Rheumatologists' awareness of and screening practices for Hepatitis B virus infection prior to initiating immunomodulatory therapy

Authors


Abstract

Objective

To assess the degree of awareness of the American College of Rheumatology (ACR) guidelines and package insert information on the screening for and management of hepatitis B virus (HBV) infection by rheumatologists in patients receiving immunomodulation drug therapies.

Method

A questionnaire survey was administered to a nationwide sample of 1,000 members of the ACR. Each participating physician answered questions regarding their awareness of the risk of HBV reactivation, familiarity with published guidelines regarding HBV reactivation, their decision process in screening patients for HBV, knowledge of antiviral treatments for HBV, personal experience with HBV reactivation, and preferred approach to prophylaxis and subsequent monitoring of those patients.

Results

Responses were highly variable with regard to awareness, screening, and treatment options. The overall response rate was 15.3%. Of those surveyed, 7.4% had seen HBV reactivation. Depending on the agent, 19–53% were aware of manufacturers' warnings for HBV reactivation within drug package inserts. Nearly three-quarters (72%) would screen for HBV reactivation regardless of the presence/absence of manufacturers' warnings. Only 69% reported performing universal screening prior to initiating therapy with biologic disease-modifying antirheumatic drugs. The majority (81%) would defer to a gastroenterologist/hepatologist to determine prophylactic therapy for HBV. Only 22% had managed patients who were given prophylaxis against HBV reactivation while receiving immunosuppressants.

Conclusion

Based on this survey, improving education among rheumatologists regarding the risks of HBV reactivation and its prevention for patients receiving immunosuppressants seems warranted. More specific consensus guidelines are recommended to achieve universal screening as the standard of care in these patients, especially with the increasing prevalence of HBV infection estimated in the US.

INTRODUCTION

The risk of chronic hepatitis B virus (HBV) reactivation in patients undergoing immunomodulating therapy is increasingly being recognized (1–10). The current body of knowledge resides predominantly in the oncology literature, in which the prevalence of chronic HBV diagnosed by a positive hepatitis B surface antigen (HBsAg) is as high as 80% in patients undergoing systemic chemotherapy (11–13). Although US-based oncologists have demonstrated a moderate degree of familiarity of HBV reactivation risk with immunomodulatory therapy (14), the US-based rheumatologists' awareness and knowledge of this risk has yet to be evaluated. Existing knowledge of HBV reactivation with immunomodulatory therapy in rheumatic diseases has originated principally from endemic areas outside the US, where up to 18% of patients have required antiviral therapy for HBV reactivation (15). In contrast, most publications from nonendemic areas are based on case reports and case series involving the use of corticosteroids or nonbiologic disease-modifying antirheumatic drugs (DMARDs), including chloroquine, azathioprine, methotrexate (MTX), cyclosporine, and sulfasalazine (5–7, 10, 15–19).

As the use of biologic agents has become more widespread in rheumatic diseases (4, 8–9, 16, 20–22), an increasing number of cases of reactivation of HBV have been reported. The use of prophylactic antiviral therapy remains unsettled in rheumatology circles, with no double-blind placebo-controlled studies having been completed to date. A recent meta-analysis did advocate the cost-effectiveness of antiviral therapy for preventing HBV reactivation in patients undergoing chemotherapy (23). Data have emerged in the rheumatology literature that recommend prophylactic lamivudine in patients receiving biologic DMARDs (etanercept, infliximab, adalimumab) (9, 20). In 2006, Calabrese and colleagues proposed prophylactic therapy with lamivudine in all HBsAg-positive patients prior to, during, and for 6 months after cessation of immunosuppressive therapy (24).

Guidelines regarding rheumatology patients and the risk of HBV reactivation secondary to biologic DMARDs, and updated nonbiologic DMARD recommendations, were published by the American College of Rheumatology (ACR) in June 2008 (25). The ACR formally recommends baseline liver aminotransaminase (aspartate aminotransferase, alanine aminotransferase) levels in all patients with rheumatoid arthritis (RA) treated with either MTX or leflunomide. Additionally, they recommend screening for HBV in high-risk patients receiving MTX or leflunomide. However, the ACR does not currently recommend HBV screening prior to biologic DMARD use. Furthermore, the ACR does not specifically delineate which screening tests to use in high-risk patients, although Calabrese and colleagues have proposed screening all patients undergoing infliximab therapy using HBsAg, hepatitis B surface antibody (anti-HBs), and anti–hepatitis B core antibody (HBcAb) testing (24). Additionally, the ACR has stated that administration of the HBV vaccine is warranted in high-risk patients prior to starting biologic DMARDs (25).

In patients with viral hepatitis, the 2008 ACR guidelines offer specific recommendations for the use of biologic DMARD therapy for RA. For those with acute HBV or HCV infection, or severe liver injury in the setting of chronic viral infection, defined as those with a Child-Pugh class B or C score (based on the presence of hypoalbuminemia, ascites, elevated total bilirubin, elevated prothrombin time, and encephalopathy) (25). The use of biologic agents is contraindicated independent of whether patients are receiving antiviral therapy.

In addition to these societal guidelines, the US Centers for Disease Control and Prevention (CDC) published guidelines in September 2008 stating that any patient receiving immunosuppressive therapy should be screened for HBV with HBsAg, anti-HBs, and HBcAb (26).

The aim of the current study was to assess the degree of awareness of these guidelines along with package insert information on the screening for and management of HBV infection by rheumatologists in patients receiving immunomodulating drug therapies.

SUBJECTS AND METHODS

Design and sample.

A descriptive cross-sectional questionnaire survey (similar to one we previously designed for oncologists) (14) was administered 4 months after the 2008 ACR guidelines and 1 month after the CDC guidelines were published. A random nationwide sample of physician members was obtained from the ACR registry (www.rheumatology.org) totaling 5,672 US-based physicians. Eligibility criteria were board certification in rheumatology and current involvement in patient care. Before initiating the survey, each physician was assigned a random 3-digit number and sorted by this random number. Ten groups of 100 physicians were then created for a total sample size of 1,000 (17.6% of all US-based physicians registered with the ACR). Physicians were randomly selected from all 50 states with at least 5 members coming from each state, ensuring a geographically diverse population. The online survey was administered in numerical order as sorted via an online survey Web site (www.survey monkey.com). Each physician received an e-mail containing an individualized link corresponding to his or her assigned number. This enabled him or her to complete the survey in a secure manner. Prior to starting data collection, each potential respondent was given the opportunity to opt out of the survey. If the physician opted out, no substitution was permitted. If they were agreeable, the respondent provided consent, which included disclosure that the information obtained was anonymous and would remain confidential. No incentives were offered to complete the survey, and the survey instrument and study design were approved by the Institutional Review Board of Georgetown University.

Survey instrument development.

The survey was formulated after an extensive review of the medical literature and drug package inserts concerning HBV infection and reactivation risks among immunosuppressed patients. The survey items were developed by a multidisciplinary team of gastroenterologists, hepatologists, and rheumatologists. To ensure understanding, comprehension, and reliability, interviews were conducted with faculty members of the Division of Rheumatology of Georgetown University who were not included in the survey. The survey cover page provided definitions for and examples of biologic and nonbiologic DMARDs and included questions about demographic characteristics. Biologic DMARDs included in the survey were abatacept (Orencia; Bristol-Myers Squibb, Princeton, NJ), adalimumab (Humira; Abbott Laboratories, North Chicago, IL), anakinra (Kineret; Amgen, Thousand Oaks, CA), etanercept (Enbrel; Immunex, Thousand Oaks, CA), infliximab (Remicade; Centocor, Malvern, PA), and rituximab (Rituxan; Genentech, San Francisco, CA). Nonbiologic DMARDs included were azathioprine (Imuran; DSM Pharmaceuticals, Greenville, NC), cyclosporine (Sandimmune and Gengraf; Novartis Pharmaceuticals, East Hanover, NJ), hydroxychloroquine (Plaquenil; Sanofi-Aventis, Bridgewater, NJ), leflunomide (Arava; Aventis, Kansas City, MO), MTX (Rheumatrex; Xanodyne, Florence, KY), and sulfasalazine (Azulfidine; Pharmacia & Upjohn, New York, NY).

A series of 10 questions was asked of each physician completing the online survey (Table 1). One question asked about awareness of published guidelines for biologic and nonbiologic DMARDs and corticosteroids. Two questions asked about the package inserts from the drug manufacturers of biologic agents, and screening practices with respect to these inserts. Three questions addressed the screening practices (prevalence, at-risk populations, screening tests) of the rheumatologists included in the survey. A free-text response box was provided after the screening-related questions to allow further clarification by the respondent. One question specifically inquired about the incidence of HBV reactivation and potential causative agents. Two questions investigated prophylactic practices. The last question inquired about the demographic characteristics of the survey respondents (i.e., practice setting, years in practice, etc.). The survey remained open for a 6-month time period. Duplicate e-mails were sent at 1, 3, and 6 months to nonresponders.

Table 1. Questionnaire survey responses*
Questions and response categoriesYesNoUnsure
  • *

    Values are the number (percentage) of respondents who gave each answer. DMARD = disease-modifying antirheumatic drug; LFTs = liver function tests; HIV = human immunodeficiency virus; IV = intravenous; HBsAg = hepatitis B surface antigen; anti-HBsAg = hepatitis B surface antibody; anti-HBcAg = hepatitis B core antigen; HBV = hepatitis B virus; GI = gastroenterologist.

  • Number (percentage) of respondents who answered “high risk.”

1) Are you aware of any published literature or American College of Rheumatology guidelines recommending screening for hepatitis B infection prior to initiating therapy with:   
   Biologic DMARDs97 (70)25 (18)16 (12)
   Nonbiologic DMARDs73 (53)43 (31)22 (16)
   Steroids16 (12)92 (67)30 (22)
2) Are you aware of drug package inserts recommending screening for hepatitis B prior to initiating drug therapy with:   
   Etanercept (Enbrel)63 (46)36 (26)37 (27)
   Infliximab (Remicade)64 (47)38 (28)34 (25)
   Adalimumab (Humira)62 (46)38 (28)36 (26)
   Anakinra (Kineret)26 (19)53 (39)57 (42)
   Abatacept (Orencia)47 (35)43 (32)46 (34)
   Rituximab (Rituxan)72 (53)27 (20)37 (27)
3) If the drug package insert does not specifically recommend to screen for hepatitis B prior to initiating drug therapy, would you still proceed with screening?98 (72)15 (11)23 (17)
4) Do you routinely screen your patients for hepatitis B infection prior to initiating therapy with:   
   Biologic DMARDs95 (69)8 (6)34 (25)
   Nonbiologic DMARDs58 (42)13 (9)66 (48)
   Steroids9 (7)64 (47)64 (47)
5) Which of the following would prompt you to screen for hepatitis B infection:   
   Elevated LFTs134 (98)2 (1)1 (1)
   Endemic area114 (83)13 (9)11 (8)
   History of hepatitis134 (99)1 (1)1 (1)
   HIV co-infection132 (97)4 (3)0 (0)
   IV drug use134 (99)2 (1)0 (0)
   Men who have sex with men127 (93)3 (2)6 (4)
   Health care workers94 (69)20 (15)22 (16)
   History of blood transfusion108 (79)14 (10)14 (10)
   Hemodialysis99 (73)22 (16)15 (11)
   Never vaccinated62 (46)37 (27)37 (27)
6) Would you use the following test(s) for hepatitis B screening:   
   LFTs88 (65)48 (35)0 (0)
   HBsAg124 (92)10 (7)1 (1)
   Anti-HBsAg70 (51)65 (48)1 (1)
   Anti-HBcAg59 (43)70 (51)7 (5)
   HBV DNA9 (7)114 (84)13 (10)
   Hepatitis panel41 (30)81 (60)14 (10)
7) Have you ever seen hepatitis B reactivation in any of your patients undergoing therapy with:   
   Biologic DMARDs10 (7)124 (91)2 (1)
   Nonbiologic DMARDs8 (6)128 (94)0 (0)
   Steroids2 (1)131 (96)3 (2)
8) In patients at risk for hepatitis B reactivation, would you consider prophylaxis:   
   Prior44 (32)37 (27)55 (40)
   Concurrent29 (21)51 (38)56 (41)
   Defer110 (81)15 (11)10 (7)
8a) How often would you monitor HBV reactivation status once prophylaxis is initiated:   
   Defer to GI/hepatology input6 (24)  
   Every 4–6 weeks2 (8)  
   Every 3 months6 (24)  
   Every 6 months5 (20)  
   Every 12 months4 (16)  
   Never1 (4)  
   Based on signs/symptoms1 (4)  
9) Have any of the following drugs been given to your patients for prophylaxis against hepatitis B reactivation: etanercept, infliximab, adalimumab, anakinra, abatacept, rituximab?30 (22)  
10) Of those using prophylaxis, which drugs were chosen?   
   Hepsera2 (7)  
   Baraclude3 (10)  
   Epivir20 (67)  
   Tyzeka0 (0)  
   Viriad5 (17)  

RESULTS

A total of 1,000 online surveys were distributed, but 96 physicians opted out of the survey prior to data collection. Of those not opting out, 138 surveys were completed (134 in their entirety), for a response rate of 15.3%. Of the rheumatologists surveyed, 51% were in academic medicine, and 49% self-identified as not being affiliated with a major academic center.

Awareness of reactivation risk and guidelines.

Although most responding rheumatologists (97 [70%]) were aware of the June 2008 ACR guidelines for the use of biologic agents in RA, only approximately half (73 [53%]) were aware of the existence of the same guidelines for nonbiologic DMARDs, despite the availability of these for a longer period of time. Regarding HBV reactivation, 10 respondents (7.4%) had seen reactivation with biologic DMARDs, and 8 (5.9%) and 2 (1.5%), respectively, had witnessed reactivation with nonbiologic DMARDs and corticosteroids (Table 2). Of the cases reported, 4 instances of reactivation were reported with infliximab use (2 during treatment of RA, and 1 case each in patients with ankylosing spondylitis and psoriatic arthritis). One case of reactivation was reported with etanercept use for RA. Among nonbiologic DMARDs used for RA, 4 cases of HBV reactivation were reported in association with MTX use, and 1 case was observed with leflunomide use.

Table 2. Demographics of the rheumatologists surveyed*
DemographicValue
  • *

    Values are the percentage unless otherwise indicated. HBV = hepatitis B virus; DMARDs = disease-modifying antirheumatic drugs; RA = rheumatoid arthritis; AS = ankylosing spondylitis; PsA = psoriatic arthritis.

Practice setting 
 Private62
 Hospital33
 Academic51
Years in practice 
 <519
 5–1015
 11–2029
 >2033
 Unknown4
Cases of HBV reactivation reported, no. 
 Biologic DMARDs 
  Infliximab (Remicade), RA, AS, or PsA4
  Etanercept (Enbrel), RA1
 Nonbiologic DMARDs 
  Methotrexate (Rheumatrex), RA4
  Leflunomide (Arava), RA1

Awareness of immunomodulating drug package inserts.

Six commonly used biologic DMARDs contain package inserts that outline the potential risk of HBV reactivation with agent administration. These include etanercept, infliximab, adalimumab, anakinra, abatacept, and rituximab. For each individual agent, the degree of awareness of the manufacturer's package inserts was highly variable. Nearly half of the rheumatologists surveyed (72 [53%]) were aware of the risk of HBV reactivation as discussed in the package insert of rituximab, but this number fell to 26 (19%) with anakinra. Regardless of package insert information, approximately three-quarters of rheumatologists (98 [72%]) felt that screening was warranted. The most common free response justifications provided by survey respondents included the publication of HBV reactivation cases in the rheumatology literature, the use of universal screening for HBV prior to beginning immunomodulating drug therapy, and concern for increased risk of hepatotoxicity with preexisting chronic HBV and hepatotoxic DMARD use.

General screening practices.

In terms of screening for HBV, 95 (69%) of these practicing rheumatologists reported universal screening prior to starting biologic DMARD therapy. However, only 58 (42%) screened universally before nonbiologic DMARD use, and only 9 respondents (7%) checked HBV status prior to corticosteroid use. Of the 42 respondents (31%) who did not screen universally, 34 did so in high-risk patients, leaving 8 respondents who did not screen in any circumstances. The rheumatologists reported screening patients with high-risk factors as follows: 134 (99%) reported screening patients with a history of hepatitis, 134 (99%) reported screening patients with a history of intravenous drug use, 134 (98%) reported screening patients with elevated liver-associated enzymes, 132 (97%) reported screening patients with human immunodeficiency virus positivity, 127 (93%) reported screening men who have sex with men, 108 (79%) reported screening patients with a history of blood-based product transfusion, 114 (83%) reported screening patients with an origin from an endemic area, 99 (73%) reported screening patients with hemodialysis, 94 (69%) reported screening health care workers, and 62 (46%) reported screening patients with nonimmune status (Table 1). When selecting an initial screening test, 124 (92%) checked the HBsAg and 88 (65%) would obtain liver associated enzymes (Table 1). Furthermore, 70 (51%) checked anti-HBs, 59 (43%) used HBcAb, and 41 (30%) would check a full hepatitis panel.

Prophylaxis trends and monitoring in high-risk patients.

More than half of rheumatologists surveyed (73 [53%]) considered prophylaxis in high-risk patients prior to or concurrent with immunomodulatory therapy. In terms of the specific antiviral agent, 110 (81%) of all respondents deferred to a gastroenterologist/hepatologist to determine the best prophylactic therapy for high-risk patients. Of those who did prescribe antiviral agents, the most commonly cited prophylactic agent was lamivudine (67%), with the second most common agent being tenofovir (17%). Few were familiar with other oral HBV agents (adefovir, entecavir, and telbivudine), with each being used ≤10% of the time as the prophylactic agent of choice.

In terms of referral patterns, >80% of respondents sent their patients to gastroenterology/hepatology. After initial consultation, however, only 24% opted to continue followup with a gastroenterologist/hepatologist. The other 76% stated they would follow the patient by themselves, with a highly variable frequency of monitoring: 8% would check liver-associated enzymes every 4–6 weeks, 24% would check every 3 months, 20% would check every 6 months, and 16% checked yearly (Table 1). Only 1 respondent felt that liver-associated enzymes did not need to be monitored, and 1 other respondent said that they would base their decisions solely on their patient's signs and symptoms.

DISCUSSION

Overall, the results of this survey suggest that awareness of the risk of HBV reactivation with biologic DMARD therapy is highly variable among rheumatologists. The majority, however, are embracing the practice of universal screening despite the lack of published consensus guidelines/recommendations. Compared with oncologists, whose universal screening rate prior to chemotherapy was reported to be <20% (14), 95 rheumatologists (69%) reported practicing universal screening prior to immunosuppressive therapy. Increased HBV screening among rheumatologists may be attributable to the increasing efforts of biologic drug manufacturers to raise awareness of HBV reactivation risk through product packaging. The recommended testing for patients undergoing immunosuppressive therapy is reported in Table 3.

Table 3. Recommended testing for patients undergoing immunosuppressive therapy*
 CDC (26)AASLD (28)ACR (25)Keeffe et al (37)Calabrese et al (24)
  • *

    CDC = Centers for Disease Control and Prevention; AASLD = American Association for the Study of Liver Diseases; ACR = American College of Rheumatology; HBsAg = hepatitis B surface antigen; NS = not specified; HBcAg = anti–hepatitis B core antibody; anti-HBs = hepatitis B surface antibody; HBV = hepatitis B virus; LMV = lamivudine; TLV = telbivudine; ADF = adefovir; ETC = entecavir.

  • High risk = individuals born in areas of high and intermediate prevalence, household or sexual contact with HBsAg-positive persons, history of intravenous drug use, persons with multiple sexual partners or a history of sexually transmitted disease, men who have sex with men, inmates of correctional facilities, individuals with chronically elevated alanine aminotransferase or aspartate aminotransferase, individuals infected with HCV or human immunodeficiency virus, patients undergoing renal dialysis, and pregnant women.

HBsAgRecommendedRecommended (only if patient is high risk)NSRecommendedRecommended
Treatment of HBsAg positivityYesYesNSYesYes
HBcAgRecommendedNSNSNSRecommended
Treatment of anti-HBc positivityNo; monitor during immunosuppressive therapyNSNSShould be considered in high-risk patientsNS
Anti-HBsRecommendedNSNSNSRecommended
Initiation of treatmentNSOnset of immunosuppressive therapyNSBefore immunosuppressive therapyBefore immunosuppressive therapy
Duration of treatmentNS6 months after end of immunosuppressive therapyNS6 months after end of immunosuppressive therapy; indefinite if pretreatment HBV DNA high; continue until HBV DNA undetectableConsultation with gastroenterologist/hepatologist
Preferred agentNSBased on anticipated length of use (<12 months: LMV, TLV; >12 months: ADF, ETC)NS Consultation with gastroenterologist/hepatologist

Our survey indicated that selective screening for high-risk individuals is >90% for those patients with baseline elevated liver-associated enzymes, a history of hepatitis, human immunodeficiency virus, a history of intravenous drug use, and are men who have sex with men. Given the high prevalence of HBV exposure (70%) and subsequent development of chronic HBV (17%) in the US in these high-risk populations (Table 4), screening is paramount to the prevention of morbidity and mortality (26). Unvaccinated patients, however, are often ignored, with only 46% of them being screened. Patients who are asymptomatic carriers to HBV have an increased risk of hepatocellular carcinoma and liver failure, and therefore the lack of screening in this at-risk population is concerning (27, 28). This could be addressed if universal screening becomes adopted as the standard of care as we recommend. The ACR guidelines currently recommend that nonimmune patients should be vaccinated prior to biologic therapy (25), but this survey did not collect data on how many rheumatologists follow this particular recommendation.

Table 4. Seroprevalence of hepatitis B virus (HBV) by region and risk factors (26), where ≥2% of the population indicates high prevalence
US high-risk populationsPercentage
  • *

    Hepatitis B surface antigen–positive for >6 months. HIV = human immunodeficiency virus.

Chronic HBV infection* 
 General population0.3
 HIV positive4–17
 Injection drug users3–6
 Men who have sex with men1–3
 Sexual contact with chronic HBV patients*3.5–9
 Household contact with chronic HBV patients*3–20
Ever infected with HBV 
 General population4.8
 HIV positive24–76
 Injection drug users20–70
 Men who have sex with men10–40
 Sexual contact with chronic HBV patients*25–59
 Household contact with chronic HBV patients*16–60

Nearly all responding rheumatologists (92%) correctly selected HBsAg as a means of screening, perhaps adhering to recommendations by Calabrese and colleagues from 2006, who recommended checking HBsAg, anti-HBs, and HBcAb (16, 24). However, these recommendations appear to have been incompletely followed because 51% and 43% of respondents selected only anti-HBs and HBcAb, respectively.

To our knowledge, this survey represents the first attempt to gather data regarding HBV reactivation in rheumatology patients undergoing immunosuppressive therapy with biologic DMARDs, and it may provide the framework for future research to establish the risk of reactivation for each individual biologic agent included in the survey. Our results suggest a HBV reactivation prevalence of 7.4% based on data from 10 of 136 respondents, representing only 15% of the rheumatologists sampled. The most commonly implicated biologic agent was infliximab, but other anti–tumor necrosis factor α agents (etanercept, adalimumab) as well as anti–B cell agents (rituximab) were also reported, all of which have been previously associated with HBV reactivation (4, 8–9, 16, 20–22). A prospective study of these agents and resultant HBV reactivation would be required to better assess the true risk of HBV reactivation in this setting, but such a study is unlikely to ever be performed given the severe outcomes that might occur.

In our survey, chronic HBV reactivation cases were reported for patients with RA undergoing both biologic and nonbiologic therapy. However, the body of literature assessing the prevalence of comorbid RA and chronic HBV is extremely limited (Table 5). To our knowledge, only 2 studies have investigated the presence of these two diseases concomitantly, with the prevalence of RA in chronic HBV patients determined to be 1–2% (29, 30). These studies neglected to stratify the data across high-risk groups. The paucity of data in this area may contribute to under-recognition of the importance of appropriate screening for chronic HBV, especially in high-risk populations. Conversely, limited data demonstrate the prevalence of chronic HBV in patients presenting with polyarthritis suggestive of RA to be no greater than the prevalence of chronic HBV in the general population, rendering serologic testing for HCV and HBV unnecessary for routine diagnostic evaluation (31). This may account for the low screening rates prior to prescribing corticosteroids or nonbiologic DMARDs (9% and 42%, respectively). Given our findings of a 5.9% prevalence of reactivation for patients undergoing nonbiologic DMARD therapy and the potentially fatal consequences of fulminant liver failure secondary to severe reactivation, we recommend universal screening for all patients starting nonbiologic DMARD therapy.

Table 5. RA prevalence in patients with chronic HBV*
 Han (41)Cacoub et al (29)Aprosina et al (42)Cesur et al (30)
  • *

    RA = rheumatoid arthritis; HBV = hepatitis B virus; NS = not specified; PMR = polymyalgia rheumatica; PM = polymyositis; PAN = polyarteritis nodosa; RP = Raynaud's phenomenon; SLE = systemic lupus erythematosus; HBsAg = hepatitis B surface antigen.

RA prevalenceNS3 (2%) of all 160 patients; 5 (16.7%) of 30 patients with extra-hepatic manifestationsNS4 (1%) of 400 subjects
SeverityRare radiographic evidence of destructionNSNSNS
Other rheumatic diseases associated with chronic HBVPMR, PMSicca syndrome, vasculitis, PAN, RP, psoriasisVasculitis, RP, PANSLE
Increased risks for extra-hepatic manifestations, including arthritisHigh concentration of HBsAgPrecore mutant HBVNSNS

The selection of the most appropriate prophylactic agent also appears to be a concern among rheumatologists. Half of the surveyed respondents would consider HBV prophylaxis only in high-risk patients prior to or concurrent with biologic agents. However, only 22% would prescribe antiviral therapy independently, with the majority preferring input from a gastroenterologist/hepatologist. Of the 22% who said they would prescribe antiviral therapy, 67% chose lamivudine. Lamivudine has been available for many years, which likely contributed to its higher usage, but is no longer considered to be first-line antiviral therapy for HBV treatment or prophylaxis due to reports and concerns of high rates of viral resistance (28, 32, 33). Although there are no recommendations for specific antiviral agent selection in the rheumatology literature, concern for lamivudine resistance in chemotherapy patients (32) can be extrapolated to that of immunosuppressed patients with rheumatic disease, and warrants the use of the newer antiviral agents entecavir or tenofovir, which are more potent and have extremely low rates of resistance to date (34–37).

Once prophylactic therapy has been instituted, the question of how frequently to monitor patients needs to be addressed. The majority of rheumatologists surveyed would defer to their gastroenterologist/hepatologist colleagues however, no formal guidelines presently exist. We concur with Calabrese and colleagues that antiviral therapy should commence prior to and continue for ≥6 months after the cessation of immunosuppressive therapy (24). The American Association for the Study of Liver Diseases published prophylaxis guidelines in February 2007 for oncology patients that echo these recommendations (28), although continuing antiviral therapy for up to 12 months after immunomodulating therapy is stopped is now being recommended (36). For patients found to have active viral replication (i.e., positive HBV DNA viral load, which should be tested in all HBsAg-positive patients along with hepatitis B e antigen and hepatitis B antibody positive status), treatment may need to be continued indefinitely due to high rates of relapse from shorter courses (37, 38).

Several areas of concern may arise from the methodology of the survey instrument. Response rates in health care research using e-mail have varied from 15% to 60%, with an average of 24% (39). Our relatively low response rate may initially be viewed as a weakness however, there is no body of literature that has determined a consensus for an acceptable response rate. In addition, it is increasingly recognized that despite a low response rate, a representative sample can still be obtained (40). In our study, the low response rate and relatively small sample size may be offset by the nationwide representativeness of the sample. Other factors that may have played a role in the relatively low response rate are the length of the survey, the specificity of the questions, and the absence of an offered incentive. Another potential concern is potential selection bias. Our results may be biased by the response of physicians with a vested interest in HBV reactivation and screening, thereby overreporting the actual prevalence of reactivation. Finally, the percentage of rheumatologists who perform screening for HBV may be under-reported if it is assumed that the majority of those choosing not to respond are in fact screening, but did not answer our survey.

In conclusion, the results of this survey and the growing body of literature regarding the morbidity and mortality associated with HBV reactivation in patients with rheumatic disease receiving immunosuppressive therapy demonstrate that improving education among rheumatologists regarding the risks of HBV reactivation and its prevention for patients receiving biologic agents is clearly warranted. Universal screening for HBV prior to initiating biologic DMARDs should be the ultimate goal to prevent HBV reactivation, which can be severe and sometimes fatal. At the present time, there remains a need for more specific, formal consensus guidelines to help make this goal as the standard of care because the current ACR guidelines make no HBV screening recommendations prior to biologic DMARD use. The confusion regarding prophylaxis and duration of followup necessitates further discussion with our rheumatology colleagues. Many patients will require chronic, ongoing HBV treatment, and it is our opinion that all high-risk patients, and those who are HBsAg positive, should be followed by a gastroenterologist/hepatologist prior to, during, and after cessation of biologic agents to determine the most appropriate agent, its duration of use, and followup, including hepatocellular carcinoma screening.

AUTHOR CONTRIBUTIONS

All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Khokhar had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Stine, Khokhar, Charalambopoulos, Shanmugam, Lewis.

Acquisition of data. Stine, Khokhar, Charalambopoulos, Shanmugam, Lewis.

Analysis and interpretation of data. Stine, Khokhar, Charalambopoulos, Shanmugam, Lewis.

Ancillary