5/15/2017. Cryptococcal Meningitis Update from Resource-limited setting perspective: Epidemiology, Treatment and Prevention.

Transcription

1 Estimated yearly cases 5/5/07 Plan: Cryptococcal Meningitis Update from Resourcelimited setting perspective: Epidemiology, Treatment and Prevention. Tom Harrison MSG Biannual meeting, Monterey September st, 06 Epidemiology Update Global burden: Park et al to Rajasingham et al On the ground data re: ongoing case numbers, outcomes Treatment Recent and ongoing antifungal trials New drugs (CSF pressure and Timing of ART) Prevention Update re cryptococcal antigen screening and preemptive therapy for those testing positive Update on the Global Burden of Disease of HIVAssociated Cryptococcal Meningitis. Rajasingham Chiller Boulware et al. in CDC Different approach: Numbers at risk, CrAg prevalence data 800, , , , ,000 00,000 00,000 00, ,800 North 54,400 Latin 70,000 7, ,500 7,00 Caribbean Western North Eastern Sub 0,000, South and East Asia Oceania Cases and deaths point estimates are lower but within confidence intervals in Park et al, 40,000 deaths per year, 7% of AIDSrelated mortality 70% in SubSaharan Africa where best estimate remains.. America America and Central Europe Africa and Saharan Middle Africa East Europe and Central Asia Southeast Asia 70% month mortality as in Park et al Region AIDS 009, :55 50; based available incidence data at the time Park et al AIDS 009; : 550

2 5/5/07 Region Total CRAG+ (95% CI) Annual burden of CCM (95% CI) Deaths from CCM (95% CI) Africa 84,600 (44,600 to,00) 9,000 (87,500 to 57,00) 0,000 (74,00 to 5,600) 5,400 (,770 to 7,770),600 (,00 to,00) 6,500 (9,800 to 56,850),000 (700 to,400),700 (6,400 5,50) WEPtoEB Pathw aystocareandoutcomesof patientswithcryptococcosis, Gauteng (50 to 70) Latin America 6,700 (,40 to 0,700) 4,000 (,840 to 6,900) North America,70 (,000 to 4,600),60 (,600 to 5,000),400 (,80 to,00) Europe 5,00 (4,000 to 6,500),800 (,80 to 4,950) Global 58,000 (9,900 to 5,800) 70,000 (7,000 to 5,000) st Asia & Pacific Caribbean E. Mediterranean Province, SouthAfrica, 05,900 (800 to,450), S. ToroSilva,,4, V. Chihota, V. Quan, A. Vassall, A. Grant, N.P. Govender,4 60 (40 to 860) London School,740 and Tropical Medicine, London, United Kingdom, 4University of the of Hygiene (,60 to,500),000 (,00 to 4,50) Poster WEPEB06,560 We aimed to evaluate pathways to care prior to hospital admission (,00 to,0) HIVassociatedcryptoccoccal meningitisinbotswana:national incidence andtemporal trendsfollowingartrollout 4,000 (95,000 to 99,000) were eligible for enrolment if they were M. Tenforde, M. Mokomane, T. Leeme, R. Patel, N. Lekwape, C. Ramodimoosi, B. Dube4, E. Williams5, K. Mokobela6, E. Tawanana, T. Pilatwe6, H. Mitchell, Slide courtesy D.ofBanda Radha Rajasingham and David Boulware, D. Moalosi, H. Stone, K. Mokgacha, H. Phillips, M. Mine, Slide courtesy of Radha Rajasingham and David Boulware J. Jarvis,0, 4 5 Friday Kingdom, University of Pennsylvania Medicine, Philadelphia, United States, London School of Hygiene and Tropical Tropical Medicine, London, United Kingdom Breaker Posters We performed a laboratorybased surveillance study to determine the incidence Author Index st WEPEB04 Pathwaystocareandoutcomesof patientswithcryptococcosis, Gauteng Province, SouthAfrica, 05, V. Chihota, V. Quan, A. Vassall, A. Grant, WEPEB 05, S. ToroSilva, N.P. Govender,4,4 Outcomesof HIVinfectedpatientswithcryptococcal meningitisin Latin London School America and onekingdom, USSite University of the of Hygiene and Tropical Medicine, London, United 4 B. Crabtree Ramírez, Y. Caro Vega, B.E. Shepherd, C. Le, P. Cahn4, B. Grinsztejn5, 7 EPEB06 C. Cortes6, C. McGowan, A. Person W Poster As yet no decrease in cases in SubSaharan Africa outside of to care prior to hospital admission We aimed to evaluate pathways South Africa despite increase in antiretroviral therapy (ART) States, were eligible for enrolment if they were access IAS Durban July 06 HIVassociatedcryptoccoccal meningitisinbotswana:national incidence andtemporal trendsfollowingartrollout 4 6 Total cases numbers unchanged, but now.. 50% patients have had ART in the past ACTA trial enrolment data; Day 5 M. Tenforde, M. Mokomane, T. Leeme, R. Patel, N. Lekwape, C. Ramodimoosi, B. Dube4, E. Williams5, K. Mokobela6, E. Tawanana, T. Pilatwe6, H. Mitchell, 8 D. Banda7, D. Moalosi6, H. Stone7, K. Mokgacha6, H. Phillips9, M. Mine, J. Jarvis,0, Friday et al, NEJM 06 74:54; Rhein et al, Lancet ID 06; 6: Kingdom, 6 but have persisting low CD4 counts due to loss from care/non adherence and/or resistance University of Pennsylvania Medicine, Philadelphia, United States, London School of Hygiene and Tropical Breaker Joe Jarvis: Botswana data, IAS Durban 06: cases increasing Posters despite ART programme as absolute numbers with CD4<00 rising due to cumulative default from care Challenge used to be keeping up Author with the epidemic, the number of Index patients progressing to <00 CD4 prior to diagnosis and Rx; now the challenge is retention in care Tropical Medicine, London, United Kingdom We performed a laboratorybased surveillance study to determine the incidence Abstract Book I WEPEB05 Outcomesof HIVinfectedpatientswithcryptococcal meningitisin Latin AmericaandoneUSSite B. Crabtree Ramírez, Y. Caro Vega, B.E. Shepherd, C. Le, P. Cahn4, B. Grinsztejn5, 7 C. Cortes6, C. McGowan, A. Person States,

3 5/5/07 Trial Recruitment in Africa simply no decrease in cases: Aetiology of Meningitis in Africa ACTA 690 AmbitionCM Cape Town, South Africa Kampala, Uganda Astro phase : Rhein et al: 5 patients seen at Mulago in one year Astro phase : >00 enrolled in 8m sites in Uganda Jarvis J et al. BMC Inf Dis 00 Durski K et al. JAIDS 0 Recent outcome data AmB in routine use: outcome data BEST: 540% 0 wk mortality wk AmB based therapy Clinical trial setting Jarvis et al CID 04 CryptoDex 4% Siddiqui et al CID online March 04 Zambia: 9% in hospital mortality with wk AmB in routine use USUAL REALITY: Malawi: Fluconazole: 0 wk mortality remains >50%, one year % survival on fluconazole. Rothe et al PLoSOne 0; 8:e67 Same high hospital casefatality ratio, AmB use in South Africa Slide courtesy Nelesh Govender

4 5/5/07 Gold Standard Therapy: Amphotericin B plus Flucytosine Vietnam trial, Jeremy Day, and colleagues IDSA Perfect et al CID 00 50:9, WHO Rapid advice guideline, WHO November, 0: wks: AmB 0.7 mg/kg/d plus flucytosine (5FC) 00 mg/kg/d Based on..van der Horst et al NEJM 997; 7:5 Supported by Brouwer et al: Lancet 004; 6:76467 Nailed by.day et al NEJM 0; 68:90 Day J et al N Engl J Med 0;68:90 Issue is: ASTRO:. Accepted gold standard wks AmB+5FC not available, feasible, safely sustainable in most African centres Lancet ID 06; 6:809. Fluconazole inferior Ongoing trials: ASTRO ACTA AmbitionCM Sertraline as affordable available safe adjunct with antifungal activity Phase : 00mg to 400 mg/d sertraline, in addition AmB 0.7 plus fluconazole 800mg/d: combined EFA more rapid ( 0.7 logcfu/ml CSF/d) than historic controls without sertraline at same sites Phase : Underway: 4

5 5/5/07 Ongoing phase III ACTA Trial: (MRC ANRS) Malawi, Zambia, Cameroon, Tanzania Built on series phase studies (0070) to define regimens: a) more effective than fluconazole, b) as active more sustainable than wks AmB Strategy : Fluconazole 00 mg /d plus flucytosine 5 mg/kg qds for weeks. Strategy : Amphotericin B (AmB) mg/kg/d + fluconazole 00 mg OR 5FC for 7 days Strategy : Amphotericin B (AmB) mg/kg/d + fluconazole 00 mg OR 5FC for 4 days Followed by standard consolidation/maintenance for all, ART at 4 wks Nussbaum et al. Clin Infect Dis 00; 50:844 UNC Project, Lilongwe, Malawi Muzoora et al: J Infect 0; 64:76 0 patients Rx 00 mg/d flucanozole plus AmB mg/kg/d 5 days: EFA 0.0 log CFU/d over weeks, 0. over 7 days week mortality: HR 0.4 (0.05.6) p=0.05 Nussbaum et al. Clin Infect Dis 00; 50:844 5

6 5/5/07 ACTA Sites 4 Countries, 9 Hospitals, funders mbio 04; 5:e0075 Murine and rabbit models CFU counts at 4 days similar whether daily doses, or days abbreviated course?amb concentrates and remains in brain tissue for prolonged time 690 now enrolled; largest Cryptococcal Meningitis RCT to date Potential Impact Strategies,, much more readily, safely sustained than. IF either shown to be as effective as, could result in a reduction in the mortality in resourcelimited settings using fluconazole from around 70% to the 540% seen with wk AmBcombination treatment to date. IF NOT as effective then substantial extra resources efforts to SAFELY deliver wks AmBcombination Rx are justified pending the Ambisome study and evaluation of new and repurposed drugs AMBITION TRIAL RATIONALE. Ambisome liposomal amphotericin in prior trials, no more effective but less toxic (renal impairment, anaemia) than conventional amphotericin B; even longer half life; excellent brain levels. Even better suited than conventional AmB to short course high dose induction treatment aimed at loading brain compartment. Precedent for such use in Visceral Leischmaniasis: Sundar S et al New Eng J Med just one 0 mg/kg dose effective: 4. IF one / few doses effective, could be cost effective due shorter admission, less monitoring; and feasible to implement 6

7 5/5/07 Antimicrob Agents Chemother 0; 56:50 NEJM 6;6 february, 00 Basis Gilead global access programme at reduced cost for VL AMBITION phase II Botswana, Tanzania. Ambisome 0 mg/kg day (single dose). Ambisome 0 mg/kg day, Ambisome 5 mg/kg day (two doses). Ambisome 0 mg/kg day, Ambisome 5 mg/kg days, and 7 (three doses) 4. Ambisome mg/kg/d for 4 days (standard dose, control arm ) All patients also: fluconazole 00 mg/d for first weeks, and then 800 mg/d until 0 weeks, and 00 mg/d thereafter. ART commenced 4 weeks after antifungal therapy. AMBITION DESIGN. Use Early fungicidal activity (EFA) as endpoint in phase to determine shortest course that is non inferior to usual daily dosing in terms clearance of infection. Also take account clinical endpoint and safety data. Choose best short regimen to take forward to phase comparison with current international standard daily conventional AmB deoxycholate 4. Review phase data after 80 patients (0 per arm) 7

8 5/5/07 90 AMBITION PHASE enrolment Gabarone Botswana, Mwanza Tanzania In addition to these ongoing efforts, we need:. Increasing access to drugs, diagnostics, manometers; CryptoMAG advocacy group flucytosine is the first priority and is coming Review underway: Phase funded (EDCTP, MRC) To start early 07. New anticryptococcal drugs. Especially: Oral, cidal drugs that don t interact with HIV/TB Rx, and that synergise with current therapies Viamet VT9, is a new compound being specifically developed for CM, and is potentially such a drug now in clinical evaluation; phase clinical studies in planning. Oral amphotericin B formulations: preclinical work ongoing Also research ongoing on repurposing drugs: In addition to sertraline tamoxifen, benzimidazoles The Novel Fungal Cyp5 Inhibitor VT9 Demonstrates Potent In vivo Activity Against Cryptococcal Meningitis with an Improved Formulation L.K. Najvar,,4 N.P. Wiederhold, A. Alimardanov, J. Cradock, X. Xu, M. Behnke, E.A. Ottinger, W.J. Hoekstra, E.P. Garvey, S.R. Brand, R.J. Schotzinger, W.R. Moore, R. Bocanegra,,4 W.R. Kirkpatrick,,4 T.F. Patterson,4 UT Health Science Center San Antonio, NIH Therapeutics for Rare and Neglected Diseases, Viamet Pharmaceuticals, Inc., 4 STVHCS BACKGROUND & OBJECTIVE Cryptococcosis is a significant cause of morbidity and mortality in immunocompromised patients. Cryptococcal meningitis is the most common manifestation of disseminated disease. This is a significant, lifethreatening invasive fungal infection in HIV/AIDS patients. RESULTS (continued) Figure. Brain tissue fungal burden data on days 8 & 5 postinoculation. From 9 th ICCC 04 From ICAAC 05 In developing countries, intravenous medications are often unavailable for the induction phase of therapy for cryptococcal meningitis. Even with appropriate therapy, morbidity and mortality in such countries remains unacceptably high. VT9 Binds Potently and Selectively to Recombinant Cryptococcal CYP5 Consistent with Its In Vitro AntiCryptococcal Activity The investigational agent VT9 is a novel fungal Cyp5 inhibitor with potent in vitro and in vivo activity against Cryptococcus spp. (00 ICAAC, Posters F850 and 85). Previous E. P. Garvey in vivo, studies W. J. Hoekstra showed highly, R. J. significant Schotzinger log, A. W. Fothergill, N. P. Wiederhold, A. G.S. Warrilow, J. E. Parker, D. E. Kelly, S. L. Kelly M848 reductions of fungal burden and up to 90% survival with 0 and 0 mg/kg twice Viamet Pharmaceuticals, Inc., Durham, NC, daily oral doses. These two doses (using a rudimentary formulation University of Texas Health Science Center at San Antonio, San Antonio, TX, of the Swansea Univ., Wales, United Kingdom parent molecule) produced mean plasma trough levels of. and.5 g/ml, respectively. Background Fig : Chemical Structure of VT9 Fig : VT9 Selective Inhibition of Crypto CYP5 Table : A new solidstate formulation has been developed to test if lower doses could Cryptococcal be as efficacious meningitis and if equivalent (CM) is thedoses second could leading have even causegreater of efficacy in a death murineinmodel HIVpositive of cryptococcus individualsmeningitis. subsaharan Africa. The most accessible therapy, highdose fluconazole monotherapy, has Agent mortality MATERIALS rates of 4060%. & METHODS Even in the U.S., mortality rates using From ECCMID 05 VT9 combination therapy range from 00%. VT9 is a novel Cryptococcus fungal CYP5 inhibitor neoformans with potent Isolate in vitro and in vivo activity Fluconazole C. The neoformans Novel clinical isolate Fungal USC 597Cyp5 was used, andinhibitor was placed into against Cryptococcus. We disclose its chemical structure (Figure VT9 Demonstrates Potent In Vivo Activity P0 brain heart infusion broth overnight in a shaking incubator. The cells were Voriconazole ) and highlight the use of tetrazole bind the active site heme then collected and washed three times in sterile saline. Contact Information: In iron, resulting Mice in Against exquisite selectivity Cryptococcal versus human CYPs while Meningitis with a Loading/Maintenance Dose Strategy Inocula concentrations were determined using a hemocytometer and N.P. Wiederhold, Posaconazole maintaining high potency of inhibition confirmed N.P. by Wiederhold, plating serial dilutions of fungal CYP5. L.K. Najvar, and colony,4 counts. A. Alimardanov, J. Cradock, X. Xu, M. Behnke, E.A. Ottinger, W.J. Hoekstra, E.P. Garvey, UTHSCSA MSC 7750 Murine Model of Cryptococcal Meningitis S.R. Brand, Fig R.J. Schotzinger, : VT9 Potently R. Bocanegra,,4 Binds Crypto CYP5 W.R. Kirkpatrick,,4 T.F. Patterson,4 770 Floyd Curl Dr., San Antonio, TX 789 T Immunocompetent ICR mice UT were Health anesthetized Science and inoculated Center San intracranially Antonio, Tel: (0) Materials & Methods NIH Therapeutics for Rare and Neglected Diseases, through a midline puncture in the cranial vault (Nguyen et al. Viamet Antimicrob Pharmaceuticals, Agents Chemother 997; 4: 0). Table. Mean Inc., (SD) 4 wiederholdn@uths brain STVHCS tissue fungal burden & plasma concentrations (μg/ml) on days 8 & 5. C CYP5 Biochemical Studies CYP5 enzymes from Cryptococcus Antifungal therapy neoformans was initiated 4 hours after inoculation and consisted of Group Control VT9 VT9 VT9 VT9 VT9 VT9 MATERIALS var. neoformans AND METHODS (Cneo) Cryptococcus Figure. Inhibition of CYP5 activity by VT9. For each the following groups: RESULTS mg/kg (cont.) mg/kg mg/kg 0 mg/kg 0 mg/kg Agent var. (Cgru) (Cgat) and Crypto CYP5, determination of potency was limited by Screening for CrAg/preemptive Rx where next? The Rationale: In Africa, many patients continue to present with CM too late headache very non specific, traditionally diagnosis based LP But the great majority are known HIV seropositive, and many even ART experienced so are known to health services Many of these cases should be preventable Through early, subclinical, diagnosis using dipstick IMMY pointofcare LFA for CrAg; as we know CrAg shed in blood (urine CSF) v early in infection, prior to symptoms So: late stage HIV patients, CD4 <00, screened for CrAg; approx 48% who screen positive are preemptively treated with antifungals (currently fluconazole) prior to ART 8

9 Proportion died /5/07 Screening for CrAg/preemptive Rx Screening for CrAg/preemptive Rx where next? The Rationale: Screening identifies those patients at risk : Jarvis et al: Clin Infect Dis 009; 48:856 Serum Antigen positive (approx 5% of CD4<00) 8% developed meningitis Serum Antigen negative (95% of CD4<00) 0/660 patients developed meningitis And according all modelling, should be highly cost effective (Meya et al CID 00 5:448, Jarvis et al PloS One 0 8:e6988) CDC led promotion of screening strategy Data enough for WHO, DoH in South Africa, MSF and others to endorse, promote, and implement screening, But we now have prospective data.. South African implementation, Nelesh Govender and colleagues Nicky Longley s study in Cape Town (Clin Infect Dis 06; 6:58) REMSTART, ORCAS trials, Tanzania (Ifekara group) Some common themes are emerging: Screening/preemptive fluconazole almost certainly does save lives But CrAg positives have continued high mortality And a significant proportion 040% CrAg positives have subclinical CM if they agree to LP A high CrAg titre predicts subclinical CM; and a high titre and subclinical CM are both associated with higher mortality We need to study more effective antifungal therapy, in the context of a bundle of diagnostic and adherence support for all Screening for CrAg/preemptive Rx REMSTART: Clinics in Dar Es Salaam, Lusaka: Sayoki Mfinanga, Duncan Chanda,,, Shabbar Jaffar. Lancet 05 85:78 Newly HIV diagnosed, CD4 <00, presenting for ART Standard of care vs CrAg screening, fluconazole (LP uptake low) plus ART adherence support lay worker weekly home visits for a month All got counselling and rapid (median 4 d) initiation of ART All TB GeneXpert sputum requested obtained in 8% of those not already on TB Rx, % new diagnosis of TB Primary endpoint all cause mortality at months 00 vs 998 enrolled Feb 0 Sept 0 Cryptococcal meningitis screening and communitybased early adherence support reduces allcause mortality among HIVinfected people initiating antiretroviral therapy with advanced disease: a randomisedcontrolled trial in Tanzania and Zambia. Sayoki Mfinanga, et al Shabbar Jaffar. Lancet 05 85:78 REMSTART Study Standard care CCS Follow up (months) Number at risk CCS Standard care Over a month followup, the number of deaths was 4(%) and 80(8%) respectively. The mortality rate was 8% (95% CI 0%, 4%) lower in the CCS arm than in standard care (p=0.004). In the CCS arm, 8 (.9%), were serum cryptococcal antigen positive All 9

10 5/5/07 CrAg positive vs CrAg negative mortality with screening/flucon: REMSTART: 0% vs % at m Longley et al: 5% vs % at m Wake, Govender et al: in progress: 7% vs % at 6 m Wake R, Sriruttan C,, Govender N. in preparation: EMBO AIDS Mycoses meeting, Cape Town July 06 Town 06 Significant proportion CrAg+ have evidence of meningitis (if agree to LP) even if asymptomatic Longley et al CID 06 Mar ;6(5):587 Cape Town: 40% (CID 06:6:58) Johannesburg: 4% (AIDS Mycoses, Cape Town 06) Tanzania: 8% (AIDS Mycoses, Cape Town 06) Risk is higher at higher blood Ag titre: Screening/preemptive Rx where next? Titre clearly associated with mortality ORCAS cut off same as for predicting subclinical CM (Bozena et al CROI 06 ) Causes of death in CrAg+ REMSTART: verbal autopsy 8/ thought CM Longley et al 7 deaths?most not CM Govender, Wake et al 6/0 deaths clinically probably CM including limited post mortem data in showing CSF CrAg positive when had been negative premortem Many CrAg +ve may benefit from more effective antifungal therapy Especially those at high titre We will be able to distinguish high and low titre through semiquantitative tests from both Immy and Biosynex Screening/preemptive Rx where next? Fluconazole plus flucytosine ( wks) is a safe effective and could be feasible and sustainable oral option for more effective therapy Consortium: Shabbar Jaffar, Nelesh Govender et al, Sayoki Mfinanga et al, Emili Letang, Joe Jarvis David Meya David Boulware et al. St George s group: REFINE STUDY: South Africa, Tanzania, Uganda Botswana CrAg positive randomised to: Fluconazole vs Fluconazole+flucytosine (first wks) In analysis explore cut off titre (if any) at which combination is needed This cut off could be targeted by semiquantitative LFA tests All participants: optimal additional diagnostics, incl. urine TB diagnostics, and ART adherence support 0

11 5/5/07 St. George s Tihana Bicanic Angela Loyse Sile Molloy Jack Adams Claire Robb Shichina Kannambath Rachel Wake Nicky Longley Joe Jarvis Wilbur Sabiiti Mat Beale Annemarie Brouwer Asna Siddiqui Emma Robertson Liverpool Shabbar Jaffar and colleagues LSHTM, University of Botswana Joe Jarvis Institut Pasteur Olivier Lortholary Tanzania Sayoki Mfinanga and colleagues South Africa Nelesh Govender and Graeme Meintjes Lilongwe, Malawi UNC project: Cecilia Kanyama Mina Hosseinipour Blantyre Rob Heyderman David Lalloo Jayne Ellis Kate Gaskell Newton Kalata Damien Ming Mary Pierse Dalitso Segula Wezzi Chimanganga Ebbie Gondwe Patrick Goodson Chikondi Jassi Christopher Kukacha Lucy Keyla Mphatso Likwinj Stanford Miyang Veronica Mwloza Limbanazo Matandika Neema Mthunthama Camilla Rothe Cecilia Sambakunsi Florence Shumba Ethel Chilima Cameroon (Yaounde/Douala) Charles Kouanfack Olivier Lotholory Sinata KoualaShiro Yacouba Mapoure Elvis Temfack Victor Sini Alain Kamden JeanGilbert Ndong Sedrick Tongo Sandrine Sa a Esther Sonkeng Philomene Vigne Margeurite Wodo. Faustin Bella Gertrude Goula Aboubakar Dang Julie Epupa Henri Essaka Eunice Tchatal Michelle Alma Reine Dombu Rine Estelle Djoukwe Clautilde Embolo Etoundi Owono Lilongwe Tanzania Cecilia Kanyama (Muhimbili/Amana/Mw Mina Hosseinipour ananymala) Chimwemwe Chawinga Sayoki Mfinanga Jacob Phulusa Sokoine Lesikari Towera Banda Ahmed Abdallah Lusungu Msumba Asteria Tumaini Blessings Banda Maliwaza Mganga Mwai Chipeta Rhoda Kalinga Nelecy Chome Kigocha Okeng o Creto Kanyemba Deus Buma Edwards Kasonkanji Asha Chande Romana Lawrence Tarimo Emily Kumwenda James Kalabashanga Wilberforce Mhango, Mary Nyange Felix Namaluweso Dickson Kileo Dorothy Sichali Scanderia Mamboya TMG Zomba Síle Molloy Adrienne Chan Angela Loyse Joep van Oosterhout Tihana Bicanic Philip Bright Neil Stone Duncan Lupiya Tom Harrison Grace Kaphale Shabbar Jaffar Loudon Silwamba Doulao Wang John Bradley Jack Adams Zambia Duncan Chanda Shabir Lakhi Aggrey Mweeba Abidan Chansa Natasha Karunaharan Muhammad Mputu Lawrence Mwenge Diana Chanda Enock Chikatula Yengwe Kakusa Marie Claire Mukasine John Mwaba Leocrisia Mwanamoonga Joseph Nglube Maria Sichone Josephine Silumba TSC Graeme Meintjes Maryline Bonnet Calice Talom Newton Kumwenda DMC Andrew Nunn Halima Dawood William Powderly Andrew Kitua South Africa: Gauteng Province Slide thanks Nelesh Govender ART coverage data derived from Actuarial Society of South Africa model

12 5/5/07 Step No AmB Step AmB Flucon 00 No 5FC Flucon FC Phase UNC Lilongwe EFA Results Nussbaum et al. CID 00; 50:8 44 Jackson et al. AIDS 0 6:670 Short course AmB: Gain in tolerability:: Muzoora et al Flu00+ AmB.0 5d AmB wks Bicanic CID 008; 47: 0 Day 4 Day 4 Laboratory parameter Day 7 Day 4 AmB 0.7 AmB.0 Decrease in hemoglobin level, > g/d, (%) Decrease in hemoglobin level, % change, mean Creatinine level >x baseline level (%) no grade III/IV anemia, ALT, no grade IV hypokalaemia, creatinine Trends in mortality: compared earlier Mbarara cohorts: and 0 wks: %, 8% Muzoora et al: J Infect 0; 64:76 Baseline characteristics and associations with 0week mortality (n=5) Variable All Alive Dead Univariate Multivariate Male (%) 67 (5%) Age (yrs) 4 (99) (98) 5 (940) Weight(kg) 5 (4659) 5 (4760) 49 (4455) < CD4 count (x0 6 /L) 5 (054) 8 (58) 9 (74) ART naïve 505 (97%) Features of CM Abnormal mental status (%) 7 (4%) 4% 44% < ,000 (5, ,000 (74,000 Baseline fungal burden(cfu/ml CSF ) 00,000 (7,00087,500) 60,000),680,000) < CSF opening pressure (cm H O) 6 (68) 7 (78) (58) 0. Initial treatment of CM Amphotericin B (median mg/kg/d) 47 (80%) 86% 68% < Fluconazole (median 00 mg/d) 06 (0%) rate of clearance (log CFU/d) 0.6 (0.0.56) 0.8 ( ) 0.8 (0.0.48) <0.00 Deaths weeks (%) 84/5 (6%) 0 weeks (%) 69/505 (%) Values are n(%) or median (IQR) Jarvis Bicanic Loyse Muzoora Meintjes et al; Clin Infect Dis 04; 58:7645 WHEN TO START ART?? Nb RCT of ART timing tell us when NOT to start Current recommendation: 46 weeks after antifungal therapy started, based: ACTG, Zolopa et al: <6 weeks (vs weeks) PLoS One 009; 4:e5575 Makadzange, al: > d (vs 0 wks) CID 00; 50:5 Boulware, al: >8 d (vs 5 wks) NEJM 04; 70:487 We currently favour around 4 weeks

13 Adjusted clearance rate (log CFU/ml CSF/d) 5/5/07 Phenotype of CD4 memory T cell CRAGspecific responses at day 0 IFN IL MIP TNF n = 7 starting ART Survived Died Jarvis Bicanic Loyse Muzoora Meintjes et al; Clin Infect Dis 04; 58:7645 IFN IL MIP TNF CSF cytokine profile in survivors and non survivors Thai study Brouwer et al. Lancet 004 6:764 Siddiqui et al. J Immunol :746 Jarvis et al. PLoS Path 05 April Jarvis Bicanic Loyse Muzoora Meintjes et al; Clin Infect Dis 04; 58:7645

14 Tropical Diseases, London School of Hygiene and Tropical Medicine, United Kingdom; 5 BotswanaUPenn Partnership, Gaborone, Botswana; 6 Research Centre for Infection and Immunity, Division of Clinical Sciences, St George s University of London, United Kingdom; 7 Infectious Disease Institute, Makerere University, Uganda; 8 Mbarara University of Science and Technology, Uganda; 9 Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam; 0 Liverpool School of Tropical Medicine, Liverpool, United Kingdom; University of Cape Town, South Africa; Mahidol Oxford Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand * These authors contributed equally. EFA: 0. ( 0.8, 0.08) EFA: 0.5 ( 0.8, 0.) EFA: 0.7 ( 0.6, 0.9) EFA: 0.5 ( 0.9, 0.0) EFA: 0.49 ( 0.5, 0.45) EFA: 0.58 ( 0.64, 0.5) Grey dots and lines display individual patients, blue lines correspond to a linear smoother assuming different rates of decline in the first and second week, respectively. EFA estimates are given as log 0 CFU/ml per day and numbers refer to the posterior median and quartiles of the estimate. FLZ: fluconazole monotherapy (n=80); FLZ+5FC: fluconazole/flucytosine combinations (n=); AmB mono: amphotericin B monotherapy (n=5), AmB+Azole: amphotericin B/azole combinations (n=44), AmB+5FC: amphotericin B/flucytosine combinations (n=4); AmB+5FC+IFN: amphotericin B/flucytosine/interferongamma combinations (n=58). 5/5/07 CID :70 CID 04 58:76 colony counts during the I trials in HIVassociated ted, it is unclear whether Association between baseline CSF immune profiles and Clinical Outcomes Jarvis et al PloS Path 05 April ty a surrogate marker for mortality in Tropical Medicine Research Unit, Mahidol University, Bangkok, Thailand * These authors contributed equally. ngal therapies in HIV associated cryptococcal meningitis. While associations between EFA and survival have been reported, it is unclear whether g, Joseph N. Jarvis 4,5, Tihana Bicanic 6, Bozena Morawski, David B. EFA Meya is a surrogate 7, Conrad marker. Muzoora 8, Chau Tran Thi Hong,9, oulware,*, Jeremy Day,,*, Thomas Ha rrison 6,* e for Tropical Medicine, Oxford University, Oxford, United Kingdom; University of Minnesota, Minneapolis, Methods MN, USA; 4 Department of Clinical Research, Faculty of Infectious and Partnership, Gaborone, Botswana; 6 Research Centre for Infection and Immunity, Division of Clinical Sciences, St George s University of London, United Kingdom; 7 Infectious Disease l for Tropical Diseases, Ho Chi Minh City, Vietnam; 0 Liverpool School of Tropical Medicine, Liverpool, Data United from Kingdom; eight randomized University controlled of Cape Town, trials and South four Africa; cohort Mahidol studies Oxford from Asia and Africa ( distinct and Africa ( distinct ct treatment and study imated based on a linear SF Cryptococcus colony Corresponding tenweek were included. cohort. Results (continued) Is early fungicidal activity a surrogate marker for mortality in the evaluation of antifungal therapies in HIV associated meningitis? Marcel Wolbers,, Katherine Huppler Hullsiek, Nhat Le Tha nh Hoang, Joseph N. Jarvis 4,5, Tihana Bicanic 6, Bozena Morawski, David B. Meya 7, Conrad Muzoora 8, Chau Tran Thi Hong,9, David G. Lalloo 0, Graeme Meintjes, Nicholas J. White,, David R. Boulware,*, Jeremy Day,,*, Thomas Ha rrison 6,* Oxford University Clinical Research Unit, Wellcome Trust Major Overseas Programme, Vietnam; Centre for Tropical Medicine, Oxford University, Oxford, United Kingdom; University of Minnesota, Minneapolis, MN, USA; 4 Department of Clinical Research, Faculty of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, United Kingdom; 5 BotswanaUPenn Partnership, Gaborone, Botswana; 6 Research Centre for Infection and Immunity, Division of Clinical Sciences, St George s University of London, United Kingdom; 7 Infectious Disease Institute, Makerere University, Uganda; 8 Mbarara University of Science and Technology, Uganda; 9 Hospital for Tropical Diseases, Ho Chi Minh City, Vietnam; 0 Liverpool School of Tropical Medicine, Liverpool, United Kingdom; University of Cape Town, South Africa; Mahidol Oxford Background Early fungicidal activity (EFA), i.e. the CSF clearance rate of quantitative yeast culture colony counts during the first 4 days of antifungal therapy, is often used as the primary endpoint in phase II trials in HIVassociated treatment and study combinations) were pooled a,b,c. For each of the distinct treatment and study combinations, EFA in the first two weeks after initiation of antifungal therapy was estimated based on a linear mixed effects model with a random intercept and slope of the longitudinal log0csf Cryptococcus colony forming unit (CFU) counts treating values below the detection limit as leftcensored. Corresponding tenweek risks of death were estimated with the KaplanMeier method. Association between EFA and mortality Results Figure. Scatter plot of EFA vs. 0week mortality Data for all treatment/study combinations. from 976 subjects contributing a total of 85 quantitative culture measurements were included. EFA estimates by treatment class R Figure. Longitudinal fungal counts and EFA estimates by treatment class in the pooled cohort. (weighted by sample size): 0.44 EFA: 0. ( 0.8, 0.08) EFA: 0.5 ( 0.8, 0.) EFA: 0.7 ( 0.6, 0.9) EFA: 0.5 ( 0.9, 0.0) EFA: 0.49 ( 0.5, 0.45) EFA: 0.58 ( 0.64, 0.5) Is early fungicidal activity a surrogate marker for mortality in the evaluation of antifungal therapies in HIV associated meningitis? Marcel Wolbers,, Katherine Huppler Hullsiek, Nhat Le Tha nh Hoang, Joseph N. Jarvis 4,5, Tihana Bicanic 6, Bozena Morawski, David B. Meya 7, Conrad Muzoora 8, Chau Tran Thi Hong,9, David G. Lalloo 0, Graeme Meintjes, Nicholas J. White,, David R. Boulware,*, Jeremy Day,,*, Thomas Ha rrison 6,* Oxford University Clinical Research Unit, Wellcome Trust Major Overseas Programme, Vietnam; Centre for Tropical Medicine, Oxford University, Oxford, United Kingdom; University of Minnesota, Minneapolis, MN, USA; 4 Department of Clinical Research, Faculty of Infectious and Results (continued) Association between EFA and mortality Figure. Scatter plot of EFA vs. 0week mortality for all treatment/study combinations. Background Early fungicidal activity (EFA), i.e. the CSF clearance rate of quantitative yeast culture colony counts during the Association between EFA and mortality Figure 4. Observed treatment effect on EFA vs. effect on 0week mortality for all randomized comparisons first 4 days of antifungal therapy, is often used as the primary endpoint in phase II trials in HIVassociated Figure. Scatter plot of EFA vs. 0week mortality for all treatment/study combinations. cryptococcal meningitis. While associations between EFA and survival have been reported, it is unclear whether from the pooled cohort. EFA is a surrogate marker. Methods Data from eight randomized controlled trials and four cohort studies from Asia and Africa ( distinct treatment and study combinations) were pooled a,b,c. For each of the distinct treatment and study combinations, EFA in the first two weeks after initiation of antifungal therapy was estimated based on a linear mixed effects model with a random intercept and slope of the longitudinal log0csf Cryptococcus colony forming unit (CFU) counts treating Rvalues below the detection limit as leftcensored. Corresponding tenweek risks (weighted by sample size): 0.04 of death were estimated with the KaplanMeier method. Results Data from 976 subjects contributing a total of 85 quantitative culture measurements were included. EFA estimates by treatment class R (weighted by sample size): 0.44 Figure. Longitudinal fungal counts and EFA estimates by treatment class in the pooled cohort. Results (continued) R (weighted by sample size): 0.44 Figure 4. Observed treatment effect on EFA vs. effect on 0week mortality for all randomized comparisons from the pooled cohort. R (weighted by sample size): 0.04 EFA: 0.5 ( 0.9, 0.0) EFA: 0.58 ( 0.64, 0.5) Grey dots and lines display individual patients, blue lines correspond to a linear smoother assuming different rates of decline in the Figure 4. Observed treatment effect on EFA vs. effect first on and second 0week week, respectively. mortality EFA for estimates all randomized are given as log 0 comparisons CFU/ml per day and numbers refer to the posterior median and quartiles of the estimate. from the pooled cohort. FLZ: fluconazole monotherapy (n=80); FLZ+5FC: fluconazole/flucytosine combinations (n=); AmB mono: amphotericin B monotherapy (n=5), AmB+Azole: amphotericin B/azole combinations (n=44), AmB+5FC: amphotericin B/flucytosine combinations (n=4); AmB+5FC+IFN: amphotericin B/flucytosine/interferongamma combinations (n=58). Mortality estimates by treatment class R (weighted by sample size): 0.04 Figure. KaplanMeier estimates of mortality by treatment class in the pooled cohort. Wolbers et al EMBO AIDS Mycoses meeting Cape Town July 06 Weighted mean of squared Somers rank correlation: R =0.07 Correlations (range between an ). individuals EFA and survival time within treatment/study combinations were also low: Weighted mean of squared Somers rank correlation: R =0.07 (range ). Mortality estimates by treatment class Conclusions Conclusions Figure. KaplanMeier estimates of mortality by treatment class in the pooled cohort. EFA is a sensitive marker for the activity of antifungal treatments and remains useful, especially for the phase II evaluation of antifungal therapies. Surrogacy of EFA for 0week mortality could not be established. Limitations of this study are that azole and azole combination treatments were more often used in the most resourcelimited settings, and that only one of the included antifungal trials was powered for mortality. EFA is a sensitive marker for the activity of antifungal treatments and remains useful, especially for the phase II evaluation of antifungal therapies. Surrogacy of EFA for 0week mortality could not be established. References Limitations of this study are that azole and azole combination treatments were more often used in the most a Jarvis JN, Bicanic T, Loyse A et al. (0). Determinants of Mortality in a Combined Cohort of 50 " resourcelimited settings, and that only one of the included PatientsWith antifungal HIVAssociated trials Cryptococcal was Meningitis: powered Implications for Improving mortality. Outcomes, Clinical " Infectious Diseases 58:76 45 b Day JN, Chau TTH, Wolbers M et al. (0). Combination Antifungal Therapy for Cryptococcal " Meningitis. N Engl J Med 68:90. c Boulware DR, Meya DB, Muzoora C, et al (04). Timing of Antiretroviral Therapy after Diagnosis of " Cryptococcal Meningitis. N Engl J Med 70: References EMBO AIDS Mycoses meeting Cape Town July 06 Correlations between an individuals EFA and survival time within treatment/study combinations were also low: a Jarvis JN, Bicanic T, Loyse A et al. (0). Determinants of Mortality in a Combined Cohort of 50 " PatientsWith HIVAssociated Cryptococcal Meningitis: Implications for Improving Outcomes, Clinical " Infectious Diseases 58:76 45 b Day JN, Chau TTH, Wolbers M et al. (0). Combination Antifungal Therapy for Cryptococcal " Meningitis. N Engl J Med 68:90. ifferent rates of decline in the r to the posterior median and mb mono: amphotericin B in B/flucytosine combinations Correlations between an individuals EFA and survival time within treatment/study combinations were also low: Weighted mean of squared Somers rank correlation: R =0.07 (range ). c Boulware DR, Meya DB, Muzoora C, et al (04). Timing of Antiretroviral Therapy after Diagnosis of " Cryptococcal Meningitis. N Engl J Med 70: