PowerPoint Slides

Screening for Cancer

English Text Screening for Cancer VideoTranscript Professional Oncology Education Screening for Cancer Time: 22:42 Therese B. Bevers, M.D. Professor, Clinical Cancer Prevention Medical Director, Cancer Prevention Center The University of Texas, MD Anderson Cancer Center Hello, I am Dr. Terry Bevers, Professor of Clinical Cancer Prevention and Medical Director of the Cancer Prevention Center at the University of Texas MD Anderson Cancer Center. Today, I am going to talk to you about screening for cancer.

Spanish Translation Exámenes preventivos del cáncer Transcripción del video Educación Oncológica Profesional Exámenes preventivos del cáncer Duración: 22:42 Dra. Therese B. Bevers Profesora de Prevención Clínica del Cáncer Directora Médica, Centro de Prevención del Cáncer Universidad de Texas, MD Anderson Cancer Center Hola. Soy la Dra. Terry Bevers, Profesora de Prevención Clínica del Cáncer y Directora Médica del Centro de Prevención del Cáncer del MD Anderson Cancer Center de la Universidad de Texas. Hoy hablaré de los exámenes preventivos del cáncer.

Screening for Cancer Therese B. Bevers, M.D. Professor, Clinical Cancer Prevention Medical Director, Cancer Prevention Center

1

Screening for Cancer Objectives

Upon completion of this lesson, participants will be able to:

The objectives of today’s lesson are that participants will be able to: describe the characteristics of populations that determine whether screening is appropriate; identify elements that determine efficacy with regard to screening tests; and identify reliable sources of screening recommendation guidelines.

Los objetivos de esta lección son que los participantes puedan describir las características de las poblaciones que determinan si es apropiado hacer exámenes preventivos, identificar los elementos que establecen su eficacia, e identificar fuentes confiables de pautas para recomendarlos.

So, why should we screen for cancer? It has been estimated that over 50% of new cancers occur at sites where screening has the potential to reduce cancer deaths. Screening results in finding cancer at an earlier stage when it is more treatable and it can also decrease the burden of suffering from cancer both morbidity and mortality.

¿Por qué debemos hacer exámenes preventivos? Se ha estimado que más del 50% de los nuevos casos ocurren en sitios donde estos exámenes podrían reducir las muertes por cáncer. El resultado es la detección en una etapa más precoz, cuando es más tratable, lo que puede reducir la carga de sufrir de cáncer, tanto su morbilidad como su mortalidad.

•Describe characteristics of populations that determine whether screening is appropriate •Identify elements that determine efficacy with regard to screening tests •Identify reliable sources of screening recommendation guidelines

Screening for Cancer Why Screen for Cancer

• Over 50% of new cancers occur at sites where screening has the potential to reduce mortality • Screening results in finding cancer at an earlier stage when it is more treatable • Screening decreases the burden of suffering from cancer (morbidity and mortality)

2

Screening for Cancer Reasons Not to Screen

• Rare cancers • Cancer does not cause significant morbidity or mortality • Early diagnosis doesn’t lead to better outcome

There are, however, reasons not to screen a person for cancer. We don’t want to be screening large populations for a cancer that is very rare and likely to occur in only few in the population. It is also not appropriate to screen if the cancer does not cause significant morbidity or mortality, or if early diagnosis doesn’t lead to a better outcome. Additionally, if there is no effective treatment, there is no reason to be looking for a cancer earlier when we don’t have anything to offer that individual.

Existen razones para no someter a una persona a exámenes preventivos. No queremos hacerlos en grandes poblaciones para detectar un cáncer infrecuente que probablemente ocurra en pocas personas. Tampoco son adecuados si el cáncer no causa morbilidad o mortalidad importantes, o si el diagnóstico precoz no lleva a un mejor resultado. Si no existe un tratamiento eficaz, no hay razón para buscar un tipo de cáncer cuando no tenemos nada que ofrecer al paciente.

So, who should we screen? We should screen asymptomatic individuals who have at least a 10-year life expectancy. Studies have shown that it takes 7 to 10 years to see the benefit from population screening.

¿A quiénes debemos hacer estos exámenes? A personas asintomáticas con una esperanza de vida de al menos 10 años. Se ha demostrado que el beneficio de estos exámenes demora de 7 a 10 años.

• No effective treatment

Screening for Cancer Who to Screen

• Asymptomatic patients • At least 10 year life expectancy

3

Screening for Cancer Who Not to Screen

• Patients with significant co-morbidities that would limit treatment of abnormal findings

There are individuals that we should not screen, those who have significant medical conditions that would limit our ability to offer them treatment options, or to be able to diagnose abnormalities. Additionally, if they’re unwilling to undergo diagnostic evaluation of the finding, or unwilling to undergo treatment, they should not be screened.

Hay personas a quienes no debemos hacer estos estudios: aquellas con condiciones médicas importantes que limitarían las opciones de tratamiento o de diagnosticar anomalías. Si rechazan una evaluación de diagnóstico de los resultados, o un tratamiento, no debemos hacerlos.

I do want to pause for a minute and distinguish between a screening and a diagnostic test. A screening test is performed on an asymptomatic individual. This is done when they don’t have symptoms, trying to find a cancer at an earlier stage. One of the most common reasons cited in my clinic for individuals not wanting to undergo a screening test, certainly like a colonoscopy, is that they don’t have symptoms. That is exactly the population that we are targeting with a screening test. Diagnostic tests, however, are done to evaluate a problem, someone who actually has a symptom, or if they have had an abnormal screening test.

Quiero hacer la distinción entre un examen preventivo y una prueba de diagnóstico. El examen preventivo se realiza en una persona asintomática, cuando no tiene síntomas, para encontrar cáncer en una etapa precoz. Una de las razones mencionadas por quienes no desean un examen preventivo, como una colonoscopia, es que no tienen síntomas. Esa es exactamente la población a la que apuntamos con los exámenes preventivos. Las pruebas de diagnóstico permiten evaluar un problema en alguien que ya tiene un síntoma o que ha tenido un resultado anormal en un examen preventivo.

• Patients unwilling to undergo diagnostic evaluation or treatment of abnormal findings

Screening for Cancer Screening vs. Diagnostic Testing

• Screening Test – Test performed on asymptomatic individuals to detect cancer at an early stage

• Diagnostic Test – Testing done to evaluate a problem (symptom, abnormal screening test)

4

Screening for Cancer

To take it a step further, screening tests usually do not diagnose cancer. They identify a potential problem that we then need to evaluate with diagnostic tests to see if it is cancer or if it is benign.

Los exámenes preventivos generalmente no diagnostican el cáncer, sino que identifican un posible problema a evaluar con pruebas de diagnóstico para comprobar si es cáncer o si es benigno.

There are at least two requirements that must be met for a screening to be considered efficacious. The test or procedure must be available to detect cancers earlier than if the cancer were detected when the individual presented with their symptoms. And we also have to have effective treatments that will change the outcome if we start the treatment earlier.

Un examen preventivo eficaz cumple al menos dos requisitos. El examen o procedimiento debe detectar el cáncer antes de que el paciente tenga síntomas y, además, debemos tener tratamientos eficaces que cambien el resultado si comenzamos el tratamiento antes.

• Screening tests usually do not diagnose cancer • If a screening test result is abnormal, more tests may be done to check for cancer – These are diagnostic tests

Screening for Cancer Effective Cancer Screening Tests

At least two requirements must be met for screening to be efficacious: •A test or procedure must be available to detect cancers earlier than if the cancer were detected as a result of the development of symptoms •Evidence must be available that treatment initiated earlier as a consequence of screening results in an improved outcome

5

In talking about the evidence that supports…

Si nos referimos a las pruebas que apoyan...

…current screening recommendations, it is helpful to understand the different levels of evidence. This particular slide reviews the levels of evidence from the highest, randomized control trials, down to the lowest, which is opinions of respected authorities.

... las recomendaciones actuales para los exámenes preventivos, hay distintos niveles de evidencia. Vemos aquí los niveles de las pruebas, desde el más alto — ensayos aleatorizados y controlados— hasta el más bajo —las opiniones de autoridades respetadas—.

Screening for Cancer

Evidence Supporting Current Screening Recommendations

Screening for Cancer Levels of Evidence 1.

Randomized controlled trials

–

Trial should be double-blinded to prevent bias

2.

Nonrandomized controlled trials

3.

Observational studies (cohort or case-control studies) – Provide indirect evidence of effectiveness of screening test and are hypothesis generating Cohort: Persons already exposed to a risk factor or not exposed are followed over time Case-control: Patients selected on whether or not they have a disease and then look at how the groups differ

4.

Descriptive studies

5.

Opinions of respected authorities

6

Screening for Cancer

A randomized controlled trial, with mortality reduction as the endpoint, is considered the gold standard for determining the usefulness of a screening test.

Un ensayo aleatorizado y controlado cuyo criterio de valoración es la reducción de mortalidad, es la norma de excelencia para determinar la utilidad de un examen.

Unlike treatment recommendations, which are commonly the result of randomized clinical trials, many prevention recommendations come from observational studies. In fact, very few are based on randomized trials.

A diferencia de las recomendaciones de tratamiento que resultan de ensayos clínicos aleatorizados, las recomendaciones de prevención provienen de estudios observacionales. Muy pocas se basan en ensayos aleatorizados.

A randomized controlled trial with mortality reduction as the endpoint is the gold standard for determining usefulness of a screening test

Screening for Cancer Differences in Available Data

Prevention •

Predominately observational studies

•

Few randomized trials

Treatment •

Primarily randomized clinical trials

? Reliability of Data

7

Screening for Cancer Screening Recommendations from Randomized Trials

Endpoint: Cancer Mortality Reduction Beneficial Outcome

Negative Outcome

• Mammography • Fecal occult blood

• Breast self examination • CXR

Screening for Cancer Most Screening Recommendations are Based on Indirect or Observational Data

• Pap smear • Prostate screening antigen (PSA)/ digital rectal exam (DRE) • Colonoscopy • Clinical breast examination

If we look at randomized trials, which had cancer mortality as the endpoint, we see that only mammography and fecal occult blood have had a beneficial outcome. And within the past week, we’ve had evidence from a large randomized controlled trial about spiral CT scans for lung cancer screening also showing a beneficial outcome. And then two screening tests, which have been shown in randomized controlled trials to have a negative outcome, and that is: breast self-examination and chest x-ray.

En los ensayos aleatorizados donde la mortalidad por cáncer es el criterio de valoración, sólo la mamografía y la prueba de sangre en materia fecal han sido beneficiosas. La evidencia reciente de un importante ensayo aleatorizado y controlado demuestra que las tomografías computadas en espiral son beneficiosas para prevenir el cáncer de pulmón. Dos exámenes preventivos con resultados negativos en ensayos aleatorizados y controlados son el autoexamen de mama y la radiografía de tórax.

Most clinicians are surprised to find that many cancer screening recommendations are based on indirect or observational data. As long as Pap smears have been around, it is thought often that we have a multitude of randomized control trials supporting their use in the clinical setting. In fact, it’s based on significant observational data that shows that women having undergone screening with a Pap smear are less likely to die from it. But it’s observational and not compared to a controlled group. Similarly, with colonoscopy, one of the most commonly recommended colorectal cancer screening tests, we do not have any randomized controlled trials supporting its use. However, we have substantial indirect evidence of its benefit.

Los médicos clínicos se sorprenden de que muchas recomendaciones de exámenes preventivos se basen en datos indirectos o de observación. Para los exámenes de Papanicolaou, prevalece la idea de que varios ensayos aleatorizados y controlados respaldan su uso en el ámbito clínico. Esto se basa en importantes datos de observación que demuestran que reduce la probabilidad de morir de cáncer. Son datos observacionales y no están comparados con un grupo controlado. Con respecto a la colonoscopia, uno de los exámenes preventivos del cáncer colorrectal más recomendados, ningún ensayo aleatorizado y controlado respalda su uso, pero tenemos pruebas indirectas sustanciales de su beneficio.

8

Screening for Cancer Challenges in Conducting RCT for Screening •

Support of scientific community

•

Funding

•

Study population

•

Informed consent

•

Study treatments

•

Participant management

•

Sample size

•

Compliance issues

•

Study duration

•

Data management

•

Toxicity/safety issues

•

Planning for the future

•

Participant recruitment

Screening for Cancer Scale of Screening Trials

Study

Sample Size

Study Duration

Shanghai

266,064

11 years

PLCO

154,000

10 years

NLST

50,000

10 years

Thomas DB et al. J Natl Cancer Inst 2002 94(19): 1445

The reason that we don’t have a lot of randomized controlled trials for the screening tests that are recommended is that there are number of challenges in conducting these trials. I would like to highlight two of them: the sample size and study duration.

No tenemos muchos ensayos aleatorizados y controlados para los exámenes preventivos recomendados porque los ensayos encuentran una serie de inconvenientes. Dos de ellos son el tamaño de la muestra y la duración del estudio.

This table shows the scale of some of the recent or ongoing clinical trials that are looking at cancer screening tests. The Shanghai Trial is a study that was done in Shanghai, China looking at breast self-examination. This study enrolled over 266,000 women. And they were followed for 11 years; clearly a massive undertaking. The PLCO, which stands for prostate, lung, colorectal, ovarian, which is looking at screening tests for each of those particular disease sites, has enrolled almost 155,000 individuals. and has followed them for over 10 years. And this trial is still ongoing. The National Lung Screening Trial or NLST is the one that I just referenced a moment ago, that is a randomized controlled trial comparing spiral CTs of the lungs to chest x-ray. This trial, which opened earlier this decade, enrolled over 50,000 individuals who were high-risk smokers, thus the need for a smaller sample size. And they have been followed for approximately 10 years to be able to get preliminary data. Clearly, given the scale of these trials, we cannot study each screening test as in depth as we would like. And we often have to look at other ways of

Esta tabla muestra la escala de algunos ensayos clínicos recientes o en curso que evalúan los exámenes preventivos del cáncer. El Ensayo de Shanghái (realizado en China) sobre el autoexamen de mama incluyó más de 266,000 mujeres con un seguimiento a 11 años; claramente, una enorme tarea. El estudio PLCO de cáncer de próstata, pulmón, colorrectal y ovario analiza los exámenes preventivos para estos sitios particulares y tiene casi 155,000 participantes seguidos por más de 10 años. Aún está en curso. El Ensayo Nacional Preventivo del Pulmón ya mencionado es un estudio aleatorizado y controlado que compara tomografías computadas en espiral de los pulmones con radiografías de tórax. Se inició a principios de esta década con más de 50,000 fumadores de alto riesgo —de ahí la necesidad de una muestra menor—. Los pacientes fueron seguidos por unos 10 años para obtener datos preliminares. Dada la magnitud de estos ensayos, no podemos estudiar cada examen preventivo con la profundidad adecuada, y con frecuencia debemos hallar otras formas de comprender sus beneficios.

9

Screening for Cancer Surrogate Endpoints

Because of the scale of screening trials, cancer mortality is often not the primary endpoint and surrogate endpoints are used: •Cancer incidence •Cancer stage •Development of precursor lesions (e.g., cervical dysplasia, adenomatous polyps)

understanding the benefits. Sometimes we do that by using surrogate endpoints. While, clearly, we would like to look at cancer mortality as the primary endpoint, because that is the most important thing is, "Do fewer individuals die from the disease by doing this test?" It’s oftentimes unrealistic. So, at times, we will look at cancer incidence. Can we actually find the cancer at an earlier stage? Or what is the difference in stage between screening and not screening that is found through the trial? And more and more we are using precursor lesions. A recent trial of virtual colonoscopy, also called CT colonography, which is now an approved cancer screening test, used adenomatous polyp identification as the endpoint, not the development of colorectal cancer or colorectal cancer mortality. Clearly, we were able to accomplish the trial with smaller numbers and in a shorter period of time by using precursor lesions as the endpoint. Cervical dysplasia is often used as the endpoint in cervical cancer screening trials, and not the development of cervical cancer or cervical cancer deaths. So, how do we evaluate a screening test?

A veces usamos criterios de valoración indirectos. Debemos considerar la mortalidad por cáncer como criterio principal, dado que es lo más importante, pero ¿esta prueba permite reducir la mortalidad? A menudo, es poco realista. Entonces, a veces tenemos en cuenta la incidencia del cáncer. ¿Podemos realmente detectarlo en una etapa más temprana? ¿Qué diferencia de estadio se registra a través del ensayo con o sin un examen preventivo? Y utilizamos cada vez más las lesiones precursoras. Un ensayo reciente de la colonoscopia virtual o por tomografía computada —un nuevo examen preventivo del cáncer— usó como criterio de valoración la identificación de pólipos adenomatosos y no el desarrollo de cáncer colorrectal o su mortalidad. Pudimos hacer el ensayo con menos participantes y en un período más breve utilizando como criterio las lesiones precursoras. En los ensayos para determinar los exámenes preventivos del cáncer cervical, el criterio suele ser la displasia cervical y no el desarrollo del cáncer cervical o las muertes. ¿Cómo evaluar un examen preventivo?

Screening for Cancer

Evaluating a Screening Test

10

Screening for Cancer CANCER OUTCOME

TEST

Screening for Cancer Sensitivity Proportion of Persons with a Condition Who Correctly Test Positive When Screened Goal: Low false negatives (less missed disease)

I am going to use this table to explain some different outcome measures or efficacy measures. The top here looks at the outcome, whether the individual actually has cancer or not. And then this ordinate looks at the outcome of the test, whether the test found a cancer or whether it did not. Using these particular variables, you can see that if the outcome was that the individual had cancer and the test was positive, that is considered a true positive. Whereas, if the individual did not have cancer and the test was negative, that is considered a true negative. However, we also see other outcomes such as the test was negative --- [I’m sorry] --- the outcome was negative, but the test was positive, we call that a false positive. Additionally, we can see where the outcome is that the individual actually has cancer, but the screening test did not find it. That’s called a false negative. Clearly, that is the worst condition and that is one we want to minimize the most with any cancer screening test. So, looking at different measures of efficacy of cancer screening tests, the first one to discuss is sensitivity. The definition of sensitivity is the proportion of persons with a condition, those who actually have cancer, who the test correctly found the cancer when it was screened. The goal of a screening test is to have low false negatives for missing less disease. The way we can calculate sensitivity is to take the true positives and divide it by the sum of true positives plus false negatives. The denominator you can see is really all of those who had the disease, but some of which, in which they did not test positive for the disease, and that was a false negative.

Utilizaré esta tabla para explicar algunas medidas con resultados o eficacia diferentes. En la parte superior tenemos el resultado, si la persona tiene cáncer o no, y el eje vertical indica el resultado del examen, si se detectó o no cáncer. Con estas variables particulares, si el paciente tiene cáncer y el examen es positivo, el caso se considera un verdadero positivo; pero si la persona no tiene cáncer y el examen da negativo, se considera un verdadero negativo. Pero también tenemos otros resultados: cuando el resultado fue negativo, pero el examen dio positivo, que es un falso positivo. Además, hay resultados en que una persona tiene cáncer, pero el examen preventivo no lo detecta. Este caso se denomina falso negativo. Por cierto, este es el peor caso y el que queremos minimizar en mayor medida con cualquier examen preventivo del cáncer.

Entre las medidas de eficacia de los exámenes preventivos, la primera es la sensibilidad, definida como la proporción de personas con una condición; son quienes realmente tienen cáncer y donde el examen preventivo lo detectó correctamente. Un examen preventivo debe tener un bajo número de falsos negativos para reducir la probabilidad de que la enfermedad pase inadvertida. Para calcular la sensibilidad, dividimos los verdaderos positivos entre la suma de los verdaderos positivos más los falsos negativos. El denominador refleja todas las personas con la enfermedad, pero incluye algunas sin un resultado positivo en el examen, los falsos negativos.

11

Screening for Cancer Specificity Proportion of Persons Without the Condition Who Correctly Test Negative When Screened Goal: Low false positives (avoid unnecessary diagnostic testing)

Screening for Cancer Positive Predictive Value Proportion of Persons With a Positive Test Who Have the Condition

Another efficacy measure is specificity. This is the proportion of persons who do not have cancer and who correctly test negative when they are screened. The goal of this is to have low false negatives. In other words, we want to avoid unnecessary diagnostic testing. If they don’t have the disease, we don't need to be putting them through additional testing. The way we can calculate this is to take the true negatives and divide it by the sum of the true negatives plus the false positives, in other words, all of those who don’t actually have the disease.

Otra medida de eficacia es la especificidad, la proporción de personas sin cáncer cuyos exámenes preventivos dan correctamente negativo. El objetivo es un bajo número de falsos negativos para evitar pruebas de diagnóstico innecesarias. Si un paciente no tiene la enfermedad, no es necesario hacer más pruebas. Para calcularla, dividimos los verdaderos negativos entre la suma de los verdaderos negativos más los falsos positivos, es decir, quienes en realidad no tienen la enfermedad.

The positive predictive value are the proportion of persons with a positive test who actually have the condition. This is calculated as the true positive divided by the sum of true positives plus false positives. In other words, all of the positive tests divided into those that were actually true positives.

El valor predictivo positivo es la proporción de personas con un examen positivo que tienen la enfermedad. Se calcula dividiendo los verdaderos positivos entre la suma de verdaderos positivos y falsos positivos. Todos los exámenes positivos se dividen entre los verdaderos positivos.

12

Screening for Cancer Negative Predictive Value Proportion of Persons With a Negative Test Who are Truly Disease-free

Screening for Cancer Biases

• Selection • Healthy volunteer • Lead time bias • Length bias

The negative predictive value are the proportion of persons with a negative test who are truly disease-free. This is true negatives divided by the sum of true negatives plus false negatives. These particular efficacy tests help us to better understand how good a cancer screening test is.

El valor predictivo negativo es la proporción de personas con un examen negativo que no tienen la enfermedad. Son los verdaderos negativos divididos entre la suma de verdaderos negativos y falsos negativos. Estos parámetros de eficacia ayudan a comprender la utilidad de un examen preventivo.

In looking at the particular studies that have examined cancer screening tests, it’s important to realize, if it is not a randomized control trial, there are a number of inherent biases, certainly selection bias, those who are selected to participate. You may be are going to a particular clinic and thus you have already selected out individuals who go to a doctor versus those who don’t. You are looking at a healthy volunteer. We tend to see those who volunteer to participate in studies tend to be healthier overall, oftentimes having healthier lifestyles, and we may be measuring that variable and not the effect of the screening test. The two, however, that I want to spend the most time on are this lead time bias and the length bias because these are some of the greatest confounders in understanding cancer screening tests.

Al observar los estudios que han analizado los exámenes preventivos del cáncer, es importante tener en cuenta —si no es un ensayo aleatorizado y controlado— que hay una serie de sesgos inherentes o de selección entre los participantes. Posiblemente se trate de una clínica particular y, por lo tanto, sólo se habrán seleccionado pacientes que consultan a un médico, que son voluntarios sanos. Por lo general, las personas que participan voluntariamente en los estudios suelen ser más sanas con estilos de vida sanos, y es posible que estemos midiendo estas variables y no el efecto del examen preventivo. Quiero destacar los sesgos por adelanto en el diagnóstico y por duración de la enfermedad, que suelen dificultar la interpretación de los exámenes preventivos.

13

Screening for Cancer Lead Time

• Period of time from when a cancer is detected by screening and when it would have presented with symptoms – Short lead time: screening not useful – Long lead time: screening more effective

• Lead time bias can occur if you look at survival time (time from diagnosis to death) and not at the impact on mortality (for screened versus unscreened individuals)

Screening for Cancer Lead Time Bias Screening Diagnosis Cancer Forms

Symptoms

Death

Cancer Forms

Symptoms

Death

Symptom Diagnosis

The lead time is the period of time from when a cancer is detected by screening and when it would have actually presented when the patient developed symptoms. Now, if there is a very short lead time, then the screening is not useful. What this is saying is the screening only found the cancer at a very short period of time, maybe weeks or months before it actually presented with symptoms. In that case, the cancer hasn’t likely to have advanced significantly and we’re not really changing the outcome. A long lead time, however, makes screening more effective. If we are able to find a cancer years before it would have presented with symptoms, then clearly we are having the opportunity to catch it at a much smaller stage before it has progressed and change the outcome. Lead time bias can occur if you look at survival time. That is the time from diagnosis to death, and not at the impact of the screening test on mortality, that is comparing the number of individuals in the screened population that died versus the number of unscreened individuals who died. I think this graph helps to illustrate that a bit. So, this is the lead time right here. This is the amount of time that we have advanced the detection of the cancer by doing the screening test here. What we are talking about is that the cancers in these two situations developed at the exact same time, but we found the cancer at an earlier point in time than if we waited for the development of symptoms. Now, in this particular scenario what we see is: death, however, occurs at the same time. So, we have not changed the outcome. The individual still died at the same time as they would have had we waited for the development of symptoms. So, mortality is not changed. However, they appear to have survived their cancer longer from the point of diagnosis, because diagnosis was with a screening test, than the diagnosis from when they presented with symptoms. So, in this situation screening -- [sorry] --- survival appears longer, but really we

El adelanto en el diagnóstico es el período entre la detección del cáncer con un examen preventivo y el momento en que el paciente hubiera desarrollado síntomas. Si el adelanto es muy breve, el examen preventivo no será útil, pues habrá detectado el cáncer un poco antes, tal vez semanas o meses, de que ocurrieran los síntomas. Probablemente el cáncer no hubiera avanzado de manera significativa y no estaríamos modificando el resultado. Un adelanto prolongado en el diagnóstico refleja un examen preventivo más eficaz. Si podemos detectar un cáncer años antes de los síntomas, habrá una clara oportunidad de atacarlo en una etapa mucho más precoz, antes de que haya progresado, y cambiar el resultado. El sesgo de duración de la enfermedad ocurre al observar el tiempo de supervivencia. Es el período desde el diagnóstico hasta la muerte, y no el impacto del examen preventivo en la mortalidad, que compara la cantidad de pacientes que murieron de la población examinada con los que murieron sin ser examinados. Este gráfico lo ilustra. Este es el tiempo de adelanto en el diagnóstico, el período en que hemos adelantado la detección del cáncer con el examen preventivo. En estas dos situaciones, el cáncer se desarrolla exactamente en el mismo momento, pero uno fue detectado antes de tener síntomas. En este escenario en particular, la muerte ocurre al mismo tiempo y no hemos cambiado el resultado. El paciente muere en el mismo momento que si hubiéramos esperado el desarrollo de síntomas y la mortalidad no varía; sin embargo, parece que hubiera sobrevivido al cáncer por más tiempo desde el momento del diagnóstico, ya que hubo un examen preventivo antes de los síntomas. La supervivencia parece mayor, pero no hemos cambiado el resultado. Este paciente en particular no se benefició con el examen preventivo, sino que supo que tenía cáncer por más tiempo. Probablemente haya recibido más tratamientos, pero el final fue el mismo.

14

Screening for Cancer Length Bias

• Slow-growing tumors are more likely to be picked up by screening whereas aggressive tumors are more likely to occur in between screening intervals • Therefore screening is more likely to pick up more indolent tumors which will have a better prognosis

haven't changed the outcome. This particular individual really did not benefit from cancer screening. He knew about the cancer longer. Yes, he probably went through more treatments, but it didn’t change the end. He probably also incurred more side effects from the treatments than --- then ---then did the individual who presented and was diagnosed based on symptoms. Additionally, he likely had more anxiety that he had to deal with during that interval than this one, as just the interval is shorter overall. Another bias is length bias. This is where we see that slow growing tumors are more likely to be picked up by a screening test. Whereas aggressive tumors are more likely to be diagnosed as an interval cancer. In other words, one that occurs in the interval between cancer screening tests. What we see is that slower growing tumors tend to have a better prognosis. They are more indolent tumors in the first place and, thus, screening is more likely to pick up more indo --- indolent tumors. And these have a better prognosis, which could actually skew the apparent benefit from the screening test making it look like that fewer people died from the screening when in fact they were less likely to die from it because they had more indolent tumors.

Es posible que también haya tenido más efectos secundarios de los tratamientos que un paciente que recibió su diagnóstico sobre la base de sus síntomas. Además, quizás sufrió más ansiedad que la necesaria durante ese intervalo que este otro paciente, donde el intervalo general fue más corto.

Otro sesgo es la duración de la enfermedad. Aquí es donde comprobamos que los tumores de crecimiento lento tienen mayor probabilidad de ser detectados por un examen preventivo, mientras que los agresivos tienen mayor probabilidad de ser diagnosticados como un cáncer de intervalo, que es el que ocurre entre exámenes preventivos. Los tumores de crecimiento lento suelen tener mejor pronóstico. Son más indolentes y suelen ser detectados con un examen de diagnóstico. También tienen un mejor pronóstico, lo cual puede distorsionar el aparente beneficio del examen preventivo. Podría parecer que menos personas murieron debido al examen preventivo, cuando de hecho tenían menor probabilidad de morir por esa causa, dado que los tumores eran más indolentes.

15

Screening for Cancer Length Bias

Screening Test

Onset Cancer

Screening Test

Diagnosis by Symptoms

Screening Test

This particular diagram helps to illustrate that. If we look at this particular situation here, this individual was actually diagnosed with the cancer before this individual was --- however, this --- [I’m sorry] --- developed cancer before this individual was. But this individual was diagnosed with cancer based on symptoms in between this test --- screening test and this screening test. This particular individual had not developed symptoms at all by the time they got to the screening test, and so that screening test found their cancer. Clearly, this is a less aggressive cancer and, thus, more likely to have a better prognosis.

Este diagrama ayuda a explicarlo. En este caso en particular, comprobamos que este paciente desarrolló cáncer antes que este otro; sin embargo, este paciente fue diagnosticado con cáncer sobre la base de los síntomas que aparecieron entre este examen preventivo y este otro. Este paciente en particular no había desarrollado ningún síntoma cuando se hizo el examen preventivo, y el examen detectó el cáncer. Claramente, este es un cáncer menos agresivo y, por lo tanto, tiene mayor probabilidad de un mejor pronóstico.

It’s important to ask, “Are screen-detected outcomes better than outcomes from clinically detected disease?” If the answer is no, we don’t need to be doing cancer screening.

Es importante preguntarse si los resultados de los exámenes preventivos son mejores que cuando la enfermedad se detecta clínicamente. Si la respuesta es “no”, entonces no es necesario un examen preventivo.

Diagnosis by Screening

Screening for Cancer Always Ask….

Are screen-detected outcomes better than outcomes from clinically detected disease?

16

Screening for Cancer Basic Tenet of Screening

Remember… • It’s hard to improve on asymptomatic

One of the basic tenets of screening is, it’s hard to improve on asymptomatic. Remember that the population that we are screening is asymptomatic individuals. So, if we are going to medically intervene on these asymptomatic individuals, we need to do so based only on strong evidence of good benefits because there will be some harms that can occur.

Uno de los principios básicos de la prevención es la dificultad para mejorar a los pacientes asintomáticos. Estamos evaluando una población de individuos asintomáticos. Si vamos a intervenir médicamente en estos pacientes asintomáticos, debemos hacerlo sólo sobre la base de beneficios contundentes, ya que puede haber aspectos perjudiciales.

As a physician, I took an oath: "First, do no harm." In reality, what I try to do is more good than harm. It’s important to realize that all interventions have harms in addition to benefits including cancer screening tests.

Como médica, hice un juramento: “En primer lugar, no hacer daño”. Lo que trato de ofrecer es más beneficio que daño. Es importante advertir que todas las intervenciones, incluso los exámenes preventivos, causan daños y beneficios.

• If we are going to medically intervene on people with no symptoms, do so based only on strong evidence of good benefits

Screening for Cancer Physician’s Oath

First, do no harm… • In reality, try to do more good than harm • All interventions have harms in addition to benefits, including screening tests

17

In looking at the harms of cancer screening:

Veamos los daños de los exámenes preventivos:

Clearly false positives and false negative are harms. These are ones that we have already talked about. False negatives can cause anxiety that, had the test actually been negative, which it should have been since the individual doesn’t have the disease, they would not have experienced. They also have to undergo testing and procedures that they would not have had to undergo. And these tests and procedures can carry risk. Conversely, false negatives may actually result in a delay in seeking medical care even if there are symptoms. Someone who has just undergone screening for a particular disease and then a couple of months later develops symptoms may say, “This isn't anything to be worried about. I just had a test and they saw nothing.” So, they don’t do anything, in which case they are now living with the disease that is clearly advancing and can lead to a worse outcome.

Es evidente que los falsos positivos y los falsos negativos tienen aspectos perjudiciales. Ya hemos hablado de ellos. Los falsos negativos pueden causar una ansiedad que, si el examen hubiera dado negativo —ya que el paciente no tiene la enfermedad—, no se hubiera experimentado. Los pacientes deben someterse innecesariamente a pruebas y procedimientos que conllevan un riesgo. Por el contrario, los falsos negativos pueden demorar la atención médica, aun con síntomas. La persona que se hace un examen preventivo de una enfermedad en particular y, dos meses después, desarrolla síntomas, puede decir: “No me debo preocupar por esto. Acabo de hacerme un examen y no encontraron nada”. Entonces no toman ninguna medida y siguen viviendo con la enfermedad, que claramente avanza y puede generar un peor resultado.

Screening for Cancer

Harms of Cancer Screening

Screening for Cancer Harms of Screening

• False positives – Screening test results appear to be abnormal even though there is no cancer – Can cause anxiety and is usually followed by more tests and procedures, which also have risks

• False negatives – Screening test results appear to be normal even though there is cancer – May result in delay in seeking medical care even if there are symptoms

18

Screening for Cancer Harms of Screening

• Possible problems from screening test itself – e.g., perforated colon due to a colonoscopy

• Overdiagnosis – Diagnosing cancers that would never have been clinically relevant – Can result in overtreatment • unnecessary surgery, chemotherapy, radiation therapy

– Length bias extreme

Screening for Cancer Net Benefit

Other possible harms are from the screening test itself. For example, during the course of a colonoscopy there is a rare, but possible, risk of perforating the colon. Another harm that we commonly talk about with cancer screening is overdiagnosis. Overdiagnosis is diagnosing cancers that never would have been clinically relevant. Of course, the biggest issue with overdiagnosis is that it leads to overtreatment or treating a cancer that never would have been clinically relevant. So, now the individual is undergoing unnecessary surgery, chemotherapy, or radiation therapy, or possibly a combination of those. This is, in fact, a length bias extreme. This is a cancer that is so indolent that they would have lived their entire life without ever developing symptoms and would have died with the disease, but not of the disease.

Otros posibles daños resultan del propio examen preventivo. Por ejemplo, al hacer una colonoscopia existe el riesgo, infrecuente pero posible, de perforar el colon. Otro daño de los exámenes preventivos del cáncer es el sobrediagnóstico: diagnosticar cánceres que nunca hubieran sido clínicamente relevantes. Por supuesto, el mayor problema del sobrediagnóstico es que lleva a tratar, a veces de manera excesiva, un cáncer que nunca hubiera sido clínicamente relevante. El paciente es sometido innecesariamente a cirugía, quimioterapia o radioterapia, o quizá a una combinación de ellas. Este es un sesgo extremo de duración de la enfermedad. Es un cáncer tan indolente que el paciente hubiera vivido toda su vida sin síntomas, y hubiera muerto con la enfermedad, pero no a causa de ella.

In looking whether we should actually perform a cancer screening test, we look at the overall net benefit, and we try to balance the benefits and harms. If the benefits outweigh the harms, then we need to consider screening.

Al analizar si realmente debemos realizar un examen preventivo, evaluamos el beneficio neto general y tratamos de equilibrar los beneficios y los daños. Si los beneficios superan los daños, consideramos un examen preventivo.

Benefits > Harms Harm s

B e ne

fits

Consider Screening

19

Screening for Cancer Net Harm

Harms > Benefits B en e

However, if the harms outweigh, or are greater than the benefits, then scream --- screening --- is likely not appropriate for that individual. Of course these benefits and harms can vary depending on an individual’s level of risk.

Si los daños son mayores que los beneficios, o los superan, el examen preventivo no es apropiado para ese paciente. Por supuesto, tales beneficios y daños pueden variar en función del nivel de riesgo de cada paciente.

There are a number of resources for cancer screening guidelines. And I have listed these for you here. MD Anderson has developed risk-based cancer screening guidelines. And you can see them at our web site at www.mdanderson.org. Additionally, the American Cancer Society, the National Comprehensive Cancer Network, and the US Preventive Services Task Force have developed cancer screening guidelines that are also reliable resources.

Hay una serie de recursos que ofrecen pautas para los exámenes preventivos del cáncer. El MD Anderson ha desarrollado pautas basadas en el riesgo, que encontrará en nuestro sitio web, www punto MD Anderson punto org. La Sociedad Americana del Cáncer, la Red Nacional Integral de Cáncer y el Grupo de Tareas de Servicios Preventivos de los Estados Unidos han desarrollado pautas que también son recursos confiables.

fits

Harm s

Screening Likely not Appropriate

Screening for Cancer Cancer Screening Guidelines MD Anderson Cancer Center -www.mdanderson.org

American Cancer Society – www.cancer.org

National Comprehensive Cancer Network – www.nccn.org

United States Preventative Services Task Force (USPSTF) – www.ahrq.gov/clinic/prevenix.htm

20

Screening for Cancer Conclusions

• Screening should make sense – – – –

Screen appropriate populations Screen when it will change outcome Review the risks/costs and benefits Primum non nocere • First do no harm

Screening for Cancer Review of Evidence for Cancer Screening Tests

In conclusion, screening should make sense. We should be screening appropriate populations and screen only when it will change the outcome. We often need to review individually, and with our patient, the risk and cost as well as the benefits. In many cases, such as prostate cancer screening, we actually need to get input from the patient about whether they actually wish to proceed with the screening because it is so equivocal. Based on the evidence, we need to take their values about the potential harms from the screening into account. And remember always: First, do no harm.

En conclusión, los exámenes preventivos del cáncer deben ser una medida sensata. Debemos hacerlos en poblaciones apropiadas y sólo si pueden cambiar el resultado. Con frecuencia tenemos que analizar individualmente, y con el paciente, el riesgo, el costo y los beneficios. En muchos casos, como en el examen preventivo del cáncer de próstata, debemos preguntar al paciente si realmente desea continuar con el examen, dado que es equívoco. Según las pruebas, debemos considerar el valor del examen preventivo frente a posibles perjuicios. Y recuerde siempre: “En primer lugar, no hacer daño”.

Here is a web site at the NCI web site www.cancer.gov that will provide evidence for the individual cancer screening test. I hope you have found this lecture today to be beneficial. I would welcome feedback from you. Thank you.

El sitio web del NCI, www punto cáncer punto gov, ofrece datos para cada tipo de examen preventivo. Espero que esta disertación le haya resultado beneficiosa. Me agradaría recibir su opinión. Gracias.

• Available at the NCI website - www.cancer.gov

21