Cost and Outcome of BehaviouRal Activation (COBRA): a randomised controlled trial of behavioural activation versus cognitive-behavioural therapy for depression

BACKGROUND

Depression is a common, debilitating and costly disorder. The best-evidenced psychological therapy - cognitive-behavioural therapy (CBT) - is complex and costly. A simpler therapy, behavioural activation (BA), may be an effective alternative.

OBJECTIVES

To determine the clinical effectiveness and cost-effectiveness of BA compared with CBT for depressed adults at 12 and 18 months' follow-up, and to investigate the processes of treatments.

DESIGN

Randomised controlled, non-inferiority trial stratified by depression severity, antidepressant use and recruitment site, with embedded process evaluation; and randomisation by remote computer-generated allocation.

SETTING

Three community mental health services in England.

PARTICIPANTS

Adults aged ≥ 18 years with major depressive disorder (MDD) recruited from primary care and psychological therapy services.

INTERVENTIONS

BA delivered by NHS junior mental health workers (MHWs); CBT by NHS psychological therapists.

OUTCOMES

Primary: depression severity (as measured via the Patient Health Questionnaire-9; PHQ-9) at 12 months. Secondary: MDD status; number of depression-free days; anxiety (as measured via the Generalised Anxiety Disorder-7); health-related quality of life (as measured via the Short Form questionnaire-36 items) at 6, 12 and 18 months; and PHQ-9 at 6 and 18 months, all collected by assessors blinded to treatment allocation. Non-inferiority margin was 1.9 PHQ-9 points. We undertook intention-to-treat (ITT) and per protocol (PP) analyses. We explored cost-effectiveness by collecting direct treatment and other health- and social-care costs and calculating quality-adjusted life-years (QALYs) using the EuroQol-5 Dimensions, three-level version, at 18 months.

RESULTS

We recruited 440 participants (BA, n = 221; CBT, n = 219); 175 (79%) BA and 189 (86%) CBT participants provided ITT data and 135 (61%) BA and 151 (69%) CBT participants provided PP data. At 12 months we found that BA was non-inferior to CBT {ITT: CBT 8.4 PHQ-9 points [standard deviation (SD) 7.5 PHQ-9 points], BA 8.4 PHQ-9 points (SD 7.0 PHQ-9 points), mean difference 0.1 PHQ-9 points, 95% confidence interval (CI) -1.3 to 1.5 PHQ-9 points, p = 0.89; PP: CBT 7.9 PHQ-9 points (SD 7.3 PHQ-9 points), BA 7.8 PHQ-9 points (SD 6.5 PHQ-9 points), mean difference 0.0 PHQ-9 points, 95% CI -1.5 to 1.6 PHQ-9 points, p = 0.99}. We found no differences in secondary outcomes. We found a significant difference in mean intervention costs (BA, £975; CBT, £1235; p < 0.001), but no differences in non-intervention (hospital, community health, social care and medication costs) or total (non-intervention plus intervention) costs. Costs were lower and QALY outcomes better in the BA group, generating an incremental cost-effectiveness ratio of -£6865. The probability of BA being cost-effective compared with CBT was almost 80% at the National Institute for Health and Care Excellence's preferred willingness-to-pay threshold of £20,000-30,000 per QALY. There were no trial-related adverse events.

LIMITATIONS

In this pragmatic trial many depressed participants in both groups were also taking antidepressant medication, although most had been doing so for a considerable time before entering the trial. Around one-third of participants chose not to complete a PP dose of treatment, a finding common in both psychotherapy trials and routine practice.

CONCLUSIONS

We found that BA is as effective as CBT, more cost-effective and can be delivered by MHWs with no professional training in psychological therapies.

FUTURE WORK

Settings and countries with a paucity of professionally qualified psychological therapists, might choose to investigate the delivery of effective psychological therapy for depression without the need to develop an extensive and costly professional infrastructure.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN27473954.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 21, No. 46. See the NIHR Journals Library website for further project information.

Clinical effectiveness, cost-effectiveness and acceptability of low-intensity interventions in the management of obsessive-compulsive disorder: the Obsessive-Compulsive Treatment Efficacy randomised controlled Trial (OCTET)

BACKGROUND

The Obsessive-Compulsive Treatment Efficacy randomised controlled Trial emerged from a research recommendation in National Institute for Health and Care Excellence obsessive-compulsive disorder (OCD) guidelines, which specified the need to evaluate cognitive-behavioural therapy (CBT) treatment intensity formats.

OBJECTIVES

To determine the clinical effectiveness and cost-effectiveness of two low-intensity CBT interventions [supported computerised cognitive-behavioural therapy (cCBT) and guided self-help]: (1) compared with waiting list for high-intensity CBT in adults with OCD at 3 months; and (2) plus high-intensity CBT compared with waiting list plus high-intensity CBT in adults with OCD at 12 months. To determine patient and professional acceptability of low-intensity CBT interventions.

DESIGN

A three-arm, multicentre, randomised controlled trial.

SETTING

Improving Access to Psychological Therapies services and primary/secondary care mental health services in 15 NHS trusts.

PARTICIPANTS

Patients aged ≥ 18 years meeting Diagnostic and Statistical Manual of Mental Disorders-Fourth Edition criteria for OCD, on a waiting list for high-intensity CBT and scoring ≥ 16 on the Yale-Brown Obsessive Compulsive Scale (indicative of at least moderate severity OCD) and able to read English.

INTERVENTIONS

Participants were randomised to (1) supported cCBT, (2) guided self-help or (3) a waiting list for high-intensity CBT.

MAIN OUTCOME MEASURES

The primary outcome was OCD symptoms using the Yale-Brown Obsessive Compulsive Scale - Observer Rated.

RESULTS

Patients were recruited from 14 NHS trusts between February 2011 and May 2014. Follow-up data collection was complete by May 2015. There were 475 patients randomised: supported cCBT (n = 158); guided self-help (n = 158) and waiting list for high-intensity CBT (n = 159). Two patients were excluded post randomisation (one supported cCBT and one waiting list for high-intensity CBT); therefore, data were analysed for 473 patients. In the short term, prior to accessing high-intensity CBT, guided self-help demonstrated statistically significant benefits over waiting list, but these benefits did not meet the prespecified criterion for clinical significance [adjusted mean difference -1.91, 95% confidence interval (CI) -3.27 to -0.55; p = 0.006]. Supported cCBT did not demonstrate any significant benefit (adjusted mean difference -0.71, 95% CI -2.12 to 0.70). In the longer term, access to guided self-help and supported cCBT, prior to high-intensity CBT, did not lead to differences in outcomes compared with access to high-intensity CBT alone. Access to guided self-help and supported cCBT led to significant reductions in the uptake of high-intensity CBT; this did not seem to compromise patient outcomes at 12 months. Taking a decision-making approach, which focuses on which decision has a higher probability of being cost-effective, rather than the statistical significance of the results, there was little evidence that supported cCBT and guided self-help are cost-effective at the 3-month follow-up compared with a waiting list. However, by the 12-month follow-up, data suggested a greater probability of guided self-help being cost-effective than a waiting list from the health- and social-care perspective (60%) and the societal perspective (80%), and of supported cCBT being cost-effective compared with a waiting list from both perspectives (70%). Qualitative interviews found that guided self-help was more acceptable to patients than supported cCBT. Professionals acknowledged the advantages of low intensity interventions at a population level. No adverse events occurred during the trial that were deemed to be suspected or unexpected serious events.

LIMITATIONS

A significant issue in the interpretation of the results concerns the high level of access to high-intensity CBT during the waiting list period.

CONCLUSIONS

Although low-intensity interventions are not associated with clinically significant improvements in OCD symptoms, economic analysis over 12 months suggests that low-intensity interventions are cost-effective and may have an important role in OCD care pathways. Further research to enhance the clinical effectiveness of these interventions may be warranted, alongside research on how best to incorporate them into care pathways.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN73535163.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 21, No. 37. See the NIHR Journals Library website for further project information.

Low-intensity cognitive-behaviour therapy interventions for obsessive-compulsive disorder compared to waiting list for therapist-led cognitive-behaviour therapy: 3-arm randomised controlled trial of clinical effectiveness

BACKGROUND

Obsessive-compulsive disorder (OCD) is prevalent and without adequate treatment usually follows a chronic course. "High-intensity" cognitive-behaviour therapy (CBT) from a specialist therapist is current "best practice." However, access is difficult because of limited numbers of therapists and because of the disabling effects of OCD symptoms. There is a potential role for "low-intensity" interventions as part of a stepped care model. Low-intensity interventions (written or web-based materials with limited therapist support) can be provided remotely, which has the potential to increase access. However, current evidence concerning low-intensity interventions is insufficient. We aimed to determine the clinical effectiveness of 2 forms of low-intensity CBT prior to high-intensity CBT, in adults meeting the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM-IV) criteria for OCD.

METHODS AND FINDINGS

This study was approved by the National Research Ethics Service Committee North West-Lancaster (reference number 11/NW/0276). All participants provided informed consent to take part in the trial. We conducted a 3-arm, multicentre randomised controlled trial in primary- and secondary-care United Kingdom mental health services. All patients were on a waiting list for therapist-led CBT (treatment as usual). Four hundred and seventy-three eligible patients were recruited and randomised. Patients had a median age of 33 years, and 60% were female. The majority were experiencing severe OCD. Patients received 1 of 2 low-intensity interventions: computerised CBT (cCBT; web-based CBT materials and limited telephone support) through "OCFighter" or guided self-help (written CBT materials with limited telephone or face-to-face support). Primary comparisons concerned OCD symptoms, measured using the Yale-Brown Obsessive Compulsive Scale-Observer-Rated (Y-BOCS-OR) at 3, 6, and 12 months. Secondary outcomes included health-related quality of life, depression, anxiety, and functioning. At 3 months, guided self-help demonstrated modest benefits over the waiting list in reducing OCD symptoms (adjusted mean difference = -1.91, 95% CI -3.27 to -0.55). These effects did not reach a prespecified level of "clinically significant benefit." cCBT did not demonstrate significant benefit (adjusted mean difference = -0.71, 95% CI -2.12 to 0.70). At 12 months, neither guided self-help nor cCBT led to differences in OCD symptoms. Early access to low-intensity interventions led to significant reductions in uptake of high-intensity CBT over 12 months; 86% of the patients allocated to the waiting list for high-intensity CBT started treatment by the end of the trial, compared to 62% in supported cCBT and 57% in guided self-help. These reductions did not compromise longer-term patient outcomes. Data suggested small differences in satisfaction at 3 months, with patients more satisfied with guided self-help than supported cCBT. A significant issue in the interpretation of the results concerns the level of access to high-intensity CBT before the primary outcome assessment.

CONCLUSIONS

We have demonstrated that providing low-intensity interventions does not lead to clinically significant benefits but may reduce uptake of therapist-led CBT.

TRIAL REGISTRATION

International Standard Randomized Controlled Trial Number (ISRCTN) Registry ISRCTN73535163.

Identifying depression with the PHQ-2: A diagnostic meta-analysis

BACKGROUND

There is interest in the use of very brief instruments to identify depression because of the advantages they offer in busy clinical settings. The PHQ-2, consisting of two questions relating to core symptoms of depression (low mood and loss of interest or pleasure), is one such instrument.

METHOD

A systematic review was conducted to identify studies that had assessed the diagnostic performance of the PHQ-2 to detect major depression. Embase, MEDLINE, PsychINFO and grey literature databases were searched. Reference lists of included studies and previous relevant reviews were also examined. Studies were included that used the standard scoring system of the PHQ-2, assessed its performance against a gold-standard diagnostic interview and reported data on its performance at the recommended (≥3) or an alternative cut-off point (≥2). After assessing heterogeneity, where appropriate, data from studies were combined using bivariate diagnostic meta-analysis to derive sensitivity, specificity, likelihood ratios and diagnostic odds ratios.

RESULTS

21 studies met inclusion criteria totalling N=11,175 people out of which 1529 had major depressive disorder according to a gold standard. 19 of the 21 included studies reported data for a cut-off point of ≥3. Pooled sensitivity was 0.76 (95% CI =0.68-0.82), pooled specificity was 0.87 (95% CI =0.82-0.90). However there was substantial heterogeneity at this cut-off (I(2)=81.8%). 17 studies reported data on the performance of the measure at cut-off point ≥2. Heterogeneity was I(2)=43.2% pooled sensitivity at this cut-off point was 0.91 (95% CI =0.85-0.94), and pooled specificity was 0.70 (95% CI =0.64-0.76).

CONCLUSION

The generally lower sensitivity of the PHQ-2 at cut-off ≥3 than the original validation study (0.83) suggests that ≥2 may be preferable if clinicians want to ensure that few cases of depression are missed. However, in situations in which the prevalence of depression is low, this may result in an unacceptably high false-positive rate because of the associated modest specificity. These results, however, need to be interpreted with caution given the possibility of selectively reported cut-offs.

Screening for anxiety disorders with the GAD-7 and GAD-2: a systematic review and diagnostic metaanalysis

OBJECTIVE

To systematically review the accuracy of the GAD-7 and GAD-2 questionnaires for identifying anxiety disorders.

METHODS

A systematic review of the literature was conducted to identify studies that validated the GAD-7 or GAD-2 against a recognized gold standard diagnosis. Pooled estimates of diagnostic test accuracy were produced using random-effects bivariate metaanalysis. Heterogeneity was explored using the I(2) statistic.

RESULTS

A total of 12 samples were identified involving 5223 participants; 11 samples provided data on the accuracy of the GAD-7 for identifying generalized anxiety disorder (GAD). Pooled sensitivity and specificity values appeared acceptable at a cutoff point of 8 [sensitivity: 0.83 (95% CI 0.71-0.91), specificity: 0.84 (95% CI 0.70-0.92)] although cutoff scores 7-10 also had similar pooled estimates of sensitivity/specificity. Six samples provided data on the accuracy of the GAD-2 for identifying GAD. Pooled sensitivity and specificity values appeared acceptable at a cutoff of 3 [sensitivity: 0.76 (95% CI 0.55-0.89), specificity: 0.81 (95% CI 0.60-0.92)]. Four studies looked at the accuracy of the questionnaires for identifying any anxiety disorder.

CONCLUSIONS

The GAD-7 had acceptable properties for identifying GAD at cutoff scores 7-10. The GAD-2 had acceptable properties for identifying GAD at a cutoff score of 3. Further validation studies are needed.

Quality of interaction between at-risk infants and caregiver at 12-15 months is associated with 3-year autism outcome

BACKGROUND

  Recent models of the early emergence of autism spectrum disorder (ASD) propose that infant intrinsic risk susceptibilities in behaviour may be amplified by interaction within the early social environment into an increasingly atypical developmental trajectory. This study examines whether 6- and 12-month parent-infant interactions in at-risk siblings differ from those with low-risk and whether--in at-risk siblings--such interactions predict later 3-year classification of ASD or no ASD.

METHOD

  Within the British Autism Study of Infant Siblings (BASIS), 6-min videotaped episodes of parent-infant free play in infants at 6-10 months (45 at-risk siblings and 47 low-risk siblings) and 12-15 months (43 at-risk siblings and 48 low-risk siblings) in a laboratory setting were rated on the Manchester Assessment of Caregiver-Infant Interaction (MACI), blind to participant information. Standard tests were administered for concurrent behavioural signs of ASD features and developmental level. Systematic consensus diagnostic classification of ASD was made at 3 years for the at-risk siblings.

RESULTS

  Parent nondirectiveness and sensitive responsiveness differed in relation to ASD/risk status (at-risk ASD, at-risk no-ASD and low-risk) at both 6 and 12 months. At 6 months, infant liveliness was lower in the at-risk groups; at 12 months, infant attentiveness to parent and positive affect were lower in the at-risk group later diagnosed with ASD. Dyadic mutuality and intensity of engagement showed a group effect at 12 months. Dyadic mutuality, infant positive affect and infant attentiveness to parent at 12 months (but not 6 months) predicted 3-year ASD outcome, whereas infant ASD-related behavioural atypicality did not.

CONCLUSIONS

  This is the first prospective evidence that early dyadic interaction between at-risk infants and their parents is associated with later diagnostic outcome in ASD. Possible explanations for these findings and their theoretical implications are considered.

Parent-infant interaction in infant siblings at risk of autism

Recent models of the early emergence of autism spectrum disorder (ASD) propose an interaction between risk susceptibility and the infant's social environment, resulting in a progressively atypical developmental trajectory. The infant's early social environmental experience consists mostly of interaction with caregivers, yet there has been little systematic study of early parent-infant interaction in infants at risk of ASD. This study examined the global characteristics of parent-infant interaction in 6- to 10-month-old infants with an older sibling diagnosed with ASD (at-risk sibs), in comparison with a group of infants with no family history of ASD (low-risk sibs). As part of the British Autism Study of Infant Siblings (BASIS), 6-min videotaped unstructured play interactions of mother-infant dyads (45 at-risk sibs and 47 low-risk sibs) were rated on global aspects of parent-infant interaction, blind to participant information. Differences in global characteristics of interaction were observed in both infant and parent contributions in the at-risk group compared to low-risk controls. In analyses adjusted for age and developmental level, at-risk sib infants were less lively, and their parents showed higher directiveness, and lower sensitive responding (as a trend after adjustment). Level of infant liveliness was independent of other interactive behaviour. Consistent with reports in previous literature in older children with autism and in other neurodevelopmental disorders, our findings may suggest that infants at genetic risk are exposed to a more directive interactive style relatively early in infancy. We discuss possible explanations for these findings and implications for further developmental study and intervention.

Comprehensive proteomic study identifies serpin and cystatin antiproteases as novel correlates of HIV-1 resistance in the cervicovaginal mucosa of female sex workers

Not all individuals exposed to HIV-1 become infected, and evidence from HIV-1 highly exposed seronegative women (HIV-1-resistant) suggests that mucosal factors in the female genital tract, the first site of contact for the virus, are playing a role. To better understand factors mediating protection from HIV-1, we performed a large clinical study using the tools of systems biology to fully characterize the cervicovaginal mucosa proteome in HIV-1-resistant women. Cervicovaginal lavage fluid was collected from 293 HIV-1-resistant, uninfected, and infected sex workers and analyzed by 2D-LC LTQ-FT-MS. Of the more than 360 unique proteins identified, 41 were differentially abundant (>3-fold cutoff) in HIV-1-resistant women. The majority of over-abundant proteins were antiproteases (>40%), some with described anti-inflammatory and anti-HIV-1 activity. Quantification of specific anti-HIV-1 antiproteases Serpin A1, Serpin A3, and Cystatin B and an epithelial antiprotease A2ML1 found them to be significantly over-abundant in HIV-1-resistant women (p = 0.004; p = 0.046; p = 0.0003; and p = 0.04, respectively). Expression levels were not correlated to sexual practices or other epidemiological factors. Mucosal antiprotease levels correlated with pro-inflammatory cytokine concentration (p = <0.0001), but independently of pro-inflammatory cytokine levels in HIV-1-resistant women including TNF-alpha, IL-1 alpha, IL-1 beta, IL-6, and IL-8. This comprehensive systems biology approach identifies mucosal serpins and cystatins as novel correlates of HIV-1-resistance. This represents the first study characterizing these factors in the female genital tract.

Functional HIV-1 specific IgA antibodies in HIV-1 exposed, persistently IgG seronegative female sex workers

Although HIV-specific cellular immune responses are found in a number of HIV highly-exposed, persistently seronegative (HEPS) cohorts, late seroconversion can occur despite pre-existing cytotoxic T lymphocytes (CTL), suggesting that a protective HIV vaccine may need to induce a broader range of HIV-specific immune responses. Low levels of HIV-specific IgA have been found in the genital tract and plasma of the majority of Nairobi HEPS sex workers and appeared to be independent of HIV-specific cellular responses. IgA purified from genital tract, saliva and plasma of most HEPS sex workers were able to neutralize infection of PBMC by a primary (NSI) clade B HIV isolate, as well as viral isolates from clades A and D, which predominate in Kenya. In addition, these IgA were able to inhibit transcytosis of infective HIV virions across a transwell model of the human mucosal epithelium in an HIV-specific manner. Preliminary work in other HEPS cohorts has suggested the recognition of different gp41 epitopes in HEPS and HIV-infected subjects. Although present at low levels, these IgA demonstrated cross-clade neutralizing activity and were able to inhibit HIV mucosal transcytosis, suggesting an important functional role in protection against HIV infection.

Lymphocyte subsets in human immunodeficiency virus type 1-infected and uninfected children in Nairobi

BACKGROUND

Reference lymphocyte subset values for African children are lacking. This study documents these values as well as their alterations associated with perinatal and postnatal HIV-1 transmission and with protection from HIV-1 infection.

METHODS

Lymphocyte subsets were determined for HIV-1-seronegative nonpregnant women and their children (controls) and for uninfected, perinatally infected and postnatally infected children born to HIV-1-seropositive mothers in Nairobi, Kenya. The mean, median and 5th and 95th percentile values for CD4+ and CD8+ lymphocyte counts and percentages were determined and compared at the age ranges birth to 3 months, 4 months to 1 year, yearly from 1 to 5 years and from 6 to 10 years of age.

RESULTS

Among control children counts differed from published values of other populations. In all age ranges, whereas the absolute values were significantly higher than adult values, the percentages were significantly lower. Children perinatally infected with HIV-1 had clearly distinguishable differences in lymphocyte subset percentages by 3 months of age, when the median CD4+ percentage was 27.9% (5th to 95th percentile, 25.7 to 30.1%) for infected vs. 35.9% (33.3 to 38.7%) for uninfected and 39.9% (37.8 to 42.2%) for control children, P < 0.001; whereas the median CD8+ percentage was 37.0% (33.1 to 41.0%) for infected vs. 27.5% (24.2 to 30.8%) for uninfected and 27.5% (24.2 to 30.8%) for control children, P = 0.001. Differences between uninfected and control children disappeared after 1 year of age.

CONCLUSIONS

Normal lymphocyte subset values among African children differ from those in other populations. Significant differences are detectable by 3 months of age in CD4+ and CD8+ lymphocyte percentages among perinatally infected infants, which may be useful as an adjunct in diagnosis. Transient differences observed among HIV-1-exposed but uninfected infants could reflect a successful immune response to HIV-1 challenge.

Human immunodeficiency virus type 1-infected women exhibit reduced interferon-gamma secretion after Chlamydia trachomatis stimulation of peripheral blood lymphocytes

Epidemiologic, animal, and in vitro models suggest an important role for interferon (IFN)-gamma in the clearance of Chlamydia trachomatis infection. IFN-gamma in the supernatants of in vitro-stimulated peripheral blood mononuclear cells (PBMC) from 22 human immunodeficiency virus type 1 (HIV-1)-infected and 73 uninfected women at high risk for C. trachomatis acute pelvic inflammatory disease (PID) was studied. PBMC were stimulated with C. trachomatis purified major outer membrane protein (MOMP) and whole elementary bodies (EBs) from the 4 predominant serovars (E, F, K, and L2) that circulate in Nairobi. PBMC IFN-gamma secretion after stimulation with C. trachomatis EBs was significantly decreased in HIV-1-infected women. Among HIV-1-infected women, CD4 T cell depletion was associated with lower IFN-gamma secretion from PBMC stimulated with either C. trachomatis MOMP or EB antigen. Decreased antigen-specific IFN-gamma production may enhance the susceptibility of HIV-1-infected women to C. trachomatis PID.

Mucosal and plasma IgA from HIV-1-exposed uninfected individuals inhibit HIV-1 transcytosis across human epithelial cells

HIV-1-specific IgA has been described in the genital tract and plasma of HIV-1 highly exposed, persistently seronegative (HEPS) individuals, and IgA from these sites has been shown to neutralize HIV-1. This study examines the ability of IgA isolated from HEPS individuals to inhibit transcytosis across a tight epithelial cell layer. A Transwell system was established to model HIV-1 infection across the human mucosal epithelium. The apical-basolateral transcytosis of primary HIV-1 isolates across this mucosal model was examined in the presence and the absence of IgA isolated from the genital tract, saliva, and plasma of HEPS individuals enrolled in both a sex worker cohort in Nairobi, Kenya, and a discordant couple cohort in Italy. In the absence of IgA, HIV-1 primary isolates were actively transported across the epithelial membrane and were released on the opposite side of the barrier. These transcytosed HIV-1 particles retained their ability to infect human mononuclear cells. However, IgA purified from the mucosa and plasma of HEPS individuals was able to inhibit HIV-1 transcytosis. Inhibition was seen in three of six cervicovaginal fluid samples, five of 10 saliva samples, and three of six plasma samples against at least one of the two primary HIV-1 isolates tested. IgA from low risk, healthy control subjects had no inhibitory effect on HIV-1 transcytosis. The ability of mucosal and plasma IgA to inhibit HIV-1 transcytosis across the mucosal epithelium may represent an important mechanism for protection against the sexual acquisition of HIV-1 infection in HEPS individuals.

Mucosal and plasma IgA from HIV-exposed seronegative individuals neutralize a primary HIV-1 isolate

OBJECTIVE

To characterize functional properties of HIV-specific IgA in samples representing both systemic and mucosal compartments of HIV-1 highly exposed persistently seronegative (HEPS) individuals.

METHODS

IgA was purified from plasma and mucosal samples from HEPS individuals and tested for the ability to neutralize infection of peripheral blood mononuclear cells (PBMC) by a non-syncytium inducing HIV-1 (clade B) primary isolate. None of these individuals had measurable HIV-1-specific IgG.

RESULTS

HIV-1-specific neutralizing activity of the purified IgA from plasma (n = 15), saliva (n = 15) and cervicovaginal fluid (CVF) (n = 14) were found in the majority of samples (73, 73 and 79%, respectively). In contrast, plasma, saliva and CVF samples of low-risk, uninfected HIV-seronegative individuals lacked neutralizing IgA, with the exception of two out of 34 (6%) saliva samples.

CONCLUSION

Mucosal and plasma IgA from HEPS individuals can neutralize HIV-1 infection.

Association of Chlamydia trachomatis heat-shock protein 60 antibody and HLA class II DQ alleles

A total of 113 female commercial sex workers had individual alleles for HLA class II genes determined by using labeled sequence-specific oligonucleotide probes to hybridize to polymerase chain reaction products of amplified DNA. Women also had microimmunofluorescent (MIF) antibody titers to Chlamydia trachomatis elementary bodies and ELISA antibody to recombinant chlamydial heat-shock protein 60 (Chsp60) determined. Women were prospectively followed at monthly intervals over 2 years for incident C. trachomatis infection and acute pelvic inflammatory disease (PID). HLA DQA1*0401 and DQB1*0402 alleles were statistically associated with increased prevalence and amount of antibody to Chsp60 but not MIF antibody. However, these alleles did not alter the risk for chlamydial PID. The potential role that HLA DQ may play in chlamydial disease pathogenesis requires further study.

The role of cytotoxic T-cells in HIV infection

HIV-specific cytotoxic T lymphocytes (CTL) are believed to play a major role in controlling virus levels through the asymptomatic period of HIV infection. For the rational design of an HIV vaccine, we need to know whether protective immunity can ever develop following HIV exposure in people who remain uninfected. We have detected HIV-specific CTL in 5/6 repeatedly exposed, persistently seronegative female sex-workers in The Gambia. Their CTL, repeatedly detected over two years, recognise epitopes presented by HLA-B35 which are cross-reactive between HIV-1 & HIV-2, suggesting they could have been primed first by HIV-2 exposure and subsequently boosted by exposure to HIV-1. Using previously identified clade B HIV-1 epitope peptides, we have now detected HIV-specific CTL in 6/15 highly exposed and apparently HIV-resistant Kenyan prostitutes, predominantly towards epitopes highly conserved between B and the Kenyan A & D clades of HIV-1. This CTL activity towards conserved virus epitopes may represent protective immunity to HIV generated in response to repeated exposure, and prophylactic HIV vaccines should aim to generate similar CTL responses.

Antibody to chlamydial hsp60 predicts an increased risk for chlamydial pelvic inflammatory disease

To determine whether serum antibody to Chlamydia trachomatis antigens alters the risk of C. trachomatis pelvic inflammatory disease (PID), 280 female sex workers were prospectively evaluated over a 33-month period for incident C. trachomatis and Neisseria gonorrhoeae cervical infection and for clinical PID. At enrollment, women were tested for antibody to C. trachomatis elementary bodies by an indirect microimmunofluorescence assay and to recombinant chlamydial hsp60 (Chsp60) by an ELISA format. At each follow-up visit, women were tested for cervical chlamydial and gonococcal infection and were identified as having clinical PID if they complained of lower abdominal pain and were found to have uterine and adnexal tenderness on pelvic examination. The data demonstrate that antibody to Chsp60 predicts a 2- to 3-fold increased risk for C. trachomatis PID.

An immunohistochemical analysis of naturally occurring chancroid

Haemophilus ducreyi is a major cause of genital ulcer disease in many developing countries and is associated with augmented transmission of human immunodeficiency virus (HIV). However, the mechanisms through which H. ducreyi produces ulceration are poorly understood. The characteristics of the host response to H. ducreyi and the pathobiology of its potential contribution to increased HIV susceptibility are not known. Chancroid ulcer biopsies from 8 patients were analyzed histologically and immunohistochemically. All biopsies had perivascular and interstitial mononuclear cell infiltrates that extended deep into the dermis. The infiltrate, which contained macrophages and CD4 and CD8 lymphocytes, was consistent with a delayed hypersensitivity type cell-mediated immune response. The recruitment of CD4 T lymphocytes and macrophages may in part explain the facilitation of HIV transmission in patients with chancroid.

Bacteriuria in a cohort of predominantly HIV-1 seropositive female commercial sex workers in Nairobi, Kenya

Although significant bacteriuria and urinary tract infection are more common in immunocompetent women than men, studies linking HIV immunosuppression with an increased risk of developing urinary infection have so far only been carried out in men. We therefore examined the relationship between bacteriuria and HIV status and CD4+cell count in a relatively homogeneous cohort of female commercial sex workers (CSW) attending a community clinic in Nairobi. Two hundred and twenty-two women were enrolled, and grouped according to HIV status and CD4 count. Group 1 were HIV seronegative (n = 52); Group 2 were HIV seropositive with CD4 + counts above 500 x 10(6)/l (n = 51); Group 3 were HIV seropositive with CD4 + counts between 201 and 500 x 10(6)/l (n = 67); Group 4 were HIV seropositive with CD4+counts below 200 x 10(6)/l (n = 52). Clinical signs and symptoms were noted and mid-stream specimens of urine obtained for culture and sensitivity. Overall 23% (50/222) had significant bacteriuria. The rates in each group respectively were 25%, 29%, 19% and 23% and there was no significant association between bacteriuria and HIV status; or between bacteriuria and level of immuno-suppression as indicated by CD4 + count. Overall 19% (30/222) of women had symptoms (frequency; dysuria; loin pain; smelly urine) or signs (fever; loin tenderness) compatible with urinary tract infection. However there was no significant association between symptoms or signs of infection and bacteriuria or HIV status. A typical range of pathogens, predominantly Enterobacteriaceae, were isolated and there were high rates of resistance to commonly used antimicrobials as well as 10% resistance to ciprofloxacin. Although high rates of significant bacteriuria can occur in highly sexually-active women, this appears unrelated to HIV infection or the level of HIV-related immunosuppression and is generally asymptomatic or clinically indistinct.

Association between cervical inflammation and cervical shedding of human immunodeficiency virus DNA

A cross-sectional study was conducted among prostitutes in Nairobi, Kenya, to determine the prevalence and correlates of cervical human immunodeficiency virus (HIV) DNA. Ninety-two HIV-seropositive prostitutes were evaluated during 137 clinic visits. Cervical HIV DNA was detected by polymerase chain reaction assay in 36 (39%) women at initial visits and in 40 (44%) women at any visit. There was a significant correlation between cervical HIV and microscopic evidence of cervical inflammation (odds ratio [OR], 7.2; 95% confidence interval [CI], 2.1-24.6). Using multivariate analysis to adjust for possible confounding, the adjusted OR for the association between cervical inflammation and cervical HIV DNA was 8.7 (95% CI, 2.0-37.2). Conditions associated with cervical inflammation are associated with the detection of HIV proviral DNA. Whether such conditions lead to increased infectivity remains to be proven.

Chlamydia trachomatis from individuals in a sexually transmitted disease core group exhibit frequent sequence variation in the major outer membrane protein (omp1) gene

60 cervical Chlamydia trachomatis infections identified by antigen detection from 51 prostitute women in Nairobi, Kenya were evaluated for sequence polymorphism in the major outer membrane protein (omp1) gene. DNA from clinical specimens was amplified by the polymerase chain reaction and cycle sequenced through variable domains (VD) 1, 2, and 4.37 (63%) samples had variant VD sequences, 19 (32%) samples had prototype VD sequences, and 4 (6%) samples had prototype VD sequences, and 4 (6%) samples contained omp1 sequences from two or more C. trachomatis strains. Among the 37 variant strains, 18 had two or fewer nucleotide substitutions in one or two VDs and represented point mutational drift variants. 19 strains had a larger number of nucleotide changes and displayed mosaic omp1 sequences that may have been generated by omp1 VD recombination. We conclude that the prevalence of C. trachomatis omp1 DNA polymorphism is substantial among prostitute women in Nairobi, Kenya and that this is the likely result of immune selection pressure.

Human immunodeficiency virus infection in long-distance truck drivers in east Africa

BACKGROUND

A cross-sectional survey was performed to determine the seroprevalence and correlates of human immunodeficiency virus (HIV) infection among long-distance truck drivers in Kenya.

METHODS

Truck drivers along the Mombasa-Nairobi highway were enrolled at a roadside research clinic. A standardized interview and serologic evaluation for HIV and syphilis were conducted.

RESULTS

We enrolled 970 truck drivers and their assistants of whom 257 (27%) had HIV antibodies. In univariate analysis, HIV infection was correlated with older age, non-Kenyan nationality, Christian religion, longer duration of truck driving, travel outside of Kenya, less frequent visits to wives, and more frequent visits to prostitutes. Uncircumcised status, history of genital ulcer disease or urethritis during the previous 5 years, and a positive Treponema pallidum hemagglutination assay for syphilis were each associated with positive HIV serostatus. Univariate correlates of uncircumcised status included younger age, non-Kenyan nationality, Christian religion, travel outside of Kenya, and less frequent visits to prostitutes. There was a significant association between uncircumcised status and 5-year history of genital ulcer disease or serologic evidence of syphilis, but not with 5-year history of urethritis. In multivariate analysis, HIV infection was independently associated with uncircumcised status (adjusted odds ratio [OR], 4.9; 95% confidence interval [CI], 2.8 to 8.4), history of genital ulcer disease (adjusted OR, 2.4; 95% CI, 1.5 to 4.1), history of urethritis (adjusted OR, 1.8; 95% CI, 1.1 to 2.9), more frequent sex with prostitutes (more than once per month; adjusted OR, 1.7; 95% CI, 1.1 to 2.8), and positive T pallidum hemagglutination assay (adjusted OR, 1.2; 95% CI, 1.0 to 1.4). The attributable risk percentage for the association between HIV and uncircumcised status was 70%, and the population attributable risk was 25%.

CONCLUSIONS

Truck drivers in east Africa are at high risk of HIV infection. The strongest correlates of HIV seropositivity were uncircumcised status and history of both ulcerative and nonulcerative sexually transmitted diseases.

Detection of HIV DNA in cervical and vaginal secretions. Prevalence and correlates among women in Nairobi, Kenya

OBJECTIVE

Factors that influence heterosexual transmission of the human immunodeficiency virus (HIV), including sexually transmitted diseases, contraceptive practices, sexual practices, HIV-related immunosuppression, and presence of cervical ectopy and the penile foreskin, have been identified through cross-sectional and prospective cohort epidemiological studies. To more directly characterize factors that influence infectivity, we conducted a study of HIV shedding from the genital tract in women.

DESIGN

Ninety-seven HIV-seropositive women attending a sexually transmitted disease clinic in Nairobi, Kenya, completed a questionnaire and underwent a physical examination and an evaluation for sexually transmitted diseases. Cervical and vaginal secretions were obtained for HIV DNA detection using polymerase chain reaction amplification.

RESULTS

Human immunodeficiency virus DNA was detected by polymerase chain reaction in 28 (33%) of 84 cervical samples and 13 (17%) of 77 vaginal samples. The prevalence of HIV was higher in specimens from the endocervix than from the vaginal wall (P = .002), and there was no correlation between presence of virus at the two sites. After adjusting for age, cervical HIV shedding was independently associated with oral contraceptive pill use (odds ratio [OR], 11.6; 95% confidence interval [CI], 1.7 to 77.6), cervical mucopus (OR, 6.2; 95% CI, 0.9 to 41.4; P = .05), cervical ectopy (OR, 5.0; 95% CI, 1.5 to 16.9), and pregnancy (OR, 4.5; 95% CI, 1.2 to 16.3).

CONCLUSIONS

Human immunodeficiency virus was detected in one third of cervical samples and one sixth of vaginal samples. The presence of HIV DNA in cervical secretions was significantly associated with oral contraceptive pill use, cervical ectopy, and pregnancy. There was a marginally significant association with cervical mucopus. The identification of factors that increase the infectivity of women suggests potential strategies for reducing heterosexual transmission of HIV.

Humoral immune response of humans to lipooligosaccharide and outer membrane proteins of Haemophilus ducreyi

The humoral immune response to purified lipooligosaccharide (LOS) and outer membrane proteins (OMP) of Haemophilus ducreyi was evaluated. Sera from chancroid-endemic (Uganda, Kenya) and -nonendemic (Canada) countries were tested by an ELISA. The response to OMPs was cross-reactive with other Haemophilus species, and elevated levels of antibody were detected in patients that did not have chancroid. The LOS component stimulated an H. ducreyi-specific immune response that was detected only in patients with chancroid. The sensitivity of the LOS ELISA was 96% (95% confidence interval, 89.9%-100%) and the specificity was 97% (95% confidence interval, 95.8%-98.2%). Thus, the anti-H. ducreyi LOS immune response is a significant diagnostic and epidemiologic indicator.

Use of an adsorption enzyme immunoassay to evaluate the Haemophilus ducreyi specific and cross-reactive humoral immune response of humans

Serodiagnosis of chancroid is limited by the cross-reactivity of Haemophilus ducreyi with Haemophilus influenzae and Haemophilus parainfluenzae. This research describes an adsorption enzyme immunoassay (EIA) that assesses the humoral immune response of North Americans and Africans to H. ducreyi. Adsorption effectively removed anti-H. influenzae and anti-H. parainfluenzae antibodies, revealing that North American control sera had no residual anti-H. ducreyi reactivity. However, African control sera still had a residual anti-H. ducreyi response. Assessment of the duration of the humoral immune response in sera from African patients with chancroid showed that the humoral antibodies persisted for up to 8 months after the diagnosis. This may explain the lack of specificity of the adsorption EIA in areas where chancroid is endemic. The detection of the humoral immune response was affected by the strain of H. ducreyi used, with indigent strains being most useful. Using H. ducreyi 35000 for Canadian sera, the sensitivity of the adsorption EIA was 100% and the specificity was 88%. For African sera, H. ducreyi strain R018 was used, and the adsorption EIA had a sensitivity of 81% and a specificity of only 23%. These data reveal that the existing humoral response in a country where chancroid is endemic differs from that in a country where it is not, and that care must be used interpreting unadsorbed humoral immune responses. The adsorption EIA approach may prove useful as an epidemiologic tool for definition of existing (past and present) levels of exposure to H. ducreyi.

Salmonella and Shigella gastroenteritis at a public teaching hospital in Nairobi, Kenya

OBJECTIVE

To measure the proportion of nosocomial diarrhea cases associated with Salmonella and Shigella species.

DESIGN

Prospective 6-month survey.

SETTING

Tertiary care center in a developing country.

PATIENTS

Pediatric and adult patients admitted with the previous 24 hours and all consenting adult or pediatric medical patients with nosocomial diarrhea.

OUTCOME MEASURES

Prevalence of Salmonella and Shigella species isolated from rectal swabs at admission and among subjects with nosocomial diarrhea.

RESULTS

Salmonella species and Shigella species were isolated from 3.0% and 2.5%, respectively, of 667 patients screened on admission. All admission Salmonella isolates were identified in children under 13 years of age; Shigella prevalence was similar for children and adults. Children with Salmonella at admission were significantly older and more likely to have diarrhea, fever, and some indicators of malnutrition than those from whom Salmonella was not isolated. Salmonella and Shigella were isolated from rectal cultures in 36 (10%) and 9 (2.5%) of 360 nosocomial gastroenteritis cases, respectively. Nosocomial cases occurred equally in adults and children. In adults, nosocomial Salmonella acquisition was associated with sharing a room with a diarrhea patient and previous institutionalization. In children, it was associated with recent antimicrobial therapy, crowding at home, and age between 6 months and 6 years. Nine (41%) of 22 nosocomial Salmonella cases in adults occurred in patients with human immunodeficiency virus-type 1 (HIV-1) infection, while none of 79 HIV-1-positive patients had Salmonella isolated at admission.

CONCLUSIONS

Salmonella is a frequent cause of nosocomial gastroenteritis in this tertiary care institution in a developing country. Risk factors appear to differ for children and adults, and HIV-1-infected subjects may be at increased risk of acquisition. Control measures feasible for the limited resources available to such institutions require evaluation.

The Nairobi STD program. An international partnership

Since 1980, Canadian, Belgian, and American scientists have been working together with colleagues in Kenya to understand the epidemiology and biology of sexually transmitted diseases (STDs) in Africa. The rapid spread of human immunodeficiency virus has dramatically increased the importance of this information. Interventions to slow the spread of all STDs and to reduce their consequences are now our major priority. However, this will require continuing research, ongoing training of scientists, epidemiologists, and technologists, and increasing financial support from industrialized societies.

Isolation of human immunodeficiency virus from genital ulcers in Nairobi prostitutes

Recent epidemiologic studies have implicated genital/anorectal ulcer disease as an important cofactor for acquisition and transmission of human immunodeficiency virus (HIV) during sexual intercourse. To better understand the mechanism for the association between genital ulcers and HIV, exudates from 62 genital ulcers of 56 HIV-seropositive prostitutes in Nairobi (Kenya) were cultured for HIV. Twenty-six ulcer cultures could not be evaluated for the presence of HIV because of bacterial or fungal contamination. HIV was isolated from 4 (11%) of the 36 remaining uncontaminated ulcer cultures (2 introital, 1 vaginal, and 1 cervical) from 4 separate women. HIV was isolated from the cervical os from only 2 of the 4 women. HIV p24 antigen was detected in exudate from 1 of the 4 culture-positive ulcers and 0 of 32 culture-negative ulcers. Genital ulcers in seropositive patients should be regarded as potential sources of HIV, which could be important in transmission of HIV during intercourse. Public health measures aimed at controlling sexually transmitted genital ulcer diseases should be an integral part of acquired immunodeficiency syndrome (AIDS) prevention programs.

Low prevalence of oral mucosal lesions in HIV-1 seropositive African women

In Nairobi (Kenya) 334 women prostitutes of whom 80.5% were HIV-1 positive, were examined for oral mucosal lesions; 15.6% of seropositive, and 4.6% of seronegative, women had oral mucosal lesions, predominantly oral candidiasis. In the seropositives, 8.6% had erythematous, 1.1% had hyperplastic candidiasis, and 0.4% had hairly leukoplakia as the only oral lesions; 0.4% had a combination of erythematous and hyperplastic candidiasis, and 1.5% had combinations of pseudomembranous and erythematous candidiasis in the presence of hairly leukoplakia. Of the 211 seropositive women for whom we knew the first date on which a positive serologic test was obtained, the likelihood of developing an oral mucosal lesion was found to be dependent on the duration of seropositivity. The low incidence of oral mucosal lesions in this population may be due to the relatively recent acquisition of HIV-1 infection.

Antimicrobial susceptibility testing and phenotyping of Neisseria gonorrhoeae isolated from patients with ophthalmia neonatorum in Nairobi, Kenya

Antimicrobial susceptibility testing, auxotyping-serotyping, and plasmid analysis were performed on 41 ocular isolates, 7 nasopharyngeal isolates, and 18 cervical isolates of Neisseria gonorrhoeae obtained during a recent treatment trial of gonococcal ophthalmia neonatorum in Nairobi, Kenya. Fourteen distinct serovar-auxotype patterns were observed with IB-1/Pro-strains which accounted for 59% of the isolates. Infection with multiple types of gonococci appeared to occur in 22% of the mothers since 4 of 18 paired maternal cervical and neonatal ocular isolates had mismatched serovar-auxotype patterns. Among 10 treatment failure isolates only 1 had a mismatched serovar-auxotype pattern. Six (15%) of the ocular isolates were penicillinase-producing N. gonorrhoeae (PPNG). Five had the 4.4-megadalton (Md) beta-lactamase plasmid and one had the 3.2-Md beta-lactamase plasmid. The 24.5-Md plasmid was found in 5 of 6 PPNG strains and in 8 of 35 non-PPNG strains (P less than 0.02). For most antimicrobial agents, PPNG and non-PPNG strains showed similar patterns of susceptibility. Ceftriaxone was the most active of the antibiotics tested, with all strains having an MIC less than or equal to 0.06 mg/liter. Among non-PPNG strains, 15 (43%) had a penicillin MIC greater than or equal to 2 mg/liter and were considered intrinsically resistant to penicillin. Overall, non-PPNG intrinsically resistant strains had greater resistance to other antibiotics than did non-intrinsically resistant strains (P less than or equal to 0.006). The Mtr phenotype was found in 53% of these strains.

Correlation of auxotype and protein I type with expression of disease due to Neisseria gonorrhoeae

To explore the correlation of outer membrane protein I (PI) type and auxotype to clinical expression of infection with Neisseria gonorrhoeae, we characterized nutritional requirements by auxotyping and PI serovars by coagglutination assay with monoclonal antibodies of 325 consecutive clinical isolates of N. gonorrhoeae. The clinical status of each infected individual was determined by chart review. The predominant auxotype required proline, citrulline, and uracil with or without hypoxanthine (Pro-Cit-Ura-[Hyx-]) and accounted for 22% of all gonococcal strains. Pro-Cit-Ura-(Hyx-) strains were recovered from seven of 15 men with asymptomatic urethral infections (P less than .01). Of the 325 strains, 33% were PIA and 67% were PIB serovars. All five isolates in disseminated gonococcal infection were PIA serovars (P less than .01). Three percent of 168 urethral infections in men were complicated by epididymitis, and 17% of 135 cervical infections in women were complicated by pelvic inflammatory disease (P less than .0005). Neither of these complications was associated with a particular auxotype or PI serovar.