Venous ulcers and risk of occult hematological or other cancers: a nationwide cohort study

Sigrun A. J. Schmidt, Katalin Veres, Anne G. Ording, Dóra K. Farkas, Karsten Fogh and Henrik T. Sørensen

To the editor:

Hematological cancers are associated with various systemic changes, including leukostasis,1 plasma hyperviscosity,2 immunosuppression,3 and hypercoagulability.4,5 Because these factors have been implicated in the pathogenesis of chronic venous insufficiency,6,7 it is possible that hematological cancers increase the risk or severity of venous ulcers. This link is supported by 2 reports describing hematological cancer patients with leg ulcers that correlated in severity with changes in the concentration of platelets8 and monoclonal γ globulin.9 Clinical or subclinical venous thrombosis is another potential cause of venous ulceration in hematological as well as other cancers.4-7 Venous ulceration due to extrinsic compression of the lower-limb vessels by a large pelvic tumor has also been reported.10 Finally, cancer-induced malnourishment11 may contribute to impaired wound healing.12,13

Based on these mechanisms, we hypothesized that cancer increases the risk of venous ulcers even at an early stage of carcinogenesis. We therefore examined whether patients diagnosed with a venous ulcer are at increased risk of occult cancer, anticipating a particularly pronounced association for hematological cancers.

We conducted this nationwide cohort study using linked data from the Danish National Patient Registry (DNPR), the Danish Cancer Registry (DCR), and the Civil Registration System (CRS). Study setting, registries, and variable definitions are described in detail in the supplemental Appendix (available on the Blood Web site).

From the DNPR, we identified all cancer-free individuals with a primary or secondary diagnosis of a venous ulcer between January 1, 1982 and November 30, 2012. We considered the date of first inpatient admission, emergency room visit, or start of outpatient clinic follow-up to be the diagnosis date. Initial data extraction revealed some diagnoses among children, which raised concern about miscoding. Therefore, we restricted to patients aged ≥40 years, among whom the condition is most frequent.14 The DNPR also provided data on prior hospital-based diagnoses of conditions associated with venous ulceration, poor healing, or edema, including venous thromboembolism, diabetes, liver disease, renal disease, chronic obstructive pulmonary disease, obesity, hypertension, congestive heart failure, varicose veins, and postthrombotic syndrome.6,7,13 We classified patients without records of any of these conditions as having a “primary ulcer.” We used the Charlson Comorbidity Index15,16 to categorize comorbidity burden in the population as none (score 0), moderate (score 1), severe (score 2), and very severe (score ≥3).

We followed patients from venous ulcer diagnosis until a record of cancer in the DCR, death or emigration in the CRS, or end of study (November 30, 2013), whichever came first. We categorized cancers as hematological or nonhematological and according to individual sites (supplemental Table 1). Because of potential malignant transformation of cutaneous areas with persistent inflammation17,18 and because skin cancers may masquerade as venous ulcers,19 we specifically constructed a subgroup of skin cancers of the lower leg.

We used indirect standardization to compute standardized incidence ratios (SIRs) of cancer with 95% confidence intervals (CIs) according to national cancer incidence rates in subgroups of age, sex, and diagnosis year. As early diagnoses likely represent occult cancers, we examined the association according to arbitrarily chosen periods of time since venous ulcer diagnosis (<90 days, 90 days to 1 year, 1 to <4 years, 4 to <7 years, 7 to <10 years, and ≥10 years; <1 year and ≥1 year for individual sites). Furthermore, we stratified results by age at venous ulcer diagnosis, sex, type of venous ulcer diagnosis, and whether the ulcer was classified as primary. Because completeness of venous ulcer diagnoses likely increased following inclusion of outpatient data in the DNPR, we also stratified by calendar period (1982-1994 and 1995-2012).

Treating death as a competing risk, we calculated the 90-day, 1-year, and overall absolute risk of cancer. Assuming that cancers diagnosed during the first year were present at the time of ulcer diagnosis, we computed the number of patients needed to examine at the time of diagnosis in order to detect 1 excess cancer as the reciprocal of the excess risk.20 Finally, to avoid potential inclusion of venous ulcers diagnosed due to cancer-related hospital contact, we performed a sensitivity analysis using the date of discharge or last outpatient visit as the start date of follow-up.

We performed analyses in SAS. The study was approved by the Danish Data Protection Agency (record number 1-16-02-1-08). Approval by an ethical review board or informed consent from patients is not required in Denmark for registry-based studies.

We followed 27 954 patients with a venous ulcer diagnosis for a median of 4.8 years (interquartile range, 1.9-9.4 years). Patient characteristics are presented in supplemental Table 2. The 90-day SIRs were 3.50 (95% CI, 2.26-5.16) for hematological cancers and 1.98 (95% CI, 1.73-2.25) for nonhematological cancers (Table 1). For hematological cancers, the SIR remained elevated even after 10 years of follow-up. SIRs were increased for the majority of individual cancers within the first year (Table 2). The majority of hematological cancers diagnosed were B-cell neoplasms (supplemental Table 3). Sensitivity and subgroup analyses supported the overall results (supplemental Tables 4-5). Specifically, we observed increased SIRs in analyses restricted to patients with “primary ulcers.”

Table 1

SIR and 95% CIs for cancer among patients with a hospital-based diagnosis of a venous ulcer, Denmark, 1982-2012

Table 2

Site-specific SIR of cancer among patients with a hospital-based diagnosis of a venous ulcer, Denmark, 1982-2012

The 90-day absolute risk was 0.91% (95% CI, 0.80%-1.02%) for any cancer, 0.09% (95% CI, 0.06%-0.13%) for hematological cancers, and 0.82% (95% CI, 0.72%-0.93%) for nonhematological cancers. The number of patients with venous ulcer needed to examine to detect one excess cancer was 137 (95% CI, 103-205), 1117 (95% CI, 649-4018), and 156 (95% CI, 114-246) for any cancer, hematological cancers, and nonhematological cancers, respectively.

Our study was limited by potential misclassification of diagnoses and lack of detailed clinical data. Inclusion of inflammatory ulcers provoked by internal malignancy21,22 is possible, although only 0.4% of patients were diagnosed with pyoderma gangrenosum or cutaneous vasculitis within the first year of follow-up, indicating a low proportion of verified misdiagnoses. However, we did observe a high risk of nonmelanoma skin cancer of the leg (SIR, 8.08; 95% CI, 5.77-11.00) within the first year, which likely represents initially misdiagnosed tumors.

Frequent health care contacts among individuals with venous ulceration cannot be ruled out, although we found no evidence of a compensatory deficit after the initial increase during follow-up. The increased risk of several smoking- and alcohol-related cancers (eg, liver and lung cancer) might suggest confounding by lifestyle factors or inclusion of ischemic ulcers. However, SIRs were also increased for other cancer types and among patients with “primary ulcers,” suggesting that the identified underlying conditions cannot fully explain the observed associations.

Omission of patients who received only general practitioner– and nurse-delivered community care may have biased our estimates toward the null23 and could limit generalizability if the association depends on determinants of hospital referral (eg, ulcer severity). However, prevalence of comorbidities in our study population was similar to that reported among patients treated in primary care.24

In conclusion, we found an association between venous ulcer and occult cancer, in particular for cancers of the hematological system. However, the absolute risks were low, limiting the clinical implications for pursuing earlier cancer diagnosis in these patients. Nonetheless, the increased risk of skin cancers of the leg reinforces current guidelines on biopsying atypical ulcers. Furthermore, the strong association with occult hematological cancer is intriguing and supports our a priori hypothesis that hematological carcinogenesis affects wound development and healing. Research into the pathophysiology, prevention, and management of venous ulcers in patients with prevalent hematological cancer may thus be warranted.


Acknowledgments: This work was supported by the Program for Clinical Research Infrastructure established by the Lundbeck and the Novo Nordisk Foundations, the Danish Cancer Society, and Aarhus University. None of the funding sources had a role in the design, conduct, analysis, or reporting of the study. The Department of Clinical Epidemiology is, however, involved in studies with funding from various companies as research grants to (and administered by) Aarhus University. None of these studies have relation to the present study.

Contribution: H.T.S. conceived the study idea and is the guarantor; S.A.J.S. directed the analyses, which were performed by K.V. and D.K.F.; S.A.J.S., K.V., A.G.O., D.K.F., K.F., and H.T.S. participated in the discussion and interpretation of the results; S.A.J.S. wrote the initial manuscript draft; S.A.J.S., K.V., A.G.O., D.K.F., K.F., and H.T.S. critically revised the manuscript for intellectual content and approved the final version; and all authors participated in designing the study.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Sigrun A. J. Schmidt, Department of Clinical Epidemiology, Aarhus University Hospital, Olof Palmes Allé 43-45, DK-8200 Aarhus N, Denmark; e-mail: saj{at}


  • The online version of this article contains a data supplement.