Drugs such as cyclosporin A and tacrolimus bind to cellular immunophilins and as a complex inhibit calcineurin (see Fig. 8.w1). This blockade dampens T cell signaling mediated by NFAT translocation and, in turn, decrease IL-2 and IFNγ production, impairing both T cell activation and proliferation. The drug sirolimus also binds an immunophilin; however, this interaction results in inhibition of the response to, rather than the production of IL-2, again blocking both proliferation and activation in some lymphocyte subsets.

Interference with cellular proliferation is another mechanism of immune suppression. Drugs such as azathioprine, or the more lymphocyte-specific inhibitor, mycophenolate mofetil prevent B and T cell proliferation by affecting DNA synthesis.

Glucocorticosteroids and their functional analogs are potent anti-inflammatory drugs with effects on all branches of the immune response. In light of their widespread use, we have included a more detailed discussion of this class of drugs below.

Glucocorticoids are powerful immune modulators

Among the pharmacological agents that dampen immune responses, the glucocorticoids have the broadest application. Glucocorticoids have pleiotropic effects that vary with both dose and duration of use; however, they are perhaps best known for their potent anti-inflammatory effect. They have been the front-line drugs for decades in the treatment of a variety of inflammatory and allergic conditions, and continue to be a major component of immunosuppressive regimens following organ transplantation.

Patients can receive glucocorticoids:

• systemically, for example, during the early period immediately post-organ transplant (Chapter 21);

• locally, for example as an inhalant for treatment of asthma (Chapter 23);

• topically, for example as for treatment of poison ivy-induced contact hypersensitivity (Chapter 23).

Glucocorticoids are naturally occurring steroids produced by the adrenal cortex. In response to chronic stress or to inflammatory cytokines, a cascade of hormone signals originating in the hypothalamus drives adrenal production of the immunomodulatory steroid, cortisol (Fig. 17.2 and see Fig. 11.17). Cortisol and its analogs are small steroid hormones that readily cross the cellular membrane and bind cytosolic glucocorticoid receptors. Activated glucocorticoid receptors enter the nucleus and can either directly bind DNA to affect gene transcription or regulate expression by disrupting other transcription factor complexes such as NFκB and AP-1 (Chapter 8).

Fig. 17.2 The hypothalamus-pituitary-adrenal axis (HPA axis)

The hypothalamus-pituitary-adrenal axis (HPA axis). Neuroendocrine signaling plays a key role in integrating feedback from the immune system to maintain homeostasis. Inflammatory cytokines trigger the hypothalamus to release corticotropin-releasing hormone (CRH). Increased CRH levels trigger pituitary production of adrenocorticotropic hormone (ACTH) that in turn signals the adrenal glands to increase glucocorticoid production. Glucocorticoids bound to their receptor can interfere with NFκB and AP-1 transcription factor complexes from binding to promoter regions. Additionally, they can interfere with transcription by directly binding to inhibitory glucocorticoid responsive elements (GRE). Under normal conditions, this signaling pathway serves as part of the negative feedback loop shutting down immune responses after infection is controlled.

The timeframe of the downstream effects of this transcriptional regulation ranges from several hours to several days, depending on the rate of protein turnover and de novo protein expression that can vary among different pathways effecting changes in the cellular response. More rapid immunosuppressive effects of glucocorticoids can also occur. In T cells, for example, treatment with steroid analogues impedes activity of the tyrosine kinases, Fyn and Lck.

Q. How would suppression of Fyn and Lck affect T cell responses?

A. During TCR signaling, both Lck and Fyn are important in activation of both phospholipase C and the MAP kinase cascades (see Fig. 8.w1). Thus, prevention of tyrosine kinase activity would suppress T cell activation in response to antigens.

Functional effects of steroid treatment

Glucocorticoids have significant effects on both the innate and adaptive branches of the immune response. There is profound suppression of inflammatory cytokine secretion (IL-1β, IL-6, IL-8, TNFα, IL-12) and chemokine expression. Both contribute to decreased recruitment of neutrophils and macrophages to sites of injury or infection. Glucocorticoids interfere with prostaglandin synthesis, COX2 production, and mast cell degranulation as well. Interestingly, while glucocorticoids enhance both phagocytosis of opsonized antigens and uptake via scavenger receptors, they reduce dendritic cell activation (upregulation of MHC class II and costimulatory B7 molecules).

Within the adaptive branch of immunity, profound downregulation of the inflammatory cytokines and the response of T cells to these cytokines preferentially shift the adaptive immune profile from TH1 toward a TH2-type (Chapter 11). In particular, glucocorticoids suppress both DC production of IL-12 and T cell expression of the IL-12 receptor. In contrast, the effect of corticosteroids on B cell responses is less profound. Thus, overall, humoral immune responses dominate cell-mediated responses during glucocorticosteroid treatment.

Q. After prolonged high steroid dosage there is a modest decrease in all immunoglobulin isotypes. Why should this be?

A. Lack of CD4⁺ T cell help for B cells will result in a general reduction in the numbers of mature B cells that develop.

While glucocorticoids are an invaluable tool, particularly for controlling inflammatory processes, they are often used for only short periods due to the risk of potent side effects. In administering any of the immunosuppressant drugs, physicians must weigh the therapeutic benefits against the risks of broad immunosuppression. Thus, the pharmacologic challenge that remains is to develop drugs that target only those immune responses involved in the disease process or organ rejection while leaving intact as much of the overall immune system functional as possible. This remains a somewhat elusive goal due to the highly integrated nature of the immune system wherein disturbance of one branch affects the function of the others.

Other causes of secondary immunodeficiencies

There are several other clinical conditions that lead to immune suppression and increased susceptibility to infection. Many chemotherapy regimens, as well as irradiation treatment for cancer, target rapidly dividing cells and cause loss of bone marrow precursor cells. Similarly, cancer metastases to the bone and leukemia involving bone marrow may decrease bone marrow output, or lead to generation of immature or atypical leukocyte populations. Major surgery and/or trauma, as well as chronic stressors or debility and advanced age all correlate with diminished immune function, in part due to the upregulation of endogenous glucocorticoids. Last, viral infections can cause loss of immune function.

Human immunodeficieny virus causes AIDS

Infection with human immunodeficiency virus (HIV) is second only to malnutrition in causing immune deficiency and is a significant cause of morbidity and mortality worldwide. HIV is a retrovirus whose primary cellular targets upon infection are CD4⁺ T cells, DCs, and macrophages. Untreated, HIV leads to depletion of the immune system or acquired immunodeficiency syndrome (AIDS), leaving the host susceptible to fatal opportunistic infections. Disease caused by normally non-pathogenic infections, such as Pneumocystis jirovecii pneumonia, cytomegalovirus retinitis, and cryptococcal meningitis occur, as do cancers such as Kaposi’s sarcoma and non-Hodgkin’s lymphoma.

Present primarily in blood, semen, vaginal secretions, and breast milk of infected individuals, HIV is primarily transmitted via unprotected sex, contaminated needles/blood products, or vertically from mother-to-child during the perinatal period. Globally, more than 30 million people are living with the virus, with 2–3 million newly infected and an estimated 1.6–2.1 million deaths each year (WHO estimates, 2009). Over 25 million people have died from AIDS since the descriptions of the first cases in 1981.

There are two main variants, HIV-1 and HIV-2:

• HIV-2 is endemic in West Africa and appears to be less pathogenic;

• HIV-1 has several subtypes (or clades), which are designated by the letters A through K, and the prevalence of the different clades varies by geographical region – over 90% of people infected with HIV-1 live in developing countries and spread is 80% by the sexual route.

HIV life cycle

HIV is a single-stranded RNA lentivirus. Each enveloped virion contains two copies of the 10-kilobase genome, each encoding nine genes flanked at each end by a long-terminal repeat sequence (LTR). The LTR are essential for integration of viral DNA into the host chromosome and also provide binding sites for initiating replication.

The HIV genome contains gag (core proteins), pol (reverse transcriptase, protease, and integrase enzymes), and env (envelope protein) genes (Fig. 17.3). In addition to these three main gene products, the virus encodes six regulatory and accessory proteins (Tat, Rev, Vpr, Vpu, Vif, and Nef). Alternatively spliced transcripts with overlapping open reading frames allow the coordinated expression of these proteins from the compact HIV genome.

Fig. 17.3 Human immunodeficiency virus and its life cycle

After attachment to CD4 and a chemokine co-receptor (usually CCR5), the virus membrane fuses with the cellular membrane to allow entry into the cell. Following uncoating, reverse transcription of viral RNA results in the production of double-stranded DNA (dsDNA). This is inserted into the host genome as the HIV provirus, by a virally coded integrase enzyme, which is a target for new antiviral medications currently in development. Cell activation leads to transcription and the production of viral mRNAs. Structural proteins are produced and assembled. Free HIV viruses are produced by viral budding from the host cell, after which further internal assembly occurs with the cleavage of a large precursor core protein into the small core protein components by a virally coded protease enzyme, producing mature virus particles, which are released and can go on to infect additional cells bearing CD4 and the chemokine co-receptor.

HIV targets CD4 T cells and mononuclear phagocytes

HIV primarily targets CD4⁺ T cells, CD4⁺ macrophages, and some dendritic cells (DC). Env encodes a 160 kDa precursor of the envelope glycoproteins and proteolytic cleavage generates gp120 and gp41. Infection of the target cells requires initial attachment of gp120 to the major receptor, CD4. Viral entry also requires additional binding through co-receptors, most commonly the chemokine receptors CCR5 and CXCR4. Once bound to the cells, interaction via gp41 mediates cell-virus fusion. Upon entry of the HIV capsid, reverse transcription of the RNA genome generates cDNA that subsequently integrates into the host DNA. These latter two steps occur primarily within activated cells.

Differences in the envelope glycoprotein sequence determine whether the virus can utilize the chemokine receptors CCR5 or CXCR4, or both. HIV variants that utilize CCR5 or CXCR4 are referred to as R5 or X4 viruses, respectively. R5 viruses, therefore, can infect memory CD4 T cells and mononuclear phagocytes expressing CCR5. R5 tropism predominates early in HIV infection, while X4, R5, and R5/X4 dual tropic variants may be found in patients during later stages. Individuals homozygous for a 32 base pair deletion in the CCR5 allele (CCR5Δ32) are highly resistant to HIV infection by R5 viruses, but remain susceptible to infection with R4 virus. Other receptors for HIV include DC-SIGN on dendritic cells and galactosyl ceramide (GalC), a major binding site for infection within the brain, gut, and vagina.

Acute symptoms occur 2–4 weeks post infection

Disease is generally the result of infection by a single virion. With infection via a mucosal surface (~80% of infections), the initial target cell is likely a tissue DC or macrophage that can then transport virus to the draining lymphoid tissue. Migration of virus to lymphoid tissue allows productive infection of an activated CD4 T cell or macrophage, inducing cytokine release and further recruitment of activated cells. Recent studies have shown the gut-associated lymphoid tissues (GALT), a site rich with memory CCR5⁺CD4⁺ T cells, is an early locus of infection. HIV infection of GALT CD4 T cells quickly leads to their depletion. Approximately 2–4 weeks post-infection, the patient experiences flu-like symptoms, with fever, swollen lymph nodes, malaise, occasionally rashes, headache, and nausea. During acute viremia plasma virus levels can reach up to 10 million copies/mL, with wide dissemination of the virus throughout tissues, often also accompanied by acute depletion of CD4 T cells in the blood and, importantly, establishment of HIV reservoirs. It is estimated that more than half of all memory CD4⁺ T cells are lost during this acute phase.

Viral latency is associated with chronic infection

Without anti-retroviral treatment, HIV levels peak 3–4 weeks post-infection, then gradually drop and plateau (Fig. 17.4). This reflects a combination of the decrease in readily available activated targets and, perhaps more importantly, control by the innate and adaptive immune responses. There is usually a moderate rebound in circulating CD4 T cell numbers at this point, though recent studies indicate GALT CD4 T cell populations do not recover. Simultaneously, the virus establishes stable viral reservoirs. This first consists of cells supporting low-level viral replication in lymphoid and other tissues, likely with efficient cell-to-cell propagation of the virus. The second reservoir is within CD4 T cells in which the HIV genome is integrated as a provirus, yet remains latent as a silent infection without viral protein transcription. Subsequent T cell activation can then stimulate virus production.

Fig. 17.4 Natural history of HIV virus

A typical course of HIV infection.

The plasma viral load after acute infection recedes (viral set point) can be an indicator for disease progression. The average period of stable infection is 10 years with mean plasma viral load of ~30 000 copies/mL. Rapid progressors can experience increases in viremia, CD4⁺ cell depletion, and onset of opportunistic disease in as soon as 6 months. Plasma viral RNA levels of >100 000 copies/mL 6 months after infection are associated with a 10-fold greater risk of progression to AIDS with 5 years as compared to patients with HIV plasma load of <100 000 copies/mL. In contrast, long-term non-progressors (LTNP) generally have a lower viral set-point and may remain asymptomatic for more than 25 years with stable CD4 T cell numbers. During this chronic infection period active immune responses keep viral levels in check. Most infected individuals are asymptomatic during this period but can still transmit infectious virus to others.

HIV infection induces strong immune responses

Both humoral and cellular immune responses develop after infection. Detectable HIV-specific antibodies are evident in the first few weeks of infection. Production of neutralizing antibodies (which prevent virus-cell fusion and correlate with protection) does not occur until at least 12 weeks after infection. Therefore, although antibodies are sufficient to drive evolution of viral epitopes, they are insufficient to prevent disease progression.

The drop in HIV viremia to the set point level coincides with the expansion of HIV-specific CD8⁺ T cells. In CD8 T cell-depleted animal models of infection, viral containment fails. Thus, as with many viruses, CD8 T cell responses are important components in the control of infection. These responses are usually against peptides derived from multiple viral proteins and comprise a considerable portion (up to 25%) of total CD8 T cells; however, neither the breadth nor the magnitude of CTL responses correlates with control of viremia. In contrast, the genetic background, specifically the HLA haplotype of the host, is important in determining viral control. Both the HLA-B*27 and HLA-B*57 class I alleles are associated with low viral set point and long-term asymptomatic control of infection. These patients are often LTNP and may have stable infection for >10 years.

HIV can evade the immune response

HIV is a rapidly mutating virus. Both error-prone reverse transcription and high recombination frequencies during reverse transcription combine with an extremely high virus production rate to generate genetic diversity. Particularly during acute infection, immune clearance via antibody and CTL recognition favor survival of virions with envelope or peptide sequence changes in regions targeted by host-protective epitopes. Additionally, the HIV protein Nef limits CTL detection of infected cells via selective downregulation of both HLA-A and HLA-B expression, reducing the surface display of viral epitopes. Viral escape, coupled with the existence of latent viral reservoirs undetectable by HIV-specific immune responses, prevents the immune system from eliminating HIV infected cells. Exacerbating this issue, the immune responses become progressively weaker over time.

Q. How does Nef mediated downregulation of only HLA-A and B, but not HLA-C and E promote viral immune evasion?

A. HLA-C and HLA-E are inhibitory signals for NK cells. Therefore Nef can decrease CTL recognition of infected cells without increasing susceptibility to NK cells (Chapter 10).

Immune dysfunction results from the direct effects of HIV and impairment of CD4 T cells

The chronic phase of HIV infection is marked by persistent generalized immune activation. Infected persons present with B cell polyclonal activation and hypergammaglobulinemia. Attachment of gp120 to mannose binding lectin and to subsets of Ig⁺ B cells contributes to this activation. Inflammatory cytokines such as IFNα, IFNγ, IL-18, IL-15, and TNFα are elevated during acute infection. Finally, increased translocation of bacteria through the gut barrier, due to extensive HIV infection within the GALT and lamina propria, increases circulating LPS levels, activating many immune effectors via Toll-like receptors. In patients with high viral load, HIV-specific CTL typically contain low levels of intracellular perforin (see Figs 10.10 and 10.12) and have a poor proliferative capacity. During this chronic phase, CD8 T cells often express PD-1, a receptor associated with programmed cell death (Chapter 8). The points above are all consistent with persistent activation and eventual exhaustion of the supply of anti-HIV CD8 T cells that worsens over time.

Of equal if not greater impact on overall immune dysfunction from HIV infection, however, is the loss of the CD4 T cells. Though circulating CD4 T cell counts often rebound (at least quantitatively if not qualitatively) to pre-infection levels after the acute phase, they do not recover within the mucosal associated lymphoid tissues. Furthermore, in untreated patients, CD4 T cell levels eventually decline though they may remain above critical levels for a period of 6 months to 10 years.

CD4 T cell loss is due to direct killing by HIV and activation-induced apoptosis. Additionally, infected CD4 T cells present both gp120 and HIV peptide:HLA complexes on the cell surface leaving them subject to anti-HIV B and CTL clearance. Ongoing loss of CD4 T cell help ultimately contributes to the collapse of the CD8 T cell response as demonstrated experimentally by the ability to rescue anti-HIV CD8 T cell functionality by replacing the lost CD4 T cell population. However, progressive immunodeficiency is a hallmark of HIV and the eventual diminished CD4 T cell levels correlate tightly with the subsequent progression to advanced disease and death.

AIDS is the final stage of HIV infection and disease

Progression to this clinical stage includes a CD4 T cell count of <200/μL. As blood CD4 T cell counts gradually decline during the chronic phase, patients become susceptible to opportunistic infection and malignancies (see Fig. 17.4). Below 500 CD4 cells/μL, less severe conditions such as oral candidiasis, recurrent herpes virus outbreaks (e.g. shingles from varicella zoster virus and anogenital herpes from herpes simplex virus), and pneumococcal infections occur. CD4 T cell levels below 200/μL are associated with increased risk of life-threatening infections and malignancies including Pneumocystis jirovecii pneumonia and Kaposi’s sarcoma, respectively. With CD4 levels below 50/μL, patients become vulnerable to additional systemic infection with organisms such as Mycobacterium avium complex. The three main organ systems affected are the respiratory system, gastrointestinal tract, and central nervous system.

Pneumonia Pneumocystis jirovecii (previously P. carinii) is the most common opportunistic respiratory infection (Fig. 17.5), but pulmonary bacterial infections, including Mycobacterium tuberculosis, also occur. Protozoa (cryptosporidia and microsporidia) are the most common pathogens isolated in patients with diarrhoea and weight loss (see Fig. 17.5). Enteric bacteria such as Salmonella and Campylobacter spp. may also afflict AIDS patients.

Fig. 17.5 Common features of late-stage HIV infection

(1) Multiple Kaposi’s sarcoma lesions on the chest and abdomen. (2) Chest radiograph of a patient with Pneumocystis jirovecii pneumonia, showing bilateral interstitial shadowing. (3) Small bowel biopsy from a patient with diarrhea caused by cryptosporidia, showing intermediate forms of cryptosporidia (small pink dots) on the surface of the mucosa. (4) Computed tomography scan of the head of a patient with cerebral toxoplasmosis. The patient presented with a history of fits and weakness of the left arm and leg. Injection of contrast revealed a ring-enhancing lesion in the right hemisphere (arrow), with surrounding edema (dark area).

Neurological complications in AIDS are due to direct effects of HIV infection, opportunistic infections, or lymphoma. AIDS-related dementia once affected between 10–40% of patients with other manifestations of AIDS, but with more effective antiviral treatment has become less common. Neurological involvement can be due to a number of pathogens. Cryptococcus neoformans is a fungus and is the most common cause of AIDS-related meningitis. Toxoplasmosis, a protozoal infection, causes cysts in the brain and neurological deficit (see Fig. 17.5). Cytomegalovirus reactivation may cause inflammation of the retina, brain, and spinal cord and its nerve roots, and a polyomavirus (JC virus), which infects oligodendrocytes in the brain, produces a rapidly fatal demyelinating disease – progressive multifocal leukoencephalopathy.

Kaposi’s sarcoma (KS), caused by infection with KS-associated herpes virus (KSHV), is the most common AIDS-associated malignancy (Fig. 17.5). KSHV infections, similar to CMV herpesvirus infections, are often asymptomatic in individuals with competent T cell immunity. With HIV co-infection, however, KSHV titers increase and KS emerges with multifocal lesions (see Fig. 17.5) of mixed cellularity often resulting in widespread involvement of skin, mucous membranes, viscera (gut and lungs) and lymph nodes. KSHV infection can also lead to development of B cell lymphomas, affecting the brain, gut, bone marrow, lymph nodes, spleen and body cavities such as the pericardial and pleural spaces. KSHV also causes two B cell lymphoproliferative diseases, multicentric Castleman’s disease and primary effusion lymphoma.

Most of the opportunistic infections as well as malignancies, such as KS and Epstein–Barr virus-associated non-Hodgkin’s lymphomas, are due to the inability of the immune response to suppress baseline levels of reactivation of latent organisms in the host and in some cases, to ubiquitous organisms to which we are continually exposed. They are difficult to diagnose and treatment often suppresses rather than eradicates them. Relapses are common and continuous suppressive or maintenance treatment is necessary.

An effective vaccine remains an elusive goal

Currently, treatment for HIV focuses on anti-viral drug cocktails that significantly reduce the patient’s viral load. Due to the rapid rate of mutation in the viral genome during replication, single drug therapy nearly always leads to rapid drug resistance. However, by providing the patient with a cocktail of anti-viral drugs, each targeting a different aspect of the viral life cycle, it is possible to prolong the period of time before plasma viral load increases and T cell counts drop. Anti-retroviral therapies, unfortunately, are not a cure, nor can they prevent transmission of the virus.

Despite increasing characterization of adaptive immune responses, the correlates of protection remain to be fully defined, and this has left the field with mostly empiric approaches. Ideally, investigators will develop a vaccine that provides sufficient protection to prevent viral transmission. Encouraging early reports of persons repeatedly exposed to HIV who never became infected suggested that adaptive immune responses, particularly HIV-specific CD8 T cell responses, might be responsible for apparent protection, but this remains somewhat controversial. Such a vaccine will almost certainly also require the induction of broadly neutralizing antibody responses; something that candidate vaccines have yet to achieve. To date there have been clinical trials of numerous candidate HIV vaccines, but most would agree that an effective vaccine remains an elusive goal.

Q. Why is it difficult to identify suitable antigens that could be used in a neutralizing vaccine?

A. The very rapid rate of HIV mutation and the high rate of virus production mean that the virus can readily mutate to evade a specific immune response, and yet may develop variants that retain infectivity.

Many in the field believe that a vaccine that protects from disease progression, while not necessarily protecting from the initial infection, is a more realistic short-term goal. Toward this goal, investigators have focused on approaches that induce cellular immune responses, particularly CD8 T cells responses. However, there is increasing evidence suggesting that optimal efficacy will also need to elicit robust CD4 T cell responses to HIV.

As no cure or vaccine is currently available, our main weapon is prevention through health education and control of infection.

Critical thinking: Secondary immunodeficiency (see p. 438 for explanations)

A 52-year-old record producer developed a severe cough with increasing shortness of breath. He also had a fever, chest pain, and malaise. For the week before presentation he complained of pain on swallowing. His past medical history included gonorrhea and genital herpes within the previous 3 years. Over the previous 2 months he had suffered from persistent diarrhea and lost 9 kg in weight from a baseline of 68 kg. He lived with his girlfriend with whom he had been having unprotected intercourse for several years. There was no history of intravenous drug abuse.

On examination he was underweight and had enlarged lymph nodes in the neck, axillae, and groin. Plaques of Candida albicans were visible in his throat. There were abnormal breath sounds in his lungs. The results of his blood tests are shown in Table 1.

Table 1 Results of investigations on presentation

Investigation	Result (normal range)
hemoglobin (g/dL)	12.8 (13.5–18.0)
platelet count (× 10⁹/L)	128 (150–400)
white cell count (× 10⁹/L)	6.2 (4.0–11.0)
neutrophils (× 10⁹/L)	5.4 (2.0–7.5)
eosinophils (× 10⁹/L)	0.24 (0.4–0.44)
total lymphocytes (× 10⁹/L)	0.75 (1.6–3.5)
T lymphocytes CD4⁺ (× 10⁹/L) CD8⁺ (× 10⁹/L)	0.12 (0.7–1.1) 0.42 (0.5–0.9)
B lymphocytes (× 10⁹/L)	0.11 (0.2–0.5)
arterial blood gases PaO₂ (kPa) PaCO₂ (kPa) pH HCO₃ base excess	7.8 (>10.6) 5.52 (4.7–6.0) 7.39 (7.35–7.45) 25.6 –0.9
ECG	normal
chest radiography	bilateral diffuse interstitial shadowing
bronchoscopy with bronchoalveolar lavage	positive for Pneumocystis jirovecii

Because of his sexual history, the patient was counseled about having a human immunodeficiency virus (HIV) test, and consented. An enzyme-linked immunosorbent assay (ELISA, see Method box 3.2, Fig. 2) was positive for anti-HIV antibodies and a polymerase chain reaction (PCR) demonstrated HIV-1 RNA in the plasma.

Examination of an induced sputum specimen revealed Pneumocystis jirovecii, which together with the positive HIV ELISA is an AIDS-defining illness. Thus a clear diagnosis of acquired immune deficiency syndrome (AIDS) was made and the patient’s P. jirovecii pneumonia was treated with oxygen by mask and parenteral co-trimoxazole. He was discharged from hospital taking oral co-trimoxazole.

Within 3 months he was seen again in accident and emergency with blurred vision and ‘flashing lights’ in his eyes. He was shown to have an infection of his retina with cytomegalovirus and was treated with injections of ganciclovir. The CD4 count at this time was 0.04 × 10⁹/L. While receiving this treatment the patient became increasingly unwell and semiconscious. Investigations at this time are shown Table 2.

Table 2 Results of investigations 3 months after presentation

Investigation	Result (normal range)
hemoglobin (g/dL)	10.4 (13.5–18.0)
platelet count (× 10⁹/L)	104 (150–400)
white cell count (× 10⁹/L)	4.1 (4.0–11.0)
neutrophils (× 10⁹/L)	4.2 (2.0–7.5)
eosinophils (× 10⁹/L)	0.24 (0.4–0.44)
total lymphocytes (× 10⁹/L)	0.62 (1.6–3.5)
T lymphocytes CD4⁺ (× 10⁹/L) CD8⁺ (× 10⁹/L)	0.03 (0.7–1.1) 0.40 (0.5–0.9)
B lymphocytes (× 10⁹/L)	0.09 (0.2–0.5)
chest radiography	minimal areas of diffuse shadowing
blood culture	negative
blood glucose (mmol/L)	7.6 (<10.0)
CSF from lumbar puncture Appearance	turbid
white cells (polymorphs/mm³)	2500
protein (g/L)	4.2 (0.15–0.45)
glucose (mmol/L)	4.5 (> 60% blood glucose)
Indian ink stain	positive for cryptococcus