Association of anti-RNA polymerase III autoantibodies and cancer in scleroderma

Abstract

Introduction: We assessed the profile and frequency of malignancy subtypes in a large single-centre UK cohort for patients with scleroderma (systemic sclerosis; SSc). We evaluated the cancer risk among SSc patients with different antibody reactivities and explored the temporal association of cancer with the duration between SSc onset and cancer diagnosis.

Methods: We conducted a retrospective study of a well-characterised cohort of SSc patients attending a large tertiary referral centre, with clinical data collected from our clinical database and by review of patient records. We evaluated development of all cancers in this cohort, and comparison was assessed with the SSc cohort without cancer. The effect of demographics and clinical details, including antibody reactivities, were explored to find associations relevant to the risk for development of cancer in SSc patients.

Results: Among 2,177 patients with SSc, 7.1% had a history of cancer, 26% were positive for anticentromere antibodies (ACAs), 18.2% were positive for anti-Scl-70 antibodies and 26.6% were positive for anti-RNA polymerase III (anti-RNAP) antibody. The major malignancy cancer subtypes were breast (42.2%), haematological (12.3%), gastrointestinal (11.0%) and gynaecological (11.0%). The frequency of cancers among patients with RNAP (14.2%) was significantly increased compared with those with anti-Scl-70 antibodies (6.3%) and ACAs (6.8%) (P < 0.0001 and P < 0.001, respectively). Among the patients, who were diagnosed with cancer within 36 months of the clinical onset of SSc, there were more patients with RNAP (55.3%) than those with other autoantibody specificities (ACA = 23.5%, P < 0.008; and anti-Scl-70 antibodies = 13.6%, P < 0.002, respectively). Breast cancers were temporally associated with onset of SSc among patients with anti-RNAP, and SSc patients with anti-RNAP had a twofold increased hazard ratio for cancers compared to patients with ACAs (P < 0.0001).

Conclusions: Our study independently confirms, in what is to the best of our knowledge the largest population examined to date, that there is an association with cancer among SSc patients with anti-RNAP antibodies in close temporal relationship to onset of SSc, which supports the paraneoplastic phenomenon in this subset of SSc cases. An index of cautious suspicion should be maintained in these cases, and investigations for underlying malignancy should be considered when clinically appropriate.

Figure 1

Graph showing the frequency of different cancer subgroups for each major systemic sclerosis-specific autoantibody subtype (anticentromere, anti-Scl-70 and anti-RNA polymerase III antibodies). ACA, Anticentromere antibody; GIT, Gastrointestinal; GU, Genitourinary; Gynae, Gynaecological; Haemato, Haematological; RNAP, RNA polymerase III. Other cancers include thyroid and brain.

Figure 2

Kaplan-Meier analysis shows three different curves for each patient group with the designated autoantibody subset. Events (defined as diagnosis of cancer) correspond to step-downs, and censored observations (defined as most recent follow-up visit) are identified by tick marks. The plot shows a significant difference (P < 0.0001) between the number and timing of events between anti-RNA polymerase III (anti-RNAP)–positive patients and anticentromere antibody (ACA)– or Scl-70-positive patients. The dotted vertical line (designated as time 0) represents the clinical onset of systemic sclerosis (SSc). Within the 36-month period prior to the onset of SSc, 13 patients with the anti-RNAP antibody, compared with only 2 patients with ACA, were diagnosed with cancer. Number at risk, number and percentage loss to follow-up are represented for the intervals between SSc onset and the 20- and 40-year periods.

Figure 3

Graphs illustrating temporal relationship of cancers (including all cancers and breast cancers) for all three major antibody reactivities and anti-RNA polymerase III antibody. (A) Frequency of all cancers (n = 129) across all three major antibody reactivities (anti-Scl-70, anticentromere and anti-RNA polymerase III (RNAP) antibodies). (B) Frequency of all cancers for anti-RNAP antibody alone (n = 38). (C) Frequency of breast cancers (n = 60) for all three major antibody reactivities. (D) Frequency of breast cancers (n = 19) for anti-RNAP antibody.