TB reactivation is an established adverse effect attributed to many anti-cancer biological agents and with TNF-α inhibitors as well [7, 8]. The incidence of TB reactivation in cancer patients is higher in hematological malignancies compared to solid tumors. While, among solid tumors, the highest incidence of TB reactivation was reported in lung cancer followed by gastric cancer, breast cancer, liver cancer, and colon cancer, respectively [9].
With the expanding use of immune checkpoint inhibitors for the management of cancer, infectious complications of immune checkpoint inhibitors became an emerging adverse effect of these agents, including TB reactivation [10].
The majority of patients infected with tuberculosis will develop a latency state with no signs of disease, with approximately, up to ten percent of those patients may develop active tuberculosis infection [11]. Containments of the infection are mediated by cytokines and the interaction between macrophages and T lymphocytes (CD4 and CD8) [12]. Immunocompromised status including, HIV, organ transplanted patients, and patients receiving immunosuppressive therapy is one of the most critical risk factors for TB reactivation [13].
The exact mechanism of TB reactivation following treatment with these agents remains unclear, with further studies is warranted. However, few preclinical studies in MTB infected PD-1-deficient mice and PD-1 blocked humans describe an increase in the IFN-α production by CD4 T cells which promote more bacterial replication and tissue destruction [6, 14, 15].
Furthermore, the role of (PD-L1/PD-1) pathway has been studied which has demonstrated its effect on M. tuberculosis infection; In mice model, PD-1 deficiency showed significant sensitivity to M. tuberculosis infection and high bacillary load after exposure to aerosol infection with M. tuberculosis. PD-1-deficient mice also showed dramatic survival reduction and lung tissue was found to be severely necrotic and inflamed in comparison to the control group [16]. On the other hand, the data about (PD-L1/PD-1) pathway and its role in the cytolytic activity of T. lymphocytes in humans is diversely contradictory [17]. However, multiple reports highlighted the reactivation of pulmonary tuberculosis infection after the use of PD-1 inhibitors [10, 15, 26, 27, 20,21,22,23,24,25,26,25].
In this paper, ICIs associated MTB infection was extensively searched by expediting all the reported cases through PubMed up to September 2019, with no language restriction applied. In general, 15 reported cases were identified retrieved from 12 articles [10, 15, 26, 27, 20,21,22,23,24,25,26,25], in addition to our case (Table 1). Data showed that all the patients were either Caucasians or Asians, aged from 49 to 87 years and with male predominance.
With respect to their oncological diagnosis, five cases had metastatic non-small cell lung cancer (NSCLC), six cases had metastatic melanoma, two cases had metastatic head and neck squamous cell carcinoma (HNSCC), one case had Hodgkin lymphoma and one case had metastatic Merkel carcinoma.
For the ICIs, eight cases were on nivolumab, six cases were on pembrolizumab, and only one case was on atezolizumab. The time to diagnosis varied among patients and ranged between 4 and 36 weeks. In all patients, no latent TB testing (LTBT) before immunotherapy was done, and it was not clear whether TB infection is primary or secondary to latent TB reactivation. TB was microbiologically confirmed in all cases and followed by anti-TB drugs initiation. ICIs were maintained in three cases and discontinued or temporarily suspended in the remaining patients.
The time to diagnosis of TB in the current case occurred after six cycles of Pembrolizumab. TB was confirmed microbiologically by PCR and AFB. The patient received her BCG vaccine as part of the local child immunization program. Our case gave a history of sick contact with a patient with active TB infection 10 years ago, but there was no documentation of latent TB or previous TB infection prior to initiation of ICPs. The mixed response noted on 14th July 2019 PET CT (Fig. 4) was not perceived as pseudoprogression-like phenomenon as overt disease progression was confirmed by 26th August 2019 CT chest and abdomen as illustrated (Fig. 8).
ICIs were not resumed in our case and carboplatin plus pemetrexed was initiated instead, as second-line chemotherapy. None of the previously reported cases has used the traditional chemotherapy as a subsequent therapy to immunotherapy; nonetheless, the outcome of TB in patients receiving cytotoxic chemotherapy for malignancies have been reported in two retrospective studies in South Korea and Japan [28, 29]. In both studies, concurrent chemotherapy was found to be effective and safe for treating cancer patients with active MTB.
In a recent Meta-analysis including United States cancer patients, the risk of active TB was 41/100,000 [30], however, it is significantly higher in high prevalence areas such as South Korea with 3.07/1000 in patients with cancer [31]. It is worth mentioning that the global prevalence of latent TB infection in 2014 was estimated to be 23.0%, while the estimate for WHO Eastern Mediterranean Region which includes the state of Qatar was 16.3 [13.4–20.5] [32].
In 2012, the incidence of tuberculosis in Qatar was 41/100,000. The majority of infected patients (90%) was non-national males [33]. Whereas, pulmonary tuberculosis represents around 46% of active tuberculosis infection [34].