Skip Navigation

Acute Hepatitis C Virus Infection in Incarcerated Injection Drug Users

  1. Barbara H. McGovern1,2,
  2. Alysse Wurcel1,3,a,
  3. Arthur Y. Kim3,
  4. Julian Schulze zur Wiesch3,4,
  5. Ioana Bica1,2,a,
  6. M. Tauheed Zaman1,3,
  7. Joerg Timm3,a,
  8. Bruce D. Walker3,4, and
  9. Georg M. Lauer3
  1. 1Lemuel Shattuck Hospital, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
  2. 2Tufts University School of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
  3. 3Partners AIDS Research Center/Infectious Disease Division, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts
  4. 4Howard Hughes Medical Institute, Chevy Chase, Maryland
  1. Reprints or correspondence: Dr. Barbara McGovern, Lemuel Shattuck Hospital, 170 Morton St., Jamaica Plain, MA 01230 (bmcgovern{at}tufts-nemc.org).
  1. Presented in part: 41st Annual meeting of the Infectious Diseases Society of America, San Diego, California, 8–12 October 2003 (abstract 592).

  • a Present affiliations: Medical Student, University of Pennsylvania Medical School, Philadelphia, Pennsylvania (A.W.); Boston Medical Center, Boston, Massachusetts (I.B.); and Zentrum für Innere Medizin, Abteilung Gastroenterologie und Hepatologie, Berufsgenossenschaftliche Kliniken Bergmannsheil, Ruhr-Universitaet Bochum, Bochum, Germany (J.T.).

Hepatitis C virus (HCV) infection accounts for 15%–20% of cases of acute hepatitis in the United States, with the leading risk factor being injection drug use [1]. Although there is a growing body of literature on the natural history of acute HCV infection in patients who have acquired infection through a variety of routes (via transfusion, needlestick injury, sexual transmission, or medical procedures) there are few prospective studies that have focused on injection drug users (IDUs) [28]. Identification of these high-risk individuals would not only add to our understanding of the early clinical, immunologic, and virologic events in this subset of patients, but it would also enable opportunities for counseling and screening for other bloodborne pathogens.

The Centers for Disease Control and Prevention (CDC) has emphasized the importance of primary and secondary HCV-prevention programs for IDUs, and in 2003, the CDC specifically targeted correctional facilities as ideal settings for these interventions [9, 10]. Prisons and jails represent a concentration of underserved, minority groups that are at high risk for HCV infection [11]. In a 1997 survey, 57% of state prisoners reported using illicit drugs in the month before their offense [9]. Because of the strong association between new injection drug use and acquisition of HCV infection, we surmised that screening for acute hepatitis would be most fruitful among patients housed in prisons or detoxification facilities [12].

Although the prevalence of chronic HCV infection in the prison system is known to be as high as 34%, there is little information about newly acquired infection in recently admitted inmates [13]. The identification of patients with acute HCV infection has taken on greater importance, because early treatment with IFN-α can lead to eradication of the virus in up to 98% of cases—a much higher rate than among persons with chronic disease [2, 14, 15]. Determination of serial HCV RNA levels may help to differentiate patients whose infections spontaneously resolve from those whose infections become persistent and require treatment [3, 4]. We surmised that the correctional system would offer the needed stability to accomplish these goals.

Lemuel Shattuck Hospital (Boston, MA), an affiliate of the Massachusetts Department of Public Health, serves both community and incarcerated patients, many of whom have a history of injection drug use. To determine the feasibility of identifying newly infected patients, we encouraged medical providers in the prison-based clinics and our affiliated substance abuse program to contact us regarding patients with symptoms, signs, and/or laboratory markers consistent with acute hepatitis. Our aims were (1) to identify and describe the clinical course of patients with acute HCV infection; (2) to offer treatment, immunization, and counseling about hepatitis C and other bloodborne diseases; and (3) to study early immunologic and virologic events in IDUs.

Methods

Study subjects. Beginning in November 2001, we gave educational seminars for medical providers on hepatitis C and requested that all personnel within the Massachusetts correctional system and our affiliated substance abuse program refer any patient with symptoms of hepatitis (i.e., fatigue, nausea, vomiting, right upper quadrant pain, anorexia, jaundice, or dark urine) or an alanine aminotransferase level >5 times the upper limit of normal. Patients were evaluated at Lemuel Shattuck Hospital by an infectious diseases specialist.

At the initial clinic visit, we obtained a full history and performed a physical examination and laboratory tests for hepatitis A, B, and C serologic findings, quantitative HCV RNA level, HCV genotype, HIV antibody, and HIV RNA and heterophile antibodies. With the permission of the patient, we obtained medical records to confirm documentation of an HCV-seronegative status.

Patients were categorized as having “definite” or “probable” acute HCV infection on the basis of the following criteria: HCV-seronegative patients with detectable HCV RNA on presentation (as determined by Quest Nichols Real-Time PCR; Quest Diagnostics) and subsequent seroconversion were categorized as having “definite” acute HCV infection; patients with abnormal transaminase levels, a positive HCV antibody test result, detectable HCV RNA, and a documented seronegative status within the prior year were also categorized as having “definite” acute HCV infection; and patients who had not undergone prior HCV testing, who had recent onset of injection drug use, and who were HCV seropositive at the time of presentation with detectable HCV RNA and an alanine aminotransferase level >10 times the upper limit of normal were defined as having “probable” acute HCV infection [3]. Information was collected regarding risk-taking behaviors.

At subsequent clinic visits, HCV RNA levels and aminotransferase levels were determined at 4-week intervals to determine the patient's virologic and biochemical outcome. If a patient had a quantitatively undetectable HCV RNA level, then a qualitative HCV RNA assay was performed (lower limit of detection, <50 IU/mL). For patients with persistent viremia >4 months after presentation, 6 months of pegylated IFN-α2b treatment was offered to patients in the community and to any inmate who would be incarcerated for an additional 6 months as part of a Human Research Review Committee—approved study.

Patients were also invited to participate in an institutional review board—approved immunology study of adaptive immune responses against HCV infection based at Massachusetts General Hospital. These protocols conform to the 1975 Helsinki guidelines for the conduct of human research and were approved by the Lemuel Shattuck Hospital Investigational Review Board, which includes a Prisoner Advocate. Patient consent forms for inmates had specific contact information for a second prisoner advocate, who was not participating in this research, to address any concerns by the participants.

Proliferation assays. Lymphocyte proliferation assays were performed, as described elsewhere, using recombinant HCV proteins representing core, NS3, NS4, and NS5 [16].

Sequencing of autologous virus. Viral RNA was extracted from plasma samples and was sequenced as described elsewhere [17].

Statistical analysis. Phylogenetic trees were generated using the maximum likelihood method available in the Phylogenetic Analysis Using Parsimony package, and bootstrap resampling analysis was undertaken with 100 replicate neighbor-joining trees. Reference sequences for genotype 1a and genotype 1b were obtained from the Los Alamos HCV database (http://www.hcv.lanl.gov).

Results

Clinical presentation of acute HCV infection in IDUs. Twenty-five patients with acute hepatitis were referred to our clinic during a 30-month period (November 2001–May 2004), including 23 newly incarcerated inmates, 1 patient from an affiliated detoxification center, and 1 patient from the community. All patients were evaluated for signs or symptoms of acute hepatitis and/or severe abnormalities in aminotransferase levels. Of the 25 referrals to our clinic, all patients had acute viral hepatitis (HCV infection, 21 patients; hepatitis B virus infection, 3 patients; and hepatitis A virus infection, 1 patient).

Of the 21 patients with acute HCV infection, 19 individuals were identified in the prison. Of the 16 inmates who acknowledged injection drug use, the mean time to onset of acute hepatitis was 44 days from the day of incarceration (range, 3–123 days). Reasons for initial evaluation varied, as follows: 13 patients had an onset of symptoms of hepatitis, 4 asymptomatic subjects were evaluated for elevated aminotransferase levels that had been detected incidentally, and 2 subjects were tested for HCV antibodies by patient request. Of the remaining 2 patients, one (subject R) was identified on the basis of an onset of symptoms while attending a substance abuse program, and another (subject C) was identified during a hospitalization for acute hepatitis.

Of the 21 HCV-infected patients in the cohort, 14 were classified as having definite acute HCV infection (tables 1 and 2). Although we classified 7 cases as probable acute HCV infection, the lack of alternative diagnoses and the mean alanine aminotransferase level (1096 IU/mL; range, 644–1553 IU/mL) support a high probability of acute HCV infection. Furthermore, 2 subjects had subsequent resolution of viremia, which provides further evidence of true acute infection in these individuals.

Figure 1

Serial measurements of hepatitis C virus (HCV) RNA levels in subjects with acute HCV infection. Serial HCV RNA levels were determined in patients with acute HCV infection. Results for subjects A, B, H, T, and U (A), all of whom had controlled HCV infection, were determined by an HCV level less than the level of detection (<600 IU/mL) in the HCV PCR assay (Roche Amplicor assay). The mean number of samples per patient was 4.2. Patients with an HCV RNA level <600 IU/mL had results verified with a qualitative HCV RNA. All subjects had a significant decrease in the HCV load (>3 logs) within 10 weeks after diagnosis. In contrast to the data from subjects depicted in panel A, HCV RNA levels in patients E, F, N, Q, and R (B) did not decrease significantly but rather increased, remained stable, or fluctuated, without any clear trend. These subjects all developed chronic HCV infection.

Figure 2

Relapse of viremia after intermittent control in subject A. Phylogenetic analysis of viral isolates recovered during the first and second episodes of viremia of subject A revealed identical viral strains during both episodes of viremia (A). A neighbor-joining tree was constructed using the PAUP package (Sinauer Associates), and bootstrap resampling analysis was undertaken with 1000 replicate neighbor-joining trees. Reference sequences for genotype 1a and genotype 1b were obtained from the Los Alamos hepatitis C virus (HCV) database. B, Longitudinal follow-up of CD4+ T cell proliferative responses in subject A. During the phase of initial viremia and subsequent control, a vigorous proliferative response, limited to HCV core, was observed. This CD4+ T cell response disappeared at the same time that the subject experienced virus rebound. SI, stimulation index.

Table 1

Demographic characteristics and classification categories for subjects with acute hepatitis C virus (HCV) infection.

Table 2

Symptoms, laboratory test results, and clinical outcomes for patients with acute hepatitis C virus (HCV) infection.

Injection drug use was the main risk factor for acquisition of HCV infection: 17 subjects reported either new onset of injection drug use (8 subjects), relapse of injection drug use (2 subjects), new paraphernalia-sharing partner (3 subjects), or changes in harm-reduction techniques (4 subjects) (table 1). Interestingly, 2 patients with definite acute HCV infection (subjects I and J) never acquired HCV infection during the 3 years that they shared needles exclusively with each other. Both subsequently developed symptoms of acute HCV infection nearly simultaneously, after subject I shared needles with a third partner for only 1 day. All patients also agreed to undergo HIV testing. One patient was found to be HIV seropositive, and 1 other patient had known HIV infection.

Of the 21 patients with acute HCV infection, only 3 individuals did not report injection drug use as a primary risk factor (table 1). One patient (subject Q), who had tested seronegative for HCV on multiple occasions in prison, became symptomatic 10 weeks after having a traumatic altercation with an HCV-seropositive inmate. The second patient had a history of intranasal cocaine use and heterosexual intercourse with an HCV-seropositive partner (subject O). One inmate denied any risk factors (subject N).

Seventeen of the 21 patients with acute HCV infection were observed for a mean of 6.3 months (range, 1–12 months), allowing us to determine the early natural history of HCV in this subgroup (table 2). Eight of the 17 patients cleared virus spontaneously. Six of 11 women had spontaneous control, compared with 2 of 7 men. Symptoms and signs have been suggested to be predictors of viral clearance. In our cohort, 6 of the 13 patients with symptoms cleared virus, whereas 2 of 4 asymptomatic patients also cleared the virus. Although the presence of jaundice has been associated with viral clearance, 4 of 10 patients with this sign developed persistent viremia. Therefore, the presence or absence of symptoms or signs did not accurately predict clearance, although we acknowledge that the cohort was relatively small.

Serial measurements of HCV RNA level in patients with acute HCV infection. Four of the patients with acute HCV infection left prison and were lost to follow-up; thus, we analyzed multiple serial samples for HCV RNA titers in a subgroup of 10 subjects 1–18 weeks after presentation. In the subjects with spontaneous clearance, we noted a ⩾4-log decrease in hepatitis C RNA over a 4–9-week interval, with 4 of 5 patients reaching an undetectable viral load by week 9 after presentation (figure 1A). In contrast, the 5 patients who developed chronic infection had either a lower slope in the viral load's decrease or no decrease at all (figure 1B). Two patients (subjects L and R) with persistent viremia also had preexisting HIV infection, a risk factor for chronic evolution of HCV [18].

Relapse of HCV viremia after initial spontaneous clearance. Two patients had recurrent viremia after initial spontaneous control and biochemical normalization. Subject A denied having any risk factors for reacquisition of HCV infection. PCR sequencing verified that the second episode of viremia represented a recurrence of the original infecting virus, as demonstrated by the presence of a virtually identical hypervariable region of E2 (figure 2A) and by full genome sequence analysis (data not shown). We also performed a longitudinal analysis of the subject's CD4+ T cell proliferative responses to HCV antigens (figure 2B). During her initial phase of successful virologic control, we detected a focused CD4+ T cell proliferative response to HCV core; this disappeared when virus reemerged.

Subject D, who initially achieved spontaneous viral control, admitted to relapse of injection drug use with an HCV-seropositive partner and was again viremic (viral load, 8388 IU/mL). Interestingly, her first episode of hepatitis C was accompanied by jaundice, but she was completely asymptomatic during her second episode of viremia. After her second exposure, she began to exhibit control of viremia, because her HCV RNA level decreased to 780 IU/mL. Her ultimate clinical outcome is unknown because she was lost to follow-up. Although we cannot rule out the possibility of viral relapse, the patient's recurrent viremia followed high-risk behavior. Unfortunately, because of the rapid clearance of HCV, we were unable to amplify virus for sequencing from the first episode of viremia for comparison with the strain that emerged later.

Response to treatment in patients with persistent HCV viremia. Five of 9 patients with persistent viremia (subjects A, I, J, N, and S) were available for long-term medical care and initiated treatment with pegylated IFN-α2b monotherapy for HCV genotype 1 infection. Subject A had a rapid virologic response at week 4 but discontinued therapy at week 7 because of adverse effects, and this subject had viral rebound. Subjects I, J, and N initiated treatment at 12, 11, and 5 months after presentation, respectively. All achieved a nondetectable HCV RNA level after 4 weeks of supervised therapy. Subjects I and N had virologic relapse 4 weeks after completion of therapy. Subjects S and J had nondetectable HCV RNA levels 24 weeks after completion of therapy.

Discussion

The CDC has issued a mandate for the prevention of HCV transmission through identification and counseling of persons at risk and through appropriate evaluation of those who are infected [10]. Although it is widely recognized that IDUs represent the highest priority group for focused interventions, implementation of such programs is difficult in this patient population.

Here, we demonstrate that the correctional system is an excellent environment for testing, counseling, and interventions for acute HCV infection. In contrast to the widely recognized high prevalence of chronic HCV infection in the prison setting [19], the relatively easy identification of acute HCV infection has not been well appreciated. Surveillance for new-onset hepatitis led to recognition of a large number of acute HCV infections, most of which were symptomatic. We suspect that these index cases were just “the tip of the iceberg,” because a prospective study demonstrated that most infections among IDUs are asymptomatic [20]. Although the CDC had recommended implementation of screening programs in the correctional setting, whether inmates would consent to such interventions was unknown [9]. However, we found that patient acceptance of medical care and screening for other bloodborne diseases, such as HIV infection, was excellent. We also used stringent criteria for assessment of elevations in the alanine aminotransferase level to avoid identification of alcohol-induced hepatitis in our screening; however, these strict criteria may have led to missed cases of acute HCV infection.

Detailed discussions with our patients also emphasized important issues regarding HCV transmission. HCV acquisition followed loss of access to needle-exchange centers, highlighting the need for wider availability of harm-reduction programs [9]. One inmate acquired acute HCV infection after a traumatic altercation. This route has been rarely described, strengthening the importance of infection-control measures and patient follow-up under these circumstances.

Although African American persons constitute >25% of the Massachusetts prison population, only white and Hispanic patients were identified during our investigation. The reason for this intriguing observation is unclear. One explanation may be related to our case definition. It is unknown whether African American patients are less likely to present with symptomatic disease or to have lower transaminase levels during the acute phase of infection than their white or Hispanic counterparts. The likelihood that different clinical presentations may be related to underlying racial differences is supported by lower treatment response rates in patients with chronic infection and lower viral clearance rates in patients with acute HCV infection [6, 7, 21, 22]. Alternatively, risk-taking behaviors may differ between these racial groups. These hypotheses would need to be studied in a rigorous prospective manner.

The structured environment of the correctional system also allowed us to study the natural history of primary HCV infection in IDUs. We found that 54% of the subjects in our cohort spontaneously cleared viremia, which is consistent with other published data [3, 4]. A significant number of patients who acquire HCV infection may initially clear viremia but become persistently infected with repeated exposures, as occurs in ongoing IDUs [18]. HCV antibody does not provide protection against recurrent infection, although a history of spontaneous virologic clearance may be favorable [18, 23].

Although no specific clinical parameter predicted viral clearance, decreasing HCV RNA levels during early infection were indicative of clinical outcome in our cohort, as demonstrated in other risk groups [3, 4]. Subjects who did not attain a nondetectable HCV RNA level or a 4-log decrease in the viral load within 10 weeks after presentation developed chronic infection. In contrast, attainment of a nondetectable HCV RNA level or a precipitous decrease in the viral load over the same period was associated with viral clearance. This held true even in an individual (subject H) who did not achieve viral suppression until 18 weeks after symptom onset.

A strategy of determining serial HCV RNA levels may assist in identification of patients whose infection will resolve spontaneously, avoiding the unnecessary drug toxicity and cost associated with treatment. Pharmaceutical costs are especially important in the correctional system, where marginal budgets are already constrained by the large burden of infectious diseases and other comorbidities. As a result of budget constraints, delays in treatment may have had a negative effect on treatment outcome in some patients. We are unaware of any other state prison system in the United States with a policy of offering therapy for acute HCV infection to inmates. Only a minority of states offer treatment for chronic HCV infection, although we and others have demonstrated the feasibility of these prison-based initiatives [2426].

Testing of HCV-specific lymphoproliferative responses may also predict spontaneous resolution or virologic relapse. The correlation of the proliferative response to outcome was well demonstrated in subject A (figure 2B). The recurrence of viremia in this individual after initial spontaneous control was preceded by evidence of waning CD4+ T cell responses. Such cases of virologic relapse highlight the importance of long-term clinical follow-up [27].

In conclusion, intensified screening for HCV infection among high-risk individuals has demonstrated that acute HCV infection is common within prison walls. The importance of including inmates in strategies for prevention of HCV infection is well recognized [28]. Management of acute infection in IDUs should be coupled with drug rehabilitation programs to prevent reinfection with HCV and other transmissible agents. Our pilot project has led to funding for a larger prospective study, in which all patients at risk—regardless of signs or symptoms—will be screened for acute HCV infection. This systematic approach will extend our knowledge of the full spectrum of acute HCV infection. It will also provide an avenue for harm-reduction counseling, testing for HIV infection, immunization for other hepatitis viruses, and early treatment interventions, as recommended by the CDC [9].

The vast majority of inmates return to the community within 2 years of incarceration [29]. Improved access to medical care and preventive services for incarcerated populations can benefit communities by decreasing rates of transmission [30]. If we look upon incarceration as a “window of opportunity,” we can assist those in need of intervention while advancing worthy public health and research initiatives.

Acknowledgments

We are indebted to the patients who participated in the study and to the many health care professionals who helped us with medical management.

Financial support. General Clinical Research Center, funded by the National Center for Research Resources of the National Institutes of Health (NIH; MO1-RR00054 to B.H.M.); Liver Scholar Award from The American Liver Foundation (G.M.L.); NIH (AI31563); and Howard Hughes Medical Institute.

Potential conflicts of interest. All authors: no conflicts.

  • Received December 4, 2005.
  • Accepted February 11, 2006.

References

| Table of Contents