Hypertensive nephropathy is associated with an increased risk of myeloma, skin, and renal cancer

Abstract

Previous studies suggest an increased cancer risk in hypertension. Patients with hypertensive nephropathy have not been studied. A national registry study was performed to assess the presence and size of this association. Clinical data and cancer diagnoses for all patients with biopsy‐proven hypertensive nephropathy between 1985 and 2015 in Denmark were extracted from four national registries and compared with age‐ and sex‐adjusted national cancer rates. The risk of cancer was twice the background population. It was raised for renal cancer (odds ratio 10.4), myeloma (13.2), skin cancer (7.9), and other/unspecified (1.8). No increase in incidence was seen until 1 year before renal biopsy and then rose rapidly. It was again normal 5 years after biopsy. Hypertensive nephropathy is associated with an increased risk of myeloma, skin, renal, and other cancers. Screening of patients with hypertensive nephropathy, in the presence of reduced renal function or significant proteinuria, may be indicated.

1. BACKGROUND

Previous studies have suggested an increased risk of cancer in patients with hypertension,1, 2, 3, 4, 5, 6, 7, 8 though not all.9 Increased risks have primarily been described for renal cancer,2, 3, 4, 5, 7, 8 but also skin, head, and neck.10 The association of several forms of cancer with glomerulonephritis is well known, and detailed reviews of the subject have been published.11, 12, 13, 14, 15 There is an increased risk of renal function deterioration in hypertension,16 characterized by the presence of hypertensive nephropathy on renal biopsy. This association is presumed to be causal, though this assumption has been challenged.17 We hypothesized that the common factor in these associations was the presence of renal involvement and that hypertensive nephropathy would be associated with increased cancer risk. The existence of a large national renal biopsy registry, covering a period of 30 years, permits an analysis of this possible link.

2. METHODS

All patients with biopsy‐proven hypertensive nephropathy resident in Denmark between the years 1985 and 2014 were included. The great majority of patients were Caucasian.

The study was an observational study in epidemiology and followed the STROBE guidelines for reporting observational studies.18

2.1. Renal biopsies

The renal biopsy information was derived from two registries:

1. Danish Renal Biopsy Registry. This registry recorded all biopsies performed in Denmark between 1985 and 1999.19

2. Since 2000 renal biopsy results have been registered by the National Pathology Data Bank (Patobank).

There were two renal biopsy registries. There was no overlap between the two registries. All SNOMED diagnoses from both registries were imported, along with available indications for renal biopsy, patient identity number, and biopsy date. The following SNOMED codes were included: M53300 (glomerulosclerosis), M52200 (arteriosclerosis), and M2210 (arteriosclerosis with fibrinoid necrosis). All biopsies with other specific diagnoses as well, for example, glomerulonephritis and diabetic nephropathy were excluded. For each patient, only the first biopsy was included. Patients <15 years old were excluded.

2.2. Cancer diagnoses

Using the patient national identity number, these patients were linked with all registered cancer diagnoses in the Danish Cancer Registry, which has been in existence since 1943. The Danish Cancer Registry has a very high degree of validity and completeness.20 ICD‐8 diagnoses were converted to the relevant ICD‐10 diagnoses. The following diagnosis groups were used: colon & rectum (C18.x‐C21.x), lung (C26.x‐C34.x), melanoma (C43.x), skin (not melanoma) (C44.x), breast (C50.0) gynecologic (C54.x‐C57.x), prostate (C61.x), renal (C64.x‐c66.x), bladder (C67.x), lymphoma (C81.x‐C88.x), myeloma (C90.x), leukemia (C91.x‐C95.x), and other/unclassified (all other Cxx.x diagnoses).

All diagnosis groups were incident at first diagnosis day; thereafter, they were regarded as prevalent. “Any” cancer diagnosis was incident on the first day of the first cancer diagnosis and was thereafter prevalent. “Any cancer” incidence and prevalence will be less than the sum of individual cancer incidences and prevalences, since later cancer diagnoses were not included in the “All” group.

2.3. Cancer epidemiology

The Danish Cancer Registry exports all cancer data to the Nordic Cancer database (Nordcan),21 which publishes yearly incidence and prevalence data, stratified by nation, diagnosis, 10‐year age groups, and sex. Data were extracted for three biopsy cohorts (1985‐1994, 1995‐2004, 2005‐2014), based on the results for 1990, 2000, and 2010, respectively. These data were used as the control group for determining expected risks of cancer for the studied patients.

2.4. Patient data

Patient sex and birthday were calculated from the national identity number. Patients entered the study 10 years before the biopsy and were lost to follow‐up 10 years after, unless censored for death or emigration. Dates of immigration, emigration, and death were extracted from the National Patient Registry. The actual period of observation was determined from these data.

Time to end‐stage renal failure (ESRD) was extracted from the Danish Nephrology Registry.

2.5. Calculations

Observation periods were as follows: 10‐5, 5‐3, 3‐1, and 1‐0 years before biopsy; 0‐1, 1‐3, 3‐5, and 5‐10 years after biopsy. Analyses of prevalent rates at biopsy used the observation period 10‐0 years before biopsy. The overall cancer incidence covered the period −10 to +10 years. These observation periods were chosen as these are the observation periods used by the NORDCAN database.

The incidence of cancer increases with age. In order to measure this, for each year after biopsy, 10% of patients were assumed to move into the next 10‐year age cohort. For instance, if the expected cancer incidence for a patient aged 45‐55 years is 1.0%/y and for a patient aged 55‐65 years 2%/y, a patient in the study in the 45‐55 year age group will have an expected cancer incidence of 1% in the year after biopsy. Three years later, his expected cancer incidence will be (0.7 × 1% + 0.3 × 2%) = 1.3%.

For each observation period, the real observation period was often shorter than the defined, due to immigration, emigration, or death. While this effect was minimal for periods before biopsy, it was substantial for periods after biopsy, particularly 3‐10 years after biopsy, and was greater in the older age groups. The real average observation period for the relevant renal diagnosis and age group (15‐44 years; 45‐64 years; >64 years) was used in calculations. For expected incidence rates before biopsy, the rate of (by definition) survivable cancers was used, the incidence rate thus being measured by changes in the prevalence rate.

By using this methodology, the number of observed incident cancer episodes per time interval and the expected incidence (%/y) were available. The observed incidence (%/y) was defined as the number of observed cases divided by the average observation period. The expected number of cases in the time interval was defined as the expected incidence multiplied by the average time period.

Patient survival and time to dialysis initiation (censored for patient death) were considered as lost to follow‐up if the patient emigrated or if the end point was not reached at the end of the study (January 1, 2015).

2.6. Statistics

Variables were compared using Student's t test. Kaplan‐Meier analysis was performed to compare cancer patient survival to patients without cancer. Categorical analysis was performed using chi‐square test. Statistical analysis of cancer risk was based on observed and expected cases. Confidence intervals and P‐values for indirect standardized incidence ratios were calculated using the Poisson model for number of incident cancer cases and population level incidence rates adjusted for age and calendar year. A significance level of <0.05 was considered significant.

The SAS program was used for analysis (SAS Institute).

3. RESULTS

3.1. Hypertensive nephropathy and prevalent cancer

Two thousand and nine hundred and seven patients had hypertensive nephropathy on biopsy. Of these, 2289 also had other specific renal diagnoses and were excluded. The remaining 618 patients with biopsy‐proven hypertensive nephropathy (either glomerulosclerosis or arteriosclerosis) as only specific renal diagnosis were included. The age was 56.1 ± 14.5 years. Twenty‐eight per cent were female.

The indication for renal biopsy was available for 428 patients (69%). The indications were as follows: uremia 386 patient (90%), proteinuria or nephrosis 126 (29%), and hematuria 14 (3%). Thirty‐eight cancer diagnoses were prevalent at biopsy (prevalence 6.1%) compared with an expected figure of 23.9 (3.9%). The odds ratio (observed cases/expected) for prevalence was 1.6 (confidence interval [CI] 1.2‐2.3), P < 0.001.

3.2. Hypertension and cancer incidence

One hundred and five (17.0%) of the 618 patients developed cancer during the period 10 years before biopsy to 10 years after. Of these, 11 patients developed two forms of cancer, in all 116 cancer diagnoses. Incidence rates are shown in Table 1 and Figure 1. The risk of cancer overall was twice the background population. It was significantly raised for the following diagnoses: renal cancer (odds ratio 10.4), myeloma (13.2), skin cancer (7.9), and other/unspecified (1.8). The increase was common for all age groups. The increased risk was similar for both males and females. No increase in incidence was seen until 1 year before renal biopsy, where it rose rapidly. The relative risk fell again 1 year after biopsy and was normal again 5 years after biopsy.

Table 1.

The association of hypertension to observed and expected cancer cases between 10 y before and 10 y after biopsy

Cancer diagnosis	Observed	Expected	Odds ratio (CI)
Colorectal	7	17.6	0.4 (0.2‐0.8)a
Lung	10	7.1	1.4 (0.8‐2.7)
Melanoma	1	2.8	0.4 (0.05‐2.5)
Breast	4	5.3	0.8 (0.3‐2.0)
Gynecologic	4	2.2	1.8 (0.7‐4.9)
Prostate	12	10.5	1.2 (0.7‐2.1)
Kidney	14	1.4	10.4 (6.0‐18.1)b
Bladder	1	5.8	0.2 (0.02‐1.2)
Lymphoma	3	2.2	1.4 (0.4‐4.3)
Myeloma	9	0.7	13.2 (6.6‐26.7)b
Leukemia	1	3.3	0.3 (0.04‐2.1)
Skin	27	3.6	7.9 (5.3‐11.7)b
Unspecified	23	13.2	1.8 (1.2‐2.8)a
Total	105	63.1	2.0 (1.6‐2.5)b
Total, age <45 y	6	2.3	2.6 (1.16.0)a
Total, age 45‐65 y	46	23.2	2.0 (1.4‐2.7)b
Total, age >65 y	53	37.0	1.4 (1.0‐2.0)a
Total, male	77	46.3	1.7(1.3‐2.2)b
Total, female	28	16.2	1.7 (1.1‐2.7)a

^a<0.05.

^b<0.001.

Figure 1.

Observed and expected cancer incidence (%/y) relative to time of renal biopsy

Two histological diagnoses were present: arteriosclerosis (394 cases, 63.8%) and glomerulosclerosis (224 cases, 36.2%). There was no significant difference in sex (30% vs 25%), but patients with arteriosclerosis were younger (54.9 ± 14.9 vs 58.2 ± 13.7 years, P < 0.01). The incidence of cancer was similar in the two groups (41 cases, 18.3% and 64 cases, 16.2%, respectively).

Age, sex, and survival after renal biopsy for cancer patients compared to patients without cancer are shown in Table 2. Patients with cancer were generally older with a shorter survival. However, for patients older than 65 years at renal biopsy, there was no difference in patient survival.

Table 2.

Age and sex of cancer cases, and median survival time compared to patients without cancer

Cancer diagnosis	No. cases	Sex (% female)	Age (y)	Median survival (y)
				Cancera	No cancer
No cancer	513	28	56.1 ± 14.5		10.8
Colorectal	7	14	70.6 ± 6.7b	3.8
Lung	10	60c	65.6 ± 11.7c	2.9b
Melanoma	1	0	55.0	–
Breast	4	100b	67.6 ± 13.7	6.4
Gynecologic	4	100b	59.4 ± 5.2	0.9b
Prostate	12	0	69.3 ± 7.9b	4.5
Kidney	14	36	63.7 ± 10.2	6.0
Bladder	1	100	74.7	9.2
Lymphoma	3	0	69.5 ± 12.8	1.7
Myeloma	9	22	65.7 ± 6.9c	1.9b
Leukemia	1	100	35.4	3.8
Skin, non‐melanoma	27	22	63.3 ± 13.3b	11.2
Unspecified	23	17	65.5 ± 8.7b	3.8
Total	105	27	64.8 ± 10.9d	4.5	10.8
Total, age <45 y	6	100	30.5 ± 7.0	8.80.07	23.9
Total, age 45‐65 y	46	22	56.3 ± 6.1	5.80.07	10.1
Total, age >65 y	53	28	72.6 ± 5.1	3.5	3.8
Total, male	77	–	64.8 ± 9.9d	4.5c	9.7
Total, female	28	–	64.8 ± 13.5d	3.8b	11.7

^aSurvival from time of biopsy.

^b<0.01.

^c<0.05.

^d<0.001.

Two hundred and seventy patients developed ESRD during the period of follow‐up. The median time to ESRD was 8.0 years. There was no difference in renal survival between patients with cancer (median 6.6 years) and those without (6.6 years).

4. DISCUSSION

In this study, we demonstrate an increased risk for cancer in patients with hypertensive nephropathy. This observation has not previously been made. As it is an observational study, no conclusions can be made as to whether hypertension causes cancer, cancer causes hypertension, or some third factor is responsible. The difference in time between the renal and cancer diagnoses is not of much use in determining etiology, since both diseases can be present sub‐clinically for many years before diagnosis. A causative role for hypertension in renal cancer is suggested by the finding that a reduction in blood pressure lowers the risk.8 This association could be the result of common factors that are known to be associated with both hypertension and cancer, for example, obesity,8, 22, 23, 24 tobacco consumption,3, 25, 26 and use of hypertensive drugs. Multiple myeloma in particular is associated with obesity.27 The effects of hypertensive drugs in particular can be difficult to separate from those of hypertension. The use of angiotensin receptor blockers (ARB) has been associated with cancer risk,28 as has the combination of ARBs with angiotensin‐converting enzyme inhibitors (ACEI),29 but a later study has been unable to confirm this association.30 Another study31 demonstrated a 14% increased risk of lung cancer in patients treated with ACEI compared to patients treated with ARB. In the present study, the risk of lung cancer was not significantly increased. Animal studies have linked diuretic use to kidney tumors in rats.32 Some studies have suggested an increased risk associated with diuretic use,3, 33, 34 but this association often disappears after correction for hypertension per se.24, 35

Uremia is characterized by general immune dysfunction.36 Since this is a recognized cause of skin cancer,37 the incidence of which was raised both in this study and in glomerulonephritis,38 this suggests that uremia may play a causative role in the development of skin cancer. Furthermore, ESRD is related to increased cancer risk,39, 40 possibly secondary to secondary degenerative cystic kidney disease.41

The present study confirms and quantifies previous findings of an increased risk of renal cancer in patients with hypertension. Possible causative mechanisms for the association of hypertension and renal cancer include formation of reactive oxygen species and chronic renal hypoxia.25

Increased risks were also noted for myeloma, skin, and other/unspecified. One previous study has shown an increased risk of skin cancer,10 but the increased risk of myeloma has not previously been described. We also found that the increased risk was not present until 1 year before biopsy, where it accelerated rapidly, and continued to be raised up to 5 years after biopsy. This temporal relationship has not been noted before, mainly because previous studies have investigated patients with hypertension per se, rather than patients with complicating renal involvement.

An epidemiological study can contribute little to a discussion of etiology, but some remarks are pertinent. The increase in cancer incidence around the time of renal biopsy may be incidental, due to increased attention to possible cancer in patients with renal problems, or vice versa.42 However, one study showed no significant increased risk of being diagnosed with renal cancer at the time of hypertension diagnosis, as defined by initiation of antihypertensive therapy.43 This possibility is particularly relevant for renal cancer and myeloma, where renal biopsy and ultrasonic investigation of the kidneys may be included in the routine diagnosis of these diseases. An earlier diagnosis would however be unlikely to affect the overall incidence, since most cancers will eventually result in clinical symptoms, and patients were followed up to 10 years after biopsy. Nor does the hypothesis explain the continuing excess risk 1‐5 years after biopsy. Most diagnoses of hypertensive nephropathy are presumptive, being characterized by the presence of a long history of hypertension, shrunken kidneys, and low levels of proteinuria. Renal biopsy will usually only be performed if other diseases are being considered (eg, glomerulonephritis), or in the presence of an uncharacteristic clinical picture, for example, rapidly progressive renal dysfunction or severe proteinuria in the presence of normally sized kidneys.

Another explanation for the simultaneous diagnosis of hypertensive nephropathy and cancer could be causal, either clinical cancer leading to a clinical exacerbation of hypertensive nephropathy, or increasing renal disease leading to an acceleration of subclinical cancer, due for example, to a concomitant reduction in immune capacity. A recent study of the incidence of cancer related to glomerulonephritis using the same database38 showed a similar pattern, though with a different distribution of cancer diagnoses, suggesting that it the renal dysfunction per se, rather than the etiology of the renal disease that is related to the cancer. Another explanation is that some patients with cancer and hypertensive nephropathy may have died before reaching a point where renal biopsy is indicated; however, this would not explain the post‐biopsy fall in incidence.

Several criticisms of this study can be raised. The number of patients is relatively small and does not permit a detailed subgroup analysis. As this was a registry‐based study, important clinical information was unavailable, such as renal function, smoking history, body mass index, medication, and degree of proteinuria. The study is limited to a small country; the results may not be applicable to other populations. The study is limited to only a small proportion of the patients with hypertensive nephropathy, and the fact that they were biopsied may be due to unknown differences between them and non‐biopsied patients. In particular, they probably had more renal impairment and/or proteinuria. Biopsies were evaluated at different centers by different pathologists; some intra‐ and inter‐individual differences in diagnoses must have been unavoidable. The presence of concurrent glomerular disease, for example, minimal change glomerulonephritis, may therefore have been missed. This will particularly be the case if the diagnosis is based on a small biopsy with a correspondingly low number of glomeruli. All cancers were assumed to be prevalent at first diagnosis, due to the difficulty of differentiating repeat registrations of the same cancer, cancer recurrence, and de novo disease. The incidence rates presented here will be an underestimate of the true incidence, particularly for skin tumors. The control group was the entire population of Denmark; perhaps 10%‐15% will have hypertension, although only a minority of these will have hypertensive nephropathy. This may have led to an underestimation of the real difference.

5. CONCLUSIONS

Hypertensive nephropathy as verified by renal biopsy is associated with an increased risk of myeloma, skin, renal, and other/unspecified cancers. Since this patient population was probably characterized by reduced renal function and/or significant proteinuria, this finding cannot be extended to all patients with hypertensive nephropathy. This study does not provide information about the risk of cancer in patients with hypertension, but without proof of hypertensive nephropathy, which can only be documented by renal biopsy, as hypertension is a common cause of renal dysfunction, which can have many other causes.

We suggest that patients with documented hypertensive nephropathy and reduced renal function to be investigated routinely for these three diseases and possibly screened clinically for malignant disease in general.

CONFLICT OF INTEREST

The authors have no conflict of interest.

AUTHORS' CONTRIBUTIONS

James Heaf involved in protocol and article preparation. Gunnar Laier involved in statistics and article review. Alastair Hansen involved in revision of renal biopsy descriptions, classification of renal diagnoses, and article review.

ETHICAL APPROVAL

The project was approved by the Danish Data Protection Agency (REG‐132‐2015) and by the Danish Patient Safety Authority (3‐3013‐1440/1). Due to the anonymized, non‐invasive, retrospective nature of the study, the requirement for specific scientific ethical committee was waived by the Danish Patient Safety Authority.

CONSENT FOR PUBLICATION

Due to the anonymized, non‐invasive, retrospective nature of the study, the Danish Patient Safety Authority (3‐3013‐1440/1) has waved the requirement for individual patient approval.

Heaf JG, Hansen A, Laier GH. Hypertensive nephropathy is associated with an increased risk of myeloma, skin, and renal cancer. J Clin Hypertens. 2019;21:786–791. 10.1111/jch.13565

DATA ACCESSIBILITY

An anonymized data file is available from the corresponding author on request. The data used in this project are available on Open Science Framework, reference DOI 10.17605/OSF.IO/ANJSP.