Successful SCID Gene Therapy in Infant with Disseminated BCG

We present a case of an infant with X-linked SCID (XSCID) and disseminated BCG infection, who enrolled on our XSCID gene therapy trial, in which newly diagnosed patients receive lentiviral-transduced, gene-corrected CD34+ hematopoietic stem cells after low-dose, targeted busulfan conditioning (NCT01512888). (1) The patient was reported in our initial publication (patient #6) and we herein describe in detail his clinical course that was complicated by disseminated BCG infection.

The patient was born in Brazil at full term to nonconsanguineous parents. He received BCG vaccination at birth despite a family history of 2 brothers dying of infections in infancy. However, the family history prompted an immunodeficiency genetic work up at 3 months of age, which found a pathogenic variant (c.903_910del; p. Glu302ArgfsX110) in the common γ-chain gene (IL2RG) confirming XSCID. Due to lack of local BCG vaccine reaction he was observed without therapy according to the European guidelines for patients with primary immunodeficiencies (PIDs) and history of BCG vaccination. (2) At 4 months of age, he developed BCG vaccine site reactivation on the right arm with axillary lymphadenopathy for which he was placed on isoniazid (INH) with complete resolution of the findings. INH was continued once the patient was transferred to St. Jude.

During pre-evaluation for gene therapy at age 9 months, he became febrile and unable to move his left arm following an accidental arm pull. Infectious work up included blood cultures and radiologic imaging. An MRI of the shoulder revealed significant edema of the rotator cuff muscles around the left scapula (Figure 1A). Acid fast bacilli (AFB) stain and culture from the muscle biopsy were negative. His restriction of movement improved on empiric treatment for myositis with cefepime and clindamycin. A follow up MRI, approximately two weeks later, demonstrated near resolution of the muscle edema but revealed an abnormal signal with medullary expansion in the left scapula, for which a cortical window and extensive curettage were performed (Figure E1). Pathology showed an exuberant proliferation of histiocytes containing AFB, described as “mycobacterial pseudotumor” (Figure E2). Rifampin (RIF), ethambutol (EMB) and azithromycin were added to INH. Culture from the bone tissue grew Mycobacterium tuberculosis complex, which was identified by PCR as Mycobacterium bovis BCG (M. bovis BCG). The organism was resistant to INH, hence treatment was modified to levofloxacin, RIF and EMB. The blood culture from the initial infectious disease work up also grew INH resistant M. bovis BCG after 4 weeks. Subsequent blood cultures and a bone marrow aspirate following initiation of multidrug antimycobacterial therapy (MAT) were negative for M. bovis BCG prior to gene therapy. Residual hypermetabolic uptake in the scapula with increased surrounding soft tissue changes were present following surgery and prior to gene therapy, but no other sites of increased metabolic activity were found on PET at that time.

Figure 1: Disseminated BCG infection.

(A) Axial fat saturated post-contrast image pre gene therapy shows extensive involvement in the muscles and scapula. (B) Resolution of MRI changes 18-months post gene therapy.

Given the excellent post-surgical recovery and reassuring clinical status, the patient was enrolled to receive gene therapy at 11 months of age. RIF was replaced with rifabutin prior to busulfan administration. The cumulative area under the curve (AUC) of busulfan was 22.9 mg x hr per liter prior to the infusion of autologous, gene-corrected CD34+ hematopoietic stem cells. Busulfan and stem cell infusion were well tolerated with no acute or long-term toxicities. In particular, we observed no liver toxicity, allowing continuation of MAT without additional alteration. Three months after cell infusion he developed fevers for 3 days that coincided with de novo production of T cells but no other signs or symptoms of immune reconstitution syndrome (IRIS). Four months after gene therapy he had significant functional T cell recovery (CD3+: 958 cells/μl, CD8+: 188 cells/μl, CD4+: 726 cells/μl, CD4+CCR7+CD45RO−: 594 cells/μl), and normal lymphocyte responses to phytohemagglutinin, concavalin A and pokeweed mitogen. He continued MAT for 20 months without any toxicity; serial ophthalmology exams and liver function tests were unremarkable. His last shoulder MRI, 18-month post gene therapy, was normal (Figure 1B). At his last follow up 2.5 years post gene therapy, the patient was clinically well with a stable, functional immune system and protective vaccine titers.

BCG vaccination is contraindicated in patients with immunodeficiency. (3) Osteitis and osteomyelitis, although rare, have been described, but documented hematogenous spread with positive blood cultures has been reported in only 1% of patients with disseminated disease. (4) Restoration of the immune system in patients with SCID, who have an active infection, is an emergency. Although matched sibling donor (MSD) transplants can be safely performed without conditioning or immunosuppression for graft versus host disease (GVHD) prophylaxis, fewer than 20% of patients have an MSD. (1) Furthermore, recovery from disseminated BCG infections following MSD transplants without conditioning and GVHD prophylaxis can vary in some patients, even when adoptive transfer of BCG immune T cells was given, and donor reconstitution occurred. (5) Since optimal alternative donor transplants necessitate a conditioning regimen and GVHD prophylaxis, these are not ideal for SCID patients with active infection due to concern of additional transplant related morbidity. (6) Thus, suboptimal treatment options outside MSD transplants as well as the lack of management guidelines for BCG infections in the US, where the vaccine is not administered, make the population particularly challenging to treat. Based on our case, gene therapy combined with sub- myeloablative busulfan presents a promising approach. In particular, GVHD prophylaxis, which slows down immune recovery in the allogeneic transplant setting, is not required. (1, 7) At present, no gene therapy for any molecular type of SCID is FDA approved. In addition, clinical studies have only been conducted for XSCID, ADA and Artemis-deficient SCID. (8)

In conclusion, our case is unique and highlights several important aspects of management of SCID patients with BCG infection. First, it reinforces the need to consider a MAT regimen as an initial therapy in SCID patients with a history of BCG vaccination even if asymptomatic since INH alone did not prevent disseminated infection and may have led to the resistance. Second, upfront and serial radiologic imaging should be considered for all patients even if asymptomatic and blood cultures for AFB should be added to the initial infectious disease work up. Third, gene therapy with low dose busulfan should be considered in patients with XSCID who lack an MSD, even in the setting of a life-threatening infection such as disseminated BCG. In particular, our patient not only recovered from busulfan and MAT without any sequelae, but also achieved full functional reconstitution including B-cell recovery, as judged by intravenous gammaglobulin independence and antibody responses to vaccination, which is unlikely post-transplant or gene therapy without conditioning for XSCID. (7, 9)

Supplementary Material

1

Figure E1: Picture of cortical window from surgical procedure. For details see text.

2

Figure E2: AFB staining of biopsy specimen. AFB staining of the biopsy specimen shows numerous acid-fast organisms; 100-fold magnification.

Clinical implications.

This case illustrates the importance of initiating antimycobacterial therapy in patients with immunodeficiency and BCG vaccination. It suggests considering gene therapy for treatment of patients with XSCID and active infections.