icon star paper   HIV Articles  
Back grey_arrow_rt.gif
Late presenters in the era of highly active antiretroviral therapy: uptake of and responses to antiretroviral therapy
  AIDS: Volume 18(16) 5 November 2004
Sabin, Caroline Aa; Smith, Colette Ja; Gumley, Helenb; Murphy, Gabrielleb; Lampe, Fiona Ca; Phillips, Andrew Na; Prinz, Bethb; Youle, Mikeb; Johnson, Margaret Ab
From the aRoyal Free and University College Medical School and the bRoyal Free Centre for HIV Medicine, Royal Free NHS Trust, London, UK.
Note from Jules Levin: I just reported to you a French study in the previous issue of the journal AIDS examining response to HAART for people over 50 compared to under 50 years of age (median age was 35 vs 57 yrs). Of note, this UK study examines responses to HAART in patients who present late, with more advanced HIV.
Objectives: To investigate the characteristics and clinical, immunological and virological outcomes for individuals presenting for care with low CD4 cell counts.
Methods: Individuals aged > 16 years presenting for care for the first time were identified between 1 January 1996 and 31 December 2002. Late presenters were those with CD4 cell count < 50×106 cells/l. Follow-up was until last contact, death or 31 December 2002.
Results: Late presenters formed 15.3% (110) of the group; they were more likely to be female (35% versus 24%), heterosexual (53% versus 38%), and of Black-African ethnicity (39% versus 27%) than other individuals. Over a median follow-up of 2.5 years, 13% of late presenters died. Ninety-nine patients started antiretroviral treatment; Of the 11 patients who did not start antiretroviral treatment, eight died within 3 months of presentation. Among those starting treatment, 87 (87.9%) achieved a viral load < 400 copies/ml and median CD4 cell counts increased from 43 × 106 cells/l at 0-2 months after presentation to 204 × 106 cells/l at 1 year. Over the first year, 71 patients attended at least one outpatient visit (median, 4.5; range, 0-39), 21 attended at least one day case visit (median, 0; range, 0-15) and 49 were admitted as an inpatient (median, 0; range, 0-4).
Conclusions: Those presenting for care with very low CD4 cell counts may make large demands on clinical resources, particularly over the first few months. While some patients do have a poor outcome on highly active antiretroviral therapy, many will benefit from this therapy and will experience good immunological and virological responses.
While the introduction of highly active antiretroviral therapy (HAART) has had a tremendous impact on the health of individuals infected with HIV, a minority of patients under care still present late with low CD4 cell counts. Despite many attempts to increase the uptake of HIV testing nationally, many individuals in the United Kingdom are still not diagnosed until their CD4 cell count has already fallen to < 200 × 106 cells/l, the lower current threshold for initiating treatment in the United Kingdom. The implications of a low CD4 cell count are widespread: as well as being at higher risk of clinical events while the CD4 cell count is low, those who start HAART with very low CD4 cell counts appear to be less likely to have a sustained virological response compared with those starting at higher CD4 cell counts.
Anecdotally, there has been some concern that the number of patients presenting late for care has increased over the past few years. Any such increase would have an impact on both the service use and treatment outcomes at our centre. Therefore, the objectives of this study were to describe the characteristics of those presenting for care at our centre with a CD4 cell count < 50 × 106 cells/l; to assess whether this phenomenon had, indeed, increased; and to describe the clinical events and resource use that occurred in this group, the uptake of antiretroviral therapy and the resulting immunological and virological outcomes.
All patients aged > 16 years who were newly diagnosed at the Royal Free Hospital, London between 1 January 1996 and 31 December 2002 were identified. The choice of 1 January 1996 as a starting point for the study reflected the fact that HAART was first introduced to the United Kingdom in this year. Information on patient demographics, AIDS events and deaths, CD4 cell counts, viral loads and use of antiretroviral therapy was extracted from the Royal Free Hospital patient database. Information on resource use (inpatient, outpatient and day case attendances) was extracted from the hospital patient activity system.
Patients whose CD4 cell count at the time of diagnosis was < 50 × 106 cells/l were defined as late presenters. Patients were excluded from the analysis if they were known to have transferred to the Royal Free from another hospital, or if the patient reported having an AIDS event that occurred more than 1 month prior to their date of diagnosis, indicating that they may have been diagnosed previously (patients presenting with an AIDS event within a month prior to HIV diagnosis were, however, included). Patients were followed from the date of HIV diagnosis until their last visit to the Royal Free, 31 December 2002 or death, whichever occurred first.
Comparison of late presenters and non-late presenters

Between 1 January 1996 and 31 December 2002, 719 individuals presented for care for the first time at the Royal Free Hospital. The median CD4 cell count at presentation for all patients was 320 × 106 cells/l (range, 0-1499). Of these, 110 (15.3%) had CD4 cell counts < 50 × 106 cells/l and were defined as late presenters. In univariable analyses, late presentation was associated with female sex (P = 0.03), older age (P = 0.0001) (median age of late presenters was 37 vs 33 for non-late presenters), Black-African ethnicity (P = 0.02) and heterosexual/other risks for HIV infection (P = 0.0003). While the proportion of patients who presented late in each year ranged from 10.4% of those presenting for care in 2001 to 18.9% of those presenting for care in 2002, these differences were not statistically significant (P = 0.49). Multivariable analyses revealed that older age [odds ratio 1.04 per year older, 95% confidence interval (CI), 1.02-1.07; P = 0.0001], heterosexual (odd ratio, 2.14; 95% CI, 1.38-3.30; P = 0.0006) and other non-sexual risks for infection (odd ratio, 3.14; 95% CI, 1.40-7.06; P = 0.006) were independently associated with late presentation.
Over half [63 (57.3%)] of late presenters had a clinical condition that was AIDS defining at the time of or within 1 week from presentation, compared with only 44 (7.2%) of those diagnosed at higher CD4 cell counts (P = 0.0001). A total of 92 different AIDS conditions were diagnosed in the 63 late presenters with AIDS over this 5-week period, compared with 51 conditions in the 44 non-late presenters with AIDS. Among those with AIDS, late presenters were more commonly diagnosed with Pneumocystis carinii pneumonia than non-late presenters (40.2% of events in late-presenters compared with 13.7% of events in non-late presenters) whereas tuberculosis (15.2% and 31.4% of events, respectively) and Kaposi's sarcoma (4.3% and 21.6% of events, respectively) were less commonly diagnosed.
Uptake of and response to antiretroviral therapy
Late presenters were followed for a median period of 2.5 years (range, 8 days to 6.9 years). Over this time, 99 patients (90.0%) started antiretroviral therapy; the median time to starting antiretroviral therapy in these patients was 22 days (range, 0-771). Therapy was with a regimen containing a protease inhibitor in 44 (44.4%), a non-nucleoside reverse transcriptase inhibitor in 38 (38.9%) and three nucleoside reverse transcriptase inhibitors in five (5.1%) (three taking zidovudine/didanosine/lamivudine, one taking didanosine/stavudine/lamivudine and one taking zidovudine/lamivudine/abacavir). The remaining 12 patients who started antiretroviral therapy started regimens including one or two nucleoside reverse transcriptase inhibitors only. Eleven of these patients presented for care in 1996 (seven) or 1997 (four) at a time when HAART had only recently been introduced; most (8 out of 12) started HAART at a later stage. Of the 11 patients who did not start antiretroviral therapy, eight died within 3 months of presentation (five from HIV-related opportunistic infections, two from HIV-related malignancies and one of an unknown cause). One further patient died 9.5 months after presentation from an HIV-related opportunistic infection; one patient remained alive 3.3 years after presentation and one patient did not return for any follow-up visits (the status of this patient remains unknown).
Of the 99 patients starting HAART, 87 (87.9%) achieved an HIV-1 RNA level of < 400 copies/ml after starting HAART, with the median time taken to achieve this being 83 days. The only factor associated with achieving a viral load < 400 copies/ml from multivariable analyses was calendar year of presentation. Compared with those who presented in 1996-1997, those presenting in 1998-1999, 2000-2001 and 2002 all achieved a viral load < 400 copies/ml more rapidly with relative hazards of 2.69 (95% CI, 1.44-5.04; P = 0.002), 3.11 (95% CI, 1.59-6.07; P = 0.0009) and 6.84 (95% CI, 3.43-13.62; P = 0.0001), respectively. Rebound rates after an initial virological response were 16.8% by 1 year and 18.9% by 2 years.
Among those who started antiretroviral therapy and were under follow-up in each period, median CD4 cell counts rose from 43 × 106 cells/l (IQR, 20-93) within the first 2 months after starting antiretroviral therapy to 204 × 106 cells/l (IQR, 136-287) at 1 year, with the proportion of patients with a CD4 cell count > 200 × 106 cells/l increasing from 2 to 53% over the same time period. As expected, in analyses in which missing values were replaced with the last available measurement (to allow for informative censoring as those with the worst response to therapy dropped out through death) the increasing trend was attenuated.
Clinical events and resource use after diagnosis
Thirty patients developed at least one new AIDS condition more than 1 week after HIV diagnosis, with 38 events occurring in total. The vast majority of these events (24) occurred in the first 3 months after presentation or before starting antiretroviral therapy. After HIV diagnosis, the most common AIDS events were tuberculosis (seven events), Pneumocystis carinii pneumonia (six events), cytomegalovirus (six events), toxoplasmosis (four events) and wasting syndrome (three events). Fourteen (12.7%) patients died, including the nine patients described above who did not receive treatment and five patients who had started treatment (two deaths from HIV-related opportunistic infections, one from HIV-related malignancy, one from a toxicity thought to be related to HAART and one from a non-HIV-related cause). Kaplan-Meier estimates of death rates at 3, 6 and 12 months after presentation were 8.4%, 10.4% and 11.5%, respectively.
Over the first year, 49 patients (44.5%) attended for at least one inpatient visit; 71 (64.5%) attended for at least one outpatient visit, and 21 (19.1%) for at least one day case visit [with 75 (68.2%) patients attending for at least one visit of any type over the year. Over the year, the patients made a total of 73 inpatient visits (mean per patient, 0.7; median, 0; range, 0-5), 1044 outpatient visits (mean, 9.5; median, 7; range, 0-46) and 99 day case visits (mean, 0.9; median, 0; range, 0-15). Inpatient bed days totalled 1710, with each stay lasting for a median of 23 days (mean, 34.9; range, 1-151). Inpatient resource use, in particular, was most heavy in the first 3 months after HIV diagnosis and then dropped off thereafter, whereas outpatient attendances remained relatively constant over the year in this group.
Despite measures to encourage earlier HIV testing in the United Kingdom, a significant proportion of patients diagnosed with HIV in each year at our centre have a very low CD4 cell count. These patients commonly present with multiple AIDS conditions, place major demands on hospital resources (particularly in the first few months after presentation) and, in some cases, prognosis is poor. However, most of these individuals start antiretroviral therapy, and clinical and surrogate marker responses appear in most cases to be reasonable.
Our findings are in line with those of other studies from this centre and the rest of the United Kingdom in that those who present late are more likely to be female, of Black-African ethnicity and older. Earlier analyses of this cohort also reported that uptake of therapy is poorer in these patient groups. Patterns of late presentation may differ to those seen in other countries, however; a recent study from Canada confirmed that late presenters were more likely to be Black and older than those who presented with higher CD4 cell counts but that men were more likely to present late than women. A study of patients from Australia reported that patients presenting with AIDS were older, more likely to be hetero- or bisexual and more likely to be born overseas than those presenting without AIDS. Results from Spain also noted that late presenters tended to be older, male and of foreign origin. Differences in the choice of a definition of late diagnosis (CD4 cell count of 200 or 50 × 106 cells/l or presentation with an AIDS diagnosis) and in the analytical methods used may go some way to explaining these differences.
We did not find that the proportion presenting late had changed in our clinic over the period of study. Earlier analyses from the Royal Free reported a decline in the median CD4 cell count at presentation between 1994 and 2000, which might be explained by an increase in the proportions of individuals (mainly Black-Africans) presenting for care with very low CD4 cell counts. In contrast, a study from another clinic in London carried out in the pre-HAART era reported that CD4 cell counts at HIV diagnosis had remained relatively stable. Clearly, the small number of late presenters at these clinics, as well as the selected nature of patients attending the larger London clinics, may explain some of these inconsistencies but it may also limit the power of these studies to detect any subtle trends that may occur.
Over half of late presenters in this study had an AIDS-defining condition at the time of presentation, and 30 patients developed at least one new AIDS condition over follow-up. The 130 AIDS events that occurred in the late presenter group represents around 50% of all AIDS events reported in the 719 patients in this study over the same period. Therefore, this small group of patients represents a major use of hospital resources. Indeed, almost half of the patients attended for at least one inpatient visit over the first year of follow-up, with a mean total inpatient stay of 15.3 days (compared with 135 inpatient admissions and a mean total inpatient stay of only 3.1 days for the non-late presenter group). Most inpatient stays occurred in the first few months while the patients were receiving treatment for their AIDS events and were being stabilized on antiretroviral therapy. It is known that the use of antiretroviral therapy and AIDS diagnoses are two of the major factors that drive costs of HIV care and the high costs of late presentation have been described previously. As expected, as a consequence of the continued virological and immunological monitoring that is required once a patient starts antiretroviral therapy, the rate of outpatient and day case visits remained relatively stable over the first year. Further analyses of the factors that are associated with high use of hospital resources are underway in this cohort and will be reported separately.
Patients who present very late may be very difficult to treat. In addition to the fact that some of these individuals may die from the AIDS event that prompted their diagnosis, the combined pill burden of treatments for opportunistic infections and for HIV may have a negative impact on adherence that may adversely affect response to therapy. Furthermore, there are often many interactions between these treatments that may adversely affect the efficacy of treatment. It should also be noted that the factors determining late presentation in this group may also determine adherence patterns: only 68% of the patients under follow-up had a booked hospital attendance over the first year of follow-up, suggesting that adherence may be poor in some of these individuals. Consequently, it may be expected that responses in this group of patients would be poorer than those reported from other studies. In our cohort, many of the patients who did not start antiretroviral therapy had died within the first few months after diagnosis (largely from HIV-related opportunistic infections). Of those who did start treatment, some patients started non-HAART regimens (largely because patients had presented in 1996 and 1997 at a time when HAART had only recently been introduced). Despite this, virological and immunological responses among those who were able to start antiretroviral therapy were good, suggesting that even in this group of patients it is possible for patients to adhere reasonably well to therapy. Although a small number of patients died rapidly in this group, survival was dramatically improved compared to that which may have been expected in the pre-HAART era. In line with other studies, virological responses to treatment were improved in patients who started treatment more recently. This is likely to be a consequence of the use of more potent antiretroviral regimens, improved toxicity management and adherence support, the use of resistance testing and other improvements in clinical care.
This study has a number of limitations that should be noted. First, our clinic is a large clinic based in North London. London and the populations served by this clinic may not represent that seen in other UK clinics. In particular, only a small proportion of HIV-infected patients at the clinic report injecting drug use as the main risk factor for infection, and the proportion of ethnic minorities is lower than seen at some other clinics. Second, the CD4 cell count threshold we have chosen for our definition of a late presenter is very low and may not be comparable with other studies. This cut-off was chosen to identify a group at particularly high risk of clinical disease and whose response to antiretroviral therapy may be particularly poor. Our definition of a late presenter does require the patients to have a CD4 cell count measurement while under follow-up at the clinic. In addition to the 719 patients presenting to the Royal Free for the first time included in this analysis, a further 57 patients presented without a CD4 cell count and could not be identified as either late or non-late presenters. Of these, only one person is known to have died although it is likely that these other patients only attended on a single occasion and they are not under active follow-up. Finally, we cannot rule out the possibility that some of these patients had previously been diagnosed and treated elsewhere. Patients may not always feel able to reveal that they had previously been diagnosed with HIV, particularly if this happened in a different country.
The importance of diagnosing patients at an early stage of HIV infection is well recognized. As well as the public health implications for prevention of transmission of HIV to sexual partners in those with high viral loads who are unaware of their infection, early diagnosis enables patients to start treatment at a time when responses are maximal. The results from our study suggest, however, that late diagnosis continues to be a problem, and that those who are diagnosed late tend to be from disadvantaged groups. These individuals present with a complex clinical picture, making them difficult to treat, and they are likely to be very heavy users of hospital resources. Continued attempts to increase early testing and to improve awareness of HIV among general practitioners and hospital doctors, who may well treat these individuals prior to their HIV diagnosis, are essential.
  icon paper stack View Older Articles   Back to Top   www.natap.org