Week 4 – NLST

“Reduced Lung-Cancer Mortality with Low-Dose Computed Tomographic Screening”

by the National Lung Cancer Screening Trial Research Team

N Engl J Med. 2011 Aug 4;365(5):395-409 [NEJM free full text]

Despite a reduction in smoking rates in the United States, lung cancer remains the number one cause of cancer death in the United States, as well as worldwide. Earlier studies of plain chest radiograph for lung cancer screening demonstrated no benefit, and thus in 2002 the National Lung Screening Trial (NLST) was undertaken to determine whether then-recent advances in CT technology could lead to an effective lung cancer screening method.

Population: adults age 55-74 with 30+ pack-years of smoking (if former smokers, they must have quit within the past 15 years)
Intervention: three annual screenings for lung cancer with low-dose CT
Comparison: three annual screenings for lung cancer with PA chest radiograph
Outcome: 1º = mortality from lung cancer, 2º = mortality from any cause and incidence of lung cancer

53,454 patients were randomized, and both groups had similar baseline characteristics. The low-dose CT group demonstrated 247 deaths from lung cancer per 100,000 person-years, whereas the radiography group demonstrated 309 deaths per 100,000 person-years. Thus a relative reduction in rate of death by 20.0% was seen in the CT group (95% CI 6.8 – 26.7%, p = 0.004). The number needed to screen with CT to prevent one lung cancer death was 320. There were 1877 deaths from any cause in the CT group and 2000 deaths in the radiography group; thus CT screening demonstrated a risk reduction of death from any cause of 6.7% (95% CI 1.2% – 13.6%, p = 0.02). Incidence of lung cancer in the CT group was 645 per 100,000 person-years and 941 per 100,000 person-years in the radiography group (RR 1.13, 95% CI 1.03 – 1.23).

Lung cancer screening with low-dose CT scan in high-risk patients provides a significant mortality benefit.

This trial was stopped early because the mortality benefit was so high. The benefit was driven by the reduction in deaths attributed to lung cancer – and when deaths from lung cancer were excluded from the overall mortality analysis there was no significant difference among the two arms. Largely on the basis of this study, the 2013 USPSTF guidelines for lung cancer screening recommend annual low-dose CT scan in patients who meet NLST inclusion criteria.

Per UpToDate, there are seven low-dose CT screening trials in progress in Europe. It is hoped that meta-analysis of all such RCTs will allow for further refinement in risk stratification, frequency of screening, and management of positive screening findings.

Of note, no randomized trial has ever demonstrated a mortality benefit of chest radiography for lung cancer screening. The Prostate, Lung, Colorectal, and Ovarian (PLCO) Cancer Screening Trial tested this modality vs. “community care,” and because the PLCO trial was ongoing at the time of creation of the NSLT, the NSLT authors trial decided to compare their intervention (CT) to chest radiography, in case the results of chest radiography in PLCO were positive (ultimately they were not).

Further Reading:
1. USPSTF Guidelines for Lung Cancer Screening (2013)
2. ClinicalTrials.gov
3. Wiki Journal Club
4. 2 Minute Medicine

Summary by Duncan F. Moore, MD

Week 3 – Dexamethasone in Bacterial Meningitis

“Dexamethasone in Adults With Bacterial Meningitis”

N Engl J Med 2002; 347:1549-1556 [NEJM free full text]

The current standard of care in the treatment of suspected bacterial meningitis in the developed world includes the administration of dexamethasone prior to or at the time of antibiotic initiation. The initial evaluation of this practice in part stemmed from animal studies which demonstrated that dexamethasone reduces CSF concentrations of inflammatory markers as well as neurologic sequelae after meningitis. RCTs in the pediatric literature also demonstrated clinical benefit. The best prospective trial in adults was this 2002 study by de Gans et al.

Population: adults with suspected meningitis

Intervention: dexamethasone 10mg IV q6hrs x4 days started 15-20 minutes before first IV abx

Comparison: placebo IV with same administration as above

primary = Glasgow Outcome Scale at 8 weeks (1 = death, 2 = vegetative state, 3 = unable to live independently, 4 = unable to return to school/work, 5 = able to return to school/work)
secondary = death, focal neurologic abnormalities, and others
subgroup analyses performed by organism

301 patients were randomized. At 8 weeks, 15% of dexamethasone patients had an unfavorable outcome (Glasgow Outcome Scale score of 1-4), vs. 25% of placebo patients (RR 0.59, 95% CI 0.37 – 0.94, p= 0.03). Among patients with pneumococcal meningitis, 26% of dexamethasone patients had an unfavorable outcome, vs. 52% of placebo patients. There was no significant difference among treatment arms within the subgroup of patients infected with meningococcal meningitis. Overall, death occurred in 7% of dexamethasone patients and 15% of placebo patients (RR 0.48, 95% CI 0.24 – 0.96, p = 0.04). In pneumococcal meningitis, 14% of dexamethasone patients died, vs. 34% of placebo patients. There was no difference in rates of focal neurologic abnormalities or hearing loss in either treatment arm (including within any subgroup).

Early adjunctive dexamethasone improves mortality in bacterial meningitis.

As noted in the above subgroup analysis, this benefit appears to be driven by the efficacy within the pneumococcal meningitis subgroup. Of note, the standard initial treatment regimen in this study was amoxicillin 2gm q4hrs for 7-10 days, not our standard ceftriaxone + vancomycin +/- ampicillin. Largely on the basis of this study alone, the IDSA guidelines for the treatment of bacterial meningitis (2004) recommend dexamethasone 0.15 mg/kg q6hrs for 2-4 days with first dose administered 10-20 min before or concomitant with initiation of antibiotics. Dexamethasone should be continued only if CSF Gram stain, CSF culture, or blood cultures are consistent with pneumococcus.

Further Reading:
1. IDSA guidelines for management of bacterial meningitis (2004)
2. Wiki Journal Club
3. 2 Minute Medicine

Summary by Duncan F. Moore, MD

Week 2 – AFFIRM

“A Comparison of Rate Control and Rhythm Control in Patients with Atrial Fibrillation”

by the Atrial Fibrillation Follow-Up Investigation of Rhythm Management (AFFRIM) Investigators

N Engl J Med. 2002 Dec 5;347(23):1825-33. [NEJM free full text]

It seems like the majority of patients with atrial fibrillation that we encounter as residents today are being treated with a rate control strategy, as opposed to a rhythm control strategy. There was a time when both approaches were considered acceptable, and perhaps rhythm control was even the preferred initial strategy. The AFFIRM trial was the landmark study to address this debate.

Population: patients with atrial fibrillation (judged “likely to be recurrent”), age 65 or older “or who had other risk factors for stroke or death,” and in whom anticoagulation was not contraindicated

Intervention: rhythm control strategy with one or more drugs from a pre-specified list and/or cardioversion to achieve sinus rhythm

Comparison: rate control strategy with beta-blockers, CCBs, and/or digoxin to a target resting HR ≤ 80 and a six-minute walk test HR ≤ 110

– primary endpoint – death during follow-up (per Kaplan-Meier estimator)
– secondary endpoint – composite end point of death, disabling stroke, disabling anoxic encephalopathy, major bleeding, and cardiac arrest
– secondary analyses – primary end point in pre-specified subgroups (e.g. age ≥ 65, comorbid CAD, etc.)

4060 patients were randomized in this multi-center RCT. Death occurred in 26.7% of rhythm control patients versus 25.9% of rate control patients (HR 1.15, 95% CI 0.99 – 1.34, p = 0.08). The composite secondary endpoint occurred in 32.0% of rhythm control patients versus 32.7% of rate control patients (p = 0.33). Rhythm control strategy was associated a higher risk of death among patients older than 65 and patients with CAD (see Figure 2). Additionally, rhythm control patients were more likely to be hospitalized during follow-up (80.1% vs. 73.0%, p < 0.001) and to develop torsades de pointes (0.8% vs. 0.2%, p = 0.007).

A rhythm control strategy in atrial fibrillation offers no mortality benefit over a rate control strategy.

At the time of publication, the authors wrote that rate control was an “accepted, though often secondary alternative” to rhythm control. Their study clearly demonstrated that there was no significant mortality benefit to either strategy, that hospitalizations were greater in the rhythm control group, and in subgroup analysis that rhythm control led to higher mortality among the elderly and those with CAD. Notably, 37.5% of rhythm control patients had crossed over to rate control strategy by 5 years follow-up, whereas only 14.9% of rate control patients had switched over to rhythm control.

But what does this study mean for our practice today? Generally speaking, rate control is preferred in most patients, particularly the elderly and patients with CHF, whereas rhythm control may be pursued in patients with persistent symptoms despite rate control, patients unable to achieve rate control on AV nodal agents alone, and patients younger than 65. Both the AHA/ACC (2014) and the European Society of Cardiology (2016) guidelines have extensive recommendations that detail specific patient scenarios.

Further Reading:
1. Cardiologytrials.org
2. Wiki Journal Club
3. 2 Minute Medicine

Summary by Duncan F. Moore, MD


“Intensive versus Conventional Glucose Control in Critically Ill Patients”

by the Normoglycemia in Intensive Care Evaluation–Survival Using Glucose Algorithm Regulation (NICE-SUGAR) investigators

N Engl J Med 2009; 360:1283-1297. [NEJM free full text]

On the wards we often hear 180 mg/dL used as the upper limit of acceptable for blood glucose, with the understanding that tighter glucose control in inpatients can lead to more harm than benefit. Interestingly, the relevant evidence base comes from ICU populations, with scant direct data in non-ICU patients. The 2009 NICE-SUGAR study is the largest and best among this evidence base.

Population: ICU patients expected to require 3 or more days of ICU-level care
Intervention: “intensive” glucose control = target glucose 81 to 108 mg/dL
Comparison: conventional glucose control = target of less than 180 mg/dL
Outcome: primary = 90-day all-cause mortality rate

6104 patients were randomized to the two arms, and both groups had similar baseline characteristics. 27.5% of patients in the intensive-control group died versus 24.9% in the conventional-control group (OR 1.14, 95% CI 1.02-1.28, p= 0.02). Severe hypoglycemia (< 40 mg/dL) was found in 6.8% of intensive patients but only 0.5% of conventional patients.

Intensive glucose control increases mortality in ICU patients.

Notably, only 20% of these patients had diabetes mellitus, suggesting that much of the hyperglycemia treated in this study (97% of intensive group received insulin, 69% of conventional) was from stress, critical illness, and corticosteroid use. For ICU patients, intensive insulin therapy is clearly harmful, but the ideal target glucose range remains controversial and by expert opinion appears to be 140-180. For non-ICU inpatients with or without diabetes mellitus, the ideal glucose target is also unclear – the ADA recommends 140-180, and the Endocrine Society recommends a pre-meal target of < 140 and random levels < 180.

Further reading:
1. ADA Standards of Medical Care in Diabetes 2016 (skip to page S99)
2. Wiki Journal Club

Summary by Duncan F. Moore, MD