Acute and chronic bone complications of sickle cell disease

INTRODUCTION — Orthopedic complications of sickle cell disease (SCD) include vaso-occlusive bone pain, osteonecrosis, and infections (osteomyelitis and septic arthritis). Individuals with SCD are functionally asplenic and are at risk for infections that may be life-threatening. Other noninfectious bone and joint complications can cause severe pain and immobility that significantly interfere with functioning and quality of life.

These complications can be challenging to distinguish, especially in individuals with chronic vaso-occlusive pain. However, it is critical to make the correct diagnosis to provide appropriate treatment.

This topic discusses the orthopedic complications of SCD, including our approaches to distinguishing among them and determining the optimal management strategy.

Separate topic reviews discuss related management issues:

●Routine care (general pediatrician) – (See "Sickle cell disease in infancy and childhood: Routine health care maintenance and anticipatory guidance".)

●Overview of specialist management – (See "Overview of the management and prognosis of sickle cell disease".)

●Transition from pediatric to adult care – (See "Sickle cell disease (SCD) in adolescents and young adults (AYA): Transition from pediatric to adult care".)

●Evaluation of acute pain – (See "Evaluation of acute pain in sickle cell disease".)

●Treatment of pain – (See "Acute vaso-occlusive pain management in sickle cell disease".)

●Use of hydroxyurea – (See "Hydroxyurea use in sickle cell disease".)

●Use of transfusions – (See "Red blood cell transfusion in sickle cell disease: Indications and transfusion techniques".)

●Hematopoietic stem cell transplant – (See "Hematopoietic stem cell transplantation in sickle cell disease".)

●New therapies under development – (See "Investigational therapies for sickle cell disease".)

CAUSES OF BONE OR JOINT PAIN — Bone or joint pain can be caused by vaso-occlusion, avascular necrosis of bone and associated fractures, or infection (eg, osteomyelitis, septic arthritis). These causes and our approach to distinguishing among them are discussed in the following sections.

Like other disease manifestations in SCD, orthopedic complications are more frequent and tend to present at an earlier age in individuals with more severe genotypes (homozygous hemoglobin SS [Hb SS] or sickle beta⁰ thalassemia) than in those with less severe genotypes (hemoglobin SC [Hb SC] disease or sickle beta⁺ thalassemia). However, exceptions may occur. (See "Overview of compound sickle cell syndromes".)

Acute vaso-occlusive pain — Acute vaso-occlusive pain episodes are the hallmark of SCD and one of the most common manifestations requiring medical attention. Pain also accounts for the majority of hospitalizations in people with SCD. Pain can present from infancy (often as dactylitis) throughout life. Acute pain can occur as often as daily or weekly, or there may be long periods without any pain events.

The mechanisms of acute vaso-occlusion and the evaluation and management of acute and chronic vaso-occlusive pain are summarized in the table (table 1) and discussed in detail separately. (See "Pathophysiology of sickle cell disease" and "Evaluation of acute pain in sickle cell disease" and "Acute vaso-occlusive pain management in sickle cell disease".)

Acute vaso-occlusion can predispose people to develop other orthopedic complications because the loss of the normal bone marrow architecture (due to the replacement of fat with hematopoietic cells needed to compensate for chronic hemolytic anemia) can lead to cortical thinning and bony collapse [1].

Pain in long bones — The long bones are the most commonly affected site for acute vaso-occlusive pain, which is likely caused by ischemic pain, as illustrated in the figure (figure 1). However, virtually any bony structure containing red (erythropoietic) marrow, including the ribs, sternum, vertebral bodies, and skull, can be affected [2]. In SCD, the red marrow is distributed throughout a greater portion of the skeleton than in individuals without SCD; chronic hemolytic anemia puts greater stress on the bone marrow to produce red blood cells (RBCs).

Pain in the ribs and sternum can also lead to splinting and hypoventilation, which can contribute to acute chest syndrome. (See "Acute chest syndrome (ACS) in sickle cell disease (adults and children)".)

Over time, repeated episodes of vaso-occlusion can lead to ischemia-reperfusion injury, inflammatory responses, bone infarcts, necrosis, degenerative changes in marrow-containing bone, and other complications such as avascular necrosis of the femoral heads or collapsed vertebral bodies, in turn leading to a chronic state of pain in addition to the more acute painful episodes. (See "Disease-modifying therapies to prevent pain and other complications of sickle cell disease", section on 'Chronic pain'.)

In addition to causing pain, expansion of the bone marrow due to chronic hemolysis and increased erythropoiesis can also lead to osteopenia and ischemia, which can affect growth and development.

A role for vitamin D deficiency in bone pain has also been suggested. (See 'Vitamin D deficiency' below.)

Dactylitis and other pain sites — Dactylitis (also called hand-foot syndrome) refers to SCD-related vaso-occlusive pain in the small bones of hands and feet; this is typically seen in infants and very young children [3]. As many as 45 percent of all children with SCD will have an episode of dactylitis by the time they reach two years of age, especially those with more severe genotypes (homozygous hemoglobin SS or sickle beta⁰ thalassemia) [4]. Dactylitis is less likely to be seen in children older than four, presumably because, as the child grows older, the red marrow in these small bones is replaced by fat and fibrous tissue.

Dactylitis presents with a painful, often symmetrical swelling of hands or feet accompanied by mild erythema and low-grade fever. Recurrent episodes of dactylitis lead to a mottled radiographic appearance of the small bones of the hands or feet. In most cases, the radiologic abnormalities disappear over time, although a permanent shortening of the involved digits may occur.

Other sites of vaso-occlusive pain include the bones of the skull and vertebral column. Like changes in the long bones and digits, vaso-occlusion in these other sites can be severe enough to cause bone infarction. (See 'Other sites of bone infarcts' below.)

Osteomyelitis and septic arthritis — Bacterial infections of bone and joints are common in children and adults with SCD. These complications may be more likely in children because the unique vasculature in immature bone is conducive to hematogenous spread, but adults can also be affected. The early loss of splenic function and chronic bone and bone marrow infarction in SCD may be predisposing factors to this increased prevalence of infection [5-8].

Presentations of acute and chronic osteomyelitis can vary significantly and can be difficult to distinguish from an acute vaso-occlusive pain episode [9]. Typical presentations include vague bone pain, inability to ambulate, soft tissue edema, skin erythema, and point tenderness to palpation. In more chronic settings, draining sinus tracts can develop. Magnetic resonance imaging (MRI) findings in osteomyelitis are shown in the figure (image 1). The image on the left (panel A) is from an eight-year-old boy who presented with fever and lower back pain that subsequently localized to the right femur, leading to assessment for osteomyelitis. The image on the right (panel B) is from a three-year-old boy who presented with fever and diffuse pain in the extremities that subsequently localized to the left upper leg. Cultures of the abscess aspirate and bone biopsy on day 5 (after initiating antibiotics) were negative.

In a meta-analysis of 227 articles focused on managing infected joints in children, no single laboratory value could identify which had septic arthritis [10]. In a retrospective analysis of 358 episodes in children with SCD who were admitted to the hospital with acute bone pain, osteomyelitis occurred in five cases (1.4 percent; two acute and three chronic osteomyelitis cases) [9]. The only laboratory value that helped distinguish osteomyelitis from acute vaso-occlusive pain was elevated C-reactive protein (CRP). However, interpretation of CRP can be challenging in individuals with SCD due to the chronic inflammation associated with SCD.

Chronic osteomyelitis in children with SCD is a major management challenge in sub-Saharan Africa. Optimal management is unclear, and care is hampered by limited resources (absence of clean water, limited antimicrobial therapy [oral, intravenous, or both], and scarce equipment and personnel for intraoperative debridement).

Despite the relatively high prevalence of chronic osteomyelitis in a busy pediatric SCD practice in Nigeria, few risk factors could be elucidated to guide optimal medical and surgical treatment. Given that the most common pathogen in a low-income setting is Staphylococcus aureus and prolonged intravenous therapy is impractical, new strategies for combined intraoperative debridement, coupled with brief antibiotics, should be considered after proper surgical training and systematic assessment of clinical outcomes [11].

The organisms responsible for osteomyelitis vary by geographic location:

●United States and Europe – In a large multinational meta-analysis of osteomyelitis in SCD, Salmonella accounted for 70 percent of bacterial isolates in the United States and 64 percent in Europe; Staphylococcus aureus accounted for 16.4 percent in the United States and 4.9 percent in Europe [12]. Other case series have also demonstrated that S. aureus, the most common cause of osteomyelitis in individuals without SCD, accounts for less than one-fourth of osteomyelitis cases in individuals with SCD [3,13-19]. Other implicated organisms include Enterococcus faecium, Enterobacter cloacae, Escherichia coli, and Pseudomonas aeruginosa [20-22].

●Sub-Saharan Africa and the Middle East – In the meta-analysis cited above, 38.5 percent of isolates in Nigeria were positive for S. aureus, 21.4 percent for Salmonella, and 34.2 percent for other Gram-negative species [12]. In Saudi Arabia, 62.1 percent of isolates were positive for S. aureus and 37.9 percent for Salmonella.

Hematogenous spread is more common than nonhematogenous spread. (See "Nonvertebral osteomyelitis in adults: Clinical manifestations and diagnosis", section on 'Hematogenous osteomyelitis'.)

Septic arthritis is less common than osteomyelitis in individuals with SCD. In a prospective study of 14 individuals with SCD and septic arthritis, S. aureus was found to be the infecting organism in 11 and Salmonella in only two [23]. Other authors did not find any predominant organism [15].

Osteonecrosis (avascular necrosis) — Osteonecrosis (also called avascular necrosis [AVN]) is a well-described complication of vaso-occlusion that affects as many as 10 percent of individuals with SCD [1].

The common pathophysiology for all forms of AVN is bone ischemia due to occlusion of blood vessel lumens, with subsequent microfractures, collapse of cancellous bone, and ultimately, collapse of the articular surface. (See "Clinical manifestations and diagnosis of osteonecrosis (avascular necrosis of bone)".)

A unique feature of AVN in SCD is the involvement of the entire epiphysis; in most other conditions associated with AVN, only weight-bearing areas are affected. Numerous mechanisms contribute to vascular occlusion in SCD, including recurrent acute vaso-occlusive pain episodes, hypercoagulability, vessel wall hyperplasia, marrow edema and hyperplasia, and fat emboli. (See "Pathophysiology of sickle cell disease".)

The epiphyses of long bones, such as in the femoral and humeral heads, the femoral condyles, and the distal end of the tibia, are common sites of AVN. The femoral heads are the most commonly affected sites due to limited collateral circulation in these areas [24-26].

Risk factors for AVN — The prevalence of avascular necrosis (AVN) increases with age. Of 2590 individuals enrolled in the Cooperative Study of SCD (CSSCD) and followed for an average of 5.6 years, 10 percent had AVN at entry [27].

People with Hb SS and concomitant alpha thalassemia were at highest risk (4.5 cases per 100 person-years, compared with 2.4 in those with Hb SS without concomitant alpha thalassemia and 1.9 cases per 100 person-years in those with Hb SC disease) [27]. As with other SCD complications, high fetal hemoglobin (Hb F) levels appear to be somewhat protective [28,29].

Other known risk factors for femoral AVN are [24,30,31]:

●Male sex

●High body mass index

●Low bone mineral density

●Leukopenia

●Recurrent vaso-occlusive pain

●Acute chest syndrome episodes

Risk factors for femoral AVN in the general population also increase the risk in children and adults with SCD, such as:

●Systemic glucocorticoid use

●Alcohol use

●Smoking

●Hyperlipidemia

●Cardiovascular disease

●Joint trauma

However, few systematic studies have been done to document the suspected risk factors.

In a large administrative dataset from the state of California that included 6237 individuals, 1356 (22 percent) developed symptomatic AVN [30]. Age was a strong risk factor for AVN; 75 percent of those with AVN were ≥30 years of age. The most common reasons for 30-day readmission in this dataset were surgery followed by acute chest syndrome.

Clinical presentation of AVN — In the prospective CSSCD, approximately 50 percent of the participants with avascular necrosis (AVN) were identified radiographically and were asymptomatic [27]. When symptomatic, the most common sites of involvement are the femoral head (hip) and the humeral head (shoulder); the area above the knee is also frequently affected.

There are case reports of AVN affecting other joints and bony structures, including the mandibular condyle, the temporomandibular joint, and the elbow. However, the prevalence of involvement of these sites is still being determined [32,33].

AVN generally progresses through a series of stages (image 2). Initial stages may be asymptomatic and may be identified incidentally in an imaging study. In other cases, AVN comes to medical attention when medullary infarcts occur in the diaphysis of a long bone and cause pain or when these infarcts lead to necrosis in the epiphysis, which can cause the subchondral bone to collapse [1]. These events cause severe pain and decreased mobility.

Stages of AVN — Various stages of AVN are used for the histological or radiographic definition of AVN.

Hip AVN can also cause back, buttock, groin, or knee pain and should be considered when patients with SCD present with chronic pain or recurrent pain in the same area that occurs over weeks to months.

AVN often affects more than one joint in the same individual. In one study, hip involvement was bilateral in 54 percent [27]. These observations support evaluating for bilateral hip involvement when treating AVN of the hip. (See 'Management of avascular necrosis' below.)

Physical examination is helpful in the diagnosis of AVN. In general, decreased range of motion and pain associated with ambulation or passive range of motion is consistent with AVN, especially if periarticular. Some individuals with AVN of the hip may have decreased ability to abduct the hip relative to the contralateral hip (assuming only one hip is affected), pain with internal or external rotation of the hip, or an antalgic gait (gait that favors weight-bearing on the unaffected side). However, these findings are not specific to AVN and cannot be used to distinguish AVN from other causes of bone or joint pain.

Scoring systems and biomarkers — There are several proposed classification and scoring systems for femoral head AVN, among them the Ficat and Arlet, Steinberg, the Association Research Circulation Osseous (ARCO), and the Japanese Investigation Committee (JIC) systems [34]. All these systems rely on radiographic findings variably incorporating plain radiographs, bone scans, magnetic resonance imaging (MRI), and computed tomography (CT). (See 'Imaging' below.)

Key elements of all these systems are the extent of the femoral head involvement, the extent of cystic and sclerotic changes, subchondral collapse, flattening of the femoral head, and extent of joint narrowing. The most commonly used systems are the Ficat and Arlet and Steinberg systems, but there is no consensus on a unified system with high inter-rater reliability and prognostic value [34,35].

The Children's Hospital Oakland Hip Evaluation Scale (CHOHES) is a derivative of the Harris Hip Score that has been modified to specifically evaluate hip disease in individuals with SCD, facilitating diagnosis of femoral head AVN [36] (see 'Femoral head' below). The CHOHES scale has three domains: a patient questionnaire, a functional assessment, and a physical examination. In a validation study, scores on the scale correlated significantly with Ficat scores from plain radiographs with strong intra- and inter-rater agreement. One limitation in the broader use of this scale is that the physical examination requires a detailed assessment of range of motion and would be best performed by clinicians or physical therapists trained in using the scale.

Given that the development of AVN is challenging to predict and difficult to diagnose with a plain radiograph in its early stages, biomarkers for early identification of bone disease have been sought. However, no clinically available biomarker can identify the subgroup of children or adults likely to have AVN.

Evaluation for AVN — We do not screen individuals with SCD for avascular necrosis (AVN) if they are asymptomatic. In contrast, all people with SCD (children, adolescents, or adults) who have intermittent or chronic hip pain or unexplained and recurrent lower back, pelvic, or knee pain should be evaluated for AVN, starting with history and physical examination and incorporating radiography and MRI as indicated, as described in a 2014 SCD guideline from the National Heart, Lung, and Blood Institute (NHLBI) [37]. This is discussed in more detail below. (See 'Role of screening for AVN' below and 'Pretreatment evaluation' below.)

AVN in specific sites

Femoral head — The femoral head (hip) is the most common site of avascular necrosis (AVN) in individuals with SCD. AVN of the femoral head can develop at any age (as early as five years of age in children). It generally presents with hip pain upon weight-bearing. However, early disease may be asymptomatic and may be diagnosed during the evaluation of the contralateral hip.

In a cohort of 6237 individuals in a discharge database from California covering 1991 to 2013, 1356 (22 percent) had femoral head osteonecrosis [30]. The median age of diagnosis was 27 years, and 23 percent of the patients had hip replacement surgery at a median age of 36 years.

Plain radiographs and MRIs are used to diagnose and stage AVN, as discussed below. (See 'Imaging' below and 'Pretreatment evaluation' below.)

Early signs and symptoms tend to be followed by progressive pain, decreased mobility, and eventual collapse, often within approximately three years [1]. The immature epiphysis can undergo remodeling despite initial necrosis and subsequent head femoral head flattening. However, AVN of the mature femoral head leads to collapse, resulting in persistent degenerative changes and pain [38].

Imaging typically shows characteristic changes in the bones and surrounding tissues, including flattening of the femoral head, the abnormal heterogeneous signal in the femoral head, areas of sclerosis, and surrounding edema (image 3) [39]. MRI images from the initial time point in the same patient (image 4) depict numerous bone infarctions and a subchondral fracture that was not apparent on the radiographs.

The natural course of AVN of the hip involves progression over time that may extend from childhood into adulthood, as illustrated in a series of radiographs from a single patient over approximately one year (image 2).

The following studies illustrate the typical presentation and natural history of AVN:

●In a study from 1991 involving 52 individuals diagnosed with AVN of the hip in childhood and followed for an average of 19 years, nearly all of the patients had bilateral involvement, with 80 percent of affected hips showing permanent damage (eg, decreased mobility, abnormal gait, and limb-length discrepancies) and causing pain [40]. The first symptom of pain had been noted at an average age of 12 years (range 7 to 15 years), and 15 of the 95 affected hips required a surgical procedure at an average of 30 years after the onset of AVN (range 18 to 32 years).

●In a study from 1991 in which 2590 individuals with SCD (primarily children and adolescents) underwent surveillance imaging with plain radiographs of the hip over a three-year interval, nearly 10 percent had evidence of AVN upon study entry [27]. Approximately one-half of the individuals diagnosed with AVN based on radiographic changes during the study had no pain or limitations to range of motion; one-fifth of these individuals subsequently developed pain or limited movement later.

●In a study from 2006, 121 individuals with SCD who had symptomatic AVN in one hip underwent imaging (radiographs and MRI) of the other hip [41]. During a mean follow-up of 14 years, 47 of 56 hips initially free of changes on plain radiographs or MRI developed symptomatic (ie, painful) AVN. Approximately one-half were diagnosed incidentally, and one-half due to the development of pain. Symptoms (pain or reduced mobility) always preceded collapse. There was an average interval of approximately three to five years until pain onset and an average of an additional five years between pain onset and collapse. By the final evaluation, 91 of 121 hips (75 percent) had intractable pain and required surgery.

●In a small case series of 16 pediatric patients with femoral head AVN, five patients showed spontaneous improvement in the degree of osteonecrosis while the other 11 had disease progression [42]. Physical therapy and core decompression, used in five patients each, had no impact on the progression of the disease. Patients who had spontaneous improvement did not differ from those who did not in Steinberg score at diagnosis, hemoglobin levels, or Hb F levels at diagnosis but were significantly younger, with a mean age of 9.9 years compared with 14.6 years in those who did not have spontaneous improvement.

Early diagnosis and intervention may help preserve function and avoid hip arthroplasty, which has a higher-than-average rate of complications in SCD, especially in individuals who are still growing. Management is discussed below. (See 'Management of avascular necrosis' below.)

Humeral head — The humeral head (shoulder) is the second-most common site for AVN in people with SCD (image 5). Like AVN of the hip, AVN of the shoulder also tends to progress from asymptomatic radiographic changes to pain and reduced mobility to collapse.

●In a prospective study of 2524 individuals with SCD, 5.6 percent had radiographic evidence of osteonecrosis in one or both shoulders at study entry [43]. Only one-fifth of the affected individuals had pain or decreased range of motion. The highest age-adjusted incidence was found in those with Hb SS and concomitant alpha thalassemia and those with Hb S beta⁰ thalassemia (4.8 per 100 person-years). Of interest, none of the individuals with Hb S beta⁺ thalassemia younger than age 25 had humeral head osteonecrosis on study entry.

●In a study of 82 adults with SCD who had symptomatic AVN of the shoulder, the collapse occurred in 86 percent over 20 years of follow-up [44]. The median interval between the onset of pain and collapse was eight years (6 months to 17 years).

As with the femoral head, early diagnosis and intervention may help to preserve function. Management is discussed below. (See 'Management of avascular necrosis' below.)

Other sites of bone infarcts — Bony infarcts can occur in unusual locations, including the parts of the skull, such as the walls of the orbits (image 6). Orbital infarction is characterized by acute periorbital pain and swelling and can be accompanied by the formation of subperiosteal or intracranial hematomas. Inflammation associated with infarction of the bone within the limited bony space of the orbit can result in an orbital compression syndrome characterized by proptosis, limitation of extraocular movement, corneal hyperesthesia, and optic nerve dysfunction. Bone infarctions in other sites of the face and skull have also been described [2,45-49].

The vertebrae are prone to changes induced by bone marrow hyperplasia, osteopenia, vaso-occlusive pain episodes, and AVN with vertebral collapse (image 7). A cohort study of 34 children and young adults with SCD revealed structural changes due to bone marrow hyperplasia in 44 percent and AVN with collapse of the vertebral bodies in 27 percent [50]. Individuals with SCD are also prone to decreased bone mineral density in the lumbar spine [51]. Vertebral bodies often have characteristic findings, including "tower vertebrae" (ie, vertebrae with compensatory elongation located next to infarcted short vertebrae [52]) and "fish vertebrae" (biconcave vertebral bodies formed by ischemia of the central growth plate of the vertebral body [53,54]). These changes can result in severe chronic back pain.

DETERMINING THE CAUSE OF BONE OR JOINT PAIN

Initial evaluation

Clinical distinguishing features — The cause of bone or joint pain in individuals with SCD can be challenging to determine; this is because presenting findings are often similar, and in some cases, more than one cause is present. Vaso-occlusive pain, bone infarction, and osteomyelitis share many clinical features, including pain, redness, edema, and limited range of motion. As a result, patients are sometimes treated for all possible causes while further evaluations are performed. All things being equal, acute pain episodes are much more common than osteomyelitis. Regardless of the suspected cause and any delays caused by additional testing, all individuals should receive adequate pain control without delay. (See 'Pain control' below.)

Routine laboratory testing in an individual with SCD who presents with pain includes a complete blood count (CBC) and reticulocyte count to assess for the stability of the hemoglobin level and compensatory reticulocytosis, as well as leukopenia or leukocytosis indicative of infection, as discussed in more detail separately. (See "Evaluation of acute pain in sickle cell disease".)

The optimal diagnostic approach to distinguish between vaso-occlusive pain, infarction, and infection remains unclear, and it is essential to keep in mind that more than one may be present.

The following features may help narrow the diagnosis. However, none of these characteristics can provide absolute diagnostic certainty and none of these should substitute for the clinical judgment of the treating specialist who has evaluated the patient in person.

Character and location of pain — For some people, the characteristics, severity, and site of vaso-occlusive pain tend to be relatively consistent between episodes. Thus, a patient's (or parent/caregiver's) report that the pain is typical for their vaso-occlusive pain should be taken into account; a patient's (or parent's) report that the pain is atypical should prompt evaluation for concurrent causes other than (or in addition to) vaso-occlusion.

Other clues to distinguishing among possible underlying causes such as vaso-occlusive pain, fracture, and osteomyelitis may include the following:

●Vaso-occlusive pain is much more common than osteomyelitis or septic arthritis; some reports suggest it is 20-fold more common [1].

●The gold standard for assessing pain is the patient's (or parent's) self-report of the location and severity of pain.

●Erythema, warmth, and local tenderness can be seen in any of these causes and cannot be used to distinguish among them.

●Low-grade fever in an otherwise well individual can be present with any of these causes and cannot be used to distinguish among them.

●Atraumatic pain that only affects a single bone (or digit, for dactylitis) is more concerning for osteomyelitis [13,14].

●Pain that responds within a few days to aggressive analgesia is more likely due to vaso-occlusion than osteomyelitis.

●Fractures are associated with a traumatic event and can be typically diagnosed by radiography.

The severity of the pain cannot be used to determine the underlying cause. Vaso-occlusive pain can vary from mild (barely interfering with everyday activities) to excruciating ("worse than breaking a leg"). (See "Acute vaso-occlusive pain management in sickle cell disease".)

The following examples illustrate aspects of pain in various locations:

●Hip – Inability to ambulate can only help distinguish between etiologies if the patient had an acute episode of trauma that is associated with fracture. Otherwise, with immobility secondary to pain, it can be difficult to distinguish between infectious versus vaso-occlusive etiologies alone. Septic arthritis often can also present with an acute atraumatic inability to ambulate secondary to pain.

●Shoulder – A published case report described a 51-year-old woman with SCD who presented with severe bilateral shoulder pain that had been progressively increasing over several years and was causing significant disability [55]. Radiographs and magnetic resonance imaging (MRI) showed avascular necrosis (AVN) in both humeral heads without joint collapse. She underwent diagnostic arthroscopy to assess the joint. Treatment included staged surgeries that consisted of the removal of necrotic bone and injection of the synthetic bone graft without injuring the joint cartilage; in both cases, she received hydration and transfusion and postoperative prophylaxis for deep vein thrombosis (DVT). This case illustrates the natural history and frequent bilateral involvement of AVN and the use of a multidisciplinary treatment plan that avoided total joint replacement and incorporated optimal preoperative, intraoperative, and postoperative care. (See 'Management of avascular necrosis' below.)

●Skull – A published case report described a 50-year-old woman with SCD and a history of occasional vaso-occlusive pain who presented with one month of unilateral, atraumatic painless swelling of the skull [56]. She was afebrile, but her white blood cell (WBC) count was elevated. MRI revealed an osteolytic defect and overlying fluid collection. She underwent craniectomy with removal of the affected bone and overlying tissue; pathology revealed chronic osteomyelitis, and culture was positive for Salmonella typhi. This case illustrates the risk of osteomyelitis in SCD and the importance of having a low threshold for culturing potentially infected bone or joint fluid, especially if there are signs of infection, such as a high WBC count. Optimal treatment decisions for salmonella osteomyelitis are ill-defined, as discussed below, and management should be discussed with an infectious disease expert and incorporate the antibiotic sensitivities of the organisms identified. (See 'Treatment of osteomyelitis and septic arthritis' below.)

●Orbit – Orbital infarction can be mistaken for orbital cellulitis but will not respond to antibiotic therapy. Conservative medical treatment is adequate for orbital bone infarction in the absence of neurologic symptoms. Still, significant hematomas or optic nerve compromise necessitate urgent surgical drainage of the hematoma for resolution [57].

Signs of infection — Fever may indicate infection, but the absence of fever does not eliminate the possibility of infection. Patients with septic arthritis and infection enclosed within the joint space can often present without fever. Signs that suggest an infection may include one or more of the following:

●Fever (any temperature ≥101.5ºF [38.5ºC] and any fever in an individual who appears acutely ill).

●High WBC count.

●Swelling of the affected area.

●Overlying skin ulceration or draining sinus.

●C-reactive protein (CRP) is likely higher in osteomyelitis than in acute pain episodes, but there is no threshold to distinguish acute vaso-occlusive pain episodes from osteomyelitis. The clinical utility of obtaining CRP is limited as a diagnostic test. Still, it may have some benefit in monitoring after the diagnosis of osteomyelitis has been made and a response to treatment is considered.

The diagnostic utility of these clinical features was evaluated in a case-control study that evaluated findings in 31 individuals with SCD who had osteomyelitis (based on discharge diagnosis as well as criteria including pain, positive blood cultures, bone aspirate culture, and typical radiographic findings) and compared them with 93 controls who had vaso-occlusive pain without osteomyelitis [58]. This found that compared with controls, individuals with osteomyelitis were more likely to have preceding fever (median, one versus no days), a longer duration of preceding pain (median, five versus two days), and swelling of the affected limb (71 versus 17 percent). Controls with vaso-occlusive pain were more likely to have more than one affected site and a lower WBC count (median, 15.6 versus 18.6/micro).

If infection is suspected, it may be necessary to treat presumptively for the likely organisms while awaiting diagnostic testing results. A decision should be made at the bedside as to whether prompt antibiotic treatment is required before obtaining a tissue sample or joint aspiration for bacterial culture and sensitivities. (See 'Laboratory testing' below and 'Treatment of osteomyelitis and septic arthritis' below.)

Response to intravenous antibiotics with clinical improvement and downward trending of inflammatory markers (eg, WBC, CRP) supports the diagnosis of osteomyelitis rather than bone infarction.

Additional testing is discussed below. (See 'Laboratory testing' below and 'Imaging' below.)

Laboratory testing — Inflammatory markers such as erythrocyte sedimentation rate (ESR) and CRP may be obtained in some instances of possible infection, but in the proinflammatory milieu seen in patients with SCD, the elevation of these markers can be challenging to interpret. However, the trend of these inflammatory markers, if serially collected, especially CRP, can be helpful in determining treatment response to antibiotic therapy.

In addition to routine testing with CBC and reticulocyte count, ESR, and CRP, additional laboratory testing appropriate for those with fever or localizing symptoms suggestive of infection may include blood cultures, urine culture, and cultures of other body fluids as appropriate. Chest radiography is appropriate if there are any findings suggestive of acute chest syndrome.

The results of cultures are essential in guiding antibiotic therapy. However, negative blood cultures do not eliminate the possibility of osteomyelitis since some individuals will clear bacteremia. In the case-control study mentioned above, 6 of 31 cases of osteomyelitis (19 percent) had a positive blood culture (three for Salmonella and one each for Haemophilus influenza, S. aureus, and Acinetobacter [58]). In the same study, 5 of 14 bone aspirate cultures were positive for Salmonella.

For individuals with joint effusion, aspiration of joint fluid for cell count, Gram stain, and cultures are considered the gold standard for testing. (See 'Treatment of osteomyelitis and septic arthritis' below.)

However, neutrophil infiltrates can be seen in synovial fluid with vaso-occlusive pain and infection, and thus, this finding is not very specific. Likewise, a negative culture does not definitively eliminate the possibility of septic arthritis, especially if the patient has received antibiotics. Unlike the evaluation in individuals without SCD, in SCD, it may be necessary to provide antibiotics before cultures can be obtained, especially in those with fever, because individuals with SCD are at risk for overwhelming sepsis due to their lack of splenic function. Delaying antibiotics for a short period (eg, one hour) is reasonable in some cases, such as those in which the aspirate will be completed within an hour, the patient monitored, and the antibiotics given immediately after the aspirate. If there is a long delay in obtaining the aspirate, antibiotics should be given as soon as blood cultures are obtained. (See "Evaluation and management of fever in children and adults with sickle cell disease" and "Septic arthritis in adults", section on 'Obtaining clinical specimens' and 'Treatment of osteomyelitis and septic arthritis' below.)

If osteomyelitis is suspected, aspiration of bone may be appropriate. However, general anesthesia carries a risk of precipitating or exacerbating acute chest syndrome; thus, invasive procedures such as bone biopsy or debridement should be performed with caution. Image-guided biopsy may be a means of avoiding general anesthesia. (See 'Imaging' below.)

Imaging — Plain radiographs often are not needed for individuals with highly typical vaso-occlusive pain that responds to analgesic therapy, and if performed, findings are often nonspecific. Likewise, MRI may show nonspecific changes with vaso-occlusive pain and osteomyelitis that cannot be used to make a definitive diagnosis. However, in other cases, MRI may reveal a joint effusion that may be aspirated or a nidus of bone infection, with significant enhancement on T1 fat-sensitive images or gadolinium.

Appropriate use of imaging in the evaluation of bone pain include the following:

●Radiographs – Plain films evaluate pain associated with trauma or limb pain. In general, it is reasonable to proceed with a radiograph (plain film) as the first imaging step to evaluate musculoskeletal pain from any etiology. However, plain films may not show any changes early in osteomyelitis and thus cannot be used to eliminate the diagnosis. (See "Hematogenous osteomyelitis in children: Evaluation and diagnosis", section on 'Radiographs' and "Approach to imaging modalities in the setting of suspected nonvertebral osteomyelitis", section on 'Imaging modalities'.)

●MRI – MRI with gadolinium contrast is the preferred imaging modality for patients whose diagnosis is unclear and to evaluate chronic or intermittent pain, especially if plain films suggest osteonecrosis. MRI can also show fluid collections that may represent septic joint effusions or infected areas of bone in septic arthritis (image 8) or osteomyelitis, respectively. The results of T1 (fat-sensitive)-weighted images or short-T1 inversion recovery (STIR) images may help distinguish bone infarct from osteomyelitis; T1 images do not show enhancement in bone infarct due to sequestration of red blood cells in the fatty marrow, but contrast enhancement in acute osteomyelitis can aid in its diagnosis [59-61]. T2 images show bone infarction that may not be apparent on plain films [1].

●Ultrasound or CT – Ultrasound guidance can be helpful when performing aspiration of deep joints such as the hip [62]. Often other joints such as the knee, elbow, wrist, shoulder, or ankle can be aspirated without ultrasound if performed by an orthopedic surgeon familiar with the topographic anatomy. Computed tomography (CT) can be helpful to guide percutaneous biopsy of the bone by interventional radiology if osteomyelitis is suspected; this can be performed under conscious sedation rather than general anesthesia, which carries a risk of precipitating or exacerbating acute chest syndrome. As an alternative to MRI, CT scans of the limb can help identify bone sequestration or involution not identified on plain films to inform surgical debridement in the setting of osteomyelitis not responsive to intravenous antibiotic administration.

●PET scanning – PET (positron emission tomography) scan or PET-CT scanning is typically not indicated in the evaluation of infection.

MANAGEMENT OF BONE AND JOINT COMPLICATIONS

Pain control — Prompt interventions to reduce pain are essential for all individuals with SCD who present with pain, regardless of the underlying cause. Pain medication should not be withheld while determining the cause of the pain. This is because most individuals only present to the hospital after they have exhausted all available at-home therapies and likely have been in significant pain for a relatively long period of time. We know of no clinical benefit of withholding pain medication for distinguishing among causes of joint or bone pain.

Additional information about patient education; home treatment of pain; our rationale for avoiding the term "crisis" for pain; and rapid triage, assessment, and treatment of pain when the patient presents for medical help, are discussed in detail separately. (See "Evaluation of acute pain in sickle cell disease" and "Acute vaso-occlusive pain management in sickle cell disease".)

Treatment of osteomyelitis and septic arthritis — Osteomyelitis and septic arthritis both require surgical input. The general principles for the treatment of sickle cell-related osteomyelitis and septic arthritis are not different from those in individuals without SCD. Still, the choice of antibiotics should take into account the different pathogenic organisms commonly found in asplenic individuals.

●Septic arthritis – Septic arthritis requires urgent formal consultation with an orthopedic surgeon and infectious diseases expert to perform needle aspiration of the joint and determine appropriate antibiotic therapy, respectively. The decision as to who will achieve the aspiration and when it will be performed is dependent on the availability of radiology and orthopedic personnel. Most importantly, a timely decision for obtaining joint fluid is appropriate when septic arthritis is part of the differential diagnosis.

The synovial fluid should be sent for white blood cell (WBC) count with differential, crystal analysis, Gram stain, and culture. A WBC count >50,000/mL and increased neutrophil count on the differential (eg, >80 percent) confirms the diagnosis of septic arthritis. Ultimately, positive cultures of the joint fluid further confirm the diagnosis and identify the organism, but these can take days for results to be available, especially for specific pathogens.

If there is suspicion of septic arthritis based on the results of the joint aspiration, operative debridement and irrigation of the joint should be performed. Surgical irrigation and debridement should not be delayed until culture results are available because devastating and permanent damage to articular cartilage can occur if septic arthritis goes untreated while awaiting culture results.

Appropriate systemic antibiotics are administered following debridement, ideally immediately after cultures are obtained; an exception is if the patient is septic-appearing or hemodynamically unstable, in which case antibiotics are given more urgently. Ideally, antibiotic administration is informed by the results from the aspirated synovial fluid or intraoperative tissue collection and is administered after surgical management.

Initial antibiotics are directed toward broad coverage of possible organisms and based on the Gram stain of the joint fluid. Antibiotics are continued when a positive joint aspiration (eg, positive culture from joint fluid and fluid WBC count >50,000/mL) confirms the diagnosis and is further informed by the culture results. (See 'Laboratory testing' above.)

●Osteomyelitis – In osteomyelitis, magnetic resonance imaging (MRI) is helpful to determine if a sequestered nidus of infection or subperiosteal abscess would benefit from a formal surgical debridement or can be managed with intravenous antibiotics only. The imaging should be reviewed by an orthopedic surgeon capable of handling a formal debridement if indicated. (See 'Imaging' above.)

As noted in a 2016 Cochrane review, there are no randomized trials to guide antibiotic selection in individuals with SCD and bone or joint infections [63]. Optimal treatment decisions for Salmonella osteomyelitis are ill-defined, as discussed below.

Ideally, sensitivities from a positive biopsy or aspiration results can inform definitive antibiotic choice. A final decision on the antimicrobial therapy and length of time for treatment should be discussed with an infectious disease expert, along with the antibiotic sensitivities of the bacteria [63]. Empiric treatment should cover S. aureus, Salmonella, and other gram-negative bacilli, as these are commonly isolated organisms from individuals with osteomyelitis in the general population [64].

Reasonable empiric treatment consists of vancomycin (dose for children: 15 to 20 mg/kg/dose every 8 to 12 hours, not to exceed 2 g per dose) and ciprofloxacin (750 mg orally or 400 mg intravenously every 12 hours). For adults with normal kidney function, vancomycin dosing is summarized in the table (table 2). Many individuals with SCD have abnormal kidney function and may warrant a different dosing approach. (See "Sickle cell disease effects on the kidney", section on 'Sickle cell nephropathy'.)

The optimal duration of antibiotic therapy for the treatment of hematogenous osteomyelitis is uncertain. In general, at least four weeks of parenteral therapy from the last major debridement are warranted; however, debridement should be avoided if possible and performed with caution due to the risks of acute chest syndrome related to the use of general anesthesia. (See "Hematogenous osteomyelitis in children: Management" and "Nonvertebral osteomyelitis in adults: Treatment", section on 'Hematogenous osteomyelitis'.)

Management of avascular necrosis

Role of screening for AVN — The utility of screening for avascular necrosis (AVN) in asymptomatic individuals has not been established. However, clinicians should have a low threshold for evaluating individuals with SCD who present with hip pain, decreased range of motion relative to the contralateral side (particularly with hip abduction) in the setting of negative orthogonal plain films, or unexplained back, pelvic, or knee pain. Clinicians should also have a low threshold for performing pelvic MRI if the radiologic imaging is equivocal, as pre-collapse AVN can be identified and treated before collapse occurs [37].

Given the range of presenting symptoms in AVN that can easily overlap with a chronic or recurrent acute vaso-occlusive pain event, a low threshold should be set to obtain a thorough history, physical examination, and imaging studies to exclude the diagnosis of early AVN.

Pretreatment evaluation — The pretreatment evaluation includes radiologic staging of the severity of AVN and assessment of the patient's overall disease status.

If AVN is suspected and traumatic injury such as a fracture of the painful hip is ruled out with radiographs, the evaluation can proceed on an outpatient basis, usually by a primary care provider or orthopedic surgeon.

Whenever possible, an individual with AVN being evaluated for surgery should have a hematology consult to address the following:

●Comorbidities that may impact the preoperative course.

●The necessity for preoperative transfusion (standard care for individuals with SCD unless contraindicated). (See "Red blood cell transfusion in sickle cell disease: Indications and transfusion techniques", section on 'Prophylactic preoperative transfusion'.)

●Postoperative medical management.

For any individual with SCD with AVN affecting one hip, assessment of the other hip is usually prudent since the condition is frequently bilateral (see 'Osteonecrosis (avascular necrosis)' above). Detection of bilateral AVN may impact the timing of the intervention, such as surgery, physical therapy, or both.

Staging can have a bearing on the best treatment modality for femoral AVN. As noted above, there are numerous staging systems in existence (see 'Stages of AVN' above and 'Scoring systems and biomarkers' above). Fundamentally, AVN of the femoral head can be categorized as pre-collapse and post-collapse of the subchondral surface. This is important because individuals with pre-collapse disease can be treated conservatively using combinations of pain control, physical therapy, and core decompression procedures to potentially delay progression to.

Initial conservative management — The management of AVN in individuals with SCD is challenging. Conservative management is usually used initially because it may be effective in some individuals and is the least invasive approach.

Conservative measures include the following [1,65]:

●Pain management with nonsteroidal antiinflammatory drugs (NSAIDs), typically required during the acute phase

●Hydration during the acute phase (bone infarction)

●Crutch-weight-bearing (<30 pounds [<13.6 kg] on the affected limb)

●Sustained physical therapy to maintain range of motion in the hip, strengthen the hip muscles, and provide gait training

We favor nonsurgical intervention with a physical therapist for a defined period (approximately six weeks) with a reevaluation of progress. Our practice is consistent with a 2014 practice guideline from the National Heart, Lung, and Blood Institute (NHLBI) of the National Institutes of Health (NIH) [37]. The physical therapy sessions should include a formal consult, with an assessment and plan, rather than self-administered home exercises. We prefer to reserve core decompression procedures for those who are clearly not benefitting from conservative therapy (see 'Core decompression' below). Concomitant humeral head AVN may limit the ability of an individual with femoral head AVN to use crutches or a wheelchair.

For the hip (and other sites of AVN), it may be very difficult to decide when to switch from conservative measures to surgery (eg, with core decompression or arthroplasty); limited evidence exists, and no consensus has been reached. In general, we prefer to try conservative measures first with close clinical monitoring, followed by core decompression in those who are not benefitting from conservative measures and arthroplasty in those not benefitting from core decompression. This is a joint decision between the hematologist and the orthopedic surgeon, along with the patient, as to whether physical therapy or orthopedic surgery should be tried initially. In some cases of late-stage AVN, physical therapy will not change the outcome, and surgery should be the initial management. However, in early-stage AVN (Steinberg Stage-I, II, or III osteonecrosis of the femoral head), physical therapy may initially be appropriate rather than core decompression [65]. Large randomized trials would help to identify the optimal strategy and timing for surgical treatment of AVN in individuals with SCD.

Once AVN has been identified, referral to an orthopedic surgeon is critical before subchondral collapse occurs, as subchondral collapse leads to permanent joint damage. (See 'Pretreatment evaluation' above.)

Principles for management of AVN of the humeral head are similar to those of the hip, with conservative medical care and physical therapy used as first-line therapy. (See "Treatment of nontraumatic hip osteonecrosis (avascular necrosis of the femoral head) in adults".)

Core decompression — Core decompression entails removal of necrotic tissue, with or without a bone graft to fill the "cored" area. The role of core decompression procedures in arresting or delaying progressive joint damage is controversial.

The following examples illustrate the range of findings in available studies:

●In a prospective case-control study involving 42 adults with SCD who had AVN of the hip, core decompression was significantly more effective than conservative therapy (provision of walking canes) in improving functional scores, especially in those treated at an early stage [66]. Of 42 hips treated with core decompression, 10 (24 percent) progressed to arthroplasty, after a mean of approximately seven years.

●In a series of 93 patients with AVN of the hip, outcomes were good with core decompression and bone grafting, allowing some with early-stage AVN to avoid other surgical interventions [67]. However, the authors have noted that revision to total hip arthroplasty (THA) may be more difficult following osteotomy because it may be more challenging to place the stem for the prosthesis. Of note, most of the patients in this cohort had underlying conditions other than SCD (only 5 of 79 were SCD-associated in the small-diameter drilling cohort).

●In a trial that randomly assigned 46 adults with SCD and AVN of the hip to receive core decompression followed by physical therapy or physical therapy alone, hip survival rates at three years were similar between the groups [65].

●A Cochrane Database systematic review revealed no evidence that adding surgical decompression to physical therapy resulted in clinical improvement in individuals with SCD [68].

These apparently contradictory findings regarding the efficacy of core decompression suggest that core decompression may provide symptomatic relief for variable periods but that more research is necessary to provide definitive information about the safety and efficacy of this procedure.

Core decompression with autologous stem cell injection — An investigational approach to the management of AVN in SCD is the injection of autologous bone marrow-derived mesenchymal stromal cells into osteonecrotic tissue. Initial experience suggests promise in reducing pain and improving function for both femoral and humeral AVN, but its long-term efficacy in arresting progression has not been established [69-71].

We do not encourage the use of autologous stem cell-based therapy for the treatment of AVN unless administered in a clinical trial setting with formal stopping rules, a Data Safety Monitoring Board (DSMB), and registration in clinicaltrials.gov. Our reservations are due to the unproven nature of this strategy, coupled with the absence of randomization, systematic evaluation, or a valid comparison group in the prior studies.

Arthroplasty (joint replacement) — Arthroplasty (ie, joint replacement) is an option for those who are not helped by conservative measures and/or core decompression.

We generally delay arthroplasty as long as possible given the risk of complications and need for revision arthroplasty in individuals who undergo arthroplasty at a relatively young age (eg, <40 years). Given the lack of uniform guidelines, ultimately, the orthopedic surgeon, patient, and family will make an informed decision, weighing the pros and cons.

Before any general anesthesia, a pretreatment evaluation by a hematologist is standard care to ensure optimal management of comorbidities, appropriate use of preoperative red blood cell transfusions, and appropriate postoperative medical management. (See 'Pretreatment evaluation' above.)

●Hip arthroplasty – Several reviews have summarized the outcomes of arthroplasty (especially hip arthroplasty) in individuals with SCD [72]:

•Some series have suggested a trend toward better outcomes with medical and surgical advances. As an example, a series from 1991 that included 27 hip arthroplasties estimated that the chance of requiring revision was 8 percent at 1 year, 20 percent within 3 years, and 30 percent within 4.5 years [27]. Studies from the 1990s also reported high rates of operative complications [73]. In contrast, a series from 2008 that included 312 arthroplasties reported failure rates of only 13 percent at 10 years and 19 percent at 15 years [74].

•One series from 2012 with midterm follow-up suggested that THA can be reasonably performed in patients under 21 years of age if it is the only option for pain control and continued mobility [75].

•In a series that included 42 patients with SCD undergoing cementless THA, implant failure rate and functional outcome scores were comparable to the general population (individuals without SCD) [76].

•A 2015 review of the Nationwide Inpatient Sample for knee and hip replacements in patients with SCD revealed a 152 percent higher rate of postoperative complications and a 20 to 42 percent longer duration of hospitalization compared with individuals who did not have SCD [77]. These results indicate some improvement in outcomes but highlight a persistently higher risk of complications for patients with SCD undergoing these procedures.

•A meta-analysis of 16 studies containing 5193 patients with SCD who underwent THA between 1988 and 2019 indicated that, compared with non-SCD patients, those with SCD had longer hospital stays, readmission rates, and medical complications [78]. One notable feature of this meta-analysis was the finding that cemented THA resulted in a higher rate of aseptic loosening and revision than cementless THA. Encouragingly, the rate of revision decreased over the timespan of the review, suggesting that outcomes of THA in patients with SCD are improving.

●Humeral head – A systematic review of the surgical management of humeral head AVN in SCD identified three retrospective studies and three case series with a total of 43 individuals who underwent core decompression, arthroscopic intervention, humeral head resurfacing, , or  [79]. In the pre-collapse stage, core decompression was the first surgical option used, but it did not clearly prevent progression of disease. There was no objective measurement of outcomes from arthroscopic intervention. Humeral head resurfacing yielded initial improvement in functional scores but frequently required revision with hemiarthroplasty or total shoulder arthroplasty due to complications. The authors of the individual studies concluded that both hemiarthroplasty and total shoulder arthroplasty resulted in improved pain, range of motion, function, and patient satisfaction. The authors of the systematic review stressed that these studies have small sample sizes, do not include randomization or control groups, and do not provide long-term survival data. Guidelines for the proper management of humeral avascular necrosis will require further systematic investigation.

General information about arthroplasty (not specific to SCD) is discussed separately. (See "Treatment of nontraumatic hip osteonecrosis (avascular necrosis of the femoral head) in adults" and "Total hip arthroplasty" and "Complications of total hip arthroplasty".)

Disease-modifying therapies (eg, hydroxyurea) — Hydroxyurea is used in children and adults with severe or recurrent vaso-occlusive pain episodes, based on strong evidence that hydroxyurea reduces the frequency of these episodes. (See "Hydroxyurea use in sickle cell disease".)

A retrospective comparison of the incidence of femoral head AVN before and after the advent of widespread hydroxyurea therapy found a significant decline in the latter cohort, suggesting that hydroxyurea could have a protective effect [30]. Confirmation of this effect will require prospective controlled studies.

There are no available studies that assess the impact of voxelotor, L-glutamine, or crizanlizumab on prevention or management of AVN. (See "Disease-modifying therapies to prevent pain and other complications of sickle cell disease".)

Hematopoietic stem cell transplantation, gene therapy, and gene editing — There are no large-scale studies of the impact of these curative therapies on the incidence or progression of avascular necrosis.

Chronic transfusion therapy — There is no evidence that regular blood transfusion therapy abates the progression of early-stage AVN in individuals with SCD. Although there is a biological basis for the possibility that regular blood transfusion therapy might potentially be of benefit, there is no randomized controlled trial demonstrating its benefit in abating the progression of AVN. In the absence of confirmed evidence that blood transfusion will attenuate progression of AVN, along with the known challenges of regular blood transfusion therapy, including eventual iron chelation and burdens to the family, we do not recommend this therapy for management of early AVN. In the Silent Cerebral Infarct Multicenter Trial (SIT), 6 of 97 patients in the observation arm developed symptomatic AVN of the hip, versus 1 of 99 in the chronic transfusion arm (IRR 0.22; p = 0.02) [80]. These data suggest that early initiation of chronic transfusion therapy could prevent symptomatic femoral head AVN, but they do not address the role of transfusions in arresting or reversing AVN that is already present. We do not typically endorse regular blood transfusion to abate progression of symptomatic AVN due to the absence of definitive evidence supporting this practice and no clear data on when transfusion could be stopped.

Bisphosphonate therapy — Data are limited on the use of bisphosphonates to prevent or treat AVN in the general population. One systematic review of zolendronic acid in 788 hips suggested some benefit in improving symptoms and delaying the need for total hip arthroplasty [81]. In a retrospective study involving 23 children with SCD treated with intravenous zolendronic acid or pamidronate for bone pain and AVN, 10 of 11 children (91 percent) with follow-up monitoring for improvement reported complete or partial resolution of pain six months after therapy without serious adverse events [82]. These results suggest that bisphosphonates may have a role in the treatment of SCD-associated AVN, but we do not recommend their routine use given the lack of strong supporting data.

VTE prophylaxis — Adults with SCD who are admitted to the hospital with an acute medical illness or who are undergoing orthopedic surgery typically require venous thromboembolism (VTE) prophylaxis due to the increased risk of VTE associated with the acute illness and/or immobilization as well as the baseline increased VTE risk conferred by SCD. Bone and joint complications that might increase the risk of VTE include acute pain episodes, infection, and chronic immobility. (See "Overview of the management and prognosis of sickle cell disease", section on 'Thromboembolism prophylaxis'.)

OTHER ASPECTS OF BONE HEALTH

Vitamin D deficiency — Vitamin D deficiency is a very common finding in individuals with SCD, with prevalences of 25 percent to as high as 98 percent depending on the report [83-88]. Several studies have also reported a high prevalence of hyperparathyroidism.

There is very little evidence regarding the efficacy of routine vitamin D supplementation in reducing rates of osteopenia, osteoporosis, bone pain, or avascular necrosis [89,90]. In the absence of high-quality data, reasonable approaches may include the following:

●Routine testing (eg, annually) for vitamin D deficiency with supplementation for those with low vitamin D levels.

●Routine supplementation with vitamin D (eg, 2000 units daily) in populations with a high prevalence of deficiency, starting at approximately four years of age.

Details of vitamin D supplementation are discussed in detail separately. (See "Vitamin D insufficiency and deficiency in children and adolescents" and "Vitamin D deficiency in adults: Definition, clinical manifestations, and treatment".)

The role of vitamin D in treating chronic pain in individuals with SCD is also discussed separately. (See "Overview of the management and prognosis of sickle cell disease", section on 'Nutrition'.)

Osteopenia and osteoporosis — Osteopenic and osteoporotic bone is susceptible to fracture. The causes are multifactorial and may include marrow expansion due to chronic hemolysis and the associated compensatory increase in erythropoietic activity that leads to widening of the medullary space and thinning of trabecular bone, along with vaso-occlusion-associated ischemia and vitamin D deficiency. Reduced bone mineralization has been reported in children and adults with SCD, especially in the lumbar spine [51,91,92].

Management includes addressing treatable factors (vitamin D supplementation if deficient, hydroxyurea for repeated episodes of vaso-occlusion, and physical therapy if indicated).

The role of other treatments such as bisphosphonates for osteopenia has not been studied in individuals with SCD. In the absence of high-quality data specifically in this population, we would treat individuals with osteopenia or osteoporosis similar to the non-SCD population. (See "Screening for osteoporosis in postmenopausal women and men" and "Overview of the management of low bone mass and osteoporosis in postmenopausal women".)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Sickle cell disease and thalassemias".)

PATIENT PERSPECTIVE TOPIC — Patient perspectives are provided for selected disorders to help clinicians better understand the patient experience and patient concerns. These narratives may offer insights into patient values and preferences not included in other UpToDate topics. (See "Patient perspective: Sickle cell disease".)

SUMMARY AND RECOMMENDATIONS

●Causes of bone or joint pain – Bone or joint pain in individuals with sickle cell disease (SCD) can be caused by vaso-occlusion, avascular necrosis of bone (AVN; osteonecrosis), fractures, or infection (osteomyelitis, septic arthritis). Dactylitis refers to vaso-occlusive pain in the small bones of the hand and feet typically seen in infants. The mechanisms, presenting findings, and natural history of these complications are described above. Like other SCD manifestations, orthopedic complications are more frequent and tend to occur at an earlier age in individuals with more severe genotypes (homozygous hemoglobin SS or sickle beta⁰ thalassemia). (See 'Causes of bone or joint pain' above.)

●Evaluation – Causes of bone or joint pain in individuals with SCD can be challenging to distinguish.

•Clinical – Findings of vaso-occlusion, infarction, and infection are often similar (pain, redness, edema, low-grade fever); in some cases, more than one cause is present. Vaso-occlusive pain is more common than the other causes, but the other causes are easy to miss. It is important to ascertain whether the pain is typical or atypical for the person's vaso-occlusive pain and whether signs of infection, such as high fever, high white blood cell (WBC) count, or overlying swelling or ulceration, are present. (See 'Clinical distinguishing features' above.)

•Laboratory – The complete blood count (CBC) and reticulocyte count as well as inflammatory markers (WBC count, erythrocyte sedimentation rate [ESR], C-reactive protein [CRP]) are used to assess changes from baseline, and cultures are used in individuals with suspected infection. Joint fluid or bone cultures are the gold standard for diagnosing septic arthritis and osteomyelitis. These may require image-guided techniques. Common pathogens include Salmonella, Haemophilus influenza, Staphylococcus aureus, and Acinetobacter. (See 'Laboratory testing' above.)

•Radiology – Magnetic resonance imaging (MRI) with gadolinium may help evaluate challenging cases, identify osteonecrosis, and delimit fluid collections. Sometimes the response to initial treatment may help narrow the diagnosis. (See 'Imaging' above.)

●Pain control – Regardless of the cause of bone pain, rapid triage and prompt and effective treatment are critical. Adequate analgesia should not be withheld while determining the cause of the pain. Additional aspects of pain management are discussed separately. (See 'Pain control' above and "Acute vaso-occlusive pain management in sickle cell disease".)

●Antibiotics – Septic arthritis and osteomyelitis are treated with empiric therapy to cover S. aureus, Salmonella, and other gram-negative bacilli until therapy can be narrowed based on culture results. (See 'Treatment of osteomyelitis and septic arthritis' above.)

●Surgery – AVN of the hip or shoulder is seen in approximately 10 percent of individuals with SCD and is often bilateral. Optimal management has not been established. Most individuals with AVN are treated with a course of conservative therapy (a period of protected weight-bearing and supervised physical therapy), and if this is ineffective, possibly with core decompression. Arthroplasty is generally reserved for those who do not benefit from or are not candidates for core decompression because of advance subchondral collapse of the femoral head. (See 'Management of avascular necrosis' above.)

●Hydroxyurea – After resolving a critical event, the indications for hydroxyurea and additional routine testing and interventions should be examined to ensure optimal preventive approaches are being used. (See "Hydroxyurea use in sickle cell disease" and "Overview of the management and prognosis of sickle cell disease".)

●Vitamin D – Individuals with SCD have a disproportionately high prevalence of low bone mineral density and vitamin D deficiency. We administer a daily dose of vitamin D (2000 units) to individuals with SCD who are >4 years of age; testing vitamin D levels and supplementation for those with deficiency is also reasonable. (See 'Other aspects of bone health' above.)

ACKNOWLEDGMENT — UpToDate gratefully acknowledges Stanley L Schrier, MD (deceased), who contributed as Section Editor on earlier versions of this topic and was a founding Editor-in-Chief for UpToDate in Hematology.

The UpToDate editorial staff also acknowledges the extensive contributions of Donald H Mahoney, Jr, MD, to earlier versions of this topic review.