Background/Purpose : The risk of Pneumocystis carinii
pneumonia (PCP) in Systemic Lupus Erythematosus (SLE) is not well-established.
Accordingly, this study aims to establish the prevalence of PCP in
hospitalized SLE patients.

Methods : This is a
retrospective cohort study evaluating the prevalence of PCP in hospitalized patients
with SLE. Data was obtained from the Columbia University Medical Center (CUMC)
Clinical Data Warehouse from January 2000 to September 2014.  HIV-positive patients and patients with renal
transplant were used as controls.

Results : Patients with SLE were identified by searching the database for an
ICD-9 of 710.0; as such, 4,800 patients were identified. Chart
review of a sample of 107 charts (roughly 3% of the identified 4,800) was used
to verify the diagnosis of SLE. Of the 107 charts reviewed, SLE case
definitions were assigned:  a) met 4
out of 11 ACR or SLICC SLE criteria = 58
patients (54%), b) has 2-3 SLE
criteria along with Lupus medications and/or Lupus auto-antibodies or met only 1
SLE criteria but had a previous diagnosis of SLE by a rheumatologist and were treated
with SLE medication = 21 patients (20%), c) had a history of Lupus (usually
“Lupus in remission”) without criteria verification, of which half had +ANA = 7
(7%).  21/107 (20%) of patients in
the sample did not have SLE but had a diagnosis of another autoimmune disease,
such as Sjogren’s syndrome, RA, scleroderma, vasculitis, or polymyositis.  We decided to consider only the first
two groups that met the most stringent SLE case definition to represent the
true SLE cases, i.e 74% as of the 4800 patients
(3,552) were extrapolated as having SLE, which is consistent with other studies
of administrative databases. 

A total of 9 cases of PCP in SLE
patients were identified. The diagnosis of PCP was made based on clinical and
radiographic evidence. PCP Gomori methenamine silver stain was negative in 8/9.

Prevalence of PCP in SLE, renal
transplant, and HIV patients is presented in the following table.

Of the 9 cases of PCP, 8 patients met SLE criteria as defined above, and
1 patient had a confirmed historical diagnosis of SLE. In addition, 3/9
patients had co-existing AIDS with a median CD4 count of 31.

All patients were treated with IV trimethoprim-sulfamethoxazole and
discharged; no deaths occurred. None
of the patients were on PCP prophylaxis prior to admission, which is not
routinely done for SLE patients at CUMC.

Conclusion :  The risk of PCP in SLE is low compared to
other immunosuppressed patients, such as HIV or renal transplant. All patients responded
to treatment with trimethoprim-sulfamethoxazole. This data does not
substantiate the need for PCP prophylaxis, except for patients with SLE-HIV.
Although rare, SLE patients admitted with bilateral infiltrates should have a low
threshold for initiating PCP treatment.