High rate of death amongst patients with HIV diagnosed late

Patients who have an AIDS-defining illness at the time of their HIV diagnosis have a high rate of mortality, an international team of investigators report in the November 30^th edition of AIDS. The researchers believe that their results show the importance of earlier diagnosis of HIV. However, they found that significant numbers of patients diagnosed late achieved an undetectable viral load and an increase in their CD4 cell count above the key level of 200 cells/mm³ after antiretroviral therapy was started.

The amount of HIV-related illness and death fell dramatically after the introduction of effective HIV treatment in the late 1990s. However, in the UK, there are still approximately 800 HIV-related deaths a year, despite the provision of free HIV treatment and care. Many of these deaths occur in patients who have their HIV diagnosed late when they are already severely ill with HIV. Last year, doctors, policy makers and activists developed an action plan to reduce the number of late HIV diagnoses.

An international team of investigators from Europe and Canada wished to obtain a better understanding of the characteristics of patients who had were already ill with an AIDS-defining illness at the time of their HIV diagnosis, their outcome (progression to another AIDS-defining illness or death) and response to HIV treatment.

Information from eight cohorts was therefore pooled. A total of 760 patients diagnosed late between 1997 and 2004 were included in the investigators’ analysis. Most of the patients (78%) were male, however they came from a broad spectrum of risk groups with 36% being defined as heterosexual, 19% as homosexual and 12% as injecting drug users. Average age at the time of diagnosis was 39 years. The data revealed that the average CD4 cell count at diagnosis was dangerously low, the median being 42 cells/mm³, and median viral load at the time of diagnosis was 120,000 copies/ml.

The two most common AIDS-defining illnesses at the time of diagnosis were PCP (35%) and tuberculosis (22%), followed by candida (12%), Kaposi’s sarcoma and toxoplasmosis (both 9%), CMV (7%) and lymphoma (4%).

Information on prognosis was available for 584 patients and a total of 110 (19%) of these individuals developed a second AIDS-defining illness after their diagnosis. A total of 125 patients, and 28% of these individuals had developed a new AIDS-defining illness prior to their death. Median CD4 cell count at the time of death was 28 cells/mm³, with median viral load being 86,000 copies/ml.

Death occurred a median of 21 days after HIV diagnosis amongst the patients who did not receive antiretroviral therapy, and a median of thirteen months after diagnosis amongst the patients who received treatment with anti-HIV drugs.

The only baseline factors associated with a higher risk of death were older age (p = 0.0001) and higher viral load (p = 0.002).

HIV treatment was started by 624 patients (82%). HIV treatment was started a median of 31 days after diagnosis with HIV. Treatment was started sooner in patients diagnosed in 2004 and in those who were older or had Kaposi's sarcoma at the time of their diagnosis. Patients with tuberculosis (TB) at the time of their diagnosis started treatment significantly later than other patients. The investigators believe that this is because the treating doctors wished to avoid the occurrence of an immune reconstitution inflammatory syndrome.

Median CD4 cell count at the time treatment was initiated was 41 cells/mm³. Two-thirds of patients started HIV treatment with a combination based upon a protease inhibitor, with 25% initiating an NNRTI-based regimen. One-in-ten patients started therapy with a triple NRTI regimen, a combination of drugs that was revealed to be suboptimal during the follow-up period.

Most (505, 89%) of the patients who took HIV treatment had at least one viral load measurement below 500 copies/ml. Viral load fell to this level a median of 90 days after antiretroviral therapy was started. The only factors significantly associated with an increased chance of achieving a viral load below this level were calendar year in which treatment was initiated (2003/04, p = 0.004), and lower baseline viral load (p = 0.01).

Viral load subsequently rebounded in 186 (37%) individuals. Factors significantly associated with this outcome included treatment with a triple NRTI regimen (p = 0.04), TB at the time of HIV diagnosis (p = 0.001) and diagnosis before 1999 (p = 0.003). Factors associated with a lower risk of rebound included treatment based upon a protease inhibitor (p = 0.02), diagnosis in 2003/04 (p = 0.03), a time when more powerful and durable HIV treatment was available, and older age at diagnosis (p = 0.0006).

Within three months of starting HIV therapy, median CD4 cell count had increased to 111 cells/mm³ and after a year of treatment, median CD4 cell count was 211 cells/mm³. Of the 587 patients whose CD4 cell count was below 200 cells/mm³ at the time of diagnosis, 58% had two consecutive counts above this level a median of one year and two months after starting treatment with anti-HIV drugs.

“Our study highlights the fact that patients who present for the first time with an AIDS event are likely to remain a clinical problem for the foreseeable future and deaths remain common in this group”, conclude the investigators, adding “steps to encourage earlier HIV testing and diagnosis among all groups and to increase clinical awareness of the possibility of an AIDS diagnosis among individuals who are not perceived to be at risk of HIV infection are required.”