Alternative Pneumocystis Pneumonia Prophylaxis in Solid Organ Transplants

ABSTRACT

Background

Despite limited data supporting use in solid organ transplant (SOT) recipients, atovaquone and dapsone are often used as alternatives to trimethoprim‐sulfamethoxazole (TMP‐SMX) for Pneumocystis jirovecii pneumonia (PJP) prophylaxis.

Methods

This single‐center, retrospective cohort study describes a multi‐organ program's experience with alternative PJP prophylaxis. Adult SOT recipients transplanted November 13, 2020 to November 13, 2022 who received non‐TMP‐SMX PJP prophylaxis and had > 1 year follow‐up were included.

Results

Among 953 SOTs performed, 333 (34.9%) recipients received alternative PJP prophylaxis (319 [95.8%] atovaquone and 14 [4.2%] dapsone). Alternative prophylaxis was initiated in 76 (22.8%) recipients without starting TMP‐SMX, mostly due to sulfa allergy (62, 81.6%). In 257 recipients who started TMP‐SMX, common reasons for switching to alternatives were hyperkalemia (105, 40.9%) and leukopenia (77, 30.0%). While 79.8% of recipients had these adverse effects resolve, only 27.3% resumed TMP‐SMX. Tolerance was high after resumption (85.7%). Barriers to accessing alternative prophylaxis included cost (25, 7.5%) and prior authorizations (26, 7.8%). There was one case of severe disseminated toxoplasmosis, one case of Nocardia infection, and no cases of PJP.

Conclusion

Alternative PJP prophylaxis carries risk of breakthrough infection and barriers to initiation. Since most recovered from adverse effects of TMP‐SMX and tolerated resumption, providers should re‐trial TMP‐SMX when feasible.

At our institution, 34.9% of solid organ transplant recipients over a 2‐year period received alternative Pneumocystis jirovecii pneumonia prophylaxis due to trimethoprim‐sulfamethoxazole (TMP‐SMX) allergies and intolerances. Those that restarted TMP‐SMX tolerated it well (85.7%). Medical interventions and established protocols can limit the use of alternative prophylaxis and reduce the risk of breakthrough infections, such as toxoplasmosis and nocardiosis.

Abbreviations

G6PD

glucose‐6‐phosphate dehydrogenase

IQR

interquartile range

PJP

Pneumocystis jirovecii pneumonia

SOT

solid organ transplant

TMP‐SMX

trimethoprim‐sulfamethoxazole

1. Introduction

A greater volume of solid organ transplantations (SOTs) is occurring, and with them, an increasing number of SOT recipients are susceptible to an array of opportunistic infections. SOT is a known risk factor for Pneumocystis jirovecii pneumonia (PJP), particularly within the early post‐transplant period for non‐lung recipients and longer for lung transplant recipients. The risk of PJP is significantly reduced in recipients receiving effective prophylaxis; without prophylaxis, 5%–15% of SOT recipients develop PJP with the highest risk being within the first 6 months post‐transplant [1, 2]. While the standard prophylaxis is trimethoprim‐sulfamethoxazole (TMP‐SMX), this agent carries with it a risk of adverse effects including hepatotoxicity, myelosuppression, hyperkalemia, renal dysfunction, and hypersensitivity reactions. Rates of these adverse effects in the SOT population are not specified empirically in the literature but are assumed to be common [2]. Attribution of adverse effects is confounded by those associated with other common post‐transplantation medications, such as those used for cytomegalovirus disease prophylaxis and maintenance immunosuppression. Similarly, rates of resolution of these adverse effects after discontinuation of TMP‐SMX, as well as the ability to retrial TMP‐SMX during the prophylaxis period, have not been well‐studied. Rates of alternative PJP prophylaxis reported range from 9% to 32% [3, 4, 5, 6, 7]. While use of alternative agents may increase the risk of developing opportunistic infections including PJP, toxoplasmosis, and nocardiosis as well as other bacterial infections such as listeriosis and urinary tract infections [3, 6, 8, 9], the data are limited in characterizing the frequency of alternative prophylaxis use and the frequency of infections while on alternative prophylaxis.

We aimed to evaluate the frequency and reason for alternative PJP prophylaxis use across all SOT recipients at our single‐center institution and to determine the proportion of recipients eligible for retrial of TMP‐SMX. We assessed for tolerability of TMP‐SMX after resumption, noted barriers to alternative prophylaxis access, and documented rates of PJP, Nocardia, and Toxoplasma infections.

2. Methods

This study was reviewed and approved by the Institutional Review Board of the University of Pennsylvania.

2.1. Study Population

Adult liver, kidney, heart, lung, pancreas, and dual‐organ recipients who were transplanted between November 13, 2020 and November 13, 2022 at the Hospital of the University of Pennsylvania who followed up at our center for at least one year post‐transplant and received alternative pneumocystis prophylaxis within the designated prophylaxis period up to the first year post‐transplant were included. Transplant recipients who had alternative PJP prophylaxis initiated beyond the routine post‐transplant prophylaxis window were excluded. Recipients who followed up at other transplant centers or had their prophylaxis initiated by outside providers were excluded.

2.2. Prophylaxis Protocol

At our institution, preferred PJP prophylaxis for SOT recipients is dosed at one single‐strength tab (400–80 mg) of TMP‐SMX daily for all transplants aside from lung transplants which are dosed at one double‐strength tab (800–160 mg) of TMP‐SMX 3 days/week. For kidney and kidney‐pancreas transplant recipients, the duration is 180 days post‐transplant, with the resumption of therapy if treated with T‐cell‐depleting therapy for rejection. Liver transplant recipients receive prophylaxis for one year post‐transplant, with similar resumption for one year for T‐cell depleting therapy. Heart transplant recipients receive prophylaxis for 1‐year post‐transplant, but prophylaxis is resumed for 3 months in the setting of lymphocyte‐depleting agents or potent immunosuppressive agents for rejection; it is also prolonged until the completion of any 4‐drug immunosuppressive regimen or chronic prednisone doses of 20 mg or greater for at least one month. Lung transplant recipients receive life‐long prophylaxis. For all those with sulfa allergies, the first preferred alternative is atovaquone 1500 mg orally once a day taken with a fatty meal; the second alternative is dapsone 100 mg orally once a day if glucose‐6‐phosphate dehydrogenase (G6PD) activity is normal. There are no institutional guidelines directing indications for and timing of the change from TMP‐SMX and none specifying a pathway to allergy referral; this is left to the discretion of the transplant team.

2.3. Data Collection

The electronic medical record EPIC and data from a medication use evaluation were used to identify transplant recipients who matched the above study criteria. The applications SlicerDicer and PennChart Reporting WorkBench were used to obtain additional clinical data. Variables were abstracted, including age, sex, race, ethnicity, prior transplant history, alternative PJP prophylaxis prescribed, organ transplanted, the reason for non‐TMP‐SMX initial start, the reason for TMP‐SMX discontinuation, retrial of TMP‐SMX, reason for intolerance of TMP‐SMX retrial, donor and recipient toxoplasma serostatus, post‐transplant infections with Toxoplasma gondii, Nocardia, or PJP, prior authorization of alternative prophylaxis prescriptions, and cost‐related barriers to initiation of preferred alternative prophylaxis agent. Hyperkalemia was defined as serum potassium ≥5.5 mg/dL. Leukopenia was defined as white blood cell count ≤ 4000/µL. A recipient was considered “tolerant of a retrial of TMP‐SMX” if levels returned to normal and remained stable after TMP‐SMX re‐initiation, independent of the use of potassium‐lowering or marrow‐stimulating agents, within the prophylaxis window, or if their other adverse effect resolved.

2.4. Measures and Outcomes

The primary goal was to describe the incidence of alternative PJP prophylaxis use in SOT recipients. Secondarily, we sought to describe the types and incidence of adverse events related to standard prophylaxis with TMP‐SMX, the proportion of those on alternative prophylaxis eligible for retrial of TMP‐SMX, the incidence of breakthrough opportunistic infections on alternative prophylactic agents, and the incidence of cost or insurance‐related barriers to initiating the preferred agent.

2.5. Data Analysis

The study was descriptive in nature and data was summarized in descriptive measures. Medians and interquartile ranges (IQRs) were reported for continuous variables. Data was collected and managed in Research Electronic Data Capture (REDCap) and Microsoft Excel [10, 11].

3. Results

A total of 953 SOTs were performed between November 13, 2020, and November 13, 2022. Of these, 377 transplant recipients had prescriptions for alternative PJP prophylaxis. Ten were excluded for receiving the alternative prophylaxis prior to transplant, 14 were excluded for TMP‐SMX being held either beyond the indicated prophylaxis window or greater than a year post‐transplant in lung transplant recipients, 13 were excluded for not actually receiving the prophylaxis, and seven did not have records available at our institution. A total of 333 (34.9%) recipients were included in the analysis (Table 1).

TABLE 1.

Baseline demographic characteristics of solid organ transplant (SOT) recipients who received non‐TMP‐SMX PJP prophylaxis after transplant.

Baseline characteristics	n (%)
Sex at Birth
Male	195 (59)
Female	138 (41)
Race
White, Non‐Hispanic	236 (68)
Black, Non‐Hispanic	56 (16)
Hispanic/Latino	31 (9)
Asian, Non‐Hispanic	10 (0.3)
Other	15 (2)
Prefer Not to Answer	16 (2)
Age at SOT, years (median, IQR)	57, 44–63
Proportion of SOTs	Total SOTs: 953
Kidney	114 (29)
Liver	106 (37)
Heart	39 (45)
Lung	39 (33)
Pancreas	2 (50)
Multi‐organ †	33 (53)
Prior SOT	23 (7)

Abbreviations: IQR, interquartile range; PJP, Pneumocystis jirovecii pneumonia; SOT, solid organ transplant; TMP‐SMX, trimethoprim‐sulfamethoxazole.

^† Pancreas‐kidney, liver‐kidney, heart‐kidney, and liver‐lung.

There were 319 (95.8%) recipients who received atovaquone and 14 (4.2%) who received dapsone. None received pentamidine. The recipients were 41.4% female at birth and 68.2% white non‐Hispanic with a median age of 57 years (IQR 44–63). As a proportion of the 953 total SOTs that occurred during the study period, these recipients received 114 (28.6%) of the 398 kidneys, 106 (37.2%) of the 285 livers, 39 (45.3%) of the 86 hearts, 39 (33.1%) of the 118 lungs, two (50.0%) of the four pancreases, and 33 (53.2%) of the 62 multi‐organ transplants. There were 76 recipients who received non‐TMP‐SMX primary prophylaxis; atovaquone was initiated in 74 (97.4%) and dapsone in two, most due to reported sulfa allergy (62, 81.6%) (Table 2). Types of allergies reported included rash, hives, itching, angioedema, and anaphylaxis; adverse effects including gastrointestinal intolerance, headache, paresthesias, and aseptic meningitis were also reported. Other reasons for not starting with TMP‐SMX included continuation of pre‐existing PJP prophylaxis with atovaquone, baseline hyperkalemia or renal dysfunction at the time of transplant, and drug‐induced liver injury and rhabdomyolysis attributed to past TMP‐SMX use. Of 257 recipients reported to be intolerant of TMP‐SMX after initiation, the most common reasons for switching to alternative prophylaxis were hyperkalemia (105, 40.9%) and leukopenia (77, 30.0%). Of the 105 recipients who had TMP‐SMX discontinued for hyperkalemia, proportions by transplanted organ during the study period were as follows: kidney (27 of 398 [6.78%]), liver (42 of 285 [14.7%]), heart (12 of 86 [14.0%]), lung (17 of 118 [14.4%]), and multi‐organ (seven of 62 [11.3%]). There were 39 recipients who had TMP‐SMX discontinued for leukopenia that had absolute neutrophil counts of less than 1000/µL (100–2860/µL). Other reasons for stopping TMP‐SMX included acute kidney injury, thrombocytopenia, hepatotoxicity, rash, allergy, and drug‐drug interactions (warfarin, dofetilide). TMP‐SMX was held at a median of 66 days (IQR 24–113). Almost all who developed hyperkalemia improved after TMP‐SMX discontinuation (100, 95.2%) with most on potassium‐lowering therapy (80, 80.0%). Despite the documented improvement of hyperkalemia in most patients, only 32 (32.0%) resumed TMP‐SMX after resolution. Those who developed leukopenia had lower rates of recovery (48, 62.3%), and 13 (27.1%) were on marrow‐stimulating agents at recovery. While 79.8% of recipients had a resolution of the reason for the TMP‐SMX switch, only 27.3% resumed it. Tolerance was high after resumption (85.7%) and the median time to resumption was 33.5 days (IQR 14–66) for those who had TMP‐SMX held. As mentioned earlier, there are financial considerations to switching recipients to alternative therapies. Weekly drug costs per average wholesale price were higher with atovaquone ($91.84) and dapsone ($21.14) compared to TMP‐SMX ($4.62), which amounts to up to $4535.44 in increased costs over the course of a prophylaxis period. 51 recipients (15.3%) had insurance‐related barriers to using alternative prophylaxis.

TABLE 2.

Alternative Pneumocystis jirovecii pneumonia (PJP) prophylaxis usage including choice of initial prophylaxis, reason for trimethoprim–sulfamethoxazole (TMP‐SMX) intolerance, barriers to alternative agent use, and patterns of adverse effect resolution with tolerance of TMP‐SMX reinitiation.

Alternative PJP Prophylaxis Usage	n (%)
SOTs on Alternative PJP Prophylaxis	333
By Organ
Kidney	114 (34)
Liver	106 (32)
Heart	39 (12)
Lung	39 (12)
Multi‐organ †	33 (10)
Initial Prophylaxis
Atovaquone	74 (22)
Dapsone	2 (1)
TMP‐SMX	257 (77)
TMP‐SMX Intolerance	257 (77)
Hyperkalemia	105 (41)
Leukopenia	77 (30)
Other	75 (29)
Alternative Agent	257 (77)
Atovaquone	245 (95)
Dapsone	12 (5)
Access Barriers	51 (15)
Cost	25 (7)
Prior Authorization	26 (8)
Hyperkalemia	105
Resolved	100 (95)
Potassium‐Lowering Therapy Use	80 (80)
TMP‐SMX Resumed	32 (32)
Leukopenia	77
Resolved	48 (62)
Marrow‐Stimulating Factor Use	13 (27)
TMP‐SMX Resumed	10 (21)
Other	75
Resolved	57 (76)
TMP‐SMX Resumed	14 (25)
Total Resumed on TMP‐SMX	56 (27)
Tolerated TMP‐SMX Resumption	48 (86)
Time to TMP‐SMX Discontinuation (median days, IQR)	66, 24–113
Time to TMP‐SMX Resumption (median days, IQR)	34, 14–66

Abbreviations: IQR, interquartile range; PJP, Pneumocystis jirovecii pneumonia; SOT, solid organ transplant; TMP‐SMX, trimethoprim‐sulfamethoxazole.

^† Pancreas‐kidney, liver‐kidney, heart‐kidney, and liver‐lung.

While toxoplasma serostatus was unavailable from many kidney and liver donors and recipients, only one lung and two heart transplants had incomplete serologies. Of the 39 heart transplant recipients with complete serologies, there were four high‐risk seropositive donor seronegative recipient mismatches. There was one case of severe disseminated toxoplasmosis in a toxoplasma serostatus mismatch heart transplant recipient who was receiving atovaquone. There was one case of Nocardia nova complex bloodstream infection in a kidney transplant recipient on atovaquone. There were no cases of PJP.

4. Discussion

Alternative PJP prophylaxis agents for recipients who have existing sulfa allergies, have not tolerated TMP‐SMX, or have developed an adverse reaction to TMP‐SMX include atovaquone, dapsone, pentamidine, clindamycin, and pyrimethamine. We describe our experience with the largest proportion of alternative PJP prophylaxis use in the literature and the largest cohort of alternative PJP prophylaxis in SOT recipients. Of nearly 1000 SOTs performed over 2 years, a full third received an alternative form of PJP prophylaxis, with most due to intolerance developing while on TMP‐SMX. This is higher than the rates currently reported in the literature [3, 4, 5, 6, 7]. Lum et al in their two‐year, multicenter experience describe 105 cases of alternative PJP prophylaxis use in the first year post‐transplant that was evenly divided between those who did not start TMP‐SMX due to allergies or intolerances reported prior to transplant and those who did not tolerate TMP‐SMX after transplant [4]. Puing et al describe a similar proportion (52%) of an alternative agent due to a reported history of allergy when reviewing 25 cases of Nocardia spp. infection that occurred in the absence of TMP‐SMX [8]. Whether the high rates of intolerance in our cohort were due to greater provider awareness of potential laboratory abnormalities or to increased use of other medications with overlapping adverse effects misattributed to TMP‐SMX is unclear.

Furthermore, the type of intolerance differed; we saw higher rates of hyperkalemia (40.9%) relative to the lower rates (3.2%–18%) reported in other studies, where neutropenia and acute renal injury were more common [4, 5, 7]. Again, it is not clear whether this reflects more early provider‐driven interventions or a decreased use of potassium‐lowering agents such as sodium zirconium cyclosilicate, patiromer, and sodium polystyrene sulfonate relative to other institutions. The higher proportion of recovery seen in those with hyperkalemia compared to recipients who developed leukopenia (95.2% vs. 62.3%) does suggest that the intervention may have occurred early during the course of hyperkalemia; however, time to medication intervention was not directly assessed. There was also a lower proportion of marrow‐stimulating agent usage relative to potassium‐lowering therapy usage at the time of adverse effect resolution (27.1% vs. 80.0%), which confounds the comparison of the proportion of recovery.

Only 4.2% of recipients receiving alternative prophylaxis received dapsone, likely due to its placement as the second choice after atovaquone at our institution; there was a considerable proportion (45%) on dapsone in Lum et al's study, highlighting the variability between transplant centers in their approaches to alternative PJP prophylaxis agents [4, 7, 12].

What remained consistent, however, were the high rates of tolerability after reinitiation of TMP‐SMX seen in this and other studies. We report an 85.7% rate of tolerance to completion of prophylaxis with TMP‐SMX up to a year post‐transplant, similar to smaller cohorts reporting 90%–100% tolerance rates [4, 5]. This may be reflective of early or unnecessary transition from TMP‐SMX and may support prompt and protocolized usage of potassium‐lowering or marrow‐stimulating agents with time‐limited avoidance of TMP‐SMX for refractory cases.

Opportunistic infections were rare, though severe when present. The frequency of infections while on alternative prophylaxis is not well‐characterized in the literature. Only a handful of investigators have described their centers’ experience with alternative PJP prophylaxis, with only one study directly comparing their use against TMP‐SMX to assess tolerability and infection risk in the kidney transplant population [13]. Our case of disseminated toxoplasmosis in a D+/R‐ heart transplant recipient while on atovaquone serves as a reminder to attend to individual transplant recipients’ risk factors for specific opportunistic and donor‐derived infections when managing prophylaxis. Most of the evidence suggesting that atovaquone is unlikely to be effective for the prevention of toxoplasmosis comes from the stem cell transplant literature, but reports of both donor‐derived and reactivated toxoplasmosis do exist in the SOT literature [14, 15, 16]. Only one other report of donor‐derived toxoplasmosis in a heart transplant recipient definitively breaking through atovaquone during the prophylaxis period appears to exist in the literature [17]. It was concerning that, in the 39 heart transplants and two heart‐kidney transplants performed at our institution during this period, only four were donor‐recipient toxoplasma antibody mismatches and of these one developed disseminated toxoplasmosis. A dapsone and pyrimethamine combination regimen should be considered in those intolerant of TMP‐SMX, with screening for G6PD deficiency prior to dapsone initiation and monitoring for complications such as methemoglobinemia, hemolysis, and drug‐drug interactions. It is important to note that even those with normal G6PD levels can develop these adverse effects [18]. Other opportunistic infections are generally rare in the literature as well. Rates of nocardiosis have been reported at 0%–3% [4, 6, 13, 19]. The highest rate was reported by Lum et al at up to 3% of recipients off TMP‐SMX developing Nocardia spp. infections [4]. Puing et al report a series of 25 Nocardia spp. infections over 22 years in immunocompromised hosts off TMP‐SMX prophylaxis of which 12 were status post intrathoracic organ transplantation, though the total number of SOT recipients off TMP‐SMX was not reported [8]. While we did not detect any Nocardia species infections in our 81 intrathoracic organ transplants on alternative prophylaxis, a bloodstream infection with Nocardia nova complex did occur in a kidney transplant on atovaquone potentially associated with a central line. The rate of 0.3% overall is lower compared to the above studies. Lastly, PJP breakthrough is exceedingly rare, with the only report in Lum et al's cohort being on pentamidine rather than atovaquone or dapsone [4]. There have been no head‐to‐head, prospective comparisons of TMP‐SMX against alternative agents for prophylaxis against PJP to date, and overall rates of breakthrough are low at 0%–1% [4, 5, 13].

Alternative agents may also be more costly, have insurance coverage restrictions, or have stricter administration considerations, thus creating additional barriers to adequate adherence and further increasing the risk of opportunistic infection.

While the majority of our cohort discontinued TMP‐SMX for intolerance after initiation, a notable proportion (22.8%) started with alternative prophylaxis. Our institution does not have a protocol in place for referral for an allergy evaluation. These results indicate that attention is needed in developing a pathway incorporating appropriate allergy screening and referral to a specialist. This, in conjunction with pathways outlining medication interventions that may attenuate adverse effects of TMP‐SMX, monitoring for resolution of these adverse effects, and appropriate re‐trial of TMP‐SMX, represents an opportunity for interdisciplinary collaboration between members of the infectious diseases, transplant pharmacy, antimicrobial stewardship, and allergy teams at individual institutions to develop protocols that optimize the care of its transplant recipients.

Our study is limited in that it occurred retrospectively at a single center. As seen in the literature, generalizability is often limited given varied prophylaxis periods protocolized at different institutions between different organs. It is additionally difficult, when assessing a recipient's reason for TMP‐SMX discontinuation and subsequent tolerance to retrial of TMP‐SMX, to exclude other confounding medications such as valganciclovir and mycophenolic acid formulations or non‐medication‐related clinical changes that impact electrolyte balance or the bone marrow. Individual variations in practice managing hyperkalemia and leukopenia may prompt different approaches to the management of these complications. The number of breakthrough infections was low; as follow‐up was limited to one year, this may underestimate the cumulative risk of breakthrough infection, particularly in lung transplant recipients on life‐long prophylaxis. All data elements were able to be captured except for toxoplasma serostatuses, not collected largely in non‐intrathoracic organ transplants, and two recipients for whom cell differentials were not obtained at the time of leukopenia measurement. We were unable to completely ascertain the number of recipients who did not tolerate atovaquone due to administration issues such as taste and timing with high‐fat meals; it is conceivable that intolerance could have contributed to a proportion of those who resumed TMP‐SMX. Additional assessments of atovaquone or dapsone‐related adverse effects and rates of other infections, such as urinary tract infections, would be helpful to expand the limited literature on this subject relevant to the increasing numbers of SOT recipients.

5. Conclusion

Alternative PJP prophylaxis agents such as atovaquone or dapsone may have barriers to initiation and a limited spectrum of activity compared to TMP‐SMX. Strategies to manage hyperkalemia and leukopenia should be considered before stopping TMP‐SMX. Due to high tolerance rates, TMP‐SMX should be reinitiated when electrolyte, hematologic, or other abnormalities have resolved.

Disclosure

Jennifer Trofe‐Clark's institution has received research funding from Veloxis Pharmaceuticals for a separate study initiative, none of which was used for the study design, performance, data analysis, or manuscript preparation of this study. Emily Blumberg has received research funding from Merck, Takeda, and Scynexis and honoraria from Biotest and Kamada; none of these were relevant to the current study.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Visual Abstract

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.