"Any sufficiently advanced technology is indistinguishable from magic.”

Arthur C. Clarke’s third adage is frequently quoted and may apply to the first experience one has with contemporary large language models (LLMs) like OpenAI’s ChatGPT 4.0,

 Depending on the prompt, these tools may appear to have human-like intelligence or a magical understanding. Speculation on the impact of LLMs or artificial intelligence (AI) more broadly in the future of health care is commonplace,

 causing excitement and concern about their use in health care settings.

Here we take a pragmatic approach to the near-term opportunities and risks to AI and LLMs in oncology. Specifically, we will focus on a few core ideas every oncologist should understand to make decisions about using these tools in their practice.

AI and LLMs represent an incremental advancement in computer science and statistics. The phrase “risk factor” is ubiquitous in medicine but was coined following the groundbreaking Framingham study.

Today, we hardly consider calculating a likelihood ratio to determine the accuracy of a physical exam predicting a phenotype or using a student 

 test to understand the difference in response between 2 populations.

Machine learning is broadly classified as a method to use a set of parameters to predict an outcome. In statistics, we call those parameters independent variables; the outcome is the dependent variable. In computer science, we use features to predict labels. The idea is consistent. Current LLMs may contain 

 of parameters. These parameters are used both to understand the prompt and to generate a natural language response to the prompt, which is why these models are also deemed as “generative” AI. The term “model” encompasses the black box that has been trained and is now ready to make predictions on new data.

The ability of AI and LLMs to impact oncology care delivery is unclear. Some believe that AI will more efficiently make care decisions with fewer mistakes. I feel that this will be unlikely in the near term. As mentioned above, each model must be trained on large datasets with labeled outcomes. Three things result from this:

Those models will be, at best, only as good as the care delivered in the training set.

Rare scenarios will be challenging to train models because of the lack of data.

Fields that advance quickly, such as oncology, will pose additional challenges as treatment recommendations based on past practice, but not incorporating recent clinical trial results, will be erroneous. Opportunities remain to improve notewriting, prior authorization, patient-facing health record portals, and others.

The 2 primary risks to AI and LLMs are confabulation and probabilistic prediction. Confabulation occurs because the model may give a result that appears to be factual but is entirely fictitious. For instance, an attorney was fined in 2023 for submitting case law that was erroneously produced by an LLM.

 In oncology, we have no tolerance for models incorrectly citing results of clinical trials or, worse, making up a clinical trial that did not occur. A second risk to AI is probabilistic prediction. When predicting an area under the curve, the same inputs will always yield the same output: deterministic output. With many of the AI models in this space, the same prompt run many times may provide very different outputs. We have not yet determined the best methods to offset this.

Simple clinical decision support, like drug interactions before prescribing chemotherapy, can be calculated in the electronic health record (EHR). Many AI models have a pay-per-query model. While EHR vendors are rapidly working to add these features, the financial impact of this is unclear as more AI models are included in clinical practice.

AI and LLMs will continue to be incorporated into the oncology practice. Like the invention and subsequent adoption of the EHR, these models will solve old problems and create new ones. The rising generation of oncologists should stay apprised as these fields grow to understand the trade-offs between opportunities and risks. I feel these tools will help us deliver better patient care, but it will take longer to optimize their benefits than many people think.

Travis J. Osterman, DO, is a medical oncologist, informatician, associate vice president for Research Informatics at Vanderbilt University Medical Center, as well as director of Cancer Clinical Informatics at the Vanderbilt-Ingram Cancer Center in Nashville, Tennessee. 

Profiles of the Future: An Inquiry Into the Limits of the Possible

OpenAI. GPT-4 is OpenAI’s most advanced system, producing safer and more useful responses. Accessed February 2, 2024. https://openai.com/gpt-4

‎Google. ‎Bard: a conversational tool by Google. Accessed February 2, 2024. https://bard.google.com

Meta. Discover the power of Llama. Accessed February 2, 2024. https://llama.meta.com

Thirunavukarasu AJ, Ting DSJ, Elangovan K, Gutierrez L, Tan TF, Ting DSW. Large language models in medicine. 

. 2023;29(8):1930-1940. doi:10.1038/s41591-023-02448-8

Sallam, M. ChatGPT utility in healthcare education, research, and practice: systematic review on the promising perspectives and valid concerns. 

. 2023;11(6):887. doi:10.3390/healthcare11060887

Cascella M, Montomoli J, Bellini V, Bignami E. Evaluating the feasibility of ChatGPT in healthcare: an analysis of multiple clinical and research scenarios. 

. 2023;47(1):33. doi:10.1007/s10916-023-01925-4

Dawber TR, Moore FE, Mann GV. Coronary heart disease in the Framingham study. 

 1957;47(4 Pt 2):4-24. doi:10.2105/ajph.47.4_pt_2.4

Neumeister L. Lawyers submitted bogus case law created by ChatGPT. A judge fined them $5,000. Associated Press News. June 22, 2023. Accessed February 2, 2024. https://apnews.com/article/artificial-intelligence-chatgpt-fake-case-lawyers-d6ae9fa79d0542db9e1455397aef381c

Articles

Speculation on the impact of large language models or artificial intelligence in the future of health care is commonplace, causing excitement and concern.

How to Make Clinical Decisions With AI in Oncology

Early-career oncologists and oncology trainees reported a multitude of issues around inadequate parental leave and support for lactating mothers, according to findings from a survey published in 

Findings from the survey demonstrated that among early-career or trainee clinicians with experience regarding parental leave and return to work after parental leave (n = 114), 79% felt overwhelmed with demands of work and home upon their return to work. Sixty percent of this group also reported feeling guilty when asking colleagues for help upon their return from parental leave and 50% felt pressured to work during their parental leave. Additionally, 40% responded that the duration of their parental leave was 4 weeks or less.

Program directors or associate program directors of training programs (n = 23) reported that their current parental leave duration is either adequate (61%) or too short (39%). Over half of the directors (52%) reported that their institution had a differing parental leave policy for trainees compared with faculty.

“I had my child during my fellowship and the more I spoke to others in training and early in their careers, I found that the experience of having inadequate parental leave, insufficient time, and lack of facilities for pumping breaks is almost universal,” Sindhu J. Malapati, MD, assistant professor of medicine, Division of Medical Oncology, Department of Internal Medicine, University of Arkansas for Medical Sciences, in Little Rock, and the study's lead author, said in an interview with 

. “I felt that the first step to fixing a problem is quantifying it, which is what I set about doing on this study.”

Investigators performed a cross-sectional study of clinicians in their early career and training, defined as the first 5 years after terminal training, regarding their knowledge of current work practices and guidelines and return-to-work experiences following parental leave. The survey was conducted online from May 5, 2021, to June 6, 2021, and was comprised of 43 questions about patient demographics, practice settings, parental leave, available resources to support parents returning to work, lactation policies and facilities, and experiences on returning to work. 

At baseline, respondents in early career and training (n = 255) consisted of trainees (71%) and early-career faculty (29%). Most respondents were female (54%), between 31 and 35 years old (58%), had children during training (51%), practiced in a university or academic setting (83%), and graduated from a US medical school (74%). 

Sixty percent of respondents were not aware of the parental leave policy of their certifying board and 40% had assigned parental leave during training. Nearly half (47%) said that they were not sure if parental leave or lactation policy documents were accessible by trainees or faculty at their current institution. Respondents were split regarding whether parental leave policies did or will play a role in the selection of their first or current job, with 37% responding yes, 45% responding no, and 18% saying that they were not sure.

Additional findings from the survey revealed that among respondents who had breastfed for any length of time (n = 65), 69% reported having access to the designated lactation room. However, only 40% felt comfortable asking for pumping breaks, 44% had an adequate duration of allotted pumping breaks to access and use the lactation room, and 66% said that pumping breaks were not mandated in their contract.

Among clinicians who took parental leave and returned to work after parental leave, only 19% reported having the option to work part time or choose lighter rotations during their transition back to work. Moreover, 32% of this group said they had additional responsibilities upon their return and only 27% said they felt they had the resources to address the transition back to work.

Notably, policy changes have been adopted by multiple governing bodies since the survey was conducted in 2021 aimed at addressing parental leave issues. In July 2022, the Accreditation Council for Graduate Medical Education applied new institutional requirements mandating that sponsoring institutions must have leave policies that include a minimum of 6 paid weeks off for medical, parental, and/or caregiver leave. In April 2023, the federal PUMP (Providing Urgent Maternal Protections) Act took effect, mandating employers to provide “a private, functional space that isn't a bathroom, and reasonable break time to pump each time you need to express milk while working.”

“There is increasing recognition of the attrition of female faculty, who are disproportionately affected by poor parental leave and lactation policies, in the early-career stage,” Malapati said. “Making the workplace more supportive for young parents with improved parental leave and lactation policies can be part of a multipronged approach to address the larger issue.”

Malapati SJ, Idossa D, Singh SRK, et al. Parent penalty: parental leave experiences of trainees and early-career faculty in oncology subspecialties. 

. 2023;19(10):899-906. doi:10.1200/OP.23.00242

ACGME answers: resident leave policies. Accreditation Council for Graduate Medical Education. September 28, 2022. Accessed February 5, 2024. https://www.acgme.org/newsroom/blog/2022/acgme-answers-resident-leave-policies/

Your breastfeeding rights. WIC Breastfeeding Support US Department of Agriculture. Accessed February 5, 2024. https://wicbreastfeeding.fns.usda.gov/your-breastfeeding-rights

Recent report details the issues early-career oncologists experience regarding parental leave.

Handling Parenting Pressures

In recognition of her innovation and empowerment, Shanta Dhar, PhD, FRSC, was named a fellow of the National Academy of Inventors for her ongoing research on combination therapeutic nanoparticles in prostate cancer.

Dhar is an associate professor of biochemistry and molecular biology and the assistant director of technology and innovation at the Sylvester Comprehensive Cancer Center’s NanoTherapeutics Research Laboratory, in Miami, Florida. The announcement was made in December 2023. 

Dhar joins Nobel Prize laureates and National Inventors Hall of Fame inductees, as well as National Academies of Sciences, Engineering, and Medicine members, in the 2023 class. The academy plans to recognize Dhar and the other fellows during an induction ceremony slated for June 18, 2024. 

In an interview with Oncology Fellows, Dhar discussed the significance of this honor and dove into the focus of her research in mitochondrial nanomedicine and the development of prodrugs with modulated mechanisms of action.

You were recently named a fellow of the National Academy of Inventors. Congratulations! What is the personal significance of this appointment to you? 

: It’s an honor to be inducted as a fellow of the National Academy of Inventors, particularly this year with this incredible list of people. Some of them are my mentors! 

To be a fellow of the National Academy of Inventors shows the appreciation of what we do day-to-day in our research toward innovation. I often tell my students, “Innovation: You often don’t see it. You have to feel it.” To be recognized as a fellow solidifies the aspect that innovation is important and that the work we are doing in the lab is innovative. I’m telling my students to keep being innovative and to start from a place where it looks like nothing is possible and then build on that.

What is your research focus and its potential clinical impact? 

[I have] a background in chemistry and targeted drug delivery, and we started thinking about how we can make targeted drug delivery a platform technology so that it can be applied to many diseases or different types of cancer. The mitochondria play a significant role in many diseases and in cancer now that we know more metabolic aspects. [Thus,] mitochondria are a standout as one of the targets to develop subplatform technologies. 

[However,] getting to the mitochondria is very challenging. In our lab we started working on how we can get things to the mitochondria of cells with a nanotechnology platform, particularly polymeric nanoparticles which can be degraded in the body and are more biodegradable and biocompatible, and can be fine-tuned for how fast or how slow the drug is released and how many drugs you can load in. 

In the context of different types of cancer we look at the pathways that are either overexpressed or altered in particular types of cancer. Once we know what mitochondrial target we have, then we develop molecules that can work on the pathway. We are developing this as a platform, using that same nanoparticle we have used from the beginning to get to the mitochondria, to then turn on and turn off particular pathways. 

[In terms of] clinical impact, these kinds of nanotechnologies are so new; [however,] all these recently approved vaccines, such as the COVID-19 vaccines, have nano components. Therefore, we feel that nano medicine is no longer foreign. However, the regulations for nano medicine, particularly for cancer, are still being developed. One challenge in the field is how to scale up this technology so that you can make off-the-shelf medicine. We are working very closely with clinicians to do work on that clinical aspect. We have been able to validate targets in clinical samples, and potentially we are looking at putting something in a real translational path.

Could you expand on the potential impact of developing nanoparticle technology for prostate cancer?

I have been working in prostate cancer for a long time. When I was a fellow at Massachusetts Institute of Technology [MIT], we developed different targeted nanoparticles that can get to the prostate cancer. [We worked on] developing polymers where you can load multiple drugs, which then can be released in a spatial temporal manner, meaning that in a particular area, 1 drug will release faster or another [will release] slower and you can have this control. 

When I started my lab back in 2010, the first-ever grant we received was a prostate cancer idea award from the Department of Defense that led us to develop this combination therapeutic nanoparticle for prostate cancer. Once I moved to The University of Miami, we received a research grant to work on this prostate cancer research from the Florida Department of Health. Within that pathway, we developed this particular molecule, which looks like shards of the fat metabolism in prostate cancer, which appears to particularly happen in advanced cancer.

However, fat metabolism also happens in the liver, so this kind of therapeutic needs to be targeted. We developed an orally administrable nanoparticle, which is targeted to prostate cancer, and we now have the ability to deliver this compound orally [for patients with this disease]. 

We are looking at how this can be combined with other kinds of therapy, whether you can also supplement this kind of nanoparticle with another nanoparticle, which will then prevent bone metastasis, which is one of the sites where prostate cancer metastases. Then, we are trying to find additional pathways that happen in advanced prostate cancer. We continue to design molecules and synthesize molecules to target those pathways.

What made you want to pursue this innovation in oncology? Where does your passion come from?

I realized the potential when I was in MIT. I worked with people who were my mentors and were great innovators. MIT celebrates innovation; [it is a place that empowers you to believe] nothing is wrong, you can ask any question, and you can do anything—and that drove the process. 

I understood how to file a patent because I had a lot of responsibility and freedom to do things [including] how to write a patent and how to work with a lawyer to get things filed. From some of my postdoc work, we were able to do a start-up and that gave me the push I needed. When I was an assistant professor in Georgia, I was able to put patents in very quickly because of that background and was able to launch my own start-up. 

We ended up running the company for a few years and received some venture funds. The focus has always been the innovation and the ability to make molecules and create new chemical entities; that is probably the core to all this innovation. I have been very lucky to have my trainees.

What are you looking forward to most as you join the National Academy of Inventors?

[I plan to] be very practical. Everyone has limited funds, and universities also have limited funds for filing patents. We are all competing for funds to propel our [research] forward. I hope that Sylvester Comprehensive Cancer Center and the University of Miami file all my ideas that I’m putting forward. 

I’m looking forward to all the upcoming meetings to interact with [people in the academy] to get to know them more, celebrate innovation, and keep pushing the research forward in this dimension. I’m [also] coleading an effort here with [The University of Miami’s] College of Engineering called Engineering Cancer Cures, where we are forming teams of cancer researchers and engineers to solve problems. 

For example, if you want to look at pancreatic cancer more closely where models are not available, can an engineer work on an [inexpensive] model and you look at the biology? These are very big ideas we are working on. I want to have a lot of focus in that area to promote more innovation here at Sylvester Comprehensive Cancer Center beyond my lab.

Sylvester cancer researcher and technology innovator named fellow of the National Academy of Inventors. News release. Sylvester Comprehensive Cancer Center. December 18, 2023. Accessed January 25, 2024. https://www. newswise.com/articles/sylvester-cancer-researcher-and-technology-innovatornamed- fellow-of-the-national-academy-of-inventors?channel=

Shanta Dhar, PhD, FRSC, discusses the significance of being named a fellow of the National Academy of Inventors for her research in prostate cancer.

Innovation Is the Answer to Making Progress Against Cancer

For years, society has been promised that the world would be changed by artificial intelligence (AI), with the medical profession no less susceptible to this hype. A sobering reminder of this optimism is a favorite quote from an article in 

“After hearing for several decades that computers will soon be able to assist with difficult diagnoses, the practicing physician may well wonder why the revolution has not occurred…”

Recent advancements with large language models and generative AI, especially with the popularity of ChatGPT, have amplified expectations of how AI will transform our lives. However, the practicing oncologist must still wonder how the revolution might affect them because its mainstream clinical presence is currently subtle at best. Here we will consider AI hype vs reality, with a specific focus on predictive modeling meant to answer medical questions regarding diagnosis, detection, prognosis, treatment choice, etc.

Predictive modeling using AI is the modern correlate of evidence-based medicine scoring algorithms that were rooted in expert opinion and accomplished through traditional statistics. Perhaps surprisingly, these simple calculators, the majority of which use stoichiometric point–assignment systems, are still the standard of care today with the Wells criteria for pulmonary embolism (PE) risk, CHA

-VASc for atrial fibrillation stroke risk, and others being taught at all levels of medical education.

Consider Phillip Steven Wells, MD, MSc, and his eponymous risk assessment tool. What is standard use throughout medicine today began in 1995 by “reviewing the literature and coming to a consensus on a scoring system” for a pilot study of 91 patients.

 He used expert opinion and literature review to identify factors that would split patients into low, moderate, and high pretest probability for PE and then tailored a scoring system that would replicate real-world risk stratification with acceptable error margins.

The model was used on a larger cohort of 529 patients in 1998

 and again in 2001, but with 1239 patients for his seminal article that cemented the importance of the Wells criteria to medical practice.

 Expectedly, there have been many updates to the Wells criteria, including AI-powered approaches. Modeling electronic health record (EHR)–derived features produced an area under the receiver operating characteristic (AUROC) performance of PE detection of 0.71 in one study,

 modeling 12-lead electrocardiograms (ECGs) alone to detect PE that was scored with an AUROC of 0.75 in another study.

 A fusion model of EHR features, ECG waveforms, and clinical notes natural language processing among 21,183 patients was the most impressive of all with an AUROC of 0.84 in a third study.

And yet, the Wells criteria system still reigns supreme despite having an AUROC decidedly lower in each of these model’s cohorts.

This is not a mistake, and understanding why the Wells criteria system is still the best choice for the average emergency department reveals the bollards holding back a surge of predictive medical models. The Wells criteria system is trusted because it leverages domain knowledge given that its scoring system is based on expert opinion, is easy to implement, has been externally validated numerous times, and accomplishes a single goal with a negative predicted value of 99.5% in appropriately selected patients.

These characteristics combine to make it very attractive to boots-onthe- ground clinicians. The highest scoring AI alternative discussed here used more than 35 independent EMH features in a large model, removing much of the domain knowledge that is captured with pointed feature selection prior to modeling, and thereby making the model cumbersome to implement or externally validate.

-VASc, in which several newer, AI-powered models significantly outperform CHA

-VASc head-to-head, but lack external validation, information on model calibration, and other necessary metrics of robustness and ultimately fall by the wayside.

Poor reproducibility in AI predictive modeling is the single greatest barrier to adoption in practice but pervades the field. A review that evaluated 86 radiologic diagnostic models found that 70 such models had decreased performance when applied to external data, 21 of which produced significantly incongruent results.

-VASc has at least 147 validation studies published to date.

 Creating standards by which to judge AI medical models is a contemporary challenge without clear guidance.

The FDA now regulates AI applications in medicine with the classification of software as a medical device (SaMD) and released an action plan addressing specific AI concerns.

 Its registry details over 690 AI/machine learning (ML)–powered medical devices, which have received 510(k) clearance, that were granted de novo request status, or given premarket approval. The rising wave of medical AI applications is clear, with a projected 30% increase in approved devices over the preceding year. Most of these devices are computer vision applications in radiology (79%), followed by various uses in cardiology (5%) and other subspecialties to a lesser degree. Only 15 devices have been approved in hematology, mostly AI-powered lab equipment like cell component analyzers, and no devices have been approved for oncology (although there are oncology-specific radiology models in the database). Of note, and why we mainly consider predictive AI models here instead of ChatGPT-like applications, is that as of late 2023, the FDA had not approved any device that leverages generative AI or large language models.

For those in favor of augmenting clinical practice using AI, the year-to-year FDA growth in AI/ML SaMD approval may be encouraging, but very few of those solutions are directly involved with predictive modeling as dreamt of by our colleague in 1987. To reach the relevance of the Wells criteria, a predictive model would likely need to satisfy all the of following:

The demographics of the training cohort should be representative of the intended use–scenario population, with special efforts to include minorities and marginalized groups.

The modeling algorithm selected should be the lowest complexity possible that is still able to achieve the desired outcome (don’t use deep learning if random forest would suffice).

While some of the features included in the model may be novel, other clinically relevant features guided by expert consensus should be included to infuse the model with domain knowledge.

The model must be evaluated with rigorous quality control metrics such as k-fold cross validation, bootstrapping, variance estimation, and other calibration analyses specific to the modeling strategy.

Iterative feature reduction should be done to decrease model complexity and allow ease of adoptability, and those remaining features should undergo importance weighting such as through Shapley statistics so that the end clinician understands why the prediction was made.

External validation must be completed at the time of publication.

It may seem unfair to compare burgeoning AI models to CHA

- VASc with its 147 validation studies. It is also true that given a good quality dataset, the most well-intending data scientist may accidently create a prediction model that is factually irrelevant but seemingly impressive with a high AUROC due to overfitting, highvariance hyperparameter tuning, and recursive training with only the incidentally highest scoring model being selected. External validation is the only true safeguard against these pitfalls, without which a predictive model hasn’t even entered the race against traditional evidence-based medicine scoring systems.

The current era in medicine will almost definitely see the big promises of AI come to fruition given the recent advancements in AI and the depth of good-quality data available in health care. Not far in the future we will see living databases that undergo nightly self-updates from EMR data streams and allow continuous retraining of models that combine clinical features, radiology, pathology, and more into true multimodal predictive machines. Getting there will require models that not only outperform their counterparts, but also demonstrate the maturity, rigor, and reproducibility that is expected in medicine. 

Jacob Shreve, MD, MS, is a Senior oncology fellow at Mayo Clinic in Rochester, Minnesota.  

Schwartz WB, Patil RS, Szolovits P. Artificial intelligence in medicine. Where do we stand? 

. 1987;316(11):685-688. doi:10.1056/ NEJM198703123161109

Wells PS, Hirsh J, Anderson DR, et al. Accuracy of clinical assessment of deepvein thrombosis. 

. 1995;345(8961):1326-1330. doi:10.1016/s0140- 6736(95)92535-3

Wells PS, Ginsberg JS, Anderson DR, et al. Use of a clinical model for safe management of patients with suspected pulmonary embolism. 

. 1998;129(12):997-1005. doi: 10.7326/0003-4819-129-12- 199812150-00002

Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and d-dimer. 

. 2001;135(2):98-107. doi: 10.7326/0003-4819-135-2-200107170-00010

Banerjee I, Sofela M, Yang J, et al. Development and performance of the pulmonary embolism result forecast model (PERFORM) for computed tomography clinical decision support. 

. 2019;2(8):e198719. doi:10.1001/jamanetworkopen.2019.8719

Silva BV, Marques J, Menezes MN, Oliveira AL, Pinto, FJ. Artificial intelligencebased diagnosis of acute pulmonary embolism: development of a machine learning model using 12-lead electrocardiogram. 

. 2023;42(7):643-651. doi: 10.1016/j.repc.2023.03.016

Somani SS, Honarvar H, Narula S, et al. Development of a machine learning model using electrocardiogram signals to improve acute pulmonary embolism screening. 

. 2021;3(1):56-66. doi:10.1093/ehjdh/ ztab101

Van der Endt VHW, Milders J, de Vries BBLP, et al. Comprehensive comparison of stroke risk score performance: a systematic review and meta-analysis among 6 267 728 patients with atrial fibrillation. 

. 2022;24(11):1739-1753. doi:10.1093/europace/euac096

Yu AC, Mohajer B, Eng J. External validation of deep learning algorithms for radiologic diagnosis: a systematic review. 

. 2022;4(3):e210064. doi:10.1148/ryai.210064

Artificial intelligence and machine learning in software as a medical device. FDA. September 22, 2021. Accessed January 31, 2024. https://www.fda.gov/ medical-devices/software-medical-device-samd/artificial-intelligence-andmachine- learning-software-medical-device 

We consider artificial intelligence hype vs reality, focusing on predictive modeling to answer medical questions about diagnosis, treatment choice, etc. 

The Medical AI Revolution

Andrea S. Porpiglia, MD, MSc, FACS, an associate professor in the Department of Surgical Oncology at Fox Chase Cancer Center in Philadelphia, Pennsylvania, has been promoted to assistant program director for the institution’s Complex General Surgery Oncology Fellowship.

Porpiglia initially arrived at Fox Chase as a surgical oncology fellow in 2013 and rejoined the institution as an instructor and surgeon in 2020 after 5 years of practice in the greater Philadelphia area. 

“One of the reasons, I decided to come back to Fox Chase was [the opportunity] to work with fellows and residents,” Porpiglia said in an interview with Oncology Fellows. “Teaching is a passion of mine, so having that opportunity to be able to do it is a true privilege. Moving forward, I want our fellowship to continue to compete with some of the bigger fellowships in the country, which we [already] do well. [I also want] to get our fellows more involved in research and more involved in national organizations. Getting them involved in leadership will help them in the future.” 

“We’re training more and more surgical oncologists, because there is a need for physicians with this unique skillset in both academia, and also in the community setting. As a result, there have been more programs coming online, to fulfil this need,” Sanjay S. Reddy, MD, FACS, associate program director, Complex General Surgical Oncology Fellowship; the Marvin S. Greenberg, MD, Chair in Pancreatic Cancer Surgery; associate professor, Department of Surgical Oncology; and codirector, Marvin and Concetta Greenberg Pancreatic Cancer Institute, all at Fox Chase Cancer Center, added in a separate interview with Oncology Fellows. “But to that end, our program has been one of the originals, storied in tradition, and has a great reputation of training well-rounded surgical oncologists.” 

Porpiglia was recognized as a 2023 Top Doctor by Philadelphia magazine and received the same distinction from Main Line Today from 2016 through 2018. She is also a member of the American Society of Clinical Oncology, the Association of Women Surgeons, the Pennsylvania Medical Society, and the Society of Surgical Oncology. 

In the interviews, Reddy and Porpiglia describe what makes the Complex General Surgical Oncology Fellowship at Fox Chase unique, outline the responsibilities of Porpiglia’s new role, and offer advice to aspiring and current surgical oncology fellows.

 What distinguishes the Complex General Surgical Oncology Fellowship at Fox Chase Cancer Center from other similar programs? 

: The Society of Surgical Oncology [SSO] fellowship is not a new fellowship. There is great tradition with this and historically it was 1 of the most challenging fellowships to obtain a position. The complexity of the cases and the limited programs made this a coveted opportunity. In 2014, the SSO decided to make this an ACGME [Accreditation Council for Graduate Medical Education]–accredited fellowship, which allowed for a more structured program and the opportunity for other programs to build and expand. [That year had] the inaugural class of the fellowship that went through the ACGME [program]. 

Our fellowship is uniquely structured, having 3 fellows per year, and 6 fellows in total. We also have a surgical oncology research fellowship, with 2 fellows total. We are all passionate about 1 thing, training the next generations of surgical oncologists, who may find their niche in academic medicine, or in a busy community practice. 

: We have 3 fellows per year and we are a dedicated cancer institution. I believe having the fellows only take care of patients with cancer [means that] we don’t have the distractions that some of the other academic sites may have with general surgery. Our fellows are able to focus on their fellowship and that alone. [Additionally], with the multidisciplinary care team that we have at Fox Chase, everyone works very closely together. The fellows are able to rotate on our medical oncology and radiation oncology colleagues in addition to having our tumor boards.

How did you know Dr Porpiglia was the right candidate for the role? 

: For any program to be successful, there has to be a team behind it. I was the associate program director for a couple of years when Jeffrey Farma, [MD, FACS,] was the program director. Then as he into other leadership positions, I became the program director of the fellowship. The role of the assistant program director is significant, and the need to problem solve and adapt is essential. It also helps when you have graduated from the program, as that perspective is priceless. There is no one better than Andrea, both of us trained at Fox Chase. I’ve known her for over 12 years, and her work ethic and dedication to the field of surgical oncology is unwavering.

What is this new role going to entail? What will the immediate contributions to the program look like? 

: This fellowship promotes surgical education, both clinically and academically. Our goal is to train broad-based, well-trained surgical oncologists. They [will be able to] go out and practice, whether they’re going to go out in the community or into academics. We’ve had a series of fellows over the years go back to their home countries and were able to bring the art of surgical oncology to their home institutions. This is 1 of the most rewarding aspects of being involved with the fellowship, seeing your trainees bring a much-needed expertise to the table.

: My role is to help with some of the administrative tasks that come with being a fellowship program director. We’re able to identify future surgical oncologists, [and need to] sort through the applications and pick which ones are going to be the best fit for our program and future stars in the field. Having gone through the program previously myself, the mentors that I was able to gain from being a fellow here I still have today and I still lean on them, even after being out [of the program] for 10 years.

What guidance would you offer to oncology fellows who are interested in pursuing a surgical focus?

: This fellowship is unique. It gives trainees who are residents who are looking to do a fellowship a way to increase their versatility. A surgical oncologist is well equipped to handle advanced complex hepatobiliary cases, [but] they could also do colorectal cases, head and neck cases, melanoma, soft tissue, breast oncology, [and so on]. Surgical oncology offers a variety of specialties within cancer surgery that fellows can train and hone their skills. It’s that broad-based training that allows our fellows to be successful in the environment of surgical oncology in this country and beyond.

Porpiglia: When you’re within a general surgery residency program, now you need to decide early on if that’s the trajectory you want to go. You need to be able to work in research—it plays an important role in our job—and you need to know everything about it, [including] how to start a project, how to see the project through, and how to get abstracts and publish. It’s helpful if the residents are able to get that [experience] during their residency program. From there, shadowing and getting mentors in the field is helpful in terms of navigating the process of applications and interviews to secure a position.

Dr Andrea Porpiglia promoted to assistant program director for complex general surgical oncology fellowship at Fox Chase Cancer Center. News release. Fox Chase Cancer Center. July 27, 2023. Accessed February 5, 2024. https://www.foxchase.org/news/2023-07-27-dr-andrea-porpigliapromoted-to-assistant-program-director-for-complexgeneral-surgical-oncology-fellowship-at-fox-chasecancer-center

Andrea Porpiglia, MD, MSc, FACS, and Sanjay Reddy, MD, FACS describe what makes the Complex General Surgical Oncology Fellowship at Fox Chase unique. 

Coming Full Circle

Reflecting on the many years since my fellowship graduation, I realize how we learn has drastically changed. When I was a fellow, we used textbooks. We did not have up-to-date information in our hands on a cell phone, we did not have robust internet searches, podcasts, tweetorials, and the like. Data were transmitted via a didactic lecture format even if that did not suit the learning needs of all participants. The concept of a flipped classroom was a radical redesign. Major learning events were usually synchronous and in person. Passive learning was the norm.

With the evolution of social media and the acceptance of virtual and asynchronous learning, however, we as educators now recognize that the playing field has changed. The way younger generations gather information and learn differs greatly from my generation. It is ironic because this article will land squarely in a print journal, which may not be your choice of media. In fact, we recognize that this publication must evolve to utilize modern educational platforms to reach our readers.

To that end, we seek to query you, our stakeholders. As an editor, I recognize that millennial and Generation Z learners tend to prefer a different format from what is being delivered by this print publication. Data demonstrate that Gen Z learners prefer active learning, by doing and often through digital methods. Millennials may appreciate micro learning, defined as learning something quickly with bite-size content rather than sitting through extended classes. Thus, we want to target how we deliver content based on your needs.

 are dedicated to artificial intelligence (AI). One can imagine that AI might be able to notice patterns and automate educational processes or responses. This could create a real-time interactive experience. While much is written about the concerns of using AI, there are likely ways to smartly integrate this tool into learning strategies that uphold the principles of learning theory and target the new generation of learners.

Again, this is where you come in. In order for this publication to meet your needs, please take this 

 to help us identify strategies that resonate with our readers so that we may best deliver our educational content to you.

Editor's Note: Jill Gilbert, MD, is a professor of medicine and director of the Hematology/Oncology Fellowship Program at Vanderbilt University Medical Center, in Nashville, TN.

Jill Gilbert, MD, explains how publications must evolve to utilize modern educational platforms to reach readers.

Vol. 16/No. 1 | Oncology Fellows

How to Make Clinical Decisions With AI in Oncology

Handling Parenting Pressures

Innovation Is the Answer to Making Progress Against Cancer

The Medical AI Revolution

Coming Full Circle

Evolving With the Times