Non-steroidal anti-inflammatory drugs (NSAIDs) and TNFis are the most widely employed therapeutic agents in the treatment of axSpA and are therefore the most studied drugs in a real-life setting in terms of their potential association with an increased risk of malignancy. In contrast, only very few data are available outside of clinical trials on IL-17 inhibitors [8, 33, 34].

The relationship between TNFis and cancer was analyzed principally in studies involving patients with RA; nevertheless, thanks to data especially from the Scandinavian registers, no increased risk of cancer related to TNFi treatment has been found in patients with SpA [23, 35, 36]. Infliximab, etanercept, and adalimumab have been commonly used to treat SpAs [37].

Data from the Swedish and Danish healthcare cancer registries together with the ARTIS and DANBIO registries were used to evaluate the risk of malignancy in TNFi-treated patients with SpA: the experts concluded that the risk in TNFi-treated and TNFi-untreated SpA cohorts was similar, with a decreased incidence of colorectal neoplasm in TNFi-treated AS subjects, which, as in RA, could depend on the impact of long-term use of NSAIDs [24].

In 2018, an observational study gathered data regarding all solid tumors and lympho- or myeloproliferative malignancies of 3321 TNFi-treated patients with PsA, AS, or IBD-related SpA selected from the Italian GISEA registry between 2003 and 2015, and compared their incidence with the one reported by the Italian Association of Medical Oncology in an age and gender-matched general population. Univariate analysis revealed that coexisting (P = 0.012) and number of comorbidities (P < 0.001), and also age at the time of initiating TNFi treatment (P = 0.001), were all linked with a higher risk of cancer. Analysis by kind of disease pointed out that the incidence of malignancies was 6.8, 6.7, and 4.4 per 1000 patient-years in AS, enteropathic arthritis, and undifferentiated SpA, respectively. The study found no discrepancy in the incidence of cancer among patients treated with individual TNFis, but it was higher in comparison to the general population, in particular in subjects who had had a previous malignancy (HR 10.6, 95% CI 4.2–27, P < 0.001) [30].

Among individuals with PsA on TNFis, no raised risk of overall malignancy was detected. However, in a recent study from the British Society for Rheumatology Biologics Register, an elevated risk of NMSC was shown among those using TNFis in comparison with the general population (SIR 2.12, 95% CI 1.19–3.50) limited to women (SIR 2.41, 95% CI 1.10–4.58) but not men (SIR 0.85, 95% CI 0.51–1.35) [13, 38].

In addition to skin cancer, the ARTIS and DANBIO registries noted a higher risk of breast cancer in patients with PsA treated with TNFis than TNFi-naive subjects (RR 1.8, 95% CI 1.1–2.9) but not in comparison with the general population (RR 0.9, 95% CI 0.8–1.1) [25]. Lange et al. found a lower risk of prostate tumor in patients with PsA treated with TNFis compared with untreated (RR 0.4, 95% CI 0.2–0.8) and the general population (RR 0.4, 95% CI 0.2–0.8) [39].

As mentioned above, the reported risk of hematological cancer in PsA is variable and this is equally true when examining the association with therapies. In most studies, no higher risk of hematologic malignancies was observed in subjects with PsA on TNFis [8, 24].

In contrast, thanks to data from the CPRD database, a retrospective cohort study observed higher risk of hematologic malignancies in PsA (IRR 3.59, 95% CI 1.96–6.60) on cs/bDMARDs (MTX, sulfasalazine (SSZ), and adalimumab) [17]. Likewise, patients from four Nordic countries (Denmark, Finland, Sweden, and Iceland) included in large prospective cohort study reported an increased risk of HL and NHL in subjects with PsA ever treated with TNFis [17].

PsA treatment with MTX and/or SSZ was associated with a higher risk of lymphoma (pooled RR 1.29, 95% CI 1.04–1.60), but subgroup analyses of cancer types and therapeutic regimens showed remarkable heterogeneity [35].

PsA subjects treated with csDMARDs were at comparably higher risk of malignancies (pooled RR 1.75, 95% CI 1.40–2.18) than those treated with bDMARDs (pooled RR 0.957, 95% CI 0.80–1.14), in particular NMSC (pooled RR 2.46, 95% CI 1.84–3.28), but not malignant lymphoid/hematopoietic malignancies and related tissue cancers (pooled RR 1.17, 95% CI 0.93–1.46), urinary organ tumors (pooled RR 1.07, 95% CI 0.90–1.27), respiratory organ cancers (pooled RR 1.00, 95% CI 0.73–1.36), breast cancer (pooled RR 0.98, 95% CI 0.85–1.14), or digestive system tumors (pooled RR 0.82, 95% CI 0.62–1.09). Neither of these studies report prior phototherapy exposure [8].

Concerning other bDMARDs used in PsA, clinical trials during up to 5 years of IL-17 (secukinumab and ixekizumab) and IL-12/23 inhibitors (ustekinumab) demonstrated stable rates of severe adverse events and malignancies, consistent with previous reports describing a favorable safety profile following shorter durations of exposure [40, 41].

Recently, the therapeutic armamentarium for the management of SpAs has also been enriched by the new class of JAKis, which have been now licensed for the therapy of both PsA and axSpA.

The JAKis underwent a review process of their safety profile by the EMA and FDA between 2022 and 2023 following the publication of the results of a head-to-head study with non-inferiority design (ORAL-SURVEILLANCE) in which a JAKi (tofacitinib) was compared with TNFis in patients with RA at high CV risk (more than 50 years of age and at least one CV risk factor) [42]. The study failed both primary endpoints (risk for major adverse cardiovascular events (MACEs) and malignancy excluding NMSC), suggesting a possible increased cancer risk associated with the use of tofacitinib in this particular population. Although no similar studies have been conducted to date for the other JAKis (baricitinib, upadacitinib, and filgotinib) and for the other therapeutic indications for which they have been licensed, a warning about a potential oncological risk has been established by the EMA and FDA for the entire class of JAKis even when used for the treatment of PsA or axSpA.

The current recommendation of the regulatory bodies is that JAKis should not be used (other than as last-resort therapy) in patients with an increased cancer risk, which, however, is not well defined with the exception of smoking habit. Further real-life observational studies will obviously be necessary to shed light on what the real role of JAKis might be in the possible increased risk of malignancy in patients with PsA or axSpA (Fig. 3).

Differential risk analysis of drugs as risk factors for cancer in spondyloarthritis (SpAs) and types of malignancy associated with psoriatic arthritis (PsA). NMSC non-melanoma skin cancer, NSAIDs non-steroidal anti-inflammatory drugs, IL-17 interleukin-17