The best answer is "It Depends". Really - it depends upon the vendor.
Traditionally IPSec VPN's have the same inherent risk as L2TP and PPTP did in regards to the spread of viruses. Typically if a remote machine becomes infected, it could easily broadcast and propagate itself to any corporate hosts due to the layer 3 nature of an IPSec tunnel (or Layer 2 for L2TP/PPTP).
I state traditionally because there are a new wave of IPSec tunnel terminators that are content aware. That is, they have the ability to not only terminate tunnel traffic but also inspect some popular protocols such as HTTP, SMTP, POP, IMAP and FTP. Still not perfect because uncommon ports can still be broadcast and MAPI is not being scanned.
SSL VPN’s on the other hand offer another layer of defense. As stated above in the thread, some remedial vendors only offer secure access to internal web servers (greatly reducing your networks exposure to Trojans and viruses). Furthermore, some vendors who have a true SSL tunneling technology actually operate at layer 5 instead – proxying all connections (remember the ole’ stateful vs. proxy firewall battles).
In this case, since all connections are proxied – an infected remote machine can only send attacks to pre-defined destinations that are allowed to be proxied; they are not able to broadcast them to your entire corporate network. Typically the destinations you are proxing are either well known mail, web or application servers and should be hardened in your DMZ. Further security could be implemented by deploying a gateway AV behind your SSL appliance and/or having strict AV software enforcement on your core servers that will be accessed via the SSL VPN.