Management of Osteonecrosis By Adult and Pediatric Sickle Cell Providers

Introduction:

Osteonecrosis (ON) is a debilitating complication in 40% of adult sickle cell disease (SCD) patients. Almost all SCD ON will collapse, develop intractable pain or require surgery in 8 to 15 years from diagnosis. Majority of treatments have not been examined in a controlled setting, leading to variations in care, impeding research to verify the efficacy of existing therapies, and development of new treatments. The study's objective is to describe the screening, diagnostic and management patterns of ON in pediatric and adult SCD patients by US and Canadian providers.

Materials and Methods:

A cross-sectional survey of SCD providers from the Sickle Cell Adult Provider Network, the American Society of Pediatric Hematology/Oncology, and the Canadian Hemoglobinopathy Association was conducted from March to July 2015. The survey had 24 questions about provider demographics, characteristics of SCD population, ON diagnosis, staging and treatment. The survey was distributed electronically over 9 weeks with 2 reminders 3 weeks apart. Descriptive statistics were calculated and logistic regression was used to identify factors (specialty, location, years of practice, number of patients, adult vs. pediatric centers) that were associated with screening, diagnostic, staging and treatment approaches.

Results:

Of the 2,475 health care providers surveyed, 78 participants responded from 63 centers with 17,232 patients (17-23% of the estimated SCD population in US and Canada). Majority of respondents were located in urban teaching centers (93%). All (100%) respondents self-identified as SCD providers: 44 adult hematology/oncology (heme/onc) specialist; 3 internists; 24 pediatric heme/onc specialist; 5 nurse clinicians; 1 pharmacist; and 1 social worker. The median time in practice was 16 years (min 1, max 40). The estimated number of unique patients under the care of the respondents was 10,697, of which 1,946 (18%) had ON. More than half of the ON patients had joint collapse (n=1,050, 54%), and 72% had joint replacement. Only (13%) of the respondents screened for ON in asymptomatic patients, and screening measures include 50% by x-ray; 38% by physical, 12% by other modalities. All (100%) respondents screened for ON in symptomatic patients: 51% using x-ray and MRI, 41% x-ray, 7% MRI. The majority of respondents (72%) did not know the ON staging methods used. Frequencies of respondents using non-surgical and surgical treatments in pre-collapse and collapsed ON are shown in Table 1. Respondents' years in practice was the only factor significantly associated with the use of screening in asymptomatic SCD patients (OR 2.9/decade in practice, p=0.030). Respondents' years in practice was also associated with a reduced likelihood of hydorxyurea (HU) use in pre-collapse ON (OR 0.44/decade more in practice, p=0.011). Patients with collapsed ON under the care of pediatric heme/onc specialists were more likely to be treated with core decompression than adult heme/onc specialist or internist (OR 10.7, p=0.016).

Discussion:

Our survey suggests that the diagnosis, classification and management of ON is highly heterogeneous, despite the 2014 NIH expert panel report on the management of SCD prior to the distribution of the survey. The lower prevalence of ON in our study compared to that in cohort-wide screening may be due to the lack of screening by surveyed providers in asymptomatic SCD patients, suggesting ON may be underdiagnosed. Despite lack of evidence for their efficacy, HU and core decompression were used by majority of respondents in pre-collapse ON. A limitation of the study is difficulty in estimating the true response rate since the number of SCD providers in US and Canada is not known, potentially impacting the results' generalizability. The variations in practice and the high prevalence of collapsed joints requiring replacement in the surveyed population confirms the need to develop a uniformed approach in screening, diagnosis, and evidence-based therapies in SCD ON.

Table 1: Non-surgical and surgical treatments used by respondents in pre-collapse and collapsed ON. Some respondents selected multiple responses

Treatments	Number of Respondents (%)
Non-surgical	Pre-collapse (n=54)	Collapsed (n=53)
Analgesia	53 (98%)	52 (98%)
Physiotherapy	49 (91%)	37 (70%)
Hydroxyurea	32 (59%)	28 (53%)
Surgical	Pre-collapse (n=38)	Collapsed (n=45)
Core decompression	33 (87%)	10 (22%)
Hip replacement	9 (24%)	43 (95%)