Our Literature Review section continues with another installment of summaries from the medical literature. Our authors have found recent articles that have direct relevance to the practice of insurance medicine. The intent of the reading list is to provide the highlights of articles, not an in-depth analysis. Contributions to the reading list are invited. Please forward your citation and summary to Michael L. Moore, MD, Associate Editor, Literature Review at Moorem1@Nationwide.com. We will acknowledge all contributors in each issue's installment.In this study of NHANES data, researchers utilized NHANES data to quantify the relationship between steps per day and mortality. NHANES participants in the 2003-2004 and 2005-2006 waves were given pedometers to wear for 1 week, removed only for sleeping, bathing or swimming. Until 2016, only pedometer ‘counts’ were available, making an analysis of steps per day impossible, but NHANES recently released step data. A pedometer count is a value that takes into account step intensity as well as frequency, such that sprinting generates more counts than walking, even if the number of steps is constant. Each of the 4840 study subjects, all over age 40, wore a pedometer for an average of 5.7 days and 14.4 hours per day. Over an average follow-up of 10 years, 1165 deaths occurred. Vital status and cause of death were assessed through the National Death Index.As a way to account for intensity, the authors used several difference cadence measures – steps per minute, including the peak cadence during a day, the average of the top 30 cadences during a day, and the average cadence during bouts of activity. There were several interesting findings regarding mortality. Firstly, there appears to be a threshold effect where, after about 12,000 steps per day, the mortality benefits of activity appear to be maximized. Activity in older individuals appear to have a larger protective effect on absolute mortality than it does in younger individuals. Also, there appears to be no significant effect of intensity once steps per day are included. Cause of death analysis showed that most of the mortality benefit resulted from improvements in cardiovascular mortality, though there was an effect on cancer mortality as well. It should be noted that it is difficult to fully exclude reverse causation in studies such as these – as debilitation and disease may cause reduced activity rather than the other way around, so findings should be viewed as associative rather than causative.In a recent article in the Journal of Insurance Medicine, this author evaluated the same data and fortunately found similar effects. There are several notable differences. The JIM study excluded participants based on the likelihood of their being offered life cover if they were to apply, controlled for a multivariate laboratory and biometric risk score, and also performed a second analysis of self-reported activity from all available NHANES waves. Submitted by Steven J. Rigatti, MDThose with the pathogenic variant HFE p.C282Y homozygote are well known to have hereditary hemochromatosis. However, the risks of hepatic malignancy and all-cause mortality are less clear including those with undiagnosed p.C282Y homozygotes, which can be identified in community genotyping. This cohort study of 451,000 participants from UK Biobank age 40-70 years follows these risks from baseline assessment 2006-2010 until 2018 (a median of 8.9 years). Men and women with these homozygous genotypes were compared to those without by information obtained from the national cancer registry, death certificate records, primary care data and hospital records to assess 2 co-primary outcomes, incident primary liver cancer and death from any cause.Of the 1294 male p.C282Y homozygotes found, 21 incident hepatic malignancies were noted, 10 of which occurred in participants without an established diagnosis at baseline. These men were determined to have a higher risk of malignancy compared to men with neither HFE variant (HR 10.5) as well as all-cause mortality (HR1.2). Using lifetable projections to age 75 for these men, the risk of hepatic malignancy was 7.2% compared to 0.6% for men without the variant. The risk of death was 19.5% for those men with the variant vs 15% those without also using lifetable projections to age 75. For women, there were just 3 incident hepatic malignancies. There were no statistically significant associations found in women participants with the p.C282Y homozygote and malignancy or death.This study of a very large community sample group, which identified HFEp.C282Y homozygotes through genotyping, provides evidence of increased risk of primary liver carcinoma and all- cause mortality, including those previously undiagnosed in men. In women however, there does not appear to be any significant associations. The authors note that further research will be needed to assess any impacts or benefits of earlier diagnosis or treatments. Submitted by Ted Gossard, MDAsymptomatic microscopic hematuria (AMH) is defined by the American Urological Association (AUA) as presence of 3 or more red blood cells (RBCs) per high-powered field (HPF). AUA guidelines recommend CT or MRI imaging despite the chance of finding malignancy may be as low at 0.68%. In women, the American Urogynecological Society and the American College of Obstetricians and Gynecologists published a committee opinion recommending avoiding evaluation in never-smoking women aged 35-50 years with less than 25 RBCs per HPF.This study is intended to validate the Hematuria Risk Index (HRI) as a risk stratification tool utilizing clinical parameters to calculate risk of urologic malignancy in patient's microscopic hematuria in 994 patients seen at a single Urology site (441 males and 583 females, average age 57.1 years). The study also looked at cost-of-testing savings by using this screening tool, but the details of these findings and arguments are not included in this review.HRI Screening Tool: 4 points for gross hematuria4 points for age greater than 501 point for positive tobacco history1 point for male sex1 point for greater than 25 RBCs/HPF on USLow-risk: 0-4 points with 0%-0.3% rate for cancer detection.Moderate-risk: 5-8 points with 1.1%-2.5% rate for cancer detection.High-risk: 9-11 points with 10.7-11.6% rate for cancer detection.In this study population of 994, there were no patients with gross hematuria, and therefore the maximum score possible was 7 (ie, no person may be considered to be high-risk in the HRI screen without having gross hematuria). The moderate-risk group (HRI = 5-7) displayed significant differences in terms of lesions detected and the grade of cancers detected. Therefore, the moderate-risk group (HRI = 5-7) was further delineated into separate cohorts: HRI = 5, HRI = 6, and HRI = 7. The total number of cancers was determined for each of the groups: low risk (HRI = 0-4), HRI = 5, HRI = 6, HRI = 7, and for the overall moderate risk group (HRI = 5-7). Tests done included CT urography, CT w/o and w/ contrast, MRI, ultrasound, and cystoscopy.The HRI model was externally validated in this patient cohort; logistic regression analysis was statistically significant for detecting genitourinary lesions in each group (P <.0001). A receiver operator curve generated for HRI and cancer detection had an area under curve of 0.885.When stratifying by risk group, lesion detection rate in the low-risk group (HRI = 0-4) was 0% (0/643), and in the moderate group (HRI = 5-7) there were escalating rates of tumor detection: HRI 5-7 = 2.96% (12/406)HRI 5 = 1.54% (4/260)HRI 6 = 3.68% (5/136)HRI 7 = 30% (3/10)In addition, 417 patients were found to have benign diagnoses with nephrolithiasis being the most common (11.3% of total cohort), with other findings including enlarged prostate, renal cysts, and urethral strictures.This study concluded that patients with asymptomatic microscopic hematuria at low risk (HRI = 0-4) could forgo further evaluation, at a cost saving of $140,490 to $242,124, without finding a single malignant lesion. The authors of this paper also argue that for evaluating moderate HRI = 5-7 patients with AMH, the combination of just renal ultrasound combined with cystoscopy would be much more cost-effective than the AUA approach of using CT/MRI scanning, missing only rare upper urinary tract lesions that were invariably only mildly malignant. Submitted by Rodney C. Richie, MD4.Kristensen JH, Basit S, Jan Wohlfahrt, et al. Pre-eclampsia and risk of later kidney disease: nationwide cohort study. BMJ. 2019;365:l1516.Pre-eclampsia is associated with multiple organ dysfunction during pregnancy. One organ that is involved is the kidneys. In this Danish study of over 1 million pregnant women from 1978 – 2015, there were 3901 cases of chronic renal disease over an 18.6 year follow-up. In those women who had suffered preeclampsia, the rate of CRF was twice that over those with normal pregnancies. In those women who suffered preeclampsia with delivery at less than 34 weeks of gestation, this risk increased to 3.9 times normal. The development of CRF was highest in the first 5 years following delivery. This data would suggest that the risk for post-delivery complications following preeclampsia persists for multiple years and does not return to baseline quickly after delivery. Submitted by Michael L. Moore, MDTransient ischemic attacks are a well-known risk factor for future ischemic strokes. This retrospective study of participants from the Framingham Study assesses the incidence of a first TIA and the subsequent risk of future stroke in this population. Additionally, the authors compared stroke rates from more recent time periods (2000-2017) with earlier periods to determine if stroke risk in patients with prior TIAs are changing.This retrospective cohort study (from the Framingham Heart Study) was conducted using prospectively collected data of 14,059 participants with no prior history of TIA or stroke followed from 1948-2017. Outcomes assessed included TIA incidence rates and proportion of stroke occurring after TIA short term (7, 30 and 90 days) and long term (1-20 years). As well, the time trends of stroke risk from prior TIA from 3 epochs (1954-1985, 1986-1999 and 2000-2017) was assessed.Overall, during the 66 years of follow-up (366,209 person-years), 435 participants experienced TIA, and the TIA incidence was 1.19/1000 person-years. These were matched to 2175 control participants without TIA. Over a median of 8.9 years of follow-up after TIA, 130 participants experienced a stroke (29.5%). Twenty-two percent of these strokes occurred within 7 days, 31% occurred within 30 days, 39% occurred within 90 days and 48.5% occurred more than 1 year after the TIA. The 90-day risk of stroke from the 3 epochs was determined to be 16.7% from 1948-1985, 11% from 1982-1999 and then dropped to 5.9% from 2000-2017.In addition to further establishing evidence of the elevated risk of stroke after TIA, the significance and an interesting finding of this study is that it presents evidence that TIAs, which have occurred in more recent time periods (2000-2017), have substantially reduced risk of future stroke relative to the earlier time periods. This would suggest that more current medical management (such as statins, platelet inhibitors) have favorably impacted clinical outcomes by reducing risk of ischemic stroke after a first TIA. Submitted by Ted Gossard, MDIn the life insurance industry, it is a common practice that visceral malignancies, regardless of site or prognosis, are used as an exclusion criteria from preferred classes. This paper quantifies mortality risks that can justify or perhaps refine this general principle.This remarkable study was performed by researchers at the American Cancer Society. It is a thorough exploration of the SEER database to quantify the risks of subsequent malignancies at least 5 years after the intial diagnosis. In all, there were 1.5 million 5-year survivors, 11.1 million years of follow-up and over 88,000 deaths among this group. A total of 30 different cancers (31 among women) were evaluated for their propensity to modify the risk of developing and dying from other malignancies. Additional cluster analysis was performed to corresponding risks for smoking and obesity-related cancers.A full presentation of these many hundreds of risk correlations is not feasible here. However, there are several highlights. Firstly, among women, the cancer type with the highest risk of both subesquent malignancy and death related to a subsequent malignancy is laryngeal cancer. The standardized incidence ratio (SIR) for a subsequent cancer was 2.48, while the standardized mortality ratio (SMR) was 4.56. Among men, it was still laryngeal cancer that had the highest SIR for subsequent malignancy (1.75), but it was biliary cancer that had the highest SMR for death from subsequent malignancy (3.82). Secondly, there were several malignancies for which the SMR was only minimally elevated, specifically thyroid cancer (SMR: 1.07, NS), kidney cancer (SMR: 1.14), testis cancer (SMR: 1.23), and colon (SMR 1.11) among men, and kidney (SMR: 1.22), thyroid (SMR: 1.01, NS), breast (SMR: 1.13), and uterine (SMR: 1.09) among women. In both sexes, cutaneous melanoma was not associated with sigificant elevation in the mortality risk of subsequent cancers. Prostate cancer was associated with a significant decrease in the risk of death from subsequent cancers (SMR: 0.90). Non-melanoma skin cancer was not studied.There are many moving parts to a study like this, so parsing out a meaningful message can be difficult, as can making rational critiques. However, it is likely that this study suffers from some of the same biases as more familiar mortality studies based on SEER data. That is to say that for screening-detected malignancies like breast and prostate cancer, the SMRs in this study likely underestimate the true risk, because people who get screened have demonstrated access to and interest in healthcare, which has an unquantifiable bias toward better overall and cancer-specific mortality. However, this study is quite useful in that it may suggest that certain visceral malignancies do not substantially increase mortality risk after at least 5 years of survival, and that survivors may have a mortality risk, which is lower than previously thought. Submitted by Steven J. Rigatti, MD7.Lim E, Kumar S, Seager M, et al. Outcomes of Renal Tumors Treated by Image-Guided Percutaneous Cryoablation: Immediate and 3- and 5-Year Outcomes at a Regional Center. Am J Roentgenology. 2020;215:242-247.This prospective UK study was used to evaluate the immediate 3- and 5-year outcomes of patients with stage TI renal cell carcinoma that had been treated with cryoablation. A total of 180 patients (29% female) were treated for T renal cell carcinoma (≤7 cm and no nodal or mets). Technical success was achieved in 98.9% of patients. The complication rate was 2.2% with no fatalities. Estimated local tumor progression free survival at 3 and 5 years was 98.3% and 94.9%, respectively. There were no distant metastases or deaths attributable to renal cell carcinoma. Glomerular filtration rate was not significantly different from pre-procedural levels.This study shows that in select individuals there is another modality for treating early renal cell carcinoma in addition to surgery and radiation. Cryoablation showed excellent long-term curative effects as well as few complications and little to no effect on renal function. Submitted by Michael L. Moore, MD8.Howlader N, Forjaz G, Mooradian M, et al. The Effect of Advances in Lung-Cancer Treatment on Population Mortality. NEJM. 2020;383:640-649.The authors used data from Surveillance, Epidemiology, and End Results (SEER) areas to assess lung-cancer mortality and linked deaths from lung cancer to incident cases in SEER cancer registries. This allowed for evaluation of population-level mortality trends attributed to specific subtypes (incidence-based mortality) from 2001 through 2016.A bit of background here may be helpful. Lung cancer is broadly separated into 2 molecularly and histologically heterogenous subtypes: non-small-cell lung cancer (NSCLC) and small-cell lung cancer (SCLC), accounting for 76% and 13% of lung cancers in the U.S., respectively. Lung cancer remains the number 1 cause of cancer death in both men and women.Death certificate data show that both types of lung cancer have slowly decreasing incidence over the time period studied: mean annual change of −2.5% from 2011 through 2015 for men and −1.2% annual change for women. But these overall trends cannot be separated into subtype-specific trends because death certificates do not record the cancer subtype. Additionally, death-certificate data includes deaths from cancers that have metastasized to the lung from other cancer sites.To address this data limitation, the SEER Program has linked mortality records to incident cancer cases, making it possible to calculate “incidence-based mortality,” which captures the population-level mortality attributable to particular tumor types reported to SEER registries. The numerator in the calculation of incidence-based mortality is the number of cancer-specific deaths among persons with a particular cancer diagnosis reported to the registry. The denominator is the general population in the general population at risk at the time of death in the SEER areas. This approach enables general population mortality to be partitioned according to characteristics that are associated with the cancer diagnosis (eg, tumor subtype or stage at diagnosis) and recorded in the SEER registries. This incidence-based method of estimating mortality has been used to assess the effect of screening, and to a lesser extent the effect of treatment, on mortality trends for several cancer types.For patients with SCLC, mortality decreased slowly and similarly among men and women – consistent with decreasing rates of smoking in the general population and the lack of any substantive change in chemotherapy for SCLC over these intervening years.For men with NSCLC the incidence decreased gradually by 1.9% annually from 2001 through 2008, and then by 3.1% annually from 2008 through 2016. For women with NSCLC, the incidence was flat from 2001 through 2006, and then started decreasing by 1.5% annually from 2006 through 2016. These decreasing trends in incidence are attributed to decreased levels of smoking in the general population, with greater numbers of men quitting compared to women.NSCLC mortality for men decreased by 3.2% annually from 2006 through 2013, and then decreased by 6.3% annually from 2013 through 2016. This translates to an improved 2-year survival among men with NSCLC from 26% diagnosed in 2001 to 35% diagnosed in 2014.NSCLC mortality in women did not improve until 2006, when it then began decreasing by 2.3% annually from 2006 through 2014, and then at a faster drop of 5.9% annually from 2014 through 2016.The overall decrease in incidences for both SCLC and NSCLC are attributed to decrease in proportion of smokers in the general population. The improved mortality in NSCLC, especially from 2014 to 2016, is attributed to the advent of targeted therapies and immunotherapies. The presence of an oncogenic driver mutation – the genes encoding anaplastic lymphoma kinase (ALK) or epidermal growth factor receptor (EGFR) – renders a tumor sensitive to targeted tyrosine kinase inhibition. In 2012, the National Comprehensive Cancer Network recommended that all NSCLC undergo genetic testing for ALK rearrangements and EGFR mutations.More recently, immune-based therapies – specifically, programmed cell death protein 1 (PD-1) and programmed death ligand 1 (PD-L1) inhibitors have substantially improved outcomes of NSCLC treatments when present (particularly at levels >50%) and treated with immunotherapies.Early low-dose CT lung imaging as screening for lung cancer in persons with a history of smoking or current smokers has not yet ramped-up sufficiently to influence any of these statistics but is felt likely to eventually lead to increased diagnosis of adenocarcinoma (a subset of NSCLC) as a result of early detection and overdiagnosis. Submitted by Rodney C. Richie, MDThere is increasing evidence that the gut microbiome plays a major role in several aspects of human health. This latest article from Chinese researchers shows a strong association between bacterial metabolites and major depressive disorders. In this study, 311 stool samples from both patients with MDD and healthy volunteers acting as controls were analyzed. A total of 47 bacterial species, 3 specific bacteriophages, and 50 fecal metabolites were significantly associated with major depressive disorder. In addition, a combination of these biomarkers very accurately predicted MDD from healthy controls. Patients with MDD were characterized by increased levels of Bacteroides and decreased levels of Blautia and Eubacterium. Gut bacterial metabolites that correlated with MDD such as gamma aminobutyrate, phenylalanine and tryptophan are ones that are known to be absorbed into the bloodstream, affect neurochemistry and have been implicated in MDD.Similar to previous studies, there is increasing evidence that the microbiome plays an overall significant role in overall health and well-being. It will be interesting and exciting to see how future clinical trials take this information into new therapeutic regiments. Submitted by Michael L. Moore, MD