending a urine albumin-creatinine ratio or a 24 hour urine test for protein. One study that examined the accuracy of spot urine albumin-creatinine (A/C) ratios in people with SCD showed that the correlation between urine albumin-creatinine ratio with 12 hour urine collections was poor; however, they did show that values of A/C ≥0.45 and <0.45 were indicative of raised and normal urine albumin excretion, respectively.²⁷ 

Once a patient is found to have micro or macroalbuminuria without any other known cause, the NHLBI guidelines recommend initiation of ACE-inhibitor therapy. This recommendation is based on several trials including a double-blind, placebo-controlled randomized trial of 22 adults with sickle cell anemia and microalbuminuria that found that captopril for 6 months significantly reduced albuminuria.²⁸  Two other observational trials demonstrated similar reductions in proteinuria with ACE inhibition. The NHLBI guidelines recommend initiation of ACE therapy even when patients are normotensive. Finally, in SCD patients who require renal replacement therapy, renal transplant should be considered. In a recent study using the Organ Procurement and Transplanting Network/United Network for Organ Sharing data, investigators showed improvements in 6 year kidney transplant outcomes in recent years (2000-2011) compared to prior years (1988-1999) with results similar to those seen for patients with diabetes who received transplants.²⁹ 

RN is a 22-year-old with hemoglobin SC disease who presents to the ED with SOB and chest pain. She reports that she was studying for her physics final and had been spending lots of time sitting in the library. She feels the pain is different from her typical vaso-occlusive pain. A CT scan of the chest reveals bilateral pulmonary emboli. She is not on birth control and does not smoke. You confirm that she is not pregnant and start her on enoxaparin.

SCD has been considered a hypercoagulable state for many years.³⁰  Activation of the coagulation cascade, increased platelet activation and impaired fibrinolysis have all been described.³⁰  The etiology of these changes is unclear. Some postulate that nitric oxide consumption from free plasma hemoglobin resulting in platelet activation plays a role, whereas other investigations have shown that abnormal phosphatidylserine exposure on the surface of sickle red cell membranes provides a site for procoagulant activity.³⁰  Several studies show elevation in a number of measures of activation of the coagulation cascade. Of note is the finding that D-dimer elevation has been seen in a number of studies with D-dimer levels being elevated in patients at steady state (not in VOC) compared with healthy controls.^31-33  Therefore, the usual algorithms based on the use of elevated D-dimer levels to diagnosis venous thromboembolism (VTE) cannot be applied to this patient population.

Recent studies have shown that the prevalence of VTE in this population is quite high. In a cross-sectional study of 404 adults with SCD the prevalence of VTE was 25%, with patients with SC or Sbeta+thalassemia having significantly higher rates of non-catheter-related thrombosis and pulmonary embolism compared to individuals with sickle cell anemia.³⁴  In this study with a median follow-up of 4.8 years after initial VTE, 24.8% of patients had a recurrent event. On multivariable analysis, patients with a history of non-catheter-related VTE were 3.6 times more likely to die than those without a history of VTE. This data suggests that the risk of thromboembolism in adults with SCD is similar to that seen in high-risk thrombophilic states, such as Protein C and S deficiency. In a study looking at the incidence of VTE in the Cooperative Study of Sickle Cell Disease, a similar increased risk of mortality was seen in those with thrombosis compared to those without.³⁵  The incidence rate in that cohort was 5.2 events/1000 patient years with a cumulative incidence of 11.3% by age 40.

Not surprisingly, this increased risk of VTE is also seen in pregnancy. When examining the risk of thromboembolism in pregnancy, a retrospective review of pregnancy outcomes in African Americans showed that 2.9% of women with SCD had thrombosis compared with 0.9% in women without SCD.³⁶  This study is consistent with a number of others that show an increased risk of VTE in pregnancy that is 1.4-6.7-fold greater than those without SCD.^37-39  This has led some to suggest the possibility that all women with SCD should receive thromboprophylaxis³⁹  during pregnancy, but this has yet to be studied and is not common practice.

Treatment for VTE for those with SCD is the same as in the general population. Questions arise as to duration of therapy once an individual has had an event and there is no data that has examined whether changes from standard guidelines should be made in this population. With at least one study showing a 25% recurrence rate, consideration for prolonged therapy should be made on an individual basis. For example, a patient with frequent hospitalizations may be at higher risk for recurrence than a patient who rarely is admitted. Similarly, as in the general population, the use of long-term anticoagulation in a patient who has had a life threatening thromboembolic event, such as a pulmonary embolism should be strongly considered as it is now clear that SCD is an independent, non-modifiable risk factor. As mentioned above baseline elevations in D-dimer levels make the use of this test for diagnosis, as well as for predicting recurrence difficult and unreliable. Finally, the use of novel oral anticoagulants in this patient population has not been studied. There is no reason to suspect that they will be less effective in individuals with SCD and their ease of use makes them appealing in a younger population who may struggle with adherence to warfarin.

KK is a 40-year-old with sickle cell anemia who is admitted for her typical vaso-occlusive crisis. She reports that since she started hydroxyurea 7 years ago she rarely has crises. She had a flood in her basement and was standing in the cold water cleaning up and she thinks this triggered her crisis, which is causing severe pain in her legs. She denies shortness of breath or chest pain and has not had a recent fever. Her exam is remarkable only for a heart rate of 110, her legs are tender without swelling, and her pulse oximeter reading is 96% on room air. Her laboratory tests show hemoglobin at 5.6 gm/dL with an absolute reticulocyte count of 300 000 K. Her internist has called you very concerned about her dropping hemoglobin and would like to transfuse.

This patient is having her typical vaso-occlusive crisis with associated increase in hemolysis. From the standpoint of her anemia, she is asymptomatic without shortness of breath and a normal pulse oximeter reading on room air and has a brisk reticulocyte response. The NHLBI guidelines recommend against transfusion in a patient with acute vaso-occlusive crisis who is asymptomatic.¹⁰  This recommendation is based on a lack of any data to suggest that transfusion will decrease the length of the crisis or provide any beneficial effect and is weighed against the many potential and sometimes life threatening complications of transfusion.

Although screening for infections, such as HIV and hepatitis, have greatly increased the safety of the blood supply, emerging organisms, such as babesiosis in the US and Trypanosoma cruzi in Latin America, have become more of a concern.⁴⁰  Although these infectious complications are rare, other risks such as alloimmunization and iron overload can contribute to significant morbidity in SCD.

The prevalence of alloimmunization has been reported to be as high as 85%.¹⁰  Phenotype matching for the most commonly implicated antigens (usually of the Rh blood system; D, Cc, Ee) has been shown to decrease the rate of alloimmunization to 0%-7%,¹⁰  but concerns over the cost effectiveness of both limited and extensive phenotypic matching have been raised.⁴¹  Alloimmunization becomes a problem when it is difficult to find blood products for patients or when patients have delayed hemolytic transfusion reactions (DHTR). DHTRs are characterized by both extravascular and intravascular hemolysis with shortened red blood cell survival, worsening anemia, and increased titers of allo and autoantibodies. When a DHTR is complicated by hyperhemolysis, patients hemolyze not only transfused cells but their own red cells as well; this is often associated with reticulocytopenia and severe life-threatening anemia. Further transfusion of even fully phenotypically matched cells is not helpful and can cause further hemolysis. In our case series (unpublished) over 19 years, we saw 19 patients with 27 episodes of hyperhemolysis. Transfusion of compatible blood during the episode worsened hemolysis in 12/17 (70.6%) cases and 4 patients died. Treatment with IVIG and steroids is recommended.⁴²  The mechanism of hyperhemolysis is unclear; autoimmune bystander hemolysis, suppression of erythropoiesis, and phosphatidylserine exposure on the surface of red cells leading to physiologic clearance by macrophages have all been implicated.^42,43 

Iron overload is another complication associated with transfusion. The prevalence of this complication has not been extensively studied in people with SCD. One single institution study described 156 adults not on chronic transfusion therapy and found that iron overload (defined as a serum ferritin ≥1500) was seen in 24% of patients with sickle cell anemia.⁴⁴  In another cohort of 254 patients followed at 3 centers, 37% were on iron chelation therapy; only 56% of these patients were on a chronic transfusion regimen.⁴⁵  These data suggest that iron overload occurs in a significant number of patients who are not on chronic transfusion therapy and who therefore may not be monitored or evaluated on a regular basis for iron overload. Frequent hospitalizations for crisis with worsening anemia might result in frequent and inappropriate transfusions, which ultimately may lead to iron overload.

The best way to decrease the risk of DHTR, hyperhemolysis, and iron overload is to avoid transfusions unless there is a clear indication for their use. NHLBI guidelines recommend matching for C, E, and K antigens for anyone with SCD requiring transfusion. It should also be standard of care for providers to ask patients where they have received red cell transfusions in the past and for the provider to communicate this to their transfusion service so that they can obtain any history of prior antibodies. The NHLBI guidelines provide recommendations for the appropriate use of red cell transfusions.^10,46 

This article describes 5 different clinical issues which play a role in the daily management of patients with SCD. Although the recently published NHLBI guidelines serve as an excellent resource to assist primary care providers and others in achieving the goal of providing high quality care for this patient population, it is limited by a lack of a strong evidence base. Therefore, as most patients do not have access to specialized sickle cell centers it is imperative that all hematologists have the knowledge and are comfortable with managing the complications of this disease.

Sophie Lanzkron, Johns Hopkins School of Medicine, Division of Hematology, 1830 E. Monument St, Suite 7300, Baltimore, MD 21205; Phone: 410-502-8642; Fax: 410-614-8601; e-mail: slanzkr@jhmi.edu.