Withdrawal symptoms as a predictor of mortality in patients HIV-infected through drug use receiving antiretroviral treatment (HAART)

Page created by Howard Lawson
 
CONTINUE READING
Withdrawal symptoms as a predictor of
  mortality in patients HIV-infected
      through drug use receiving
   antiretroviral treatment (HAART)

 Michel L123, Giorgi R4, Villes V5, Poizot-Martin I6,
 Dellamonica P7, Spire B5, Carrieri M.Patrizia5
 1INSERM, U669, Paris, France ;
 2Univ Paris-Sud and Univ Paris Descartes, UMR-S0669, Paris, France ;
 3AP-HP, Emile Roux Hospital, Centre de Traitement des Addictions, Limeil-Brévannes, France ;
 4 LERTIM, Faculté de Médecine, 27 Bd Jean Moulin, 13385 Marseille Cedex, France ;
 5 Health and Medical Research National Institute, Research Unit 912 (INSERM UMR912), Marseilles, France
 6 Southeastern Health Regional Observatory (ORS PACA), Marseilles, France
 7 Sainte Marguerite Hospital, Marseilles, France

 8 Archet Hospital, Infectious Diseases Service, Nice, France
Background:
    HIV progression in IDUs
No difference
N   diff      iin d
                  death
                     th rates
                           t existed
                                  i t d iin the
                                            th
pre-HAART era
IDUs receiving HAART have an increased
risk of pre-AIDS death with respect to other
people living with HIV (PLWH)
This increased risk is not attributable to poor
                                              p
response to HAART but rather to HCV, drug
use and social instability
                         y ((Mocroft 2002).
                                          )
Self-reported heroin and cocaine use have
been associated with a development of new
Aids defining conditions (Lucas 2006).
Background:
Life events and HIV progression
Higher cumulative average stressful life
events predict faster progression to AIDS in
gay men (Leserman 2002)

More lifetime
      lif i    trauma significantly
                       i ifi     l predict
                                      di
faster all-cause and AIDS-related mortality
i HIV-infected
in      i f    d men andd women (Leserman
2007).

Studies investigating to what extent life
events experience may be predictive of an
increased risk of death in IDUs are sparse.
Objective

To explore the role that life events
                              events’
experience and particularly drug-
related events may play on the risk of
death in HIV-infected IDUs.
Methods:
       MANIF 2000 Study
   the MANIF-2000
Multidisciplinary cohort to study social and
behavioral characteristics of patients HIV-
infected through injecting drug use since
1995
467 patients enrolled (criteria: clinical stage
A or B and CD4>300 mm3)
Participation of 12 hospital centers in two
French regions : Provence-Alpes-Côte
d’Azur and the Ile de France
Causes of death were retrieved either by   y
death certificates or by hospital records.
Methods : Questionnaires (1)
Standardized medical forms to collect clinical,
treatment and laboratory data
Face to face interviews to collect psycho
Face-to-face                       psycho-social
                                          social
information including patient’s personal
experience with HIV infection,
                     infection care and
adherence to HAART
Self-administered questionnaires collecting
experience of negative life events, social
conditions, prison history, access to OST, drug
and alcohol use and depressive symptoms
(CES-D)
Methods : Questionnaires (2)
Negative life events included:
• Financial problems
• Negative relational events
• Violence
• Hospitalizations
• Family member illnesses or accidents
• D
  Drug-related
        el ted problems
                 oble   : Detoxification,
                          Deto ifi tio death
                                          de th
  from overdose of a family member or a friend,
  experience of withdrawal symptoms
Methods: Participants
     467 patients enrolled

           294 patients
               p
receiving HAART during follow-up

 accounting for 1430 person-years
     (max follow up= 9 years)
Statistical methods
Kaplan Meier and Cox model to study the effect of
possible baseline predictors and time dependant
covariates on the risk of death. Variables with
p
Results
During follow-up 26 deaths (1.8 deaths per 100 PY)
 occurred in IDUs receiving HAART
                             HAART. These were
 attributed as follows:
  -   2 HIV
  -   8 Liver disease
  -   2 Suicides
  -   1 Road accident
  -     C di        l problems
      4 Cardiovascular     bl
  -   4 “Unspecified” at home or in a car
  -   5 other
         th causes (aggression,
                      (       i    medullar
                                      d ll aplasia,
                                             l i sepsis,
                                                     i
      hypopituarism, unknown)

  41 patients reported negative experience with
  withdrawal symptoms in 60 follow-up visits
Characteristics of participants at
       ART enrollment (n=294)

                                           n (%)
Age (median [IQR])                      35 [32-38]
Men                                     215 (73.1)
High school certificate                63 (21.4)
Employment                              92 (31.3)
                                           (31 3)
Depressive symptoms                    116 (39.5)
Injecting
  jec g d drug
            ug use                       41 ((16.5)
                                               6 5)

Drug-related problems *                 20 (11.6)
Drug possession *                         8 (4.7)
Withdrawal symptoms (WS) experience*     19 (11.0)

* Negative life events
Clinical characteristics of
   participants at HAART enrollment
                 (n=294)
                                          n (%)
Clinical stage A                        125 (42.7)
                                             (42 7)
                B                       156 (53.2)
                 C                       12 (4.1)
CD4 /mm3        (median [IQR])   390 [300-530]

U d t t bl viral
Undetectable i l lload
                     d                   99 (33.7)
                                            (33 7)

No OST                                  194 (66.0)
                                             (66 0)
Buprenorphine                           60 (25.8)
Methadone                               36 (14.1)
Morphine sulfate                        4 (1.3)
Variables eligible for the
multivariate Cox model (p
Kaplan-Meier
                  Kaplan Meier survival estimates, by withdrawal symptoms
         00
       1.0

                    stset fipoint0 ,id(id) failure(statut)
                    sts graph, by(qv41brok)
       0.75
0.50
0.25
0.00
0

              0          20                 40                   60         80   100
                                                 analysis time

                           No                                         Yes
Factors independently associated
      with the risk of death

                                Adjusted
                                Adj  t d HR
                                  [95% CI]        P-value

Viral load ≥ 10000 copies/ml    4.7 [2.0-10.9]
Why are withdrawal symptoms
       predictive of survival ?
No significant difference in WS occurrence when
comparing
       i    methadone
                h d     anddbbuprenorphine
                                      hi   substituted
                                             b i     d
patients.
The incidence of withdrawal symptoms was
significantly
  g         y higher
                g    in the 1995-1998 p
                                      period than in the
period after 1999
IR[95%CI] =27
          =27.5[16.6;
              5[16 6; 45.6]
                      45 6] - 1995
                              1995-1998
                                   1998
IR[95%CI] =10.6 [7.3; 15.4] - 1999
 Pharmacosurveillance data confirm that there was an
increase in dosage prescription for both methadone and
buprenorphine when comparing the two periods.
Why are withdrawal symptoms
       predictive of survival ?
Withdrawal symptoms may be a proxy of possible
interactions with HAART requiring an increase in OST
dosages

Withdrawal symptoms in patients receiving OST may be
a sign of under-dosing or inadequate care.

Underdosed substituted patients may use opioids to
manage withdrawal symptoms but they may die from
overdose because of their reduced tolerance to opioids.
Conclusions (1)
Experience of withdrawal symptoms is a
predictor
   di t off survival
                i l iin IDUs
                        IDU on HAART
even after adjustment for HIV status (VL)

This association did not change even:
ƒ   after adjustment for adherence and initial CD4
ƒ   after exclusion of the two deaths due to suicide
ƒ   after multiple imputation of missing data of the other
       possible predictors
Conclusion (2)
Withdrawal symptoms in IDUs living with HIV
mainly reflects difficulties in the management
of opioid dependence
          dependence.

This association needs to be investigated
more thoroughly in other studies

Such symptoms should be detected and
managed more effectively to properly deliver
comprehensive care to IDUs on HAART and
prevent possible drug-related deaths
Thanks to :
The French National Agency for Aids Research
(ANRS, France), the charity organisation ECS-
SIDACTION (France), and the Departmental
Council (Bouches-du-Rhône, France).

Manif 2000 study group : C Boirot, AD Bouhnik, MP Carrieri, JP
Cassuto, M Chesney, P Dellamonica, P Dujardin, H Gallais, JA Gastaut, P Kurkdji, G
Lepeu, C Marimoutou, D Mechali, JP Moatti, J Moreau, M Nègre, Y Obadia, I Poizot-
Martin, C Pradier, D Rey, C Rouzioux, A Sobel, B Spire, F Trémolières, V Villes, E
Vi
Vincent,
      t D Vl
          Vlahov
             h

All PLWH who accepted
                 p    to p
                         participate
                                p    in
the study
You can also read