The prevalence of chronic kidney disease in cancer patients
The prevalence of chronic kidney disease (CKD) is reported to be high in patients with malignancy, being 33% and 27% according to Dogan et al. and Launay-Vacher et al., respectively. But the prevalence of CKD stage 3 or more is 20% after age 60 years, and it is 45% after age 70 years; furthermore, cancer is also more common in older people. Thus the percentages reported by Launay-Vacher and colleagues may simply show the coincidence of age, CKD, and cancer. Unfortunately, the Renal Insufficiency and Anticancer Medications (IRMA)-1 lacks any information relative to the timing of CKD prevalence relative to diagnosis and treatment of cancer.
However, the IRMA-1 study provided us with some important findings. It showed that 50% to 60% of the subjects had a reduced glomerular filtration rate (GFR < 90 mL/min/1.73 m 2 ), whereas serum creatinine (sCr) was normal in most patients. These results emphasize not only that CKD incidence is high in subjects with malignancy, but also that creatinine is not sufficient to monitor such a condition. In France, a prevalence of a GFR below 90 mL/min/1.73 m 2 was 53% in IRMA-1 and 50% in IRMA-2, within a cohort of 5000 subjects with different malignancies. According to the Kidney Disease Improving Global Outcomes (KDIGO) definition, the prevalence of CKD stage 3 to 5 (GFR < 60 mL/min), excluding renal replacement therapy, was also high, reaching 12% in IRMA-1 and 12% in IRMA-2, respectively. Huang et al. reported that despite having normal preoperative concentrations of sCr, 87% of renal cancer patients had an estimated GFR (eGFR) less than 90 mL/min/1.73 m 2 . This study was performed in a cohort of 662 subjects with a renal cortical tumor, undergoing either partial or radical nephrectomy; in addition, an eGFR less than 60 mL/min/1.73 m 2 was reported in 26% of these patients. Unfortunately, as for the IRMA report, the relative role of age has not been adequately considered as a possible confounding factor. Prevalence of CKD ranged from 16% to 25% in patients with malignancy in Belgium, the United States, and Japan. In the IRMA-1 study, CKD was also highly prevalent, reaching approximately 50%, in breast, colorectal, lung, ovarian, or prostate cancers. The Belgian Renal Insufficiency and Anticancer Medications (BIRMA) study was a large (1218 patients), national, multicenter and retrospective study, performed to evaluate the prevalence of CKD in Belgian patients with malignancy. Differently from the IRMA-1 study, one of the specific aims of BIRMA was also to describe the type and dosage of the antineoplastic agents prescribed according to kidney function. Elevated sCr (≥ 1.2 mg/dL) was found in 15% of patients, but eGFR was below 90 mL/min/1.73 m 2 in 64% of them. Overall, 79% of patients ( n = 1087) were administered at least one drug requiring dose adjustment because of kidney function, and 78% received at least one drug known to be nephrotoxic. Notably, 56% of CKD patients treated with chemotherapy agents requiring dose adjustment in case of reduced kidney function had no dose reductions. This study proved that prevalence of CKD is high in patients with malignancy and may be underestimated.
CKD is more than just a measurement; it appears to affect survival. Yang et al. reported that 32% of patients with newly diagnosed cancer had CKD; in addition, renal function was inversely related to all-cause mortality. Indeed, eGFR below 60 mL/min/1.73 m 2 was an independent predictor of mortality, as compared with eGFR 60 mL/min/1.73 m 2 or higher, and it was influenced by cancer site. After adjustment for possible confounders, eGFR less than 60 mL/min/1.73 m 2 at the time of diagnosis was associated with a higher mortality risk among patients with both hematologic malignancies and gynecologic cancers.
Wong et al. studied a cohort of 3654 subjects and assessed the relation between eGFR and risk of cancer. They found that in men, but not in women, with an eGFR lower than 55 mL/min/1.73 m 2 , the risk for cancer was significantly higher. In particular, lung and urinary tract cancer risk increased by 29% for each 10 mL fall in eGFR. Based on the aforementioned findings, it appears that CKD itself is a risk factor for cancer; in breast, colorectal, lung, ovarian, and skin cancers, the prevalence of CKD was increased. Others have confirmed that breast, cervix, colon, and kidney cancers are more common in CKD patients than in the general population. Thus in addition to the observed increase in CKD prevalence in cancer patients, CKD is a risk factor for several malignancies. However, not all solid tumors appear to be equally represented in this population. A retrospective cohort study of 1,190,538 adults assessed the association between eGFR level and the risk of incident cancer; during 6,000,420 person years of follow-up, 76,809 incident cancers were identified in 72,875 subjects. After adjustment for time-updated confounders, lower eGFR was associated with an increased risk of renal cancer, with an adjusted hazard ratio (HR) of 2.28 (95% confidence interval [CI], 1.78–2.92) for an eGFR less than 30 mL/min/1.73 m 2 . The authors also observed an increased risk of urothelial cancer at an eGFR less than 30 mL/min/1.73 m 2 , but no significant associations between eGFR and other cancers. Finally, CKD conferred an increased cancer-specific mortality in patients with kidney and urinary tract cancers. In ESRD patients on dialysis, the observed increased risk for renal parenchymal cancer is related to the development of acquired renal cystic disease, which increases with time on dialysis.
Factors with a potential negative effect on kidney function are summarized in Table 33.1 . Thus the relationship between the kidney and cancer appears to be bidirectional. For example, preexisting CKD may impact the bioavailability and/or safety profile of an anticancer drug, potentially leading to different and sometimes suboptimal treatment choices; on the other hand, it is also possible that unexpected renal effects of a novel anticancer drug may lead to progressive kidney injury or to worsening of preexisting CKD.
|Other drugs (e.g., NSAIDs, bisphosphonates)||Direct nephrotoxicity|
|Radiation therapy that includes kidney||Fibrosis, vascular injury|
|Contrast medium||Direct nephrotoxicity|
|Paraneoplastic renal manifestations||Immune-mediated mechanisms|
|Nephrectomy||Loss of nephrons|
|Tumor infiltration||Kidney infiltration (loss of nephrons)|
Assessment of kidney function in patient with chronic kidney disease
The direct measurement of GFR is complex, cumbersome, and time-consuming to perform on an everyday basis. However, sCr may be used to estimate GFR in subjects with stable kidney function, considering that precise GFR measurement is not required in most clinical settings. However, in some clinical situations, measuring GFR may be very useful for adjusting the doses of cancer medications, especially those with a narrow therapeutic index. Inulin clearance is the gold standard for the evaluation of GFR, but it is used only in clinical research settings; alternative methods, radioactive or nonradioactive (iohexol, iothalamate, ethylenediaminetetraacetic acid [EDTA] or diethylenetriaminepentaacetate), represent simpler and less cumbersome methods, as compared with inulin clearance. At present, the most commonly used methods for GFR assessment are the measurement of creatinine clearance (CrCl), and estimation formulas based upon sCr, such as the Cockcroft-Gault formula, the Modification of Diet in Renal Disease study formula, and the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) formula. Unfortunately, sCr is an unreliable marker during acute changes in renal function and its use to assess true kidney function has several limitations. A significant decline in GFR can be observed before it is reflected in an increased sCr; indeed, up to 50% of renal function could be lost before significant changes in sCr might be recorded. In addition, sCr does not reflect renal function during acute changes, until a steady state has been reached, which may require several days.
Concerning the different equations available to estimate GFR, the CKD-EPI formula is currently recommended to assess kidney function for screening and diagnosis of CKD, according to the KDIGO guidelines. However, this formula has not been validated in cancer patients. It should also be underscored that people older than 65 years often (but not always) present with a decreased GFR, because other comorbidities may influence kidney function. As reviewed earlier, a reduced GFR is of particular importance in cancer patients for its clinical implications.
A recent publication evaluating GFR estimating equations in cancer patients noted that body surface area (BSA)-adjusted CKD-EPI method appears to be the most accurate published model to estimate GFR in patients with cancer. BSA-adjusted CKD-EPI, based on the analysis of data from 2582 cancer patients using 51 Cr-EDTA GFR measurement as the gold standard, was found to be the most accurate and least biased published model to estimate GFR. The authors also developed a new model that further improves the estimation of GFR and allows calculation of predictive confidence intervals for this estimation. The new model has been implemented as an online tool found at the following link: tavarelab.cruk.cam.ac.uk.easyaccess1.lib.cuhk.edu.hk/JanowitzWilliamsGFR/. This new model to estimate GFR may represent a new standard of care and should be further examined along with BSA-adjusted CKD-EPI in clinical onconephrology practice.
In the field of cancer pharmacology, until recently, different studies have been performed using different estimation methods for kidney function. This highlights the need for a common language.
Finally, these assessments and adjustments are made on the basis of the apparent GFR. Many drugs are excreted via renal tubular epithelial transporters; measurement of that tubular excretory function is not done in clinical practice or in research. This is a potentially significant gap in our pharmacologic understanding and clinical care.
The balance between toxicity and efficacy in chronic kidney disease cancer patients treated with anticancer agents
Anticancer chemotherapy-related acute kidney injury (AKI) and CKD have been well described over the past decades; despite this, literature suggesting how to modify the doses of these agents in patients with underlying renal dysfunction, and those on dialysis, appears to be controversial, and too often not evidence-based.
Moreover, little is known about the appropriate use of targeted agents and immune checkpoint inhibitors in this population, leading to complex and nonevidence-based decisions in these settings. After decades of use of common cytotoxic drugs, clinicians versed in cancer care and its complications are well aware of the main toxicities of these cytotoxic agents, but less is known for targeted agents, and especially immunotherapy. On the other hand, novel anticancer agents that have recently entered clinical practice have a wide array of previously unrecognized and ill-defined adverse events. Ultimately, these toxicities must be readily recognized and managed by those providing care for patients exposed to these drugs. This includes understanding risk factors for drug-induced kidney injury, appropriate drug dosing for the patient with CKD and those on dialysis, the clinical manifestations of drug nephrotoxicity, and the optimal management of nephrotoxic complications. Whenever oncologists ask their nephrology colleagues to assess the degree of kidney function to provide insight into dosage adjustment of anticancer therapy, a thorough knowledge of the specific metabolism of each anticancer agent and of its pharmacokinetic and pharmacodynamic properties is mandatory, to decide “if” the drug should be administered, “when” it is appropriate to administer it, and to “what extent” dosage adjustment should be used in the setting of underlying kidney disease. This approach must be accurate, as unnecessary treatment interruptions and drug dose reductions may be associated with suboptimal cancer therapy, thus hampering the clinical benefits of cancer therapy. Optimal management of underlying CKD and its complications, and prevention of further kidney damage from other exogenous nephrotoxins (e.g., contrast medium, nonsteroidal antiinflammatory drugs, and bisphosphonates) in cancer patients with preexisting CKD, is also key to minimize drug-related complications. Patients with CKD and those on dialysis should not be undertreated for their neoplastic disease because of the fear of drug-induced adverse events.
As already discussed, reduced GFR has been reported to be associated with reduced overall survival (HR, 1.27) , and increased cancer-related mortality, , with cutoff values for the significance of the effect of 60 (13,14), 70 (12), or 75 mL/min/1.73 m 2 . This association has been related to the impact of reduced GFR on the pharmacokinetics of anticancer drugs, leading to overexposure compared with normal renal function and dose-related toxicities. Recently, Chen et al. showed that patients with metastatic colorectal cancer and unrecognized kidney impairment at baseline experienced more toxicities, more often, leading to more treatment discontinuations and reduced time to progression.
The dose of anticancer drugs in CKD patients should be adjusted to avoid severe toxicities. In addition, using chemotherapeutics with potential nephrotoxicity will also require specific monitoring and, when available, specific prevention strategies to reduce the risk of kidney damage, especially in patients with preexisting CKD. Some authorities advise that the noncorrected eGFR be used for adjustment of drug dosages, not the eGFR standardized to 1.73 m 2 BSA. In this perspective, specific and simple guidelines for monitoring each drug should be developed.
To date, only patients with adequate kidney function (sCr < 1.5 × the upper limit of normal) have been included in registered randomized controlled trials. A brief survey of the clinicaltrials.gov website showed that only 25% of active clinical cancer trials allowed subjects with an elevated sCr level. Thus we often lack an evidence base for patients with a CrCl less than 30 mL/min. This leads to limitations on use of some very effective cancer therapies. But this conservative recommendation may sometimes be inappropriate. For example, because large molecular weight monoclonal antibodies are cleared by the reticuloendothelial system and not metabolized or excreted by the kidneys; their clinical use should not be limited to patients with normal kidney function. Although pharmacokinetic studies in patients with impaired renal function are sorely needed, as suggested by a recent European Medicine Agency guideline, it is also clear that harmonization of kidney function assessment is mandatory, not only in early clinical trials, but also within pivotal trials of novel anticancer agents. Excluding patients with CKD and ESRD from these trials, especially from those that generate specific pharmacokinetic data, makes the results obtained in these studies not transferrable to the large number of patients with cancer with an underlying kidney disease. For these reasons, nephrologists should be involved in the design of the next generation of oncologic trials, especially when renal toxicity is expected to be an issue. Furthermore, the respective scientific societies should work together to accomplish this goal. Indeed, the full development of onconephrology as a new subspecialty is important because a number of key questions in the field remain not only unanswered, but also never asked.
Finding the right dose of a given anticancer agent in a patient with kidney impairment is key to minimize the risk of increased toxicities and to increase the likelihood of a benefit from the oncologic treatment, either in terms of response or of survival gain ( Fig. 33.1 ).