he COVID- 19
pandemic haschanged manythings during thepast year and ahalf, and it’s had an indelibleeffect on healthcare. During theshutdown of 2020, when doctors’ offices were not open, telehealth boomed. Even thoughdoctors and clinics are takingin-person appointments again,things are not expected to getback to normal, especially whenit comes to sales reps interactions with physicians. Expertssay while telehealth appointments will decrease overall, incertain therapeutic categoriesthey could stay the same or evenincrease. And as many officesstill remain closed to sales reps,pharma will need to engage withphysicians and patients in theseonline spaces.
The impact of COVID- 19
and the telehealth boom
Figures from McKinsey & Com-
pany show how physician ex-
pectations of how to interact
with pharma have been reset
by telehealth. McKinsey reports
that in the first quarter of 2020
compared with the same period
in 2019, there was a 50 percent
increase in telemedicine and en-
during utility in some categories.
And 42 percent of physicians adjusted their prescribing habits,with one example being preferring to write scrips for medicinesthat could be administered bythe patients themselves. In April
2020, overall telehealth utilization for office visits and outpatient care was 78 times higherthan in February 2020, but hassince stabilized at levels 38 timeshigher than before the pandemic, at levels ranging from 13 to 17
percent across all specialties.
“This utilization reflects morethan two-thirds of what we anticipated as visits that could bevirtualized,” wrote McKinsey’sOleg Bestsennyy, Greg Gilbert,Alex Harris, and Jennifer Rostin July.
These experts say they are seeing differences in the uptake oftelehealth across all specialties,with the highest penetration inpsychiatry at 50 percent, andsubstance abuse at 30 percent.
Rounding out the top five are
endocrinology and rheuma-
tology at 17 percent each, and
gastroenterology at 16 percent.
Only 2 percent of visits in ortho-
pedic surgery and opthalmology
were telehealth ones.
As of April 2021, 84 percentof physicians were offering virtual visits and 57 percent wouldprefer to continue offering virtual care.
In the meantime, HCP pref-
erences for how they choose to
engage with pharma have also
evolved, McKinsey reports, with
75 percent of physicians indicating they will have restrictedaccess in 12 months. Sixty-eightpercent of the physicians McKinsey surveyed said they preferto be engaged by remote onlyor by a hybrid engagement ofremote and low-frequence inperson interactions. And 50 percent of the HCPs who said theywanted to be interacted with ina hybrid manner said they onlywanted to be engaged one tothree times a year.
Additionally, 75 percent ofthe physicians surveyed byMcKinsey reported a mismatchbetween preferred and actualchannel mix in engagementswith pharma. These physicianssaid they were unsatisfied withthe quality of the offering frommany pharmaceutical companies, stating that there are manyunmet needs existing relatedto scientific collaboration andpeer-to-peer exchange.
Genevieve P. Kanter, Ph.D.,
writing in the September 8,
2020 issue of JAMA, observed
that, “The upheavals brought
about by COVID- 19 have led to
growth of telehealth and other
services that have been advances
in how health care is delivered.”
She adds that these changes
“also present an opportunity to
reset the boundaries and norms
for physician-industry interac-
“As clinics and hospitals, as
well as drug and device firms,
regroup, policy makers, physi-
cians and patients should be in-
tentional in re-establishing rules
for doctors’ engagement with
industry,” she says.
According to Dr. Kanter, theswitch to video meetings andonline presentations has madedrug promotion far less costlyfor companies, with “no travelexpenses for sales reps and doctors, no fine dining events, andlow fixed costs of producing video content that can reach a largeaudience. History and economics tell us that the lower the costsof doing something, the morefirms will engage in it.”
Riding the wave
For ePocrates, athenahealth’smobile medical reference app,the increase in telehealth andonline interactions has produced an increase in interest inwhat the company can do frompharmaceutical clients.
10•MEDADNE WSAUGUS T 2021potentially largemarket is starting tocoalesce around theendocannabinoidpharmaceutical market – and itisn’tdue to the makers of CBDoil for wellness purposes.
Even as more states are approving medical marijuanaand dispensaries have boomedduring the pandemic, the lackof clinical data around “natural”CBD and other plant-derivedcompounds – along with thefederal government’s statutorydisapproval of any product thatcan get a user high – means thatthe companies who want to advance drugs that work with theendocannabinoid system havea difficult scientific road aheadof them. They must formulatedrugs that can target the receptors that trigger the “good” effects without triggering the bad(see Q&A with Yuval Cohen,CEO of CorbusPharmaceu-ticals Holdings, for an explanation of the endocannabinoidsystem and what it does).Creating these drugs is difficult – but not impossible. GWPharmaceuticals managedto strike gold from the greenwith the U.S. approval of Epidiolex in June 2018 for severeforms of epilepsy. The drug isthe first plant-derived cannabinoid prescription medicine andthe only FDA-approved form ofcannabidiol (CBD). Epidiolexwas initially approved by theFDA in June 2018 for the treatment of seizures associated withLennox-Gastaut syndrome andDravet syndrome in patients 2years of age and older. In July2020, the FDA approved theproduct for the treatment of seizures associated with tuberoussclerosis complex (THC) in patients 1 year of age and older.While Epidiolex containsCBD, it does not contain thecompound that can get usershigh, THC. The drug achievedsales of more than $500 millionduring 2020.
“Epidiolex sales increased by
over 70 percent in 2020 despite
the challenges of COVID- 19,
reflecting the positive impact
this medicine has on patients as
well as the performance of our
commercial team,” stated Jus-
tin Grover, GW’s CEO, in Janu-
ary. “We remain encouraged by
our patients’ experience on this
product, as demonstrated by
high persistence and refill rates.
This, combined with our expansion of payer coverage andthe recently approved tuberoussclerosis complex indication,leads us to expect continuedstrong growth in 2021 in boththe U.S. and Europe.”GW has been seeking FDAapproval for a second canna-bis-derived product, Sativex(nabixomols), for the treatmentof spasticity due to multiplesclerosis. Unlike Epidiolex, Sativex contains THC in additionto CBD, as well as minor cannabinoids and other non-cannabi-noid components. Phase III trials for Sativex are under way inthe United States, with resultsexpected later in 2021.
GW’s success made it thesubject of an acquisition. ThecompanybecamepartofJazzPharmaceuticalsduring May
2021. “The addition of GW further diversifies our commercial portfolio and innovativepipeline with therapies that arecomplementary to our existingbusiness, including Epidiolex, ahigh-growth commercial product with near-term blockbusterpotential,” says Bruce Cozadd,chairman and CEO of Jazz.
Unlike GW/Jazz, other companies are optingto head downthe synthetic route rather thannaturally derived route, withlab-created versions of CBD andTHC.Thegrandaddyofthesedrugs is Marinol (dronabinol),approved in 1985 for treatinganorexia associated with weightloss in patients with AIDS, andnausea and vomiting associated with cancer chemotherapyin patients who have failed torespond adequately to conventional antiemetic treatments.
Other companies are choos-
ing to go a different route all
together by creatingsmall-mol-
ecule drugs that target specific
Two companies – Corbus andEmerald Health Pharmaceuticals – have drugs in development for indications suchas dermatomyositis, systemicsclerosis, obesity, and cancer(see chart for more info).
Corbus’ lead drug is lenaba-
sum, an oral, non-immuno-
suppressive preferential can-
nabinoid- 2 agonist. The drug
failed the primary endpoint of
its Phase III trial for systemic
sclerosis in September 2020. Dr.
Cohen is not discouraged, how-
ever, tellingMed Ad News that
lenabasum did show some effect
and the company will be discuss-
ing the results with FDA and de-
termining next steps.medadnews
Although the endocannabinoid system has been intensely studied,there are still few drugs approved or in development that interact with it.
By Christiane Truelove • firstname.lastname@example.org
CANNABINOID DRUG PIPELINE SNAPSHOT
COMPANY PRODUCT INDICATION PHASEOFDEVELOPMENTJazz Pharmaceuticals(via GW Pharmaceuticals)
Seizures associated withLennox-
Gastaut syndrome, Dravet syndrome,
Note: Jazz Pharmaceuticals completedthe acquisition of
GWPharmaceuticals andthe company’s cannabinoid-based
prescription medicines and product pipeline on May 5, 2021
Nabiximols MSspasticity PhaseIIISpinalcordinjuryspasticity PhaseIII(planned)
Additional Cannabinoids Schizophrenia Phase IIAdditional Cannabinoids Autismspectrumdisorders Phase IIAdditional Cannabinoids Neonatalhypoxic-ischemic encephalopathy Phase IAdditional Cannabinoids Neuropsychiatrytargets Phase IUndisclosedtargets Cannabinoids Preclinical
CorbusPharmaceuticalsHoldings Lenabasum Dermatomyositis PhaseIIISystemicsclerosis PhaseIIISystemiclupuserythematosus PhaseIICystic;brosis PhaseIIbCB1InverseAgonists Metabolicdisorders PreclinicalCB2Agonists Cancer Preclinical
EmeraldHealthPharmaceuticals EHP-101(vce-004.8) Multiplesclerosis PhaseIISystemicsclerosis(scleroderma) PhaseIIEHP-102(vce-003.2) Parkinson’sdisease PreclinicalHuntington’sdisease Preclinical
Beyond medical marijuana
AUGUST 2021 MEDADNEWS• 15
harma’s Class of 2019 ismaking the rare beware.
Four of the top six 2019
launches in 2020 salestarget rare diseases. Trikafta, forcystic ;brosis, generated nearly $4billion in sales for Vertex last year andmight hit $5 billion this year. P; zer’sVyndaqel andVyndamax are the;rst and only medicines approvedby FDA to treat the cardiomyopathyof wild-type or hereditary transthyretin-mediated amyloidosis.Alexion’s Ultomiris, the ; rst and onlylong-acting C5 complement inhibitoradministered every eight weeks,is approved for bothparoxysmalnocturnal hemoglobinuria andatypical hemolytic uremic syndrome,and is being studied in amyotrophiclateral sclerosis and generalizedmyasthenia gravis. And perhapsmost extraordinary of all, Novartis’Zolgensma has transformed care forspinal muscular atrophy, a disorderwhose worst iterations were nearlyan inevitable death sentence just afew years ago. All are expensive – at$2.125 million per dose in the UnitedStates, Zolgensma might just bethe most expensive drug productever marketed anywhere – and allhave rewritten the care possibilitiesfor their respective targets, not tomention the scope of what’s possiblefor drug researchers.
ivacaftor and ivacaftor) was ;rst
approved by FDA in October 2019
for the treatment of cystic ; brosis in
people ages 12 years and older who
have at least one F508del mutation
in the cystic ; brosis transmembrane
conductance regulator (CFTR) gene,
the most common CF-causing
mutation. With this approval, for the
;rst time, about 6,000 people with
CF ages 12 years and older who have
one F508del mutation and one min-
imal function mutation (F/MF) had
a medicine that targets the under-
lying cause of their CF. Additionally,
about 12,000 people withone or
two F508del mutations who were
currently eligible for one of Vertex’s
three other FDA-approved CF medi-
cines became eligible for Trikafta.
“Today marks a milestone for CF
patients, their families and Vertex,”
said Je;rey Leiden, M.D., Ph.D.,Ver-
tex’s chairman, president, and CEO,
upon announcement of the approv-
al.“After a 20-year journey together,
we have received FDA approval of
Trikafta: a single breakthrough med-
icine with the potential to treat up
to 90 percent of all people with CF in
the future. I want to personally thank
the hundreds ofVertex scientists who
have been working onthis program
for nearly 20 years – many of whom
have dedicated their entire careers to
changing the course of this disease;
the CF Foundation which has pro-
vided support, encouragement and
help throughout the journey; and
most importantly the thousands of
patients, caregivers, doctors and ad-
vocates who have courageously and
persistently worked side-by-side with
us to get to where we are today.”
In July 2020, Vertex announced
results of a global Phase III study of
Trikafta in people with cystic ; brosis
ages 12 years and older who have
one copy of the F508del mutation
and one gating mutation (F/G) or
one copy of the F508del mutation
and one residual function mutation
(F/RF).The study met its primary end-
point of mean absolute within-group
change in percent predicted forced
expiratory volume in 1 second
(ppFEV1) from baseline through 8
weeks of treatment, demonstrating a
statistically signi;cant 3. 7 percentage
point improvement in ppFEV1 in
patients treated with Trikafta com-
pared to their baseline after a 4-week
run-in of treatment on ivacaftor or
tezacaftor/ivacaftor. The study met
all secondary endpoints, including
a statistically signi; cant mean with-
in-group reduction of 22. 3 mmol/L
Phase III studies with Trikafta.
“The results of this study demon-
strate that the triple combination
provides signi; cant additional
bene; t compared to existing CFTR
modulator therapy for F/G and F/
RF patients and adds to the robust
body of evidence supporting the
bene; t of this medicine for patients
with at least one F508del mutation,”
said Carmen Bozic, M.D., executive VP,
Global Medicines Development and
Medical A;airs, and Chief Medical
O;cer at Vertex.
In September 2020, Vertexannounced that FDA had accepted three supplemental NewDrugApplications for Trikafta, Symdeko,and Kalydeko.These regulatory submissions were intended to expandthe labels for Trikafta, Symdeko, andKalydeko to include additional rareCFTR mutations, allowing peoplewith cystic ;brosis not previouslyeligible for these medicines an opportunity to bene;t from treatmentthat targets the underlying cause oftheir disease. In addition, the submissions may also allow certain peoplewith CF who are currently eligiblefor Kalydeco to become eligible forSymdeko or Trikafta andcertain people currently eligible for Symdekomay become eligible for Trikafta.The regulatory submissions werebased on data from an in vitro cellassay showing that these rare CFTRmutations respond to one or more ofthese CFTR modulator regimens.
Also in September 2020, Vertex
announced the completion of a
global Phase III study of Trikafta in
children ages 6 through 11 years old
with cystic ;brosis who have either
two copies of the F508del mutation
or one copy of the F508del mutation
and one minimal function muta-
tion. Based on the results, Vertex
announced its intention to submit a
supplemental New Drug Application
to FDA inthe fourth quarter of 2020,
with additional global regulatory
submissions to follow.
In December, FDA expanded the
eligibility for Trikafta to include peo-
ple with cystic ; brosis ages 12 years
and older with certain mutations in
the cystic ;brosis transmembrane
conductance regulator (CFTR) gene
that are responsive to Trikafta based
on invitro data. Symdeko and
Kalydeco also received approvals to
include additional responsive muta-
tions in people withCF ages 6 years
and older and age 4 months and
older, respectively. These approvals
allow more than 600 people with CF
not previously eligible for these med-
icines an opportunity to potentially
bene;t from treatment that targets
the underlying cause of their disease.
In April 2019, FDA approved AbbVie’s Skyrizi (risankizumab-rzaa), aninterleukin- 23 (IL- 23) inhibitor, forthe treatment of moderate-to-severeplaque psoriasis inadults who arecandidates for systemic therapy orphototherapy. In clinical trials, Skyriziproduced high rates of durable skinclearance. Results from the trialsshowed most people (82 and81 percent) treated with Skyrizi achieved90 percent skinclearance (PASI 90) atone year, with the majority (56 and
60 percent) achieving complete skinclearance (PASI 100).The approval of Skyrizi was supported by results from AbbVie’s global Phase III psoriasis program, whichassessed the safety and e;cacy ofSkyrizi in adults with moderate-to-severe plaque psoriasis across four randomized, placebo and/or active-con-trolled pivotal studies: ultIMMa-1,ultIMMa- 2, IMMhance, and IMMvent.The co-primary endpoints of thesestudies were Psoriasis Area andSeverity Index (PASI 90) andstaticPhysician Global Assessment [sPGA]score of clear or almost clear [sPGA0/1] at 16 weeks versus placebo.InultIMMa-1 and ultIMMa- 2 at16 weeks, PASI 90 was achieved in75 percent of people treated withSkyrizi, compared to 5 and 2 percentreceiving placebo, respectively.PASI 100 was achieved in36 and51 percent of people treated withSkyrizi, compared to 0 and 2 percentreceiving placebo, respectively.InultIMMa-1 and ultIMMa- 2 at oneyear (52 weeks), 82 and 81 percent ofpeople treated with Skyrizi achievedPASI 90, and 56 and 60 percentachieved PASI 100, respectively.Anintegrated analysis of ultIM-Ma-1 and ultIMMa- 2 showed mostpeople treated with Skyrizi whoachieved PASI 90 and PASI 100 at
Treatments for rare diseasesdominate the list of best-sellersamong pharma’s Class of 2019.
By Joshua Slatko • email@example.com
The rares inherit the earth
Quarterly sales of Trifakta
Vertex’s fixed dose combination cysticfibrosis product Trikafta generated$3.86 billion in sales in 2020 andanother $2.45 billion in the first halfof 2021.
AbbVie’s autoimmune drug Skyriziearned $1.59 billion in its first full yearon the market and has added another$1.25 billion in the first half of 2021.
Quarterly sales of Skyrizi
By Abidur RahmanVP, innovation at Intouch Group
he American Heart Association Get with the Guidelines–Heart Failure Risk Scorepredicts the risk of deathin patients admitted to the hospital. Itassigns three additional points to anypatient identified as “nonblack,” therebycategorizing all black patients as beingat lower risk. The AHA does not providea rationale for this adjustment.
The Kidney Donor Risk Index,implemented by the national KidneyAllocation System in 2014, uses donor characteristics, including race, topredict the risk that a kidney graft willfail. The race adjustment is based onan empirical finding that black donors’kidneys perform worse than nonblackdonors’ kidneys, regardless of the recipient’s race. The developers of the KDRIdo not provide possible explanations forthis difference. If the potential donor isidentified as black, the KDRI returns ahigher risk of graft failure, marking thecandidate as a less suitable donor.The Vaginal Birth after Cesarean(VBAC) algorithm predicts the riskposed by a trial of labor for someonewho has previously undergone cesarean section. It predicts a lower likelihood of success for anyone identifiedas African American or Hispanic. Thestudy used to produce the algorithmfound that other variables, such asmarital status and insurance type, alsocorrelated with VBAC success. Thosevariables, however, were not incorporated into the algorithm.
The STONE score predicts the likelihood of kidney stones in patients whopresent to the emergency departmentwith flank pain. The “origin/race” factor adds three points (of a possible 13)for a patientidentified as “nonblack.”The developers of the algorithm did notsuggest why black patients would beless likely to have a kidney stone. Aneffort to externally validate the STONEscore determined that the origin/racevariable was not actually predictive ofthe risk of kidney stones.
Each of these tales appeared in “Hidden in plain sight – Reconsidering theuse of race correction in clinical algorithms,” an article published in the August 27, 2020 issue of the New EnglandJournal of Medicine.
Large-scale data analysis and artificial
intelligence have transformed every
square inch of health care, very much
for the better. But examples like those
mentioned are the “through a glass,
darkly” of our data analytics and AI
revolution. Just as seeing acelebrity put
her imprimatur on a brand sometimes
grants that brand more respect than it
deserves, it’s easy to get carried away by
the hype surrounding an extraordinary
new technology. We need to be wary,
though, to avoid being blindsided by the
challenges that must be addressed in or-
der to derive the full benefit of that new
technology – in this case, the challenges
of bias. Humans, of course, have biases,
and sometimes data does too.
For example, it’s much easier to collect data from large, well-funded healthsystems with robust record keeping thatmostly serve well-offpatients with easyaccess to care than it is to collect datafrom small, under-resourced hospitalsthat mostly serve poorer, underservedpopulations. Fed with data that’s mostlysourced from the well-funded and thewell-off, an AI algorithm is bound todraw conclusions that simply won’t apply to the underserved, harder-to-reachpart of the population. By doing the veryjob it is designed to do – uncoveringrelevant trends in large data sets – evena well-designed AI will only magnify thebiases present in its source data. Basedin part on a lack of robust data fromAfrican-Americans, some AI systemshave concluded thatAfrican-Americanpopulations are often healthier thanCaucasian populations – a conclusionthat is demonstrably false but has stillhad a significant impact on real care onthe ground. The price of such biaseddata and conclusions can be substandard care for the very populations whocan least afford it.
Race and ethnicity and social classaren’t the only sources of potential bias,either. Depending on the circumstances,the people collecting data or building anAI algorithm may have a bias towardsa specific action, or against one. Theymight be consciously or subconsciouslyfavoring a diagnosis of X because Xmeans more dollars for their researchproject or brand or health system. Theymight be overlooking certain data because it has the opposite effect, itdoesn’tsupport the goal that a particular company might be promoting. And once thatsort of bias leaks into the analysis, theentire exercise becomes self-fulfilling.The price to be paid for that sort of biascan be millions of dollars in researchfunding wasted, or – worse – patientswith Y being treated for X, or not beingtreated at all.
The COVID pandemic has offered
an education in this. AI and large-scale
data analysis tools have been a godsend
for public health authorities trying to
track incidence and forecast for resource
allocation, not to mention the scien-
tists developing treatments, especially
in the United States. Butit’s been very
easy to see that the data coming outof
different countries, or even different
parts of the same country, have reflected
varying levels of bias. Politicians want
to encourage a sense of normalcy or
promote tourism; and so researchers or
analysts, dependent on those politicians
for funding, may consciously or uncon-
sciously alter the parameters of what a
COVID-related death is. People, a great
many people in a great many places,
have suffered and died due to that kind
All this is why AI equity must be amatter of priority for anyone and everyone in our industry who touches AI ordata analysis.
AI equity refers to how effective AIis in a broad range of scenarios with avariety of population groups and demographics. Achieving it is a challenge thatall AI engineers and datascientists inthe health care space must face, and thesooner the better.
It’s easier said than done, of course. AIsystems have to be trained by humanswho are sometimes biased. And theneven if the training is as unbiased asitcan possibly be, AI systems have tobe fed data, which is also sometimesbiased.
So whatcan we do about it?
Every one of those standards and scoresabove were created by teams of expertsin their respective fields. But all thoseexperts may not have been expert indata science, and they almost certainlydid not reflect the ethnic, demographic, or sociological complexity of thepatient pools for whom the standardswere being set. And in the companysetting, it’s quite common for AI ordata analytics teams to include expertswho know all about the technology, butnot so much about the demographicsand needs of the patient population inquestion. So we need to be sure thatthe standard-setting body or team isdiverse enough in itself to represent thefull population over which the standardwill carry, and that the body has the dataand AI expertise within itself to knowwhat data and AI bias looks like.
Whenever we are building AI models
that will touch patient care, we need to
be sure that every bitof data we put into
them – whether during the training of
the model or its use – mustbe of good
quality and truly representative of the
population in question. Plenty of natural
roadblocks exist in the way of achieving
this, the well-funded versus under-re-
sourced health system issue being only
one of the more prominent. But as a
matter of basic fairness and justice, we
must overcome those roadblocks.
The danger of bias should be on theminds of AI, data, and analytics teams atall times, from the creation and trainingof any AI tools through testing and validation and the data collection processand review of any conclusions thatmight be reached. That’s what good scientists do, after all; they look atevery-thing with a healthy dose of skepticism.Always be asking yourself, “Is the datareally complete? Are there missing pieces in the data? Will this really representand serve the full patient population?”The search for bias should begin withthe beginning of every project, and endwith … well, never.
Yes, transparency in how AI is beingtrained, how and why data is being collected, how conclusions might be implemented, every step of the process. Internally this means everyone on the teamhas the opportunity to call out any biasthey might see along the way. Externallyit means building trust with any patientcommunity that might be touched.Some groups may be less likely to sharepersonal health information with datacollectors than others, a variance thatcan easily introduce bias into a data set.So we must be very clear about what weare collecting, why we are collecting it,and whatthe privacy policies are. If weare to ask our patient population to trustus with their most personal information,we need to trust them back.None of this is easy, and all of it requires some rowing against the current.Someone is always going to push to finish the project more quickly and cheaply. The aggressive seeking of bias andAI equity will almost inevitably makeprojects more time- and resource-inten-sive than companies expect. But in theend it is good business as well as goodethics to do so. Better trained AI toolsand better data means more patientsgetting the right treatment and fewerpatients getting the wrong one. Itmeansa patient population that has observedin their own lives how serious you areaboutseeking out what’s unique aboutthem. It means fulfilling the immensepotential of AI and data science to improve human health. medadnews
SPECIALFEATURE HEALTHTECHNOLOGYData and AI scientists must strive to eliminatebias from anything that touches patient care.
will see you ...
Telehealth boomed during thepandemic, and experts believethat even with the world openingback up, pharma companieswill have to continue to shiftand strengthen tactics to reachphysicians and patients virtually.