BCG for patients with non–muscle-invasive bladder cancer (NMIBC)

Approximately 82,000 new cases of bladder cancer are reported annually in the United States, and among them, 80% are NMIBC. ¹ Intravesical instillation of BCG is the preferred approach for high-risk NMIBC, and an option for intermediate-risk NMIBC per the American Urological Association (AUA) guidelines. ^{Reference Chang, Boorjian and Chou2} Although risk categories differ somewhat between AUA, the European Association of Urology (EUA), and the US National Comprehensive Cancer Network (NCCN) guidelines, all recommend BCG as the preferred therapy for intermediate- and high-risk NMIBC. Following transurethral bladder tumor resection and staging, patients who qualify for BCG undergo a 6-week induction cycle. At each weekly visit, a vial of BCG is diluted and instilled into the bladder for contact time of 90–120 minutes. If the patient tolerates therapy and response is demonstrated on repeat cystoscopy after the 6 weeks of therapy, a maintenance phase may be continued for up to 1 year for intermediate-risk patients or up to 3 years for high-risk patients. ^{Reference Chang, Boorjian and Chou2–Reference Lenis, Lec and Chamie5} Although the optimal timing and schedule for maintenance remains unknown, a typical schedule is weekly for 3 weeks during months 3 and 6, then every 6 months thereafter. ^{Reference Chang, Boorjian and Chou2} BCG therapy is associated with complete resolution in ∼80% of patients at the end of induction and 55% at the end of 3 years. ^{Reference Lenis, Lec and Chamie5} Due to worldwide BCG shortages since 2019, the AUA recommends prioritizing BCG for high-risk NMIBC patients undergoing induction therapy and considering reduced doses (splitting vials among qualifying patients) if necessary. ^6–8 Few alternatives exist for high-risk patients who cannot receive BCG due to the shortage or those who fail to respond to BCG. ^{Reference Lenis, Lec and Chamie5}

Infectious risks to patients receiving intravesical BCG

Infectious risks to patients receiving BCG have been well documented in the literature and are reported with low incidence (<1%), theorized to be associated with traumatic catheterization during procedure or to systemic spread within immunocompromised hosts. ^{Reference Lamm9–Reference Gonzalez, Musher and Brar16} Infections secondary to BCG may result in localized or disseminated BCG with various organ involvement. Patients may present with infection symptoms early (within 3 months) or late (>1 year) after BCG administration. ^{Reference Gonzalez, Musher and Brar16} Although there are no treatment recommendations for postintravesical BCG infections, treatment has generally been provided for symptomatic patients.

Risks to healthcare personnel related to preparation or administration of BCG

Few publications have documented direct risks to HCP related to intravesical administration, despite use spanning almost 50 years. Three case reports detail hand infections after needlesticks in HCP who were preparing or administering BCG for intravesical administration (Table 1). ^{Reference Atiyeh, Wazzan, Kaddoura, Saghir, Zaatari and Hamdan17–Reference Mundinger, Douglas and Higgins19} Use of needleless systems for all biologic transfers is a best practice that may reduce needlestick exposures. ²⁰ The other documented HCP BCG infections were skin-test conversions due to secondary transmission from a pediatric patient who developed disseminated BCG and respiratory infection after being treated in the same healthcare facility where BCG was prepared, but who did not receive BCG. ^{Reference Waecker, Stefanova, Cave, Davis and Dankner21} Contact tracing of 46 HCP revealed 2 new skin-test conversions. Details are insufficient to determine a specific risk that led to these secondary transmission events.

Table 1. Case Reports of Infections in HCP Related to Preparing or Administering Bacillus Calmette-Guérin (BCG) Therapy

Note. HCP, healthcare personnel.

^a Underwent 4 months of ethambutol, pyrazinamide, isoniazid, rifampin plus 2 months of treatment with isoniazid and rifampin.

HCP infections may be underrepresented in the literature for several reasons. First, BCG-related infection may be culture-negative or not identified to the species-level within the M. tuberculosis (MTB) complex. For BCG cases that are culture positive, clinical laboratories may not routinely provide genotyping. ²² Genotypic reporting from the National Tuberculosis Genotyping Service may be delayed, and phenotypic susceptibilities (eg, intrinsic resistance to pyrazinamide) may be the first clue that differentiates M. bovis from other members of the MTB complex. Second, the lack of required reporting of M. bovis infections (when identified to species level) decreases recognition of this and nontuberculous mycobacterial infections. ²³ A review of 118 BCG cases reported to the US National Tuberculosis Surveillance System between 2004 and 2015 postulated that confusion between case definitions of the US Council of State and Territorial Epidemiologists (CSTE) and Centers for Disease Control and Prevention (which specifically discourages reporting of M. bovis) contributed to reporting of only some cases of BCG infection. ^{Reference Wansaula, Wortham and Mindra24} Potential lack of recall or recognition of BCG preparation or administration activities as a potential risk exposure in HCP may also contribute to underreporting. Temporal delays from exposure to disease development due to latency decrease the likelihood that BCG would be linked to infection, should it occur. Finally, for HCP from countries where BCG is employed as a vaccination strategy, infections may be classified as endogenous reactivation of the vaccine strain rather new exposure. For these reasons, we expect there have been additional HCP cases that were not captured and/or could not be linked to BCG exposure.

Risks to patients treated at the facility where BCG is prepared or administered

Several case reports detail suspected or confirmed BCG infections related cross contamination within the healthcare facility where BCG was prepared or administered (Table 2). ^{Reference Waecker, Stefanova, Cave, Davis and Dankner21,Reference Coppes, Oliveria, Howes, Pusic, Gold and Richardson25–Reference Aqua, Holdsworth, Burd, Jacob, Ray and Schetchter31} Serious BCG infections, including meningitis and disseminated disease, have been reported among immunocompromised patients who received chemotherapy or other products mixed in the same biological safety cabinets (BSC) as BCG. ^{20,Reference Wansaula, Wortham and Mindra24,Reference Coppes, Oliveria, Howes, Pusic, Gold and Richardson25,Reference Stone, Vannier, Storch, Nitta and Zhang27} Some investigations never identified gaps in preparation or administration procedures, despite extensive review. ^{Reference Waecker, Stefanova, Cave, Davis and Dankner21,Reference Coppes, Oliveria, Howes, Pusic, Gold and Richardson25,Reference Shope, Garrett and Waecker26} Aerosol contamination in the pharmacy was considered the most likely cause in these instances, but the same administrator or shared equipment could not be excluded.

Table 2. Case Reports of Infections in Patients Who Did Not Receive Bacillus Calmette-Guérin (BCG) Therapy Due to Suspected Cross Contamination During BCG Preparation

Note. ALL, acute lymphocytic leukemia; AFB, acid-fast baccilli; BAL, bronchoalveolar lavage; BCG, Bacillus Calmette-Guérin; CSF, cerebrospinal fluid; HCP, healthcare personnel; HIV, human immunodeficiency virus; IAI, intrabdominal infection; OR, operating room.

^a 28 suspected patients were identified. Only 10 patient samples were assessed; all were negative. Also, 8 living patients had symptoms that resolved spontaneously. Although 1 patient died 13 months after receiving chemotherapy, BCG OncoTICE–related infection was not confirmed.

Vos et al ^{Reference Vos, de Haas and Verbrugh28} assessed Dutch hospital pharmacy procedures and identified gaps in biosafety procedures. Preparers did not routinely change gloves, the same BSC was used for BCG and other hazardous preparations in succession, and the BSC was not routinely cleaned after each preparation. Spills were managed with tap water rather than a proven disinfectant and decontaminant. Dust collected from the spill trough of the BSC and the pharmacy technician’s gloves were PCR-positive for MTB complex DNA. At the time, OncoTICE was supplied in an ampule rather than a vial, which increased spill and aerosolization risk.

Another case report detailed cross contamination in the operating room. ^{Reference Aqua, Holdsworth, Burd, Jacob, Ray and Schetchter31} A case patient with multiple S. aureus breast abscesses grew BCG from one sample, though the patient had never received BCG. Through tracing, the case was linked with a source patient with large BCG intra-abdominal abscess due to complication of intravesical BCG administration drained in the same operating room 48 hours earlier. Retraining of staff on proper disinfection was required after 2 gaps were identified in operating-room turnover cleaning; investigators could not determine whether any equipment was shared. Improper disinfection and use of protective equipment contributed to 9 cases of BCG infection that occurred at outpatient clinics providing central venous catheter care and BCG instillation.Reference Meije, Martínez-Montauti and Caylà ²⁹ These cases highlight the need for reinforcement, accountability, and periodic re-education to ensure that appropriate cleaning and disinfection are performed.

Guidance concerning BCG preparation and administration

Several groups have statements on BCG safe handling. The TICE BCG package insert recommends: (1) a dedicated area and BSC when available and (2) other operations should be ceased or fully segregated while BCG is prepared. ³² The latter recommendation is impractical for the contemporary pharmacy clean room. Currently, the US National Institute for Occupational Safety and Health (NIOSH) lists BCG as a “hazardous” drug. ³³ As a NIOSH hazardous drug, the US Pharmacopeia (USP) chapters 797 and 800 have additional specific requirements for the appropriate transport, storage, handling, and preparation of BCG. These regulations resulted in removal of BCG preparation from many urology clinics into pharmacy cleanrooms. ^{Reference Gomella34} The AUA and the Society of Urologic Nurses and Associates (SUNA) provides guidance on administration and waste disposal for HCP within the urology clinic and patient education to decrease risk of transmission (Table 3). ³⁵ Issues that remain concerning handling of BCG in clinics and health systems are presented in Figure 1.

Figure 1. Clinic and health-system issues concerning handling of BCG.

Table 3. AUA and SUNA Universal Protocol to Reduce Healthcare Personnel Exposure to Hazardous Drugs ³⁵

Note. AUA, American Urological Association; BCG, Bacillus Calmette-Guérin; SUNA, Society of Urologic Nurses and Associates.

If the hazardous drug designation were lifted for BCG, as suggested in a prior draft document, ³⁶ care would be needed to support facility-level risk assessment of BCG. Facility-level risk assessment is the current standard for gene and cellular biologics and is a systematic process by which individual facilities designate the safe and proper handling of an agent. ^{Reference Petrich, Marchese, Jenkins, Storey and Blind37} A consistent, clear framework is needed to evaluate risks of these therapies to patients, HCP, and other patients who might be exposed. For an assessment of BCG, multidisciplinary teams including pharmacy, nursing, urology, infection prevention, and waste management should work together to develop a plan for patient, HCP, and environmental safety. Many health systems are developing and implementing biosafety categories and associated risk-mitigation policies to standardize operational procedures for biologics. ^{Reference Penzien38} Our medical center has chosen to prepare BCG and novel biologics in a segregated room or dedicated BSC away from oncolytic preparations as infrastructure allows. A deactivation, decontamination, and cleaning procedure is conducted between preparations to prevent cross contamination using solution dwell times that have efficacy against BCG.

In conclusion, BCG has been the standard immunotherapy for intermediate- and high-risk NMIBC for decades. Infection risk of intravesical BCG to patients receiving the therapy is low and well characterized. The risks to HCP and other patients treated at facilities where BCG is prepared and administered are rare and potentially limited by recognition and reporting in the medical literature. BCG poses biohazard and infection risks that warrant handling and preparation aligned with elements of USP 800, even if no longer deemed ‘hazardous’ by NIOSH standards. Consistent use and regular re-education on best practices for infection prevention are important to maintain safety and mitigate risks of this highly effective therapy.