International Journal of Rheumatic Diseases 2014
ORIGINAL ARTICLE
Diagnostic dilemma of paraneoplastic arthritis: case series Bunyamin KISACIK,1 Ahmet M. ONAT,1 Timucin KASIFOGLU,2 Yavuz PEHLIVAN,1 Omer N. PAMUK,3 Ediz DALKILIC,4 Salim DONMEZ,3 Sule Y. BILGE,2 Sedat YILMAZ,5 Hakan ERDEM,5 Ridvan MERCAN,6 Mehmet A. OZTURK,6 Cemal BES,7 Mehmet SOY,7 Sukran ERTEN,8 Veli COBANKARA,9 Soner SENEL,10 Fatma A. ONER,11 Haner DIRESKENELI,11 Sema YILMAZ,12 Ayten YAZICI,13 Hakan EMMUNGIL,14 Kenan AKSU,14 Seval KUL,1 Gozde Y. CETIN15 and Mehmet SAYARLIOGLU15 Departments of Internal Medicine, Division of Rheumatology, 1Gaziantep University Faculty of Medicine, Gaziantep, 2Osmangazi University Faculty of Medicine, Eskisehir, 3Trakya University Faculty of Medicine, Edirne, 4Uludag University Faculty of Medicine, Bursa, 5Gulhane Military School of Medicine, Ankara, 6Gazi University Faculty of Medicine, Ankara, 7AbantIzzetBaysal University of Medicine, Bolu, 8Ministry of Health, Ataturk Hospital, Division of Rheumatology, Ankara, Departments of Internal Medicine, Division of Rheumatology, 9Pamukkale University Faculty of Medicine, Ankara, 10Cumhuriyet University Faculty of Medicine, Denizli, 11Marmara University Faculty of Medicine, Istanbul, 12Selcuk University Selcuklu Faculty of Medicine, Konya, 13Division of Rheumatology, Sakarya Research and Training Hospital, Sakarya, Departments of Internal Medicine, Division of Rheumatology, 14 Ege University Faculty of Medicine, Izmir, and 15Sutcu Imam University Faculty of Medicine, Kahramanmaras, Turkey
Abstract Objectives: Paraneoplastic arthritis (PA) may mimic rheumatic diseases. While presenting the demographic and laboratory features of the patients diagnosed with PA, this study also aims to provide possible appropriate tools to differentiate the PA cases from early rheumatoid arthritis (ERA). Methods: Sixty-five patients with PA (male/female: 43/22) from 15 different rheumatology clinics and 50 consecutive patients with ERA (male/female: 13/37) fulfilling the 2010 American College of Rheumatology (ACR) criteria for the diagnosis if the RA from Gaziantep Rheumatology Early Arthritis Trial (GREAT) as controls who were diagnosed at least 12 months before, were enrolled into study. Results: Mean ages of the patients with PA and ERA were 50.2 15.3, and 42.7 12.3, respectively, and the mean ages of the patients with PA were significantly higher than the ERA. Unlike the ERA patients, in our case series PA was predominantly observed among males. Oligoarthritis was significantly higher in solid tumors in contrast to ERA (P = 0.001). Polyarthritis and symmetric arthritis were significantly higher in the ERA group in contrast to all malignancies (P = 0.001). Rheumatoid factor (RF) and anticyclic citrullinated peptide antibody (anti-CCP) positivity were significantly higher in the ERA group (each P = 0.001). Lactic dehydrogenase levels of hematologic malignancies were significantly higher than other groups (each, P = 0.001). Conclusions: ERA patients had more symmetric joint involvement than PA; laboratory markers could be also an alternative where there is high RF and anti-CCP positivity with antibody levels among the ERA patients. Finally, the demographic features can be used as differentiating factors; ERA was seen predominantly among females aged 40–59 years which refers to young adults. Key words: early rheumatoid arthritis, malignancy, paraneoplastic arthritis.
Correspondence: Dr. Bunyamin Kisacik, MD, Gaziantep Universitesi Tip Fak€ ultesiRomatolojiUnitesi, 27100 Sahinbey, Gaziantep, Turkiye. Email:
[email protected]
INTRODUCTION Paraneoplastic syndromes (PS) are a group of rare disorders associated with malignant diseases, particularly
© 2014 Asia Pacific League of Associations for Rheumatology and Wiley Publishing Asia Pty Ltd
B. Kisacik et al.
independent from the location and the size of the tumor. The disease is more often among older patients, and generally coexists with cancers of the lung, breast, ovaries or lymphatic nodes. Clinical, radiological and laboratory findings play important roles in the diagnosis of PS. The initiation of PS has to be limited in the first 2 years before the diagnosis of cancer.1 Mostly the symptoms of the PS appear before the malignancy and they may mimic various diseases, including rheumatic and skin related disorders. The pathogenesis is still unknown; however, hormones, cytokines, immunoglobulin and several mediators are blamed in the etiology. The association of the rheumatologic diseases and malignancies are well known, such as dermato/polymyositis, hypertrophic pulmonary osteo-arthropathy and systemic sclerosis. As in other diseases, the musculoskeletal manifestations of malignancy may be observed before, during or after the diagnosis of cancer.2 Occasionally, the symptoms of PS may mimic nearly all of the rheumatologic diseases. Paraneoplastic arthritis (PA) also known as cancer-associated arthritis is one of the most prevalent and important ones that should be kept in mind among the other paraneoplastic syndromes. Today, there is not an agreed definition for early rheumatoid arthritis (ERA) but it is defined as ‘a disease with symptoms of less than 3 years’ in randomized clinical studies.3 We herein focused our attention: (i) to present the demographic and laboratory features of patients diagnosed with PA; and (b) on how we can differentiate the PA cases from ERA?
MATERIALS AND METHODS The inclusion criteria for the diagnosis of PA were defined with a literature review. Patients aged over 45 years, rapid disease onset, symmetric or asymmetric disease mostly affecting the lower extremities and usually sparing the small joints of the hands at least 2 years earlier than the diagnosis of PA, supporting laboratory findings, poor or no response to diseasemodifying anti-rheumatic drugs (DMARDs) prior to the detection of malignancies, were the agreed criteria. After a retrospective review of data on file within the last 5 years from January 2007 to January 2011, 75 patients meeting the agreed inclusion criteria for PA were reported. All cases were reviewed by all researchers involved and 10 reported cases were excluded from the study because of not meeting the inclusion criteria. Sixty-five cases (male/female: 43/22) from 15 different rheumatology clinics were finally randomized
2
to the study. Most of the PA cases had been misdiagnosed as ERA; this why we preferred to choose ERA patients as controls. In order to compare clinical and laboratory manifestations of malignancies and early arthritis, 50 consecutive patients with ERA (male/ female: 13/37) fulfilling the 2010 American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) criteria for the diagnosis of RA4 from Gaziantep Rheumatology Early Arthritis Trial (GREAT) as controls who were diagnosed at least 12 months before, were also enrolled into the study. In order to categorize similar malignancies, we divided the PA group into two sub-groups, as hematologic malignancies and solid tumors. Patients’ duration of symptoms, final diagnosis, coexisting musculoskeletal symptoms, rheumatologic markers and other laboratory findings, including anti-nuclear antibody (ANA), anti-cyclic citrullinated peptide (anti-CCP), rheumatoid factor (RF), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) and lactate dehydrogenase (LDH) were recorded to a special form generated for this study. This study was approved by the local Ethics Committee.
Statistical analysis The Statistical Package for Social Sciences (SPSS) was used to analyze the data (SPSS Inc., Chicago, IL, USA). One-way analysis of variance (ANOVA) and post-hoc tests were used for multiple comparisons. In addition, the chi-square and Fisher’s exact tests were used for categorical variables, and Student’s t-test was used to compare mean values. P-values < 0.05 were considered as statistically significant.
RESULTS Mean ages of the patients with PA and ERA were 50.2 15.3, and 42.7 12.3 years, respectively and the mean ages of the patients with solid tumors were significantly higher than the ERA (55.3 11.6 vs. 42.7 12.3, respectively). Unlike the ERA patients, in our case series the PA were predominantly observed among males. Demographic, clinical and laboratory findings of all groups are presented in Table 1. Lung cancer was the most frequently observed malignancy among the solid tumors and leukemia was the most frequent type among hematological cancers (Tables 2,3). While PA patients’ responses to DMARDs were very limited and most of the cases needed prolonged steroid doses, the ERA patients responded adequately to conventional DMARDs.
International Journal of Rheumatic Diseases 2014
Diagnostic dilemma of paraneoplastic arthritis
Table 1 Demographic and laboratory features of all groups
Age (years) Gender (male/female) Polyarthritis Oligoarthritis Monoarthritis Symmetric arthritis Vasculitic lesion The time between onset of symptom and diagnosis ESR (mm/h) CRP (mg/L) LDH (U/L) ANA positivity RF positivity Anti-CCP positivity
All malignancies (n = 65)
Hematologic malignancies (n = 26)
Solid tumors (n = 39)
ERA (n = 50)
P
50.2 15.3 43/22 22 31 12 6 5 5.1 4.8
42.4 16.9 17/9 10 10 6 2 2 4.8 5.4
55.3 11.6 26/13 12 21 6 4 3 5.4 4.4
42.7 12.3 13/37 45 5 0 40 0 5.5 3.3
< 0.05* NA < 0.05** < 0.05***
58.5 31.9 65.1 85.5 611 1126 10 15 7
55 34.2 76.2 115.4 943 1626 4 4 1
60.9 30.6 57.1 55.4 345 239 6 11 6
31.5 22.1 17.3 24 311 108 7 24 29
NA NA < 0.05+ NA < 0.05 + + < 0.05 + ++
NA
*Hematological malignancies vs. solid tumors P = 0.004, **polyarthritis of rheumatoid arthritis vs. other malignancies P < 0.001, ***oligoarthritis and monoarthritis of malignancies groups significantly higher than ERA P = 0.001, P = 0.009 respectively, +LDH levels of hematological malignancies significantly higher than other groups P = 0.003, ++, +++ RF and anti-CCP significantly higher compared malignancies group each P < 0.001. ERA, early rheumatoid arthritis; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; LDH, lactic dehydrogenase; ANA, antinuclear antibodies; RF, rheumatoid factor. Anti-CCP, anticyclic citrullinated peptide antibody.
Table 2 Hematologic malignancies and histology
Table 3 Solid tumors types in detail and histology
Hematologic malignancies (n = 26)
Solid Tumors (n = 39)
Histology
Lung cancer (17)
Adenocarcinoma (7), epidermoid (3), unknown (5), bronchogenic (1), small cell carcinoma (1) Adenocarcinoma (2), unknown (6) Transitional cell carcinoma (1), unknown (2) Adenocarcinoma (2) Unknown (2) Adenocarcinoma
Acute myeloblastic leukemia (7) Acute lymphoblastic leukemia (4) Chronic lymphoblastic leukemia (1) Myelodisplastic syndrome (4) Lymphoma (7)
Histology M3 (2), M4 (2), M5 (3)
Hodgkin lymphoma (1) Non-Hodgkin lymphoma (6)
Multiple myeloma (3)
Concerning the rheumatologic findings, oligoarthritis was significantly higher in solid tumors and hematological cancers in contrast to ERA (P = 0.001, P = 0.02, respectively). The allocation of monoarthritis was similar in both malignancy groups; however, it was significantly lower among the ERA patients (each, P = 0.001). Polyarthritis was significantly higher in the ERA group in contrast to all malignancies (P = 0.001). Excepting the significantly higher oligoarthritis in hematological malignancies, patients with hematologic malignancies and solid tumors had similar joint involvement profiles (P > 0.05). Joint involvements in all malignancies are shown in Table 4. The time between the onset of symp-
International Journal of Rheumatic Diseases 2014
Breast cancer (8) Urinary bladder cancer (3) Prostate cancer (2) Pancreas cancer (2) Endometrial cancer (1) Bone cancer (1) Colon cancer (1) Coledoc Cancer (1) Thyroid Cancer (1) Renal cell cancer (1) Primary unknown (1)
Sternal chondrosarcoma Adenocarcinoma Unknown Papillary carcinoma Adenocarcinoma
tom and diagnosis was 5.1 4.8 months for PA (hematologic malignancies 4.8 5.4, solid tumors 5.4 4.4) and 5.5 3.3 months for ERA. The time between the onset of symptom and diagnosis was not significant (all P > 0.05). There was not a valuable difference in terms
3
B. Kisacik et al.
Table 4 Joint involvement in malignancies
Hand and wrist Elbow Shoulder Ankle Knee Hip
All of paraneoplastic group n = 65
Hematologic malignancies n = 26
28 (43.1) 8 (12.3) 2 (3.1) 35 (53.8) 39 (60) 1 (1.5)
13 (50) 4 (15.4) 1 (2.6) 11 (42.3) 16 (61.5) 1 (3.8)
Solid tumors n = 39 15 (38.5) 4 (10.3) 1 (3.8) 24 (61.5) 23 (59) 0
of symmetric involvement between the hematologic malignancies and solid tumors (7.6% vs. 10.2). The laboratory findings of ERA compared with PA, RF and anti-CCP positivity were significantly higher in ERA group (each P = 0.001). The CRP and ESR levels of almost all PA patients were elevated and it was significantly higher than the ERA group (P < 0.0001). Moreover, ANA positivity was 15.4% (14% in ERA patients, P = 0.810), RF positivity was 23.1% (48% in ERA patients, P = 0.001) and anti-CCP positivity was 10.8% (58% in ERA patients, P = 0.001). Lactate dehydrogenase levels of the hematologic malignancies group were significantly higher than the solid tumors and ERA group (each, P = 0.001). However, the difference between ERA and solid tumor groups concerning the serum LDH levels was not statistically significant.
DISCUSSION Musculoskeletal manifestations of malignancy may mimic rheumatic diseases. This is why the accuracy of diagnosis in all diseases is vitally important for patients and physicians. To-date, there have been different smaller case series of paraneoplastic arthritis reported in the literature; however, the presented reports do not provide sufficient data to reach relevant differentiating symptoms between PA and ERA. To the best of our knowledge, the present study includes the largest cases series reported in the literature. This article provides patients diagnosed with malignancy and whose baseline complaints were mimicking ERA-like symmetric and asymmetric polyarthritis, oligoarthritis and monoarthritis. Leukemia, lymphoma and lung cancer were the leading diseases in the PA group. While leukemia and lymphomas were concurrently seen with symmetric arthritis, solid tumors were simultaneously observed with asymmetric mono-oligoarthritis. After a comprehensive analysis, we have observed
4
some parameters to differentiate ERA from PA. First, increased LDH levels might be a valuable feature to differentiate hematological malignancies from other cases. Second, ERA patients were differentiated from PA in terms of joint involvement; the ERA patients had more symmetric joint involvement. Third, laboratory markers, including RF and anti-CCP positivity, could also be alternative tools to differentiate ERA from PA. Finally, demographic features can be used as differentiating factors; ERA was seen predominantly among females aged 40–59 years, who are commonly regarded as young adults. Musculoskeletal symptoms might especially be observed in hematologic malignancies. Morel et al. showed in their PA series that solid tumors had predominantly been seen more commonly than hematologic malignancies (solid tumors 77%, hematologic malignancies 23%); additionally, most of their PA patients had lung cancer, leukemia and lymphoma.5 The causes of PA could be diverse in adult and pediatric subgroups. Goncßalves et al. investigated pediatric PA patients and they concluded that most of their patients had leukemia; however, solid tumors are very rare in their study population.6 In our PA group solid tumors had a frequency of 58.1% and hematological malignancies were 41.9%. Similarly leukemia, lymphoma and lung cancers were the most frequent disorders in the present study. Additionally, breast cancer and myelodisplastic syndrome were also common. Paraneoplastic syndromes are rarely comorbid with rheumatic diseases, such as remitting seronegative symmetrical synovitis with pitting edema (RS3PE) syndrome and polymyalgia rheumatica (PMR). However, both in RS3PE and PMR, symptoms are more atypical and patients had poor steroid response. Further, none of the cases had been diagnosed in our case series.7,8 Unless a patient whose arthritis onset presented with findings suggestive for a malignancy, a detailed screening for cancer is not recommended.1,9,10 Therefore, clues indicating PA such as older age, asymmetric arthritis, especially in the lower limbs, and aggressive onset and/or absence of RF, should be carefully evaluated. We have found that while most of the hematologic malignancy subgroup presented with symmetric arthritis, the solid tumor patients had oligoarthritis. Moreover, the onset of PA was acute in all patients with malignancies. The elevation of acute phase reactants is a characteristic of rheumatic diseases. However, they could also be increased in patients with malignancies and the discrimination of both conditions depending on the acute phase reactants is unreliable and our study has similar
International Journal of Rheumatic Diseases 2014
Diagnostic dilemma of paraneoplastic arthritis
results with the literature. On the other hand, tumor markers could be nonspecific in the diagnosis of early cancers and these markers might also be positive in rheumatologic diseases. LDH levels have been recognized as a marker of cell turnover and its secretion is elevated in liver diseases, myocardial infarction and hemolysis.10 Goncßalves et al. and Wallendal et al. have shown that LDH is increased in malignant arthritis among infant patients.6,11 In our study we observed that elevated LDH level was a useful marker for differentiation of malign hematological diseases in contrast to solid tumors and ERA. Therefore, it has to be kept in mind that higher LDH levels might indicate hematological malignancies in patients with arthritis. Arthritis is the leading finding and has been reported rarely at the onset of these diseases.12 The mechanism of PA in lympho-proliferative disorders is the direct synovial involvement or a reaction of the adjacent tissue. Leukemia has a higher prevalence of PA than other hematological malignancies.13 The presence of joint swelling in leukemia is more frequent in children than in adults, and in acute rather than in chronic forms.14 Goncßalves et al. found 0.25% of malignancies among 3528 pediatric cases whose first admission were musculoskeletal symptoms. Leukemia and lymphoma were predominant in their study.6 Gur et al. revealed that 5.8% of the leukemia patients presented with arthritis and complaints were observed 3.2 months prior to the diagnosis of leukemia.15 In our series, the patients were diagnosed with malignancy after 4.6 months following the emergence of PA. Besides other findings, the present study prevails with the number of PA cases having anti-CCP positivity. Up to date there have been only two case reports demonstrating anti-CCP antibody positive paraneoplastic syndromes. Larson et al. and Kumar et al. reported cases of paraneoplastic polyarthritis with positive anti-CCP antibody test, which had not been reported before.16,17 In our case report we have presented seven malignant cases out of 65 (10.7%) having anti-CCP positivity. Since this article primarily attempts to demonstrate differentiating parameters and demographic features of our study, we were unable to provide treatment regimens of patients from different clinics, and this could be regarded as a limitation. Consequently, paraneoplastic syndromes may mimic rheumatologic diseases. In order to minimize the confusion for patients with arthritis, RF and anti-CCP negativity, atypical rheumatologic symptoms, acute onset of the complaints, and elevated LDH levels can be important to differentiate PA patients. Although this
International Journal of Rheumatic Diseases 2014
is the largest case series in patients with PA, still the number of patients is not enough to generalize conclusions. Further studies with larger cohorts are needed to verify the findings of our study.
CONFLICT OF INTEREST The authors have conflicts of interest to declare regarding this work.
REFERENCES 1 Butler RC, Thompson JM, Keat AC (1987) Paraneoplastic rheumatic disorders: a review. J R Soc Med 80 (3), 168–72. 2 Fam AG (2000) Paraneoplastic rheumatic syndromes. Baillieres Best Pract Res Clin Rheumatol 14 (3), 515–33. 3 Amjadi S, Khanna D, Park GS, Bulpitt KJ, Wong WK, Paulus HE (2004) Dating the “window of therapeutic opportunity” in early rheumatoid arthritis: accuracy of patient recall of arthritis symptom onset. J Rheumatol 31 (9), 1686–92. 4 Kay J, Upchurch KS (2012) ACR/EULAR 2010 rheumatoid arthritis classification criteria. Rheumatology (Oxford) 51 (Suppl. 6), vi5–9. 5 Morel J, Deschamps V, Toussirot E et al. (2008) Characteristics and survival of 26 patients with paraneoplastic arthritis. Ann Rheum Dis 67 (2), 244–7. 6 Goncßalves M, Terreri MT, Barbosa CM, Len CA, Lee L, Hilario MO (2005) Diagnosis of malignancies in children with musculoskeletal complaints. Sao Paulo Med J 123 (1), 21–3. 7 Brooks RC, McGee SR (1997) Diagnostic dilemmas in polymyalgia rheumatica. Arch Intern Med 157 (2), 162–8. 8 Schaeverbeke T, Fatout E, Marce S et al. (1995) Remitting seronegative symmetrical synovitis with pitting oedema: disease or syndrome? Ann Rheum Dis 54 (8), 681–4. 9 Naschitz JE, Yeshurun D, Rusher I (1995) Rheumatic manifestations of occult cancer. Cancer 75, 2954–8. 10 Marshall T, Williams J, Williams KM (1991) Electrophoresis of serum isoenzymes and proteins following acute myocardial infarction. J Chromatogr 569 (1–2), 323–45. 11 Wallendal M, Stork L, Hollister JR (1996) The discriminating value of serum lactate dehydrogenase levels in children with malignant neoplasms presenting as joint pain. Arch Pediatr Adolesc Med 150 (1), 70–3. 12 Ueno Y, Manabe T, Shimizu S (1995) Non-Hodgkin’s lymphoma and polyarthritis. Br J Rheumatol 34, 293–7. 13 Evans TI, Nercessian BM, Sanders KM (1994) Leukemic arthritis. Semin Arthritis Rheum 24, 48–56. 14 Jonsson OG, Sartain P, Ducore JM, Buchanan GR (1990) Bone pain as an initial symptom of childhood lymphoblastic leukaemia: association with nearly normal hematologic indexes. J Pediatr 117, 233–7. 15 Gur H, Koren V, Ehrenfeld M, Ben-Bassat I, Sidi Y (1999) Rheumatic manifestations preceding adult acute leukemia:
5
B. Kisacik et al.
characteristics and implication in course and prognosis. Acta Haematol 101 (1), 1–6. 16 Larson E, Etwaru D, Siva C, Lawlor K (2011) Report of anti-CCP antibody positive paraneoplastic polyarthritis and review of the literature. Rheumatol Int 31 (12), 1635–8.
6
17 Kumar S, Sethi S, Irani F, Bode BY (2009) Anticycliccitrullinated peptide antibody-positive paraneoplastic polyarthritis in a patient with metastatic pancreatic cancer. Am J Med Sci 338 (6), 511–2.
International Journal of Rheumatic Diseases 2014
