Untitled

Levy: JAMA, Volume 276(15).October 16, 1996.1219 JAMA Journal of the American Medical Association

Volume 276(15) 16 October 1996 p 1219

Controversies: Does the CD4⁺ Cell Count Reflect Clinical Efficacy? [Letters: In Reply]
Levy, Jay A. MD
University of California School of Medicine, San Francisco (Levy).

Outline

REFERENCES

In Reply.-The response by Drs Kinloch-de-Loes and colleagues to my Controversies reiterates the difficulty in knowing whether surrogate markers give a true insight into a clinical benefit. I maintain my premise that drug therapy should mimic the natural cellular immune responses associated with low viral RNA to control HIV and sustain a long-term asymptomatic course. Kinloch-de-Loes et al emphasized neutralizing antibodies, but these most likely have no effect on a clinical course: the cellular immune response is the major component in preventing progression to disease. My emphasis has been on approaches to increase this cellular immune activity, which can control HIV by a variety of mechanisms, particularly suppression of virus replication. [1] This comment, however, does not negate the conclusion by Kinloch-de-Loes et al (with which I agree) that a treatment regimen needs to be judged by its clinical effectiveness, not by any particular measure, although t! he reason for the effectiveness may give insight into other approaches to control HIV infection.

My comments on the monthly improvement of CD4⁺ cells were meant to emphasize that measuring this factor alone, particularly with such small increases in number after therapy, has not provided convincing evidence of clinical relevance. To cite the "statistical" importance of the results described is merely to underline my concern that statistics should not be expected to give conclusive answers to biological phenomena. The relationship of low CD4⁺ cell counts to the clinical state has been well established, but whether this factor has merit under treatment has been questioned by several investigators beside me. Moreover, to comment that most HIV-infected patients would rather have a treatment-induced excess of 0.137x10⁹/L CD4 (⁺) cells than not may be true only if there is no substantial toxicity associated with that therapy. Once again, emphasizing CD4⁺ cell counts without long-term clinical benefit could be misleading, as! we learned from the Concorde Study. [2]

Finally, I do not understand the reference of Kinloch-de-Loes et al to Rose's "prevention paradox," which does not let us understand whether the improvement in 1 surrogate marker reflects the real reason for a clinical benefit. Their example, an overall reduction of diastolic blood pressure of 5 mm Hg, appears to be a surrogate marker for what may reflect a more meaningful factor that has not been considered in the analyses. What does an average say about individual indexes? In the case of HIV, an increase in the CD4⁺ cell count at the expense of quality of life with no long-term clinical benefits does not have much merit. One major concern is that this induced increase in CD4⁺ cells could recruit and therefore deplete the pool of CD4⁺ cells within lymphoid tissues. Thus, a momentary improvement in CD4⁺ cell count and clinical state may take place, but not the long-term clinical benefit that might be achieved by allowing or inducing ! the immune system to respond to the infection.

Jay A. Levy, MD

University of California School of Medicine, San Francisco

REFERENCES
1. Levy JA, Mackewicz CE, Barker E. Controlling HIV pathogenesis: the role of noncytotoxic anti-HIV activity of CD8⁺ cells. Immunol Today. 1996;17:217-224. [Medline Link] [Context Link]

2. Seligmann M, Warell DA, Aboulker JP, et al. Concorde: MRT/ANRS randomized double-blind controlled trial of immediate and deferred zidovudine in symptom-free HIV infection. Lancet. 1994;343:871-881. [Context Link]

Acquired Immunodeficiency Syndrome; Biological Markers; CD4 Lymphocyte Count; CONTROVERSIES (Fontanarosa PB, ed); HIV Infections; Surrogate Markers; Viral Burden; Zidovudine