Polymyalgia rheumatica (PMR) and giant cell arteritis (GCA) are two closely related diseases that frequently coexist. Although their pathophysiology is not fully understood, they are considered rheumatic diseases with similar characteristics, such as onset in patients over 50 years old, predominance in females, marked inflammation with elevated acute-phase reactants, and treatment with glucocorticoids and even interleukin-6 receptor inhibitors.1,2 In fact, there is a growing tendency that views both conditions as part of a single disease spectrum, constituting the giant cell arteritis-polymyalgia rheumatica spectrum disease (GPSD).3

Across Europe, GCA and PMR are frequent in older adults. In Spain, recent nationwide data report a GCA incidence of 7.42 (95% CI, 6.57–8.27) cases per 100,000 persons older than 50.4

The dysregulation of the immune system inherent in rheumatic diseases has been associated with an increased incidence of neoplasms.5 Diseases such as rheumatoid arthritis,6 systemic lupus erythematosus,7 or dermatomyositis8 have shown a consistent association with cancer development. The data on the GPSD are varied and contradictory, with studies showing both positive9–11 and negative12–14 associations. However, malignancies can mimic GPSD (particularly PMR) and must be ruled out in selected cases such as atypical presentation, inadequate response to therapy or refractory course.15

Specific characteristics of the GPSD, such as its partially unknown pathophysiology, onset in elderly individuals, and variability in chronicity and maintenance of a pro-inflammatory state, make establishing a causal relationship between GPSD and cancer even more challenging.

The term paraneoplastic syndrome describes symptoms derived from a neoplasm, not caused by direct tissue invasion, but rather by substances released by tumor cells.16 Paraneoplastic syndromes are typically thought to precede the onset of neoplasia by around two years,17 with cancer being diagnosed mainly within the first year after the clinical onset. The fact that studies have found that patients develop neoplasms predominantly within the first year following GPSD diagnosis has led to speculation that GPSD may be a paraneoplastic syndrome. PMR has been described as a manifestation of an underlying neoplasm that improves with treatment.18 However, a recent review found insufficient evidence to categorize PMR as a true paraneoplastic syndrome.19

Our study aims to describe the relationship between the GPSD and cancer by calculating the crude incidence ratio of neoplasms in our population and comparing it with the Spanish population. Additionally, we aim to interpret its potential role as a paraneoplastic syndrome, particularly regarding the time criterion (cancer incidence within two years of diagnosis).

We present a study of 334 Spanish patients diagnosed within the GPSD. This is one of the first studies on cancer and GPSD in the Spanish population from a Mediterranean region, which is often underrepresented in large-scale epidemiological studies because of the higher incidence in northern European populations.

A total of 16 neoplasms were diagnosed throughout the follow-up period. When comparing our neoplasm rate with that of the Spanish population, we found a higher number of neoplasms in women diagnosed within the GPSD compared to the general population. Although statistical significance was not reached in men or in the ratios calculated after the first two years post-diagnosis, we observed a tendency indicating more neoplasms in our population compared to the general population. None of the variables selected for the bivariate analysis and Poisson model showed a statistically significant association with the incidence of neoplasms.

The most common cancer types diagnosed in our sample were related to the digestive system and prostate cancer, which is consistent with the study by Muller et al.10 that also reported an increased number of prostate cancer cases.

In our sample, we observed two peaks of cancer detection: one at the time of diagnosis and another more than five years after diagnosis. The pathophysiology of these two entities is not clearly established; however, a significant proportion of patients achieve remission without treatment or with low doses of glucocorticoids. Therefore, we might hypothesize whether long-term cancers are a result of a chronic carcinogenic inflammatory state or are simply because we are dealing with an elderly patient population in which the development of neoplasms is expected. Indeed, of the five cancers that developed beyond five years of follow-up, four occurred in patients over 80 years of age.

There are studies examining the relationship between GPSD and cancer, some focusing on the entire spectrum,9,13,14 others only on PMR,10,11,20 and others only on GCA,12,21 using registry data or cohorts from the same study centers as comparator groups. Recently, there has been a growing tendency to consider PMR and GCA as part of the same disease spectrum. Our study considers GPSD in terms of its initial phenotype, with none of the phenotypes being linked to cancer development. As expected, given its higher prevalence, 92% of patients presented a polymyalgic phenotype at disease onset, which was associated with 93.7% of cancer cases.

One of the aims of our study was to compare the incidence of neoplasms in our sample with the reference Spanish population. After adjusting for sex and age, we obtained a standardized neoplasm ratio for women of 4.03 (95% CI 1.53–6.53), showing a statistically significant higher number of neoplasms in the GPSD population compared to the Spanish population. Although statistical significance was not achieved, neoplasm rates were numerically higher in the case of men and during the first two years after diagnosis. Nearly 70% of the cancer cases were diagnosed within the first two years of follow-up, especially during the first year. This could be due to more in-depth examinations in patients presenting other abnormalities after an initial contact, such as unexplained iron deficiency anemia or elevated prostate-specific antigen levels. On the other hand, this may also introduce a diagnostic lead-time bias. So, we must consider the possibility that our patients underwent closer follow-up and more thorough evaluation compared to the reference population.

These findings are consistent with other studies showing a positive association between GPSD and neoplasms.9–11 This has clinical relevance and implications for clinical practice, with physicians needing to consider the necessity of cancer screening from the time of diagnosis. The rheumatologist played a direct role in the diagnosis of 5 (31.3%) of the recorded neoplasms by requesting additional tests and referring the patients to other specialists.

Even though the chronic inflammation associated with GPSD in the long term could contribute to the development of neoplasms, numerous studies have discussed the possibility that GPSD might be a paraneoplastic syndrome. The fact that cancer cases frequently appear within the first year of follow-up in many studies9,10 could reinforce the idea that in some cases, GPSD acts as a clinical manifestation of an underlying occult neoplasm, and we may, in fact, be dealing with prevalent, not incident, cancer cases. Furthermore, it has been suggested that the paraneoplastic role of PMR is especially linked to hematologic neoplasms,16 which are less prominent in our series (only one case). Nevertheless, we did not find sufficient evidence to affirm that GPSD is indeed a paraneoplastic syndrome. The rapid improvement of patients with accessible, effective, and inexpensive treatment complicates the design of studies to differentiate between causality and coincidence.19 Our findings suggest that GPSD may not strictly fulfill the criteria for paraneoplastic syndrome but warrant further study regarding its long-term cancer risk.

The limitations of our study are mainly due to its retrospective nature, with the possible loss of associated information. Additionally, our sample size is relatively small compared to larger registries, which makes it difficult to analyze specific types of cancer associated with GPSD. This fact reduces statistical power and makes it difficult to draw definitive conclusions. Prior studies with larger samples9,10 provide essential comparative data, highlighting the need for further studies with greater statistical power. Therefore, while our findings aligned with previous reports suggesting an increased risk of cancer in GPSD, we emphasize that these results should be interpreted with caution due to the small sample size.

Another potential limitation of the study is the COVID-19 pandemic. GLOBOCAN's estimates rely on pre-pandemic data, which may affect the reliability of their estimates.22 The pandemic affected the number of cancer diagnoses, so the actual number may be underestimated, potentially altering the estimates. In our sample, eight cancers were diagnosed before March 15, 2020 (the start of Spain's lockdown), and eight were diagnosed afterward. Among cases that developed cancer in the long term (more than five years), two patients were diagnosed with cancer before the pandemic, while three were diagnosed with GPSD before the pandemic and developed cancer posteriorly. It is possible that a diagnostic delay occurred in these patients.

On the other hand, even though our study does not have a comparator group from the same center, it does combine individual case review of clinical histories with Spanish general data provided by GLOBOCAN.

Even though in other regions, patients with GPSD, especially those with PMR, are treated in primary care and not followed by a rheumatologist, in our center, they are usually referred from the moment of diagnostic suspicion. A possible limitation of our approach could be the clinical heterogeneity inherent to considering the entire spectrum. However, despite the potential heterogeneity of the sample, in our study, patients with PMR are typically diagnosed by a rheumatologist due to this referral system. Therefore, we included the entire spectrum, as patients are evaluated, followed, and classified within it by a rheumatologist.

Future studies with larger sample sizes could confirm the possible relationship between neoplasms and GPSD and help to understand the underlying mechanisms.

Our finding of a higher incidence of neoplasms in women compared to the general population and the fact that most cancer cases appear within the first year after diagnosis should raise awareness about the potential presence of an underlying neoplasm in some cases. Therefore, close follow-up should be conducted to enable early cancer detection, especially during the first two years.

No fundings declared.

The authors declare that they have no conflicts of interest.

During the preparation of this work the author) used ChatGPT (OpenAI) in order to improve readability and to support translation into English. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the published article.