# 120 - 227 Pneumocystis Infections ### 227 Pneumocystis Infections Treatment and Prognosis  Scedosporium and Lomentospora species are intrinsically resistant to AmB, echinocandins, and some azoles. Voriconazole has been the agent of choice for S. apiospermum, and posaconazole has been increasingly used for this infection (Table 226-1). L. prolificans is resistant in vitro to almost every commercially avail­ able antifungal agent; the addition of agents such as terbinafine to a voriconazole regimen has been attempted because in vitro data sug­ gest possible synergy between these two agents against some strains of L. prolificans. Mortality rates for invasive S. apiospermum infection are ~50%, but those for invasive L. prolificans infection remain as high as 85–100%. The novel antifungal agents fosmanogepix and olorofim have shown in vitro activity against both S. apiospermum and L. prolificans and have demonstrated efficacy in preclinical mouse models of these infections. Ongoing clinical trials for these agents will help determine their clinical efficacy in infected patients. YEAST INFECTIONS In addition to the far more common human pathogenic yeasts Candida and Cryptococcus, which are discussed in Chaps. 221 and 222 the yeast-like fungus Trichosporon has emerged as a significant opportu­ nistic pathogen among immunocompromised patients. ■ ■TRICHOSPORONOSIS Etiologic Agent, Epidemiology, and Pathogenesis  The genus Trichosporon encompasses many species, some of which cause localized infection of hair and nails; white piedra is a superficial infection of the hair and scalp caused by T. ovoides that is characterized by the appear­ ance of white nodules along the hair shaft. The most common species responsible for invasive infection is Trichosporon asahii, although other species also can cause disseminated disease. Trichosporon species grow as yeast-like colonies in vitro; in vivo, however, hyphae, pseudohyphae, and arthroconidia can be seen in addition to yeast forms. These yeasts are commonly found in soil, sewage, and water and in rare instances can colonize the human skin and the gastrointestinal and respiratory tracts. Most infections follow inhalation or entry via central venous catheters. The presence of foreign material is a risk factor for trichospo­ ronosis, as both catheter-associated infections and prosthetic cardiac valve infections have been described. Systemic infection occurs almost exclusively in immunocompromised hosts, especially those who have hematologic malignancies, are neutropenic, have received a solid organ or hematopoietic stem cell transplantation, or are receiving cortico­ steroids. Severe neutropenia is a major risk factor for disseminated disease. Myeloid phagocytes and their oxidative burst are critical for protection against Trichosporon, as illustrated by the development of invasive trichosporonosis in patients with CARD9 deficiency or CGD. Clinical Manifestations  Disseminated trichosporonosis resem­ bles invasive candidiasis, and fungemia is often the initial manifesta­ tion of infection typically presenting with fever. Pneumonia, skin lesions, and sepsis are common, although many patients can present without clinically apparent lung infection. The skin lesions in dissemi­ nated disease begin as papules or nodules surrounded by erythema and progress to central necrosis. Renal involvement is common in dissemi­ nated disease and may cause funguria and hematuria. A rare chronic form of infection mimics hepatosplenic candidiasis, now known as chronic disseminated candidiasis. PART 5 Infectious Diseases Diagnosis  The diagnosis of systemic Trichosporon infection is established by growth of the organism from infected tissues or from blood. Histopathologic examination of a skin lesion showing a mixture of yeast forms, arthroconidia, pseudohyphae, and hyphae can lead to an early presumptive diagnosis of trichosporonosis. The presence of arthroconidia in particular can help differentiate trichosporonosis from candidemia. The serum cryptococcal antigen latex agglutination test may be positive in patients with disseminated trichosporonosis because T. asahii and Cryptococcus neoformans share polysaccharide antigens. Treatment and Prognosis  Rates of response to AmB have been disappointing, and many Trichosporon isolates are resistant to AmB in vitro. Voriconazole has been the antifungal agent of choice (Table 226-1). The mortality rates for disseminated Trichosporon infection have been as high as 70% but are decreasing with the use of voricon­ azole; however, neutrophil recovery is vital in overcoming this infec­ tion, and mortality rates with persistent neutropenia remain very high despite antifungal therapy. ■ ■FURTHER READING De Almeida Junior JN, Hennequin C: Invasive Trichosporon infec­ tions: A systematic review on a re-emerging fungal pathogen. Front Microbiol 7:1629, 2016. Hospenthal D: Uncommon fungi and related species in Mandell, Douglas, and Bennett’s Principles and Practice of Infectious Diseases, 10th ed, Blaser MJ et al (eds). Philadelphia, Elsevier, 2025. Neoh CF et al: Scedosporiosis and lomentosporiosis: Modern perspec­ tives on these difficult-to-treat rare mold infections. Clin Microbiol Rev 37:e0000423, 2024. Nucci M et al: Fusariosis. Semin Respir Crit Care Med 36:706, 2015. Revankar SG et al: A Mycoses Study Group international prospec­ tive study of phaeohyphomycosis: An analysis of 99 proven/probable cases. Open Forum Infect Dis 4:ofx200, 2017. Alison Morris, Henry Masur Pneumocystis Infections ■ ■DEFINITION AND DESCRIPTION Pneumocystis is an opportunistic pathogen that is an important cause of pneumonia in immunocompromised hosts, particularly those with HIV infection (Chap. 208), organ transplants, or hema­ tologic malignancies and those receiving high-dose glucocorticoids or certain immunosuppressive monoclonal antibodies. Pneumocystis was discovered in rodents in 1909 and was initially believed to be a protozoan. Because Pneumocystis cannot be cultured, understanding its biology has been limited, but molecular techniques have demon­ strated that the organism is actually a fungus. Formerly known as Pneumocystis carinii, the species specific to humans has been renamed Pneumocystis jirovecii; the species that is specific to rats is Pneumocys­ tis carinii; there are multiple variants of Pneumocystis that are specific to other animals. ■ ■EPIDEMIOLOGY Pneumocystis jirovecii pneumonia (PCP) came to medical atten­ tion in the early 1950s when pathologists in Czechoslovakia rec­ ognized Pneumocystis in the alveolar exudates of infants involved in nursery outbreaks of interstitial pneumonia, outbreaks that had been described in Europe since the 1920s. Among adults, PCP was rarely recognized until the populations of immunosuppressed adults increased due to the development of immunosuppressive thera­ pies for solid-organ transplantation, bone marrow transplantation, cancer, and autoimmune disorders, and until the development of better pulmonary diagnostic techniques such as bronchoscopy. In 1981, PCP was first reported in men who had sex with men and in intravenous (IV) drug users who had no obvious cause of immuno­ suppression. These cases were subsequently recognized as the first cases of what came to be known as the acquired immunodeficiency syndrome (AIDS) (Chap. 208). The incidence of PCP increased dramatically as the AIDS epidemic grew: without PCP chemoprophylaxis or antiretroviral therapy (ART), 80–90% of patients with HIV/AIDS in North America and Western Europe ultimately developed one or more episodes of PCP. While its incidence declined with the introduction of anti-Pneumocystis prophylaxis and durably effective ART, PCP has continued to be an important cause of AIDS-associated morbidity in the United States and Western Europe, particularly in individuals who do not know they are infected with HIV until they are profoundly immunosuppressed and in people living with HIV (PLWH) with CD4+ T lymphocyte counts of <200/μL who are not receiving ART or PCP prophylaxis. PCP also develops in HIV-uninfected patients who are immuno­ compromised secondary to congenital immunodeficiencies, hemato­ logic or malignant neoplasms, stem cell or solid-organ transplantation, and treatment with immunosuppressive medications. The incidence of PCP depends on the degree and duration of immunosuppression. PCP is increasingly reported among individuals receiving tumor necro­ sis factor α inhibitors and certain (but not all) immunosuppressive monoclonal antibodies for autoimmune, rheumatologic, or neoplastic diseases. In many health care systems, PCP occurs more often due to non-HIV-related immunosuppression than due to HIV infection. While clinical disease due to Pneumocystis in immunocompetent hosts has not been clearly documented, studies have shown that Pneu­ mocystis organisms can cause subclinical infection among children and adults who are not immunocompromised and can be associated with pulmonary pathology. The relevance of these organisms to acute or chronic clinical syndromes, such as chronic obstructive pulmonary disease (COPD), in immunocompetent patients is being investigated. In some developing countries, the incidence of PCP among PLWH has been reported to be lower than that in more industrialized coun­ tries. This lower incidence may be due to competing mortality from infectious diseases such as tuberculosis and bacterial pneumonia, which typically occur before patients become immunosuppressed enough to develop PCP. Geographic variations in Pneumocystis expo­ sure and underdiagnosis attributable to lack of diagnostic resources also may explain the apparent lower frequency of PCP in some countries. ■ ■PATHOGENESIS AND PATHOLOGY Life Cycle and Transmission  The life cycle of Pneumocystis likely involves both sexual and asexual reproduction. The organism exists as a trophic form, a cyst, and a precyst. Studies in rodents show that immunocompetent animals can serve as reservoirs for respiratory transmission of P. carinii (the infecting species in rats) to immunocom­ petent and immunosuppressed rats. Human Pneumocystis is thought to be transmitted by a respiratory route as well. P. jirovecii, like all pneu­ mocystis species, is host-specific. Thus, humans are not infected, for example, by P. carinii (rodents) or P. oryctolagi (rabbits), but are only infected by P. jirovecii, which is from other humans. Serologic and molecular studies have demonstrated that most humans are exposed to P. jirovecii and infected early in life. It was historically thought that Pneumocystis pneumonia usually developed from reactivation of latent infection. However, molecular evidence makes it clear that children and adults can develop PCP from primary infection or reinfection. The source of infection is thought to be either healthy or immunosuppressed individuals who themselves experi­ enced recent infection or reinfection, or immunosuppressed persons with clinical PCP. Nosocomial outbreaks among immunosuppressed persons occur in inpatient and outpatient settings. The utility of drop­ let or airborne isolation for preventing transmission from patients with PCP to other immunosuppressed individuals has been debated; no clear evidence exists, although it seems prudent to isolate patients with active PCP from other immunosuppressed patients using at least droplet precautions. Role of Immunity  Defects in cellular and/or humoral immunity predispose to development of PCP. Such defects may be congenital, or they may be acquired as a result of HIV infection or treatment with certain immunosuppressive drugs such as glucocorticoids, fludarabine, temozolomide, temsirolimus, rituximab, or alemtuzumab. CD4+ T cells are critical in host defense against Pneumocystis. Among PLWH, the incidence of PCP is inversely related to the CD4+ T-cell count: at least 80% of cases occur at counts of <200/μL, and most cases actually develop at counts of <100/μL. HIV viral load is another factor that predisposes patients to PCP. Clinicians must recognize that PCP can occur at CD4+ T-cell counts >200/μL in persons with HIV infection, but such occurrences are uncommon, especially with CD4+ T-cell counts substantially higher than 200/μL. CD4+ T-cell counts are less useful in predicting the risk of PCP in patients who are immunosuppressed for reasons other than HIV infec­ tion. In these populations, a CD4+ T-cell count <200/μL is a sensitive indicator of susceptibility to PCP. However, a CD4+ T-cell count >200/μL in many populations does not imply protection. Lung Pathology  Pneumocystis has a unique tropism for the lung. Organisms are presumably inhaled into the alveolar space after being exhaled by another human. Clinically apparent pneumonia occurs only if an individual is immunocompromised. Pneumocystis prolifer­ ates in the lung, provoking a mononuclear cell response. The alveoli become filled with proteinaceous material, and alveolar damage results in increased alveolar-capillary injury and surfactant abnormalities. Stained lung sections typically show foamy, vacuolated alveolar exudates composed largely of viable and nonviable organisms (Fig. 227-1A). Interstitial edema and fibrosis may develop, and organisms can be seen in the alveolar space with silver or other stains. The organisms can also be seen when tissue is subjected to colorimetric or immunofluorescent staining (Fig. 227-1B–D). ■ ■CLINICAL FEATURES Clinical Presentation  PCP presents as acute or subacute pneu­ monia that may initially be characterized by a vague sense of dyspnea alone, but that subsequently manifests as fever and nonproductive cough with progressive shortness of breath. Patients may ultimately progress to respiratory failure and death. Extrapulmonary manifesta­ tions of PCP are rare, but can include involvement of almost any organ, most notably the lymph nodes, spleen, and liver. CHAPTER 227 Physical Examination, Oxygen Saturation, and Imaging  The physical examination findings in PCP are nonspecific. Patients have decreased oxygen saturation—at rest or with exertion—that, without treatment, progresses to severe hypoxemia. Patients may initially have a normal chest examination and no adventitious sounds, but later develop diffuse rales and signs of consolidation. Pneumocystis Infections Laboratory Findings  The results of routine laboratory tests are nonspecific in PCP. Serum levels of lactate dehydrogenase (LDH) are often elevated as a result of pulmonary damage; however, a normal LDH level does not rule out PCP, nor is an elevated LDH value specific for PCP. The peripheral white blood cell count may be elevated in rela­ tion to the patient’s baseline values, but the increase is usually modest and the baseline may have been below usual normal limits due to HIV infection. Hepatic and renal function are typically normal. Radiographic Findings  Although the initial chest radiograph may be normal when patients have mild symptoms, the classic radiographic appearance of symptomatic PCP consists of diffuse bilateral interstitial infiltrates that are perihilar and symmetric (Fig. 227-2A)—another finding that is not specific for PCP. The interstitial infiltrates can prog­ ress to alveolar filling (Fig. 227-2B). High-resolution chest computed tomography (CT) shows diffuse ground-glass opacities in virtually all patients with PCP, often before a routine chest radiograph becomes abnormal (Fig. 227-2C). A normal chest CT essentially rules out the diagnosis of PCP. Pneumatoceles and pneumothoraces are characteris­ tic chest radiographic findings, especially in patients with HIV infection (Fig. 227-2D). A wide variety of atypical radiographic findings have been described, including asymmetric patterns, upper-lobe infiltrates, mediastinal adenopathy, nodules, cavities, and effusions. ■ ■DIAGNOSIS The optimal sample for a specific microbiologic diagnostic examina­ tion depends on how ill the patient is and what resources are avail­ able. Before the 1990s, diagnoses of PCP were usually established by open lung biopsy; later, transbronchial lung biopsy was employed. Hematoxylin and eosin (H and E) staining of pulmonary tissue A B PART 5 Infectious Diseases C D FIGURE 227-1  Direct microscopy of Pneumocystis pneumonia. A. Transbronchial lung biopsy stained with hematoxylin and eosin shows eosinophilic alveolar filling. B. Methenamine silver–stained bronchoalveolar lavage (BAL) fluid. C. Giemsa-stained BAL fluid. D. Immunofluorescent stain of BAL fluid. demonstrates a foamy alveolar filling and a mononuclear interstitial infiltrate (Fig. 227-1A). This appearance is pathognomonic for PCP even though the organisms cannot be specifically identified with this H and E stain. Since human pneumocystis has never been success­ fully cultured, the diagnosis is typically established in lung tissue or pulmonary secretions by staining of the cyst and or trophozoite—e.g., with methenamine silver (Fig. 227-1B), toluidine blue O, or Giemsa (Fig. 227-1C)—or by staining with a specific immunofluorescent antibody (Fig. 227-1D). Since the 1990s, bronchoalveolar lavage (BAL) has become the most common method for obtaining a sample of respiratory secretions in which to detect Pneumocystis organisms. The demonstration of organ­ isms in BAL fluid is almost 100% sensitive and specific for PCP. The organisms are identified in pulmonary secretions with the specific stains indicated above for lung biopsy. While expectorated sputum or throat swabs have very low sensitivity, an induced sputum sample obtained and interpreted by an experienced provider can be highly sensitive and specific; however, the sensitivity is dependent on both the characteristics of the patient, the quality of the sputum sample, and the experience of the center conducting the test and is widely variable (55–90%). Many laboratories offer polymerase chain reaction (PCR) test­ ing of respiratory specimens for Pneumocystis in preference to direct microscopy of stained respiratory secretions. However, PCR tests are so sensitive that it is difficult to distinguish patients with colonization (i.e., those whose acute lung disease is due to some other process but who have low levels of Pneumocystis DNA in their lungs) from those with acute pneumonia due to Pneumocystis. The PCR cycle number is probably helpful, i.e., the lower the cycle number, the more DNA was detected. However, PCR tests for PCP are not standardized and speci­ men collection can vary considerably resulting in variable quality of the specimens assessed. Therefore, there is no specific PCR threshold that can be deemed to be especially convincing. Such PCR tests on appropriate samples may be more useful for ruling out a diagnosis of PCP if they are negative than for definitively attributing the disease to Pneumocystis if they are positive. There has been considerable interest in serologic tests, such as assays for (1→3)-β-d-glucan, a component of the fungal cell wall. Serum (1→3)-β-d-glucan levels are frequently elevated in patients with PCP. However, serum or BAL (1→3)-β-d-glucan levels are not perfectly sensitive or highly specific for PCP. There are increasing numbers of studies of serum PCR tests for Pneumocystis, but such serum tests are not yet useful for establishing the presence or absence of PCP. ■ ■COURSE AND PROGNOSIS Untreated, PCP is invariably fatal. Patients with HIV infection often have an indolent course that may present early as mild exercise intoler­ ance or chest tightness without fever or cough and a normal or nearly normal posterior–anterior chest radiograph. However, this process progresses over days, weeks, or even a few months to fever, cough, dif­ fuse alveolar infiltrates, and profound hypoxemia. Some patients with HIV infection and most patients with other types of immunosuppres­ sion have more acute disease that progresses over a few days to respira­ tory failure. Rare patients also develop distributive shock due to PCP. A few unusual patients present with extrapulmonary manifestations in the skin or soft tissue, retina, brain, liver, kidney, or spleen. Extrapul­ monary disease is nonspecific in presentation and can be diagnosed only by histology. When there is extrapulmonary clinical disease in a A B D C FIGURE 227-2  Radiographs in Pneumocystis pneumonia. A. Posterior–anterior chest radiograph showing symmetric interstitial infiltrates. B. Posterior–anterior chest radiograph showing symmetric alveolar infiltrates (courtesy of Alison Morris). C. Computed tomography (CT) image demonstrating symmetric interstitial infiltrates and ground-glass opacities. D. CT image showing symmetric interstitial infiltrates, ground-glass opacities, and pneumatoceles. patient with PCP, the priority is to determine what other concurrent infectious or neoplastic process might be present, given the rarity of extrapulmonary pneumocystosis. Factors that influence mortality risk of PCP include the patient’s age and degree of immunosuppression as well as the presence of preexisting lung disease, the need for mechanical ventilation, and the development of a pneumothorax. With advances in supportive criti­ cal care, the prognosis for patients with PCP who require intubation and respiratory support has improved and now depends to a large extent on comorbidities and the prognosis of the underlying disease. Some patients do not respond to therapy for 4–8 days, and thus sup­ portive care for a minimum of 10 days is a reasonable consideration if such support is compatible with the patient’s wishes and the prog­ nosis of comorbidities. In fact, if patients of any level of severity are treated with specific therapy but without corticosteroids, they often deteriorate during the first few days, presumably due to enhanced inflammation induced by dying organisms. Patients whose condi­ tion continues to deteriorate after 3 or 4 days or have not improved after 7–10 days should be reevaluated to determine whether other infectious processes are present (either having been missed on initial evaluation or having developed during treatment), whether initial anti-Pneumocystis treatment has failed, or whether noninfectious processes (e.g., congestive heart failure, pulmonary emboli, pulmo­ nary hypertension, drug toxicity, or a neoplastic process) are causing pulmonary dysfunction. CHAPTER 227 Pneumocystis Infections TREATMENT P. jirovecii Pneumonia The treatment of choice for PCP is trimethoprim-sulfamethoxazole (TMP-SMX), given either IV or PO for 14 days to non-HIVinfected patients with mild disease and for 21 days to all other patients (Table 227-1). TMP-SMX, which interferes with the organ­ ism’s folate metabolism, is at least as effective as alternative agents and is better tolerated. TMP-SMX can cause leukopenia, hepatitis, rash, fever, elevation of potassium and creatinine, and anaphylactic and anaphylactoid reactions. Patients with HIV infection have an unusually high incidence of hypersensitivity to TMP-SMX. Moni­ toring of serum drug levels is useful if renal function or toxicities are issues in order to enhance the likelihood that therapy will be effective and toxicity will be avoided. Maintenance of a 2-h post­ dose serum sulfamethoxazole level of 100–150 μg/mL has been associated with a successful outcome. Resistance to TMP-SMX cannot be measured by organism growth inhibition in the labora­ tory because human Pneumocystis cannot be successfully cultured. However, mutations in the target gene for sulfamethoxazole that confer in vitro sulfa resistance when found in other organisms have been recognized in Pneumocystis. The clinical relevance of these mutations for the response to therapy is uncertain. Sulfadiazine plus pyrimethamine, an oral regimen more often used for treatment of toxoplasmosis, also is highly effective. TABLE 227-1  Treatment of Pneumocystis Pneumoniaa DRUG(S) DOSE, ROUTE ADVERSE EFFECTS First-Choice Agent TMP-SMX TMP (5 mg/kg) plus SMX (25 mg/kg) q6–8h PO or IV (i.e., 2 double-strength tablets tid or qid) Fever, rash, cytopenias, hepatitis, hyperkalemia Alternative Agents Atovaquone 750 mg bid PO Rash, fever, hepatitis Clindamycin plus Primaquine 300–450 mg q6h PO or 600 mg q6–8h IV 15–30 mg qd PO Hemolysis (G6PD deficiency), methemoglobinemia, neutropenia, rash Pentamidine 3–4 mg/kg qd IV Hypotension, azotemia, cardiac arrhythmias (torsades des pointes), pancreatitis, dysglycemias, hypocalcemia, neutropenia, hepatitis Adjunctive Agent Prednisone or methylprednisolone 40 mg bid × 5 d, 40 mg qd × 5 d, 20 mg qd × 11 d; PO or IV Peptic ulcer disease, hyperglycemia, mood alteration, hypertension aTreatment can be administered for 14 days to non-HIV-infected patients with mild disease and for 21 days to all other patients. Abbreviations: G6PD, glucose-6-phosphate dehydrogenase; TMP-SMX, trimethoprim-sulfamethoxazole. Either intravenous pentamidine or the combination of clindamy­ cin plus primaquine is are options for patients who cannot tolerate TMP-SMX and for patients in whom treatment with TMP-SMX appears to be failing. Pentamidine must be administered IV over at least 60 min to avoid potentially lethal hypotension. Adverse effects can be severe and irreversible and include renal dysfunction, dysgly­ cemia (life-threatening hypoglycemia that can occur days or weeks after initial infusion and be followed by hyperglycemia), neutropenia, and torsades de pointes. Clindamycin plus primaquine is effective, but primaquine can be given only by the oral route—a disadvantage for patients who cannot ingest or absorb oral drugs and primaquine can cause hemolysis in patients with glucose-6-phosphate dehydro­ genase deficiency and can cause methemoglobinemia. Oral atova­ quone is also a reasonable option for patients with mild disease who have no impediments to absorbing an oral drug that requires a highfat meal for optimal absorption. There is some evidence for activity of echinocandins against the cyst form (but not the trophozoite form) of Pneumocystis, but the role for echinocandins as part of combination therapy is currently uncertain, and these drugs should never be used as single drug therapy for PCP. PART 5 Infectious Diseases A major advance in therapy for PCP was the recognition that glucocorticoids could improve survival rates among PLWH with moderate to severe disease (initial room air Po2 <70 mmHg or alveolar–arterial oxygen gradient ≥35 mmHg). Glucocorticoids appear to reduce the pulmonary inflammation that occurs after specific therapy is started and organisms begin to die, eliciting inflammation. Therapy with glucocorticoids should be the stan­ dard of care for patients with HIV infection and moderate to severe PCP. Therapy with glucocorticosteroids is also probably effective for patients with other immunodeficiencies who have moderate to severe PCP. This treatment should be started for moderate or severe disease when therapy for PCP is initiated, even if the diagnosis is suspected but has not yet been confirmed. If PLWH or HIVuninfected patients are receiving high-dose glucocorticoids when they develop PCP, there are theoretical advantages to decreasing the steroid dose to improve immune function, but there is no convinc­ ing evidence on which to base any specific strategy. No definitive trials have identified the best therapeutic algorithm for patients in whom TMP-SMX treatment for PCP is failing. If no other treatable infectious or noninfectious processes are detected and pulmonary dysfunction appears to be due to PCP alone, many authorities would switch from TMP-SMX to either IV pentamidine or IV clindamycin plus oral primaquine. Clindamycin-primaquine is certainly less toxic than IV pentamidine. Some authorities would add the second drug or drug combination to TMP-SMX rather than switching regimens. If patients are not already receiving them, glucocorticoids should be added to the regimen; the dosage and regimen are usually chosen empirically and depend on what glu­ cocorticoid regimen (if any) the patient was receiving when PCP therapy was begun. For PLWH who present with PCP before the initiation of ART, ART should be started within the first 2 weeks of therapy for PCP in most situations. Immune reconstitution inflammatory syndrome (IRIS) can occur, and the decision to initiate ART thus requires considerable clinical judgement that factors in the severity of pneu­ monia, the response of PCP to therapy, and concurrent medical conditions. ■ ■PREVENTION The most effective method for preventing PCP is to eliminate the cause of immunosuppression by withdrawing immunosuppressive therapy or treating the underlying cause (e.g., HIV infection). Patients who are susceptible to PCP benefit from chemoprophylaxis during the period of susceptibility. For patients with HIV infection, CD4+ T-cell counts are a reliable marker of susceptibility, and counts <200/μL are an indi­ cation to start or continue prophylaxis (Table 227-2). For patients who are immunosuppressed as a result of factors other than HIV infection, CD4+ T-cell counts <200/μL are a plausible but not absolute marker of susceptibility. However, counts >200/μL are not a reliable marker of protection from PCP. For these patients who are immunosuppressed due to causes other than HIV infection, the period of susceptibility is usually estimated on the basis of experi­ ence with the underlying disease and the specific immunosuppressive regimen. Cessation of prophylaxis has been associated with clusters of cases in certain patient populations, such as solid-organ transplant recipients, where the period of susceptibility is not well-defined. Patients receiving a prolonged course of high-dose glucocorticoids TABLE 227-2  Prophylaxis of Pneumocystis Pneumonia DRUG(S) DOSE, ROUTE COMMENTS First-Choice Agent TMP-SMX 1 tablet (double- or single-strength) qd PO Incidence of hypersensitivity is high. Rechallenge for non-lifethreatening hypersensitivity; consider dose-escalation protocol. Alternative Agents Dapsone 50 mg bid or 100 mg qd PO Hemolysis is associated with G6PD deficiency. Dapsone plus Pyrimethamine plus Leucovorin 50 mg qd PO Leucovorin ameliorates cytopenias due to pyrimethamine. 50 mg weekly PO 25 mg weekly PO Dapsone plus Pyrimethamine plus Leucovorin 200 mg weekly PO Leucovorin ameliorates cytopenias due to pyrimethamine. 75 mg weekly PO 25 mg weekly PO Pentamidine 300 mg monthly via Respirgard II nebulizer Aerosol may cause bronchospasm. Pentamidine is probably less effective than TMP-SMX or dapsone regimens. Atovaquone 1500 mg qd PO Requires fatty meal for optimal absorption. Abbreviations: G6PD, glucose-6-phosphate dehydrogenase; TMP-SMX, trimethoprim-sulfamethoxazole.