Scolaris Content Display Scolaris Content Display

Cochrane Database of Systematic Reviews Protocol - Intervention

Optimal timing for discontinuation of Pneumocystis jiroveci pneumonia prophylaxis in adult patients on highly active antiretroviral therapy (HAART) for HIV infection

Authors' declarations of interest

Version published: 18 January 2012 Version history

https://doi.org/10.1002/14651858.CD009556

This is not the most recent version

view the current version 12 January 2017
Collapse all Expand all

Abstract

This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:

To assess the impact of discontinuation of Pneumocystis jiroveci prophylaxis on the morbidity and mortality due to Pneumocystis jiroveci pneumonia in HIV patients who show a stable rise in CD4 counts to >200cells/mm3for three or more months following initiation of HAART.

Background

Description of the condition

Pneumocystis jiroveci pneumonia(PJP), formerly called Pneumocystis carinii pneumonia (PCP) (Stringer 2002) is one of the major causes of illness and death in immunocompromised individuals (Baez‐Escudero 2004; Morris 2004; Sepkowitz 1993) PJP is one of the commonest opportunistic infections in HIV/AIDS patients with a mortality rate of 15‐20% with standard treatment (John 2009; Morris 2004)

Pneumocystis jiroveci pneumonia is caused by Pneumocystis jiroveci a fungus (Stringer 2002). Infection starts in the lungs where the organism multiplies in the alveoli leading to exudate production, A type 2 pneumocyte hyperplasia occurs with infiltration of the lungs by mononuclear cells. Increased permeability of the alveolar capillary and non‐cardiogenic pulmonary edema occur as desquamation of the alveolar lining cells occurs. Further destruction of the lung tissue can lead to pneumatocoeles and pneumothorax. A larger number of pneumocystis jiroveci and fewer inflammatory cells are found in HIV positive patients than in HIV negative patients. Common sites of extrapulmonary lesions are lymph nodes, spleen, liver and bonemarrow.Pneumocystis jiroveci pneumonia is characterised by dyspnea, non‐productive cough and fever. Chest X‐rays show bilateral infiltrates with extrapulmonary lesions occurring in less than 3% of the patients (Thomas 2004; Thomas 2007). 

First line treatment of PJP is Trimethoprim‐sulfamethoxazole. Alternative regimens include pentamidine, trimethoprim‐dapsone, atovaquone and primaquine and clindamycin. Corticosteriod therapy is recommended PO2<70% (CDC 2009; John 2009). World Health Organisation (WHO) and Centre for Disease Control (CDC) both recommend Pneumocystis jiroveci prophylaxis in HIV patients with CD4 counts <200cells/mm3 and this has been associated with a decrease in the morbidity and mortality due to Pneumocystis jiroveci pneumonia in HIV patients  Trimethoprim‐sulfamethoxazole is the recommended first line drug for chemoprophylaxis against Pneumocystis jiroveci pneumonia. Alternative regimens include, dapsone,dapsone plus pyrimethamine, atovaquone or aerosolized pentamidine. Primary prophylaxis is done in patients without a history of Pneumocystis jiroveci pneumonia while secondary prophylaxis refers to prophylaxis in patients with a history of Pneumocystis jiroveci pneumonia (CDC 2009; Morris 2004; WHO 2006).

Closely related to PJP prophylaxisis, is prophylaxis against Toxoplasmosis gondii. This protozoan infection has an estimated Worldwide seroprevalence in humans of 30‐65% and is usually characterised by a latent asymptomatic state in immunocompetent people. Although symptomatic infection has been reported following primary infection, in immunosuppressed individuals, it is usually caused by a reactivation of the latent infection.Transmission is through ingestion of raw or partly cooked meat with toxoplasma cysts as well as ingestion of contaminated cat fecal matter. It commonly manifests as encephalitis and dissemination to the retina, lungs, heart, skin and liver have been reported (Astrid 2000)

The recommended first line treatment of Toxoplasmosis is pyrimethamine plus sulfadiazine plus leucovorin. Alternative regimens include pyrimethamine plus clindamycin plus leucovorin,Trimethoprim‐sulfamethoxazole, atovaquone plus either pyrimethamine plus leucovorin or sulfadiazine, atovaquone alone, azithromycin plus pyrimethamine plus leucovorin. CDC recommends primary and secondary chemoprophylaxis against Toxoplasmosis gondii in HIV patients with CD4 counts <200cells/mm withTrimethoprim‐sulfamethoxazole as the first line drug choice. Alternative chemoprophylactic regimen are also similar to PJP prophylaxis and include; dapsone‐pyrimethamine plus leucovorin, atovaquone with or without pyrimethamine/leucovorin (CDC 2009).

PJP prophylaxis has also been associated with a decrease in the overall morbidity with studies reporting a decrease in the incidence of community acquired pneumonia, HIV associated diarrhoea and malaria in HIV patients (Stefan 1999, Jonathan 2004, Watera 2006) .

Description of the intervention

The initiation of Highly Active Antiretroviral Therapy (HAART) in HIV management has been associated with an increase in the CD4 counts and a decrease in HIV viral loads of patients making them less susceptible to opportunistic infections (Autran 1997; Mocroft 2003). Studies have shown that prophylaxis against specific opportunistic infections can be safely discontinued in HIV patients who respond to HAART with a stable increase in CD4 counts (El‐Sadr 2000; Martínez 2000). Observation studies and randomised controlled trials have reported that safe discontinuation of Pneumocystis jiroveci prophylaxis in HIV patients who have shown a steady increase in CD4 counts to >200cells/mm3 following initiation of HAART can be carried out (Dworkin 2000; Furrer 1999; Lopez 2001). The Centre for Disease Control in their weekly mortality and morbidity report recommended the discontinuation of Pneumocystis jiroveci pneumonia prophylaxis in HIV patients whose CD4 counts rise to >200cells/mm3 for more than 3 months (CDC 2009). However, reports of Pneumocystis jiroveci pneumonia occurring in HIV patients with CD4 counts above 200cells/mm3 question the safety of this intervention (Crothers 2003; Killingley 2006).

How the intervention might work

The susceptibility of HIV patients to developing Pneumocystis jiroveci pneumonia has been associated with a decrease in the CD4 counts to <200cells/mm3(CDC 2009, Morris 2004). Using the CD4 count as a marker for changes in the level of immunity, a stable rise in the immunity of HIV patients following initiation of HAART reduces the chances of an individual developing Pneumocystis jiroveci pneumonia thus making it possible to safely discontinue Pneumocystis jiroveci prophylaxis in these patients. This would not only decrease the pill burden of the patients but also decrease the cost of HIV management, potential for drug toxicity, drug interactions and development of drug resistant pathogens.

Why it is important to do this review

The high morbidity and mortality rates associated with Pneumocytis jiroveci pneumonia in HIV patients as well as reports of Pneumocystis jiroveci pneumonia in HIV patients with CD4 counts above the 200cells/mm3 threshold necessitate the examination of the evidence for the safe discontinuation of both primary and secondary Pneumocystis  jiroveci prophylaxis in HIV patients who show a stable increase in CD4 counts in response to initiation of HAART.

Objectives

To assess the impact of discontinuation of Pneumocystis jiroveci prophylaxis on the morbidity and mortality due to Pneumocystis jiroveci pneumonia in HIV patients who show a stable rise in CD4 counts to >200cells/mm3for three or more months following initiation of HAART.

Methods

Criteria for considering studies for this review

Types of studies

All randomised controlled studies relevant to the study question.

Types of participants

HIV positive patients 15 years and above, who are on highly active antiretroviral therapy and either primary or secondary Pneumocystis Jiroveci pneumonia prophylaxis whose CD4 count has risen >200/mm3for at least three months.

Types of interventions

Discontinuation of either primary or secondary Pneumocystis Jiroveci pneumonia prophylaxis compared to continuation of Pneumocystis Jiroveci prophylaxis. All Pneumocystis Jiroveci pneumonia prophylaxis regimens will be included.

Types of outcome measures

Primary outcomes

  • Morbidity due to Pneumocystis jiroveci pneumonia confirmed by history, chest x‐ray and bronchoalveolar lavage and sputum analysis.

  • Death secondary to Pneumocystis jiroveci pneumonia.

Secondary outcomes

  • Morbidity due to toxoplasmosis.

  • Death secondary to toxoplasmosis

  • Occurrence of other HIV‐related events according to stages B and C of the CDC classification system for HIV infected adults and adolescents (CDC 1993).

  • All cause mortality

  • Ocurrence of adverse events

Search methods for identification of studies

To identify studies both electronic and manual searches will be done. All languages will be included.

Electronic searches

The search for relevant published trials will be done in:

  • Cochrane Central Register of Controlled Trials (CENTRAL)

  • EMBASE (from 1995 to the search date),

  • MEDLINE (from 1995 to the search date),

  • CINAHL (from 1995 to the search date),

  • LILACS (from 1995 to the search date),

  • WHO Global Health Library ( http://www.globalhealthlibrary.net)

Ongoing trials will be searched for in:

  • Current Controlled Trials register (www.controlled‐trials.com)

  • Center Watch(http://www.centerwatch.com/)

  • WHO International Clinical Trials Registry Platform (ICTRP) (http://www.who.int/ictrp/en/)

  • The National Institutes of Health Clinical Trials Registry (http://www.clinicaltrials.gov/)

  • ACTIS (AIDS Clinical Trials Information service at (http://www.actis.org/)

Relevant abstracts will be obtained from the following conferences:

  • Conference on Retroviruses and Opportunistic Infection (CROI)

  • International AIDS Conference

  • International AIDS Society Conference on on HIV Pathogenesis Treatment and Prevention

The above conferences will be searched by means of the AEGIS database of HIV/AIDS conferences (http://www.aegis.com/search). As the AEGIS database contains conferences up to 2008, we will also search the conference web sites for more recent conference abstracts.

We will also search NLM GATEWAY (http://gateway.nlm.nih.gov)

Search strategies for MEDLINE and Cochrane Central Register for Controlled Trials

See Appendix 1 and Appendix 2 respectively.

Searching other resources

Bibiliographies of published studies and reviews will be searched for references to randomised controlled trials addressing the study question.

Information regarding unpublished or incomplete studies will be obtained through correspondence with the respective investigators and effort will be made to obtain information about unpublished or incomplete studies from other experts in the field.

Data collection and analysis

Selection of studies

The eligibility of the studies identified will be judged independently by two authors (EM,TY).

The titles and abstracts of the studies identified using the search strategy will be screened to identify those that are likely to meet the inclusion criteria. Full text copies of those studies initially thought to fulfil the inclusion criteria will be obtained and further assessed using an eligibility form based on the inclusion criteria to determine whether they actually meet the inclusion criteria of the study. Reasons for excluding potentially relevant studies will be given in the 'characteristics of excluded studies'. Where necessary, the authors of the relevant studies will be contacted for clarification. Any disagreements will be resolved by discussion with the third author (UO).

Data extraction and management

The relevant data will be independently extracted by two authors (EM,UO) using a pre‐piloted data extraction form and entered into Review Manager 5. The data extracted will be summarized in the relevant tables, namely; table of included studies, table of ongoing studies, table of excluded studies and the table of studies awaiting assessment.The data will then be cross checked against the original full text study articles to ensure accuracy and where necessary, the study authors will be contacted for clarification or to provide missing information.

Data extracted will include:

Characteristics of the trial

  • Study design (randomised or quasi randomised, type of randomisation)

  • Duration and location of the trial

  • Treatments used, dose of treatment and duration of treatment

  • Length of follow‐up

Characteristics of participants

  • Number of participants

  • Age and sex of participants

  • CD4 counts

Outcomes

  • Description of outcomes as presented in the trials

  • Time points when outcomes were assessed

Assessment of risk of bias in included studies

The risk of bias in the included studies will be independently assessed by two authors (EM,OU) as described in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Any disagreements will be resolved by discussion with the third author (UO).The risk of bias data will be presented in tables and figures using the scales described in the Cochrane Handbook with judgments of 'low risk of bias', 'high risk of bias' and 'Unclear risk of bias'.

The risk of bias of each included study will be assessed in terms:

  • the method of generation of the randomisation allocation sequence;

  • the method of allocation concealment;

  • blinding of participants and personnel

  • blinding of outcome assessors;

  • incomplete outcome data reporting dealing with study participants lost to follow up and whether intention to treat analysis was carried out;

  • selective reporting of outcomes;

  • presence of other possible sources of bias including baseline differences in the different arms of the clinical trials as well as funding of the clinical trials.

Measures of treatment effect

This study is dealing with dichotomous outcomes (presence or absence of morbidity or mortality) and summary statistics will therefore be reported as risk ratios (RR) with 95% confidence intervals (CI).

Unit of analysis issues

Though the majority of the studies to be included in the review are likely to be individually randomised and analysed, some of the studies may have cluster randomisation. If cluster randomised trials are included in the review, they will be assessed to ensure that the analysis adjusted for the cluster randomisation method used. However where this has not been done, attempts to adjust the results for clustering will be done by multiplying the standard errors of the estimates by the square root of the design effect where design effect = 1+(M‐1)x ICC. M= average cluster size., ICC= Intra‐cluster correlation coefficient (Higgins 2011b). This will require obtaining the average cluster size and intra‐cluster correlation coefficient either from the full print copies of the studies or the authors of the studies.

Dealing with missing data

Where participant drop outs lead to missing data, an intention to treat analysis will be performed. Clinical trial authors will be contacted where necessary to provide missing statistics. Where participant outcome data is not located, a poor outcome will be assumed and a sensitivity analysis carried out to assess the sensitivity of the results obtained to changes in these assumptions.

Assessment of heterogeneity

Assessment of heterogeneity between the trials in the review will be done using the chi‐squared test for heterogeneity with a 10% level of significance, the I2 test using a value of 50% and above to represent substantial to considerable heterogeneity (Deeks 2011) and forest plots will be checked to assess the degree of overlap of the confidence intervals.

Assessment of reporting biases

The likely hood of small study effects such as publication bias will be assessed by constructing a funnel plot and examining it for asymmetry. However, funnel plot asymmetry may also be due to poor methodological quality of the included studies, heterogeneity, it may occur by chance and may even be artefactual.

Data synthesis

When no statistically significant heterogeneity between trials is detected, a fixed‐effect meta‐analysis model will be used. A random‐effects meta‐analysis model will be used if statistically significant heterogeneity that cannot be explained by subgroup analysis or sensitivity analysis is detected. However, if substantial heterogeneity is detected (high I2 value, small Chi2 P value) or when a pooled meta‐analysis result is considered to be meaningless due to a high degree of heterogeneity, meta‐analysis will not be carried out and a forest plota of the results with the pool effect suppressed will be used and the assessment of heterogeneity statistics reported and a narrative of the results given.

Subgroup analysis and investigation of heterogeneity

Subgroup analysis for unit of randomisation will be done (i.e. individual randomised controlled trials versus cluster‐randomised controlled trials with adjustment for clustering) as well as subgroup analysis for primary Pneumocystis jiroveci prophylaxis versus secondary Pneumocystis jiroveci prophylaxis, age, sex and CD4 count levels.

Sensitivity analysis

Sensitivity analyses will be done to explore the effects of assumptions made where there is missing outcome data and also to explore the effect of excluding studies with a high risk of bias as described in the Cochrane hand book of systematic reviews for interventions.