Infection DOI 10.1007/s15010-015-0745-2

REVIEW

The Cologne–Bonn cohort: lessons learned Jürgen Kurt Rockstroh 

Received: 22 January 2015 / Accepted: 5 February 2015 © Springer-Verlag Berlin Heidelberg 2015

Abstract  Much of our knowledge about HIV infection has been obtained from cohort studies, including description of the natural history of infection, identification of CD4 count and viral load as good surrogate markers of clinical progression, identification of co-factors [including older age and viral infections (CMV, HCV)] for progression of HIV-related disease and assessment of impact of highly active antiretroviral therapy on clinical outcomes. The Cologne–Bonn cohort was founded by Gerd Fätkenheuer and Bernd Salzberger after introduction of combination antiretroviral therapy in 1996 and has delivered important findings which have helped to improve treatment strategies as well as quality of overall care in HIV infection in these two cities. Indeed, the first pivotal paper from the cohort reported on an unexpectedly high rate of virological treatment failure of protease inhibitor therapy in an unselected cohort of HIV-infected patients. The subsequent analysis of risk factors for virological failure initiated the development of more potent HIV combination therapy. This review summarizes some of the major findings and contributions from the Cologne–Bonn cohort since 1996. Keywords  HIV · Cohort · ART

J. K. Rockstroh (*)  Department of Medicine I, University Hospital Bonn, Sigmund‑Freud‑Str. 25, 53105 Bonn, Germany e-mail: [email protected]‑bonn.de J. K. Rockstroh  German Centre for Infection Research (DZIF), Partner Site Bonn-Cologne, Bonn, Germany

Introduction Acquired immune deficiency syndrome (AIDS) was first described as a new clinical entity in 1981 [1, 2]. Initial reports were based on an unusual increase in the incidence of Kaposi sarcoma and pneumocystis pneumonia, diseases that were considered at that time to occur rarely. In 1983 the human immunodeficiency virus (HIV) was identified as the true cause of AIDS [3]. In 1987 the first antiretroviral agent, AZT (zidovudine, Retrovir®), was licensed for the treatment of HIV. With the introduction of first drugs for treatment of HIV, the need for observational cohorts which were followed longitudinally emerged, allowing to assess treatment outcome beyond randomized clinical trials in real-life patient populations. As registrational trials for new antiretroviral to the very day mostly only have a follow-up between 96 and 144 weeks, cohort studies allow for longer follow-up and also due to much larger patient numbers to better detect rare drug-related severe adverse events and long-term impact of HIV therapy on well-defined clinical end points. When highly active antiretroviral therapy was introduced in 1996 allowing for the first time complete viral suppression, the idea was born under the leadership of Gerd Fätkenheuer and Bernd Salzberger to build up the Cologne–Bonn cohort to study the impact of these new mostly HIV protease inhibitor-based therapies and their impact on clinical outcome [4]. Subsequently, many clinically relevant questions have been addressed within the cohort mostly aiming at better understanding the disease course in the respective cohort, but also as a quality control and direct feedback upon changes in clinical care. Data from the cohort have also been fed into larger cohort efforts (such as COHERE, ClinSurv, ART-CC) allowing answering meaningful clinical questions in ten thousands of patients. This review aims at summarizing some of the

13

J. K. Rockstroh

main findings from the Cologne–Bonn cohort and to outline some of the most relevant clinical messages from this endeavor. This review is dedicated to Gerd Fätkenheuer’s 60th birthday.

Which factors determine virological failure in HIV patients receiving PI‑based HAART? After introduction of HIV protease inhibitors in 1996, the first major study within the Cologne–Bonn cohort was the determination of the virological treatment failure rate with protease inhibitor therapy in unselected patients and to further assess underlying risk factors [4]. A total of 198 HIVinfected patients treated with protease inhibitors in 1996 were included in the analysis. A total of 226 treatment episodes with protease inhibitors were evaluable (saquinavir, 83; ritonavir, 47; indinavir, 96). The overall rate of virological treatment failure was very high with 44 % (saquinavir, 64 %; ritonavir, 38 %; indinavir, 30 %) indicating that the implementation of this first PI-based combination therapy was facing considerable challenges. In additional multivariate analysis, the following independent risk factors for virological failure were found: CD4 cell count, pre-treatment with antiretroviral drugs (number) and protease inhibitor (compound). The relative risk reduction for each CD4 cell count increase was 0.997 (P  = 0.012), 2.64 for pre-treatment with one or two drugs versus no drug (P = 0.05), 2.97 for pre-treatment with more than two drugs versus no drug (P  = 0.05) and 4.62 for treatment with saquinavir versus indinavir (P = 0.001). The conclusion of the paper was that response to antiretroviral combination therapy in normal clinical practice may considerably differ from results of randomized clinical trials. From the viewpoint of 2014, these results make even more sense as we by now have understood the initial limitations of first-generation HIV protease inhibitor: on the one hand, the low bioavailability of the compounds (in particular true for saquinavir) and on the other the tremendous negative impact on treatment response of the presence of NRTI-associated resistance-conferring mutations which in many patients led to a functional monotherapy with the new HIV PI, because due to resistance no activity from concomitant NRTI therapy was to be expected. In subsequent years, new HIV drugs were added as they became available to failing or recycled other HIV drugs, which ultimately led to repeated functional monotherapy cycles, which then ended in broad 3-drug class resistance. Fortunately, the advent of new drug classes with new mechanisms of action and the effort to only design salvage therapies with at best three fully active HIV drugs have helped to eventually control HIV replication even in this more challenging patient population. Introduction of the concept of boosting protease inhibitors (addition of low-dose

13

ritonavir as a pharmacoenhancer to guarantee efficacious trough drug levels of the companion PI) overcame the low bioavailability issues and led to once daily PI/r regimens with a high genetic barrier to resistance. Indeed, as a consequence of the high virological failure rate observed in the Cologne–Bonn cohort and following a first report on boosting protease inhibitors with low-dose ritonavir in the animal model, a first prospective trial in the Cologne–Bonn region was initiated which explored the safety and efficacy of a 2NRTI + indinavir/ritonavir dual PI regimen, as improved PK profile of indinavir through ritonavir drug interactions promised much better virological efficacy and, because of lower Cmax, no risk for the development of indinavir-associated kidney stones [5, 6]. Indeed, in 72 weeks of followup no virological failure with emergence of PI resistance and no case of kidney stones were reported [7].

Which impact does successful combination antiretroviral therapy have on overall survival in HIV? With treatability of HIV clearly one of the most important questions were, how does the prognosis of HIV change under successful control of HIV replication, was the survival benefit true for all patients and what were the predictive factors? Was life expectancy matched up to the general population? This is one of the good examples where answers can only be obtained from sufficiently large cohort collaborations. Therefore, data from the Cologne–Bonn cohort and the larger ART Cohort Collaboration were used to compare mortality in patients starting ART in nine countries of Europe and North America with the corresponding general population, taking into account their response to ART [8]. Eligible patients were enrolled in prospective cohort studies participating in the ART Cohort Collaboration. The ratio of observed to expected deaths from all causes [standardized mortality ratio (SMR)] was calculated, measuring time from 6 months after starting ART, according to risk group, clinical stage at the start of ART, and CD4 cell count and viral load at 6 months. Expected numbers of deaths were obtained from age-, sex- and country-specific mortality rates. Among 29,935 eligible patients, 1,134 deaths were recorded in 131,510 personyears of follow-up. The median age was 37 years; 8,162 (27 %) patients were females, 4,400 (15 %) were injecting drug users (IDUs) and 6,738 (23 %) had AIDS when starting ART. At 6 months, 23,539 patients (79 %) had viral load measurements ≤500 copies/ml. The lowest SMR, 1.05 [95 % confidence interval (CI) 0.82–1.35], was found for men who had sex with men (MSM), started ART free of AIDS, reached a CD4 cell count of ≤350 cells/microL and suppressed viral replication to ≤500 copies/ml by the sixth month. In contrast, the SMR was 73.7 (95 % CI

The Cologne–Bonn cohort

46.4−116.9) in IDUs who failed to suppress viral replication and had CD4 cell counts <50 cells/microL at 6 months. The percentage of patients with SMRs <2 was 46 % for MSM, 42 % for heterosexually infected patients and 0 % for patients with a history of injection drug use. The corresponding percentages for SMRs >10 were 4, 14 and 47 %. These data clearly demonstrated for the first time that under combined antiretroviral therapy, overall survival increased substantially and that for many patients excess mortality was moderate and comparable with patients having other chronic conditions. Furthermore, it was suggested, that excess mortality might be prevented by earlier diagnosis of HIV followed by timely initiation of ART. Indeed, since then guidelines for the initiation of antiretroviral therapy have globally moved to much earlier time points. Further cohort collaboration again including data from the Cologne–Bonn cohort even demonstrated, that mortality patterns in most non-IDU HIV-infected individuals with high CD4 counts on cART were now similar to those in the general population [9]. This has had a very strong impact on how patients are counseled about their prognosis after the initial HIV diagnosis and has helped to create a much better prognosis under successful HIV therapy.

Which challenges remain in the era of HIV treatability? With the advent of potent combination antiretroviral therapy with simplified single-tablet regimens and longer halflives and drastically improved short- and long-term tolerability, the question obviously becomes how have the causes of death changed in the HIV-seropositive patient population and which comorbidities that need to be addressed in clinical care are emerging. In a recent analysis from the Cologne–Bonn cohort, the causes of death were evaluated [10]. The causes of death from the Cologne–Bonn cohort between 2004 and 2010 were systematically recorded using the CoDe algorithm (The Coding Causes of Death in HIV Project). In 3,165 patients followed from 2004 to 2010, 182 deaths occurred (5.7 %, 153 males, 29 females). The median age at the time of death was 47 years (range 24–85 years). The most frequent causes of death were AIDS-defining events (n  = 60, 33 %), with non-Hodgkin lymphoma (NHL) (n = 29, 16 %) and infections (n = 20, 11 %) being the leading entities in this category. Non-AIDS malignancies accounted for 16 % (n  = 29), non-HIVrelated infections for 10 % (n  = 18), cardiovascular diseases for 7 % (n = 14), suicide or accident for 4 % (n = 7) and liver diseases for 3 % (n  = 5) of deaths (unknown n = 47, 26 %). Although the majority of patients (92.5 %) were on antiretroviral therapy (ART), only 50 % were virologically suppressed (HIV-RNA <50 copies/mL) and 44 %

had a decreased CD4 + count (<200/μL) at their last visit before death. These results clearly demonstrate that even in an era of combination antiretroviral therapy, one-third of death causes in the Cologne/Bonn cohort remain AIDS related, mostly due to late HIV diagnosis and presentation in late CDC stages with advanced disease. This emphasizes the need for improved testing strategies leading to earlier diagnosis and treatment of HIV infection. The most worrisome increase in non-AIDS death causes is observed for malignancies which are well in line with other cohort data, underlining the need for better malignancy screening in an increasingly aging HIV population [11].

Which impact does immune recovery have on long‑term survival in HIV‑seropositive patients on HIV therapy? Some HIV-infected individuals initiating combination antiretroviral therapy (cART) with low CD4 counts achieve viral suppression, but not CD4 cell recovery. Therefore, a recent cohort study aimed at identifying the risk factors for failure to achieve a CD4 count >200 cells/μL after 3 years of sustained viral suppression and to explore the association of the achieved CD4 count with subsequent mortality [12]. The analyses included treated HIV-infected adults from two large international HIV cohorts, who had viral suppression (≤500 HIV type 1 RNA copies/mL) for >3 years with CD4 count ≤200 cells/μL at the start of the suppressed period. Logistic regression was used to identify risk factors for incomplete CD4 recovery (≤200 cells/μL) and Cox regression to identify associations with mortality. Of 5,550 eligible individuals, 835 (15 %) did not reach a CD4 count >200 cells/μL after 3 years of suppression [12]. Increasing age, lower initial CD4 count, male heterosexual and injection drug use transmission, cART initiation after 1998 and longer time from initiation of cART to start of the virally suppressed period were risk factors for not achieving a CD4 count >200 cells/μL. Individuals with CD4 ≤ 200 cells/μL after 3 years of viral suppression had substantially increased mortality (adjusted hazard ratio, 2.60; 95 % confidence interval, 1.86–3.61) compared with those who achieved CD4 count >200 cells/μL. The increased mortality was seen across different patient groups and for all causes of death. In conclusion, suppressed HIVpositive individuals on cART who do not achieve a CD4 count >200 cells/μL have substantially increased long-term mortality. These findings help to provide the evidence why earlier HIV therapy is recommended in older patients. Also, it highlights that once the immune system is destroyed, complete reconstitution can be very difficult to achieve again stressing the need to start HIV therapy at earlier CD4 counts and to prevent late presentation.

13

J. K. Rockstroh Table 1  Patient characteristics of the Cologne/Bonn cohort in 2014 (personal communication Janne Vehreschild) Parameter

(n)

(%)

Patients, total Women On ART Below the limit of detection (<50 copies/ml) Median age (years) Risk factor for HIV transmission  Occupational exposure  Blood transfusion  Hemophilia  Heterosexual contacts  Men who have sex with men  I.V. drug use  Pattern II  Pre- or perinatal infection  Unknown CDC stage  A1  A2  A3

2,736 554 2,437 2,035 47 (19−88) 8 7 17 96 561 1,401 148 254 7 236

100.0 20.2 89.1 74.4 0.29 0.26 0.62 3.51 20.51 51.22 5.41 9.29 0.26 8.73

955 580 103

34.92 21.21 3.77

 B1  B2  B3  C1  C2

240 114 33 308 282

8.78 4.17 1.21 11.26 10.31

 C3

120

4.39

The Cologne/Bonn cohort in 2014 Looking at the current main patient characteristics of the Cologne/Bonn cohort in Table 1, it becomes clear that the median age has increased quite impressively since the beginning of the HIV epidemic. Indeed, it can be estimated that by the end of 2015, almost half of the entire cohort will be ≥50 years old. Therefore, consequent comorbidity management becomes a central part of HIV management, including assessment of cardiovascular risk, osteoporosis workup and neurocognitive testing. The epidemiology of HIV infection in Germany remains unchanged and shows the highest prevalence of HIV infection in men who have sex with men. Nevertheless, the 20 % women in the Cologne/Bonn cohort also underline that the number of women living with HIV within the cohort has increased quite steadily. Reassuringly, most patients are on combination antiretroviral therapy and well suppressed. In a recent study, the clinical course of a cohort of 247 treatment-naïve HIV-infected patients newly presenting to an HIV clinic within the Cologne/Bonn cohort was followed for a median of 4.2 years [13]. Twenty-two percent had clinical signs of

13

HIV infection (CDC stage B or C), 32 % had CD4 cells below 200 per microliter and 56 % had CD4 cells below 350 per microliter. Thus, a high proportion of this patient population can be classified as late presenters, comparable to data reported by the largest European cohort initiative COHERE (51.7–57.3 %) [14]. This underlines again the need for earlier HIV diagnosis and earlier access to HIV therapy. With regard to about 25 % of patients within the Cologne/Bonn cohort (see Table 1) who have already developed a CDC-classified AIDS event (CDC stage C), it becomes clear that a substantial proportion of patients received their HIV diagnosis mostly only after first AIDS manifestations. Unfortunately, these late presenters have a considerably worse prognosis than those HIV-positive individuals who start their HIV therapy before development of symptomatic disease or advanced immunodeficiency. Interestingly, further cohort collaborative work (including the patients from the Cologne/Bonn cohort) has demonstrated that there is a persistent role of CD4 count at baseline and at 12 months after starting combination antiretroviral therapy in predicting AIDS, non-AIDS infection and nonAIDS malignancy deaths [15]. Also, lack of viral suppression on ART was associated with AIDS, non-AIDS infection and other causes of death, clearly emphasizing that the current goal of HIV therapy can only be complete virus suppression.

Conclusions Overall, the establishment of the Cologne/Bonn cohort by Gerd Fätkenheuer and Bernd Salzberger back in 1996 after first introduction of highly active antiretroviral therapy has paved the way for better understanding of the clinical outcome of HIV in the era of combination antiretroviral therapy and has helped to determine which factors influence disease outcome. The quality of care and success of HIV therapy have dramatically increased with the help of the cohort data. Therefore, continuation of the cohort appears instrumental as a well-validated tool of quality control for HIV management in the region. Acknowledgments  This work was supported by the DZIF TTU HIV Project 05.803 and the German Center for Infection Research (DZIF). Conflict of interest  JKR received honoraria for speaking at educational events or consulting from Abbvie, Bionor, BMS, Boehringer, Gilead, Janssen, Merck, Tibotec and ViiV.

References 1. Gottlieb MS, Schanker HM, Fan PT, et al. Pneumocystis Pneumonia−Los Angeles. MMWR Weekly. 1981;30(21):250–2.

The Cologne–Bonn cohort 2. Gottlieb MS, Schroff R, Schanker HM, et al. Pneumocystis carinii pneumonia and mucosal candidiasis in previously healthy homosexual men: evidence of a new acquired cellular immunodeficiency. N Engl J Med. 1981;305:1425–31. 3. Barre-Sinoussi F, Chermann JC, Rey F, et al. Isolation of a T-lymphotropic retrovirus from a patient at risk for AIDS. Science. 1983;220:868–71. 4. Fätkenheuer G, Theisen A, Rockstroh J, et al. Virological treatment failure of protease inhibitor therapy in an unselected cohort of HIV-infected patients. AIDS. 1997;11(14):F113–6. 5. Kempf DJ, Marsh KC, Kumar G, et al. Pharmacokinetic enhancement of inhibitors of the human immunodeficiency virus protease by coadministration with ritonavir. Antimicrob Agents Chemother. 1997;41(3):654–60. 6. Rockstroh JK, Bergmann F, Wiesel W, et al. Efficacy and safety of twice daily first-line ritonavir/indinavir plus double nucleoside combination therapy in HIV-infected individuals. German Ritonavir/Indinavir Study Group AIDS. 2000;14(9):1181–5. 7. Lichterfeld M, Nischalke HD, Bergmann F, et al. Longterm efficacy and safety of ritonavir/indinavir at 400/400 mg twice a day in combination with two nucleoside reverse transcriptase inhibitors as first line antiretroviral therapy. HIV Med. 2002;3(1):37–43. 8. Collaboration Antiretroviral Therapy Cohort, Zwahlen M, Harris R, May M, et al. Mortality of HIV-infected patients starting potent antiretroviral therapy: comparison with the general population in nine industrialized countries. Int J Epidemiol. 2009;38(6):1624–33. 9. Collaboration of Observational HIV Epidemiological Research Europe (COHERE) in EuroCoord, Lewden C, Bouteloup V, De

Wit S, et al. All-cause mortality in treated HIV-infected adults with CD4 ≥ 500/mm3 compared with the general population: evidence from a large European observational cohort collaboration. Int J Epidemiol. 2012;41(2):433–45. 10. Ehren K, Hertenstein C, Kümmerle T, et al. Causes of death in HIV-infected patients from the Cologne–Bonn cohort. Infection. 2014;42(1):135–40. 11. Smith CJ, Ryom L, Weber R, et al. D:a:D Study Group. Trends in underlying causes of death in people with HIV from 1999 to 2011 (D:A:D): a multicohort collaboration. Lancet. 2014;384(9939):241–8. 12. Engsig FN, Zangerle R, Katsarou O, et al. Antiretroviral Therapy Cohort Collaboration (ART-CC) and the Collaboration of Observational HIV Epidemiological Research Europe (COHERE) in Euro Coord. Long-term mortality in HIV-positive individuals virally suppressed for > 3 years with incomplete CD4 recovery. Clin Infect Dis. 2014;58(9):1312–21. 13. Platten M, Linnemann R, Kümmerle T, et al. Clinical course and quality of care in ART-naïve patients newly presenting in a HIV outpatient clinic. Infection. 2014;42(5):849–57. 14. Mocroft A, Lundgren JD, Sabin ML, Monforte A, Brockmeyer N, Casabona J, et al. Risk factors and outcomes for late presentation for HIV-positive persons in Europe: results from the collaboration of observational HIV epidemiological research Europe study (COHERE). PLoS Med. 2013;10(9):e1001510. 15. Ingle SM, May MT, Gill MJ, et al. Antiretroviral Therapy Cohort Collaboration. Impact of risk factors for specific causes of death in the first and subsequent years of antiretroviral therapy among HIV-infected patients. Clin Infect Dis. 2014;59(2):287–97.

13